Plant Lipids 9781071613610, 9781071613627

Citation preview

Methods in Molecular Biology 2295

Dorothea Bartels Peter Dörmann Editors

Plant Lipids Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-by step fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

Plant Lipids Methods and Protocols

Edited by

Dorothea Bartels and Peter Dörmann Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany

Editors Dorothea Bartels Institute of Molecular Physiology and Biotechnology of Plants (IMBIO) University of Bonn Bonn, Germany

Peter Do¨rmann Institute of Molecular Physiology and Biotechnology of Plants (IMBIO) University of Bonn Bonn, Germany

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-1361-0 ISBN 978-1-0716-1362-7 (eBook) https://doi.org/10.1007/978-1-0716-1362-7 © Springer Science+Business Media, LLC, part of Springer Nature 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover Caption: Mass spectrometry (MS) imaging of phosphatidylcholine (PC) from a cotton (Gossypium hirsutum, variety ‘Coker 312’) embryo. From left to right: light microscopy image of section used in MS imaging (scale bar ¼ 1 mm), PC 34:2 (max mol% ¼ 55), PC 36:3 (max mol% ¼ 35), colorimetric scale bar representing mol% from green (low) to red (high). This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface Lipids have diverse roles in plant cells such as establishing the membrane bilayer or serving as energy storage or signaling molecules. Acyl lipids are the most abundant group of lipids and represent a structurally broad family of fatty acid-derived compounds. In addition, plants contain considerable amounts of sterol lipids and sphingolipids which are mainly found in the plasma membrane, and they have been implicated in developmental and defense-related processes. The compartments of the plant cell are characterized by the occurrence of specific lipid classes, and each plant tissue has its specific lipid profile. In accordance with their functions, lipids are present in very different amounts in the cells, ranging from highly abundant structural and storage lipids to minor amounts in the case of signaling lipids. Many lipid classes like membrane glycerolipids and triacylglycerol are characterized by the existence of numerous molecular species with characteristic fatty acid patterns. The abundance of the molecular species of a lipid class represents an important consideration to determine the strategy for their analysis. The focus of this protocol series is on analytical methods to study complex lipid mixtures from plants and algae. This book has assembled 26 chapters, which all together cover a broad range of state-of-the-art methods and technologies. Isolation of lipids from plants requires special care because these molecules are highly nonpolar, they require organic solvents for extraction, and lipase activities need to be inhibited to prevent lipid degradation. Thin-layer chromatography represents a well-established technique for obtaining an overview of the composition of a crude lipid extract. Many lipids including fatty acids can be derivatized to increase their volatility for analysis by gas chromatography (GC) or gas chromatography-mass spectrometry (GC-MS). Lipidomics is an emerging technology which involves the structural identification and quantification of the molecular species of a lipid class in a whole plant organ or in a tissue. Progress in lipidomics has been driven in particular by advances in mass spectrometry (MS). Two alternative strategies are currently employed for high-throughput lipid analyses, i.e., direct infusion MS (shotgun lipidomics) and liquid chromatography-mass spectrometry (LC-MS). Furthermore, imaging methods based on MALDI-MS techniques have been developed to localize different lipids to the tissue level. Next to MS, nuclear magnetic resonance (NMR) spectroscopy represents a highly important technique for structural analysis of lipids. Different protocols have been developed for the isolation and characterization of lipids from specific tissues or subcellular compartments. Two chapters describe the isolation and quantification of extremely low abundant signaling lipids of the phosphoinositide family. While the analysis of interactions of lipids with proteins is a highly important research field for systems biology and for putting lipids in a metabolic context, these methods are covered in only two chapters and we refer here to protocols in other publications including this series. Analyses of mutants and genomic sequences of plants and algae have contributed toward elucidating lipid biosynthesis pathways and cellular lipid interaction networks. Therefore, overviews are presented on lipid databases, which include lipid structural information and pathway analyses. All chapters have been written by experts in the field, and the methods have been optimized and thoroughly tested in the respective laboratories. The methods have been designed or optimized for lipids from plants or algae. The required chemicals and equipment as well as the experimental steps are described in detail, followed by extensive

v

vi

Preface

troubleshooting notes. As part of the general workflow, some procedures such as lipid isolation, thin-layer chromatography, or gas chromatography are described in more than one chapter. This allows for comparisons of the methods and demonstrates the possible variations in the different laboratories. The methodologies range from simple to technically demanding approaches. Some of the simple approaches can be used for mass screening strategies of mutants, or they can be integrated into protocols for student classes. However, some methods such as MS or NMR analyses require dedicated, expensive equipment and can only be performed in specialized laboratories. We trust that the protocols will be useful not only for experienced researchers but that also undergraduate, graduate, or Ph.D. students will use these protocols as a reliable guide for their experiments. One of our objectives is to encourage scientists with no or little experience in lipid experiments to approach this research field experimentally, as lipid research will be highly important for the understanding of plant metabolism, both on the structural and signaling level. We thank all contributing authors for excellent co-operations. We also thank Prof. John M. Walker, University of Hertfordshire, UK, for guidance in assembling the chapters and Anna Rakovsky from Springer Nature for assistance in getting the chapters ready for publication. Bonn, Nordrhein-Westfalen, Germany

Dorothea Bartels Peter Do¨rmann

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PART I

BASIC METHODS OF LIPID ISOLATION AND ANALYSIS

1 Lipid Isolation from Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jesse D. Bengtsson, James G. Wallis, and John Browse 2 Three Methods to Extract Membrane Glycerolipids: Comparing Sensitivity to Lipase Degradation and Yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Samira Mahboub, Zachery D. Shomo, R. Maxwell Regester, Mahaa Albusharif, and Rebecca L. Roston 3 Thin-Layer Chromatography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Georg Ho¨lzl and Peter Do¨rmann 4 Lipid Analysis by Gas Chromatography and Gas Chromatography–Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mathias Brands, Philipp Gutbrod, and Peter Do¨rmann 14 5 C-Tracing of Lipid Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hari Kiran Kotapati and Philip D. Bates 6 Methods of Lipid Analyses for Microalgae: Charophytes, Eustigmatophytes, and Euglenophytes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Masako Iwai, Shiori Shibata, Hiroyuki Ohta, and Koichiro Awai

PART II

v xi

3

15

29

43 59

81

MASS SPECTROMETRY AND NMR ANALYSIS

7 Direct Infusion Mass Spectrometry for Complex Lipid Analysis. . . . . . . . . . . . . . . Katharina Gutbrod, Helga Peisker, and Peter Do¨rmann 8 Fatty Acid Composition by Total Acyl Lipid Collision-Induced Dissociation Time-of-Flight (TAL-CID-TOF) Mass Spectrometry. . . . . . . . . . . . . . . . . . . . . . . . Pamela Tamura, Carl Fruehan, David K. Johnson, Paul Hinkes, Todd D. Williams, and Ruth Welti 9 Targeted Analysis of the Plant Lipidome by UPLC-NanoESI-MS/MS . . . . . . . . Cornelia Herrfurth, Yi-Tse Liu, and Ivo Feussner 10 Mass Spectrometry-Based Profiling of Plant Sphingolipids from Typical and Aberrant Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rebecca E. Cahoon, Ariadna Gonzalez Solis, Jennifer E. Markham, and Edgar B. Cahoon 11 Analysis of Free and Esterified Sterol Content and Composition in Seeds Using GC and ESI-MS/MS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Richard Broughton and Fre´de´ric Beaudoin

vii

101

117

135

157

179

viii

12

13

14

Contents

Techniques for the Measurement of Molecular Species of Acyl-CoA in Plants and Microalgae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Richard P. Haslam and Tony R. Larson Quantification of Acyl-Acyl Carrier Proteins for Fatty Acid Synthesis Using LC-MS/MS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Lauren M. Jenkins, Jeong-Won Nam, Bradley S. Evans, and Doug K. Allen Structural Analysis of Glycosylglycerolipids Using NMR Spectroscopy . . . . . . . . 249 Wiebke Knaack, Georg Ho¨lzl, and Nicolas Gisch

PART III LIPID ISOLATION AND ANALYSIS FROM PLANT TISSUES CELL COMPARTMENTS AND ORGANELLES 15

16 17 18 19

Analysis of Extracellular Cell Wall Lipids: Wax, Cutin, and Suberin in Leaves, Roots, Fruits, and Seeds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Johanna Baales, Viktoria V. Zeisler-Diehl, and Lukas Schreiber Isolation of Lipid Droplets for Protein and Lipid Analysis. . . . . . . . . . . . . . . . . . . . Patrick J. Horn, Kent D. Chapman, and Till Ischebeck Isolation of Plastoglobules for Lipid Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Denis Coulon and Claire Bre´he´lin Isolation of Mitochondria for Lipid Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Se´bastien Leterme, Morgane Michaud, and Juliette Jouhet Collection and Analysis of Phloem Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Susanne Hoffmann-Benning

PART IV 20

21

22

23

24

275 295 321 337 351

LIPID SIGNALING, LIPID-PROTEIN INTERACTIONS, AND IMAGING

Analyses of Inositol Phosphates and Phosphoinositides by Strong Anion Exchange (SAX)-HPLC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debabrata Laha, Marı´lia Kamleitner, Philipp Johnen, and Gabriel Schaaf Analysis of Phosphoinositides from Complex Plant Samples by Solid-Phase Adsorption Chromatography and Subsequent Quantification via Thin-Layer and Gas Chromatography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Larissa Launhardt, Monique Matzner, Mareike Heilmann, and Ingo Heilmann Studying Lipid–Protein Interactions Using Protein–Lipid Overlay and Protein–Liposome Association Assays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Guido Ufer, Peter Do¨rmann, and Dorothea Bartels Investigations of Lipid Binding to Acyl-CoA-Binding Proteins (ACBP) Using Isothermal Titration Calorimetry (ITC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ze-Hua Guo and Mee-Len Chye In Situ Localization of Plant Lipid Metabolites by Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry Imaging (MALDI-MSI) . . . . . . . . . . Drew Sturtevant, Mina Aziz, Trevor B. Romsdahl, Chase D. Corley, and Kent D. Chapman

365

379

391

401

417

Contents

PART V 25 26

ix

LIPID DATABASES

Plant Lipid Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 Peter Do¨rmann Lipid Pathway Databases with a Focus on Algae . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 Naoki Sato and Takeshi Obayashi

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

469

Contributors MAHAA ALBUSHARIF • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA DOUG K. ALLEN • Donald Danforth Plant Science Center, St. Louis, MO, USA; Plant Genetics Research Unit, USDA-ARS, St. Louis, MO, USA KOICHIRO AWAI • Department of Biological Science, Faculty of Science, Shizuoka University, Shizuoka, Japan; Research Institute of Electronics, Shizuoka University, Hamamatsu, Japan MINA AZIZ • Department of Biological Sciences, University of North Texas, Denton, TX, USA; BioDiscovery Institute, University of North Texas, Denton, TX, USA JOHANNA BAALES • Department of Ecophysiology, Institute of Cellular and Molecular Botany, University of Bonn, Bonn, Germany DOROTHEA BARTELS • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany PHILIP D. BATES • Institute of Biological Chemistry, Washington State University, Pullman, WA, USA FRE´DE´RIC BEAUDOIN • Plant Sciences Department, Rothamsted Research, Harpenden, UK JESSE D. BENGTSSON • Institute of Biological Chemistry, Washington State University, Pullman, WA, USA MATHIAS BRANDS • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany CLAIRE BRE´HE´LIN • CNRS, Laboratoire de Biogene´se Membranaire, UMR 5200, Univ. Bordeaux, Villenave d’Ornon, France RICHARD BROUGHTON • Institute of Aquaculture, University of Stirling, Stirling, UK JOHN BROWSE • Institute of Biological Chemistry, Washington State University, Pullman, WA, USA EDGAR B. CAHOON • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA REBECCA E. CAHOON • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA KENT D. CHAPMAN • Department of Biological Sciences, University of North Texas, Denton, TX, USA; BioDiscovery Institute, University of North Texas, Denton, TX, USA MEE-LEN CHYE • School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China; State Key Laboratory of Agrobiotechnology, The Chinese University of Hong Kong, Shatin, Hong Kong, China CHASE D. CORLEY • Department of Biological Sciences, University of North Texas, Denton, TX, USA; BioDiscovery Institute, University of North Texas, Denton, TX, USA; Department of Molecular Genetics, University of Texas Southwestern Medical Center, Dallas, TX, USA DENIS COULON • CNRS, Laboratoire de Biogene´se Membranaire, UMR 5200, Univ. Bordeaux, Villenave d’Ornon, France; Bordeaux INP, Talence, France PETER DO¨RMANN • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany BRADLEY S. EVANS • Donald Danforth Plant Science Center, St. Louis, MO, USA

xi

xii

Contributors

IVO FEUSSNER • Department of Plant Biochemistry, Albrecht-von-Haller-Institute for Plant Sciences, University of Go¨ttingen, Go¨ttingen, Germany; Service Unit for Metabolomics and Lipidomics, Go¨ttingen Center for Molecular Biosciences (GZMB), University of Go¨ttingen, Go¨ttingen, Germany; Department of Plant Biochemistry, Go¨ttingen Center for Molecular Biosciences (GZMB), University of Go¨ttingen, Go¨ttingen, Germany CARL FRUEHAN • Kansas Lipidomics Research Center, Division of Biology, Kansas State University, Manhattan, KS, USA NICOLAS GISCH • Division of Bioanalytical Chemistry, Research Center Borstel, Leibniz Lung Center, Borstel, Germany ZE-HUA GUO • School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China KATHARINA GUTBROD • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany PHILIPP GUTBROD • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany RICHARD P. HASLAM • Department of Plant Science, Rothamsted Research, Harpenden, UK INGO HEILMANN • Charles Tanford Protein Center, Department of Cellular Biochemistry, Institute for Biochemistry and Biotechnology, Martin-Luther-University Halle-Wittenberg, Halle (Saale), Germany MAREIKE HEILMANN • Charles Tanford Protein Center, Department of Cellular Biochemistry, Institute for Biochemistry and Biotechnology, Martin-Luther-University Halle-Wittenberg, Halle (Saale), Germany CORNELIA HERRFURTH • Department of Plant Biochemistry, Albrecht-von-Haller-Institute for Plant Sciences, University of Go¨ttingen, Go¨ttingen, Germany; Service Unit for Metabolomics and Lipidomics, Go¨ttingen Center for Molecular Biosciences (GZMB), University of Go¨ttingen, Go¨ttingen, Germany PAUL HINKES • Waters Corporation, Milford, MA, USA SUSANNE HOFFMANN-BENNING • Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI, USA GEORG HO¨LZL • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany PATRICK J. HORN • Department of Biology, East Carolina University, Greenville, NC, USA TILL ISCHEBECK • Albrecht-von-Haller-Institute for Plant Sciences and Go¨ttingen Center for Molecular Biosciences (GZMB), Department of Plant Biochemistry, University of Go¨ttingen, Go¨ttingen, Germany MASAKO IWAI • School of Life Science and Technology, Tokyo Institute of Technology, Yokohama, Japan LAUREN M. JENKINS • Donald Danforth Plant Science Center, St. Louis, MO, USA; Plant Genetics Research Unit, USDA-ARS, St. Louis, MO, USA PHILIPP JOHNEN • Department of Plant Nutrition, Institute of Crop Science and Resource Conservation, Rheinische Friedrich-Wilhelms-University Bonn, Bonn, Germany; BASF SE, Limburgerhof, Germany DAVID K. JOHNSON • Computational Chemical Biology Core, University of Kansas, Lawrence, KS, USA JULIETTE JOUHET • CNRS, CEA, INRAE, IRIG, Laboratoire de Physiologie Cellulaire & Ve´ ge´tale, Universite´ Grenoble Alpes, Grenoble, France MARI´LIA KAMLEITNER • Department of Plant Nutrition, Institute of Crop Science and Resource Conservation, Rheinische Friedrich-Wilhelms-University Bonn, Bonn, Germany

Contributors

xiii

WIEBKE KNAACK • Division of Bioanalytical Chemistry, Research Center Borstel, Leibniz Lung Center, Borstel, Germany HARI KIRAN KOTAPATI • Institute of Biological Chemistry, Washington State University, Pullman, WA, USA DEBABRATA LAHA • Department of Biochemistry, Indian Institute of Science, Bengaluru, Karnataka, India TONY R. LARSON • Department of Biology, University of York, York, UK LARISSA LAUNHARDT • Charles Tanford Protein Center, Department of Cellular Biochemistry, Institute for Biochemistry and Biotechnology, Martin-Luther-University Halle-Wittenberg, Halle (Saale), Germany SE´BASTIEN LETERME • CNRS, CEA, INRAE, IRIG, Laboratoire de Physiologie Cellulaire & Ve´ge´tale, Universite´ Grenoble Alpes, Grenoble, France YI-TSE LIU • Department of Plant Biochemistry, Albrecht-von-Haller-Institute for Plant Sciences, University of Go¨ttingen, Go¨ttingen, Germany SAMIRA MAHBOUB • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA JENNIFER E. MARKHAM • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA MONIQUE MATZNER • Charles Tanford Protein Center, Department of Cellular Biochemistry, Institute for Biochemistry and Biotechnology, Martin-Luther-University Halle-Wittenberg, Halle (Saale), Germany MORGANE MICHAUD • CNRS, CEA, INRAE, IRIG, Laboratoire de Physiologie Cellulaire & Ve´ge´tale, Universite´ Grenoble Alpes, Grenoble, France JEONG-WON NAM • Department of Bioindustry and Bioresource Engineering, Sejong University, Seoul, South Korea TAKESHI OBAYASHI • Tohoku University, Sendai, Miyagi, Japan HIROYUKI OHTA • School of Life Science and Technology, Tokyo Institute of Technology, Yokohama, Japan HELGA PEISKER • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany R. MAXWELL REGESTER • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA TREVOR B. ROMSDAHL • Department of Biological Sciences, University of North Texas, Denton, TX, USA; BioDiscovery Institute, University of North Texas, Denton, TX, USA REBECCA L. ROSTON • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA NAOKI SATO • Department of Life Sciences, University of Tokyo, Tokyo, Japan GABRIEL SCHAAF • Department of Plant Nutrition, Institute of Crop Science and Resource Conservation, Rheinische Friedrich-Wilhelms-University Bonn, Bonn, Germany LUKAS SCHREIBER • Department of Ecophysiology, Institute of Cellular and Molecular Botany, University of Bonn, Bonn, Germany SHIORI SHIBATA • Department of Biological Science, Faculty of Science, Shizuoka University, Shizuoka, Japan ZACHERY D. SHOMO • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA ARIADNA GONZALEZ SOLIS • Department of Biochemistry and Center for Plant Science Innovation, University of Nebraska-Lincoln, Lincoln, NE, USA

xiv

Contributors

DREW STURTEVANT • Department of Biological Sciences, University of North Texas, Denton, TX, USA; BioDiscovery Institute, University of North Texas, Denton, TX, USA; Department of Molecular Genetics, University of Texas Southwestern Medical Center, Dallas, TX, USA PAMELA TAMURA • Kansas Lipidomics Research Center, Division of Biology, Kansas State University, Manhattan, KS, USA GUIDO UFER • Institute of Molecular Physiology and Biotechnology of Plants (IMBIO), University of Bonn, Bonn, Germany; Bayer AG Research & Development, Monheim, Germany JAMES G. WALLIS • Institute of Biological Chemistry, Washington State University, Pullman, WA, USA RUTH WELTI • Kansas Lipidomics Research Center, Division of Biology, Kansas State University, Manhattan, KS, USA TODD D. WILLIAMS • Mass Spectrometry Laboratory, University of Kansas, Lawrence, KS, USA VIKTORIA V. ZEISLER-DIEHL • Department of Ecophysiology, Institute of Cellular and Molecular Botany, University of Bonn, Bonn, Germany

Part I Basic Methods of Lipid Isolation and Analysis

Chapter 1 Lipid Isolation from Plants Jesse D. Bengtsson, James G. Wallis, and John Browse Abstract Analysis of plant lipids provides insights into a range of biological processes, from photosynthetic membrane function to oil seed engineering. Many lipid extraction protocols are tailored to fit a specific lipid class. Here we describe a procedure for extraction of glycerolipids from vegetative tissue. This procedure is designed for 1 gram of tissue per sample but maybe scaled for larger samples. Key words Chloroform–methanol, Glycerolipids, Lipases, Lipid extraction, Plant lipid

1

Introduction Plant lipids are essential components of every tissue type and as such, accurate measurements of their composition are important to a thorough understanding of plant biology. Lipid analysis can be a routine procedure which greatly amplifies a plant scientist’s toolbox. Critical to this analysis is understanding that plant tissues, particularly leaves, are replete with a range of lipases that complicate lipid extraction. Additionally, the fatty acid substituents of lipids can oxidize unless proper care is taken to protect polyunsaturated fatty acids from damage [1]. Phospholipases present in vegetative tissues can rapidly alter the lipid composition, generating artifacts which can be difficult to identify. Products of lipase activity may be misidentified as other lipids of interest, leading to embarrassing errors [2]. The problem begins with tissue disruption, releasing phospholipases which frequently remain active in the organic solvents used for lipid extractions, even retaining activity at low temperatures (
16  C) in some cases [3, 4]. Detection of intermediates in lipid synthesis such as lysophospholipids, free fatty acids, and phosphatidic acid as large fractions of the lipid sample may indicate phospholipase activity [1]. Inactivation of this potential lipase activity is essential for successful plant lipid extraction. A straightforward way to achieve this is to use a solvent system containing formic or acetic acid at

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021

3

4

Jesse D. Bengtsson et al.

cold temperature. Boiling the sample in isopropanol with 0.01% butylated hydroxytoluene (BHT) as antioxidant before proceeding with the lipid extraction also reduces artifact production [5]. Lipid extractions typically use chloroform, methanol, and water, in proportions that produce a monophasic mix. A classic example of this is the Bligh and Dyer method designed for the high water content of fish tissue [6]. It is important to note that a protocol designed for plant vegetative tissue may not need the addition of water, due to the high-water content of leaves, while lipid extraction from dry seeds will require added water. Generating a monophasic mix permits efficient lipid extraction from the tissue. This is followed by addition of salts to obtain a phase separation leading to partition of the lipids. Nonlipid components segregate into the water-rich layer while most lipids partition to the chloroform layer. However, highly polar lipids may partition partially into the aqueous phase, in an equilibrium responsive to negative charges on the lipid. The addition of acids such as phosphoric acid produces mildly acidic conditions that protonate charged lipids and neutralize their charge, so that they partition more efficiently into the chloroform layer (Fig. 1a). Chloroform–methanol–water procedures readily extract glycerolipids, but some sphingolipids are not readily extracted, and a more involved extraction than discussed here is required [7]. Here we describe a simple method for extracting glycerolipids from vegetative plant tissue. This procedure limits oxidation of fatty

Fig. 1 Lipid isolation and quantification by GC separation of FAMEs. (a) Phase separation of an extract from an Arabidopsis thaliana leaf sample provides for partitioning of the lipids into the lower chloroform layer which is green due to the presence of chlorophyll. (b) Separation of FAMEs derived from Arabidopsis leaf lipids by GC with 17:0 as internal standard allows quantification of the total glycerolipids in the sample

Lipid Isolation from Plants

5

acids by keeping temperatures low, and inhibits lipase activity with acetic acid. Like both the Folch [8], and Bligh and Dyer [6] methods, we use chloroform, methanol, and water (present in the vegetative tissue) along with acetic acid to generate a monophasic solution. We induce phase a separation by adding aqueous KCl and phosphoric acid. Lipids partition to the lower chloroform layer. Neutral and most polar lipids readily partition into the chloroform layer, but highly polar lipids such has lysophosphatidic acid maybe partially soluble in the aqueous methanol and water layer. Quantitative extraction of these lipids is achieved by back extraction with chloroform. Other lipids, particularly some intermediates such as acyl-ACPs and acyl-CoAs, are soluble in the water–methanol layer and may bind to protein at the water–chloroform interface [9, 10]. Estimating the quantity of glycerolipids from a sample is accomplished by derivatizing the substituent fatty acids to methyl esters. The fatty acid methyl esters (FAME) can be separated and analyzed by gas chromatography with a flame ionization detector (GC-FID). Use of an internal standard when derivatizing fatty acids esterified to glycerolipids allows for easy quantification of FAMEs. An internal triacylglycerol standard such as triheptadecanoylglycerol (17:0-TAG) normalizes for the efficiency of fatty acid derivatization. Detection of FAMEs by FID provides a nearly linear response to the carbons in the acyl chains, allowing for reliable quantification of the sample (Fig. 1b).

2

Materials

2.1 Preparing Solvents

1. Fume hood. 2. Glass graduated cylinder (1 L). 3. Storage bottles with solvent-resistant polypropylene caps. 4. Chloroform (HPLC grade). 5. Methanol (HPLC grade). 6. Glacial Acetic acid. 7. Phosphoric acid. 8. Potassium chloride. 9. Toluene (HPLC grade). 10. Butylated hydroxy toluene (BHT). 11. Sulfuric acid. 12. 2 L Erlenmeyer flask.

2.2 Harvesting Tissue and Storage

1. 50 mL Falcon tubes (e.g., Fisher Scientific, Cat # 14-432-22). 2. Two tube racks accommodating 50 mL Falcon tubes. 3. Liquid nitrogen.

6

Jesse D. Bengtsson et al.

4. Vacuum insulated Dewar flask for liquid nitrogen. 5. Two insulating foam boxes. 6. Sharp scissors. 7. Balance with sufficient precision for your samples. 8.
80  C freezer. 2.3 Tissue Homogenization

1. Mortar and pestle. 2. 30 mL solvent-resistant fluorinated ethylene/propylene polymer Nalgene tubes. 3. 10 mL glass pipettes and safety bulb or pipette plunger. 4. Liquid nitrogen. 5. Solvent mixture chloroform–methanol–glacial acetic acid (10:10:1, v/v/v) (12 mL per sample).

2.4

Lipid Extraction

1. Refrigerated centrifuge that can accommodate solventresistant tubes. 2. 25 mL borosilicate tubes. 3. Teflon lined caps for 25 mL glass tubes. 4. Glass Pasteur pipettes. 5. Pipette bulb for glass Pasteur pipettes. 6. Vortex mixer. 7. Clinical centrifuge suitable for 25 mL glass tubes. 8. Organic extraction solution chloroform–methanol–acetic acid (5:5:1, v/v/v) (5 mL per 1 g tissue sample). 9. Extraction buffer 1 M KCl, 0.2 M H3PO4 (6 mL per 1 g tissue sample). 10. Gas evaporation manifold equipped with a heating block able to heat samples at 40  C. 11. Nitrogen gas.

2.5 Quantification of Total Lipids

1. Gas chromatograph with spilt/splitless injector and flame ionization detection (see Note 1). 2. Carbowax capillary GC column (e.g., 10, 0.53 mm, 15 m, Sigma-Aldrich).

SUPELCOWAX

®

3. Triheptadecanoylglycerol (17:0-TAG) standard. 4. Toluene (HPLC grade). 5. Butylated hydroxy toluene (BHT). 6. Pipettes (P200) or Drummond Digital Microdispenser (see Note 2). 7. 8 mL glass tubes. 8. Teflon-lined caps for glass tubes.

Lipid Isolation from Plants

7

9. Hot water bath set to 85  C. 10. Hexanes (mixture of hexane isomers, e.g., Sigma-Aldrich HX0296) or n-hexane. 11. GC vials (1 mL) with caps. 12. GC vial inserts (300 μL).

3

Methods

3.1 Preparation of Solvents (Protocol Designed for Samples of 1 g or Less)

1. The solvent mixtures should be prepared 1 day before extracting lipids. Because the mixtures contain chloroform, a known carcinogen, use a fume hood and wear gloves, goggles and a lab coat when working with chloroform. 2. Prepare chloroform–methanol–glacial acetic acid (10:10:1, v/v/v) (see Note 3). Prepare sufficient volume for 12 mL per sample with some to spare. Using a glass graduated cylinder and working in a fume hood, mix the appropriate volumes of chloroform, methanol and acetic acid together into a borosilicate bottle, then cap the bottle and store at
20  C. 3. Using a glass graduated cylinder and working in a fume hood, prepare chloroform–methanol–acetic acid (5:5:1, v/v/v). Prepare sufficient volume to use 5 mL per sample, with some to spare. 4. Prepare 1 M KCl, 0.2 M H3PO4 (6 mL per g of sample). For 1 L: Add to 74.6 g of KCl to 500 mL of deionized water in a 2 L Erlenmeyer flask. Add a magnetic stir bar and place on a stir plate, gently stirring the sample. Using a graduated cylinder measure 35.3 mL of 85% (w/v) phosphoric acid. Add the phosphoric acid to the Erlenmeyer flask containing KCl in 500 mL of deionized water. When the KCl has completely dissolved adjust volume to 1 L in a glass graduated cylinder and store in a tightly capped glass bottle. 5. Prepare 0.005% BHT in toluene (see Note 4). Measure 500 mL of toluene with a glass graduated cylinder. Transfer to a glass bottle with solvent-resistant cap and add 25 mg of BHT. Allow BHT to dissolve in toluene.

8

Jesse D. Bengtsson et al.

6. Prepare 2.5% (v/v) sulfuric acid in methanol. Appropriate attire for this exothermic reaction includes lab coat, googles and gloves. In a fume hood very slowly add 25 mL sulfuric acid to 975 mL of methanol. Danger: Failure to add the sulfuric acid slowly can result in rapid boiling and splashing of sulfuric acid, potentially burning exposed tissue (see Note 5). Store the reagent in a capped glass bottle at room temperature. 7. Prepare the internal standard. Dissolve 100 mg of 17:0-TAG in 200 mL of toluene with 0.005% (w/v) BHT. This provides a final concentration of 0.5 μg standard per μL. Store in Teflon-capped glass bottle at
20  C until use. Warm to room temperature prior to use. 3.2 Harvesting Tissue for Storage

1. Appropriate attire includes lab coat, googles and gloves. Liquid nitrogen used in this process is dangerous and will burn skin on contact. 2. Label Falcon tubes. 3. Put a tube rack in an insulating foam box and place labeled 50 mL Falcon tubes into the tube rack. 4. Pour approximately 4 cm of liquid nitrogen into the box. Liquid nitrogen can burn skin. Carefully pour the liquid nitrogen, avoiding skin contact. 5. Fill the Falcon tubes with 2 to 3 cm of liquid nitrogen; do not cap (see Note 6). 6. Place a second tube rack in a second foam box without liquid nitrogen. 7. Cut approximately 1 g of tissue per sample and quickly weigh it using the balance. 8. Transfer the tissue from the scale to a Falcon tube containing liquid nitrogen. 9. Place the tube containing the weighed sample into the second box, not immersed in liquid nitrogen. 10. Allow the liquid nitrogen in the tube to completely evaporate, then immediately cap each tube while the sample is still frozen and place it in the tube rack immersed in liquid nitrogen. Proceed to the next sample. 11. Store samples at
80  C until extraction.

Lipid Isolation from Plants

3.3 Tissue Homogenization

9

1. Label 30 mL solvent-resistant Nalgene tubes and place on ice. 2. Place the bottle of chloroform–methanol–glacial acetic acid (10:10:1, v/v/v) on ice in the fume hood. 3. Chill the mortar and pestle by filling the mortar with liquid nitrogen, placing the pestle in the mortar (see Note 7). 4. When the liquid nitrogen stops boiling the mortar and pestle are ready to use. 5. Transfer the first sample from the Falcon tube to the liquid nitrogen in the mortar. 6. Homogenize the frozen sample with the pestle, initially crushing the sample gently into smaller pieces, then grinding with the pestle until the sample is powdered (see Note 8). It is best to rapidly grind the tissue when much of the liquid nitrogen has evaporated, leaving tissue with the appearance of snow. 7. Transfer 5 mL of chloroform–methanol–acetic acid (10:10:1, v/v/v) to the mortar with a 10 mL glass pipette and manual pipette pump (see Note 9). Do not mouth pipette. 8. Resuspend the tissue in the solvent mixture and pour into a labeled 30 mL solvent-resistant fluorinated ethylene propylene tube. 9. Rinse the mortar and pestle with an additional 5 mL of chloroform–methanol–acetic acid (10:10:1, v/v/v) and combine the wash with the sample in the solvent-resistant tube. 10. Rinse the mortar and pestle again with 2 mL of chloroform– methanol–acetic acid (10:10:1, v/v/v) and combine the material into the solvent-resistant tube. 11. Wipe the mortar and pestle with a paper towel and cool again with liquid nitrogen before proceeding to the next sample. 12. Store the samples at
20  C at least 4 h (or overnight) to allow for a complete extraction of lipids.

3.4

Lipid Extraction

1. Cool the centrifuge and rotor to 4  C. 2. Place the bottle of chloroform–methanol–acetic acid (5:5:1, v/v/v) on ice in the fume hood. 3. Label a second set of 30 mL solvent-resistant tubes and place on ice. 4. Label a set of 25 mL glass tubes. 5. Remove the samples from the
20  C freezer and centrifuge at 4000  g at 4  C for 5 min to pellet the debris. 6. Taking care not to disturb the pelleted debris, transfer the supernatant to a new labeled tube using a glass Pasteur pipette (see Note 10).

10

Jesse D. Bengtsson et al.

7. Wash the pellet by adding 5 mL of cold chloroform–methanol– acetic (5:5:1, v/v/v) acid to the 30 mL tube containing the pellet, cap and mix thoroughly with a vortex mixer. 8. Centrifuge at 4000  g and 4  C for 5 min. 9. Combine the wash supernatant with the supernatant from the sample, taking care to only transfer the supernatant to the new tube. Discard the pellet. 10. Create a phase separation by the addition of 6 mL 1 M KCl in 0.2 M H3PO4 to the new tube (see Fig. 1a). Cap and vortex the tubes. Centrifuge at 4000  g and 4  C for 5 min to aid the phase separation. Transfer the bottom chloroform layer to a clean labeled 25 mL glass tube, being careful to take only the chloroform layer. Back-extract the aqueous layer with an additional 5 mL of chloroform. Cap and vortex the tube. Centrifuge and combine the lower chloroform layer with the first extraction, being careful to take only the chloroform layer. 11. Evaporate the chloroform under nitrogen in the gas evaporation manifold, with the heating block set to 40  C. 12. Resuspend in 500 μL toluene, 0.005% BHT (see Note 11). Close with a Teflon-lined cap. Gently vortex the sample to dissolve any lipids on the tube wall. 13. For storage keep the sample at
20  C where it will be stable for at least 6 months. 3.5 Quantification of Total Lipids by GC Analysis of Fatty Acid Methyl Esters

1. Warm the 17:0-TAG standard in toluene to room temperature (see Note 12). 2. Using either a P200 or 100 μL Drummond microdispenser place 100 μL of internal standard into an 8 mL glass tube as a standard-only control. 3. Using either a P200 or 100 μL Drummond microdispenser take a 50 μL aliquot of lipid extract and place in a labeled 8 mL glass tube, add 100 μL (50 μg) of 17:0-TAG internal standard. 4. Using 10 mL glass pipettes, add 1.5 mL of 2.5% sulfuric acid in methanol to each of the 8 mL glass tubes (see Note 13). 5. Close tightly with a Teflon-lined cap. 6. Check that the water bath is at 85  C and place the rack of samples in it. After 5 min check tubes to make sure the caps are tight, and no solvent is evaporating. If solvent appears to be evaporating replace evaporated solvent and recap the tube before returning it to the bath.

Lipid Isolation from Plants

11

7. Continue heating at 85  C for 1 h in the water bath. 8. Cool to room temperature. 9. Add 200 μL of hexanes and 1 mL of water and recap the tubes. 10. Shake or vortex the tubes for 30 s, then centrifuge at less than 1000  g for 2 min (see Note 14). 11. Label 1 mL GC vials. 12. Using either a P200 or 100 μL Drummond microdispenser remove 70–100 μL of the hexane (top) layer to a glass insert, place in the labeled GC vial, and cap. 13. Prepare the GC for analysis. 14. FAMEs can be separated by using a carbowax capillary column (e.g., SUPELCOWAX ® 10, 0.53 mm, 15 m) (Fig. 1b) (see Note 15). 15. An example of GC conditions is as follows: 1 μL sample injected. Start temperature 190  C for 1 min. Temperature ramp increases by 7  C per min to 250  C. 16. Use software of choice to output peak areas. 17. Quantify total lipid based on the area of the 17:0 standard peak (see Note 16). μg/area ¼ 50 μg/area of 17:0. Total μg in sample ¼ (sum of all peak areas except 17:0 peak area)  μg/area. μg/μL fatty acid in lipid extraction ¼ (total μg in sample/ 50 μL) (see Note 17).

4

Notes 1. Using a GC-MS to calculate peak areas based on the total ion chromatogram may also be used to quantify the amount of FAMEs in the lipid sample. 2. If a dispensing pipette is used, it must have solvent-resistant parts and seals. 3. Glacial acetic acid is highly concentrated and should be handled carefully. 4. 500 mL of toluene is sufficient to make the internal standard and to resuspend 600 samples in 500 μL of toluene. 5. Sulfuric acid is very corrosive and will readily dissolve clothing. Therefore, care and appropriate lab attire are absolutely essential when working with it.

12

Jesse D. Bengtsson et al.

6. If the tubes are capped with liquid nitrogen still present, the tubes will explode from increased internal pressure, likely dispersing your sample and risking injury from plastic shrapnel. 7. Alternatively, lipase activity maybe be inactivated by hot isopropanol. Heating the sample in isopropanol (with 0.01% BHT added to prevent oxidation) to 85  C for 15 min will efficiently inactivate lipases. After the sample(s) cool to room temperature the extraction can proceed as described. 8. A ground glass homogenizer can be used with the isopropanol inactivation in lieu of a mortar and pestle. The ground glass homogenizers are particularly useful when working with small samples such as young Arabidopsis seedlings [11]. 9. To avoid solvent dripping from the pipette while using the 10 mL glass pipette, prime the pipette by pipetting the solvent mixture up and down to increase vapor pressure of the solvents in the pipette. 10. The pellet is sensitive to agitation, so be very careful when transferring the supernatant. If the pellet becomes dislodged, recentrifuging the sample is the simplest solution. 11. If samples of more than 1 g are extracted, scale up the amount of solvent mixtures appropriately. To evaporate solvent from a larger extraction, use a rotary evaporator (rotovap) rather than nitrogen gas. It is best to pull a very slight vacuum with the sample flask outside the rotovap water bath. After the sample stops boiling under weak vacuum, increase to full vacuum, then slowly lower the sample into the water bath, which should be set to 35  C. 12. The internal standard will partially precipitate from solution at
20  C, so it is important to warm the standard to room temperature before use to ensure the concentration of standard is accurate. 13. Glass tubes may break when centrifuged at greater than 2000  g leaving behind fine glass particles which present a danger. 14. For large numbers of samples, it is convenient to have the reagents for making FAMEs in bottles with chemical-resistant pump dispensers. 15. Many polar GC columns will readily separate FAMEs and can be used to estimate the amount of lipid in the extraction; consult manufacturer’s guides. 16. The 17:0-TAG has nearly the same formula weight as three 17:0-methyl ester molecules. However, for an extremely precise measure of the internal standard it is necessary to account for the slight weight difference.

Lipid Isolation from Plants

13

17. A GC-FID response to the mass of acyl chains is for the most part linear, although there are small differences in detector response to chain length, degree of unsaturation and branching. If an exceedingly high degree of precision is required, response factors may be applied. Response factors for long chain fatty acids (16 to 20 carbons) range from 0.97 to 1.02, and can be found in William Christie’s book “Lipid Analysis,” along with a more detailed explanation of the underlying principles [1].

Acknowledgments This work was supported, in part, by grants from USDA-NIFA (#2018-67013-27459) and NSF (#IOS-1555581). JDB was supported by the WSU NIH training grant in Protein Biotechnology. References 1. Christie WW (2003) Lipid analysis: isolation, separation, identification and structural analysis of lipids, 3rd ed. edn. Bridgewater/Oily Press, Manchester 2. Roughan PG, Slack CR, Holland R (1978) Generation of phospholipid artefacts during extraction of developing soybean seeds with methanolic solvents. Lipids 13(7):497–503. https://doi.org/10.1007/BF02533620 3. Harfoot CG (1978) Lipid metabolism in the rumen. Prog Lipid Res 17(1):21–54. https:// doi.org/10.1016/0079-6832(78)90004-6 4. Cabot MC, Lumb RH (1981) The activity of a low temperature lipase in the larvae of Sarcophaga bullata (Diptera: Sarcophagidae). Comp Biochem Physiol Part B: Comp Biochem 68 (2):325–328. https://doi.org/10.1016/ 0305-0491(81)90106-1 5. Kates M (1957) Effects of solvents and surfaceactive agents on plastid phosphatidase C activity. Can J Biochem Physiol 35(2):127–142. https://doi.org/10.1139/o57-016 6. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37(1):911–917. https://doi. org/10.1139/y59-099

7. Markham JE, Li J, Cahoon EB, Jaworski JG (2006) Separation and identification of major plant sphingolipid classes from leaves. J Biol Chem 281(32):22684–22694. https://doi. org/10.1074/jbc.M604050200 8. Folch J, Ascoli I, Lees M, Meath J, LeBaron F (1951) Preparation of lipid extracts from brain tissue. J Biol Chem 191(2):833–841 9. Larson TR, Graham IA (2001) Technical advance: a novel technique for the sensitive quantification of acyl CoA esters from plant tissues. Plant J 25(1):115–125. https://doi. org/10.1111/j.1365-313X.2001.00929.x 10. Kopka J, Ohlrogge JB, Jaworski JG (1995) Analysis of in vivo levels of acyl-thioesters with gas chromatography/mass spectrometry of the butylamide derivative. Anal Biochem 224(1):51–60. https://doi.org/10.1006/ abio.1995.1007 11. Lunn D, Smith GA, Wallis JG, Browse J (2018) Development defects of hydroxy-fatty acidaccumulating seeds are reduced by castor acyltransferases. Plant Physiol 177(2):553–564. https://doi.org/10.1104/pp.17.01805

Chapter 2 Three Methods to Extract Membrane Glycerolipids: Comparing Sensitivity to Lipase Degradation and Yield Samira Mahboub, Zachery D. Shomo, R. Maxwell Regester, Mahaa Albusharif, and Rebecca L. Roston Abstract Glycerolipids form the largest fraction of all membrane lipids and their composition changes quickly during plant development, the diurnal cycle, and in response to hormones and biotic or abiotic stress. A challenge to accurate glycerolipid measurement is that lipid-degrading enzymes tend to remain active during extraction, and special care must be taken to ensure their inactivation. Multiple extraction methods have arisen to cope with this challenge but only a few comparative studies are available in the literature. Here we compare three commonly used methods for lipase inactivation and lipid extraction from two different plant tissues. The first method employs formic acid in an organic solvent for inactivation followed by immediate separation of the organic phase, while the second uses the same acidic solvent, but expands the time of lipase inactivation and lipid extraction by incubation at low temperature. The third method includes a boiling step of the tissue in isopropanol for enzyme inactivation. The first method is the fastest for lab conditions with few samples, the second and third are convenient with large sample numbers, including field work. The first two methods are commonly followed by lipid derivatization and gas chromatography, while the third avoids acids and is thus preferable for lipidomics approaches. We directly compare the methods on both Arabidopsis thaliana and Sorghum bicolor leaf tissues and measure the relative abundances of glycerolipid species formed by lipase activity. We conclude that each method provides intact lipid extracts of similar quality, if performed according to the protocols described below. Key words Glycerolipids, Arabidopsis thaliana, Sorghum bicolor, Lipid extraction, Lipase inhibition

1

Introduction A successful lipid extraction from plant tissue is achieved by inactivation of resilient enzymes including lipases and phosphatases, and efficient extraction of lipid classes with differing hydrophobicity from the tissue. When the tissue is lysed, the lipids are exposed to lipases and phosphatases that can lead to lipid degradation. The enzymes responsible for the degradation are notoriously resilient to inactivation [1] and can alter the lipid profile [2, 3]. The efficiency of the lipid extraction from the tissue can also be a concern. The membrane lipids in a plant cell differ widely in their hydrophobicity,

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021

15

16

Samira Mahboub et al.

and during extraction associate to different extents with the aqueous fraction that is not retained for analysis [4]. Both lipase inactivation and extraction efficiency are crucial for measuring the lipid profile that is an accurate representation of the lipidome at the time of extraction. Currently, many different methods are used for lipid extraction [5–8]. Modern protocols include mechanisms to inactivate resilient enzymes, broadly based on a manuscript by Bieleski (1964) that addressed concerns of phosphatase activity during lipid extraction [1]. There are multiple types of lipases in plant tissues (Fig. 1), and their common products include diacylglycerol (DAG), phosphatidic acid (PA) and free fatty acids (FFA) [9]. A number of extraction techniques, including some protocols which focus on analysis of multiple types of metabolites through mass spectrometry, use a solvent containing methanol and methyl tert-butyl ether to extract metabolites without an appropriate enzyme inactivation step [10]. While these methods do extract lipids from plant tissues, there is strong evidence suggesting the resulting profile of lipids is artificially modified by lipase activity [2, 3]. Here, we give step-by-step instructions for three methods of lipid extraction and lipase inactivation that produce reliable lipid profiles obtained from leaves of two plant species, Arabidopsis thaliana and Sorghum bicolor. The first method according to N+

OH HO

4: PA

OH O

HO O

1: DAG

2: FFA O

O

O O

O

O

O

O O

O

O OP O O 3: DAG O

5: FFA

O O O

Fig. 1 Lipase activity during lipid extraction increases levels of diacylglycerol, free fatty acids, and phosphatidic acid. (1) Galactolipase activity leads to production of diacylglycerol (DAG). (2) Triacylglycerol lipase activity produces free fatty acids (FFA) and DAG. (3) Phospholipase C cleaves the glycerol-phosphate bond producing DAG. (4) Phospholipase D removes head groups attached to the phosphate of a phospholipid, producing phosphatidic acid (PA). (5) Phospholipase A1 and A2 cleave the sn-1 or sn-2 acyl chains from the glycerol backbone, producing FFAs

Leaf Glycerolipid Extraction

17

Wang and Benning [8] is performed with tissue samples kept on ice and quickly disrupted at room temperature. It uses an extraction solvent with formic acid to inactivate the lipases; this is followed by an immediate phase extraction that causes the lipases to separate into the aqueous layer, leaving the lipids in the organic phase [8]. This method is convenient for routine sampling in a lab setting, and suitable for a large number of samples for simultaneous extraction, or a fine time-course experiment. The second method, based on a personal communication from J. A. Browse (see Chapter 1, for a similar method), uses the same solvent system to inactivate lipases. However, tissue samples are incubated in acidic solvent for 24 h at low temperatures. The Browse method is based on a study showing that phosphatase activity is substantially decreased at 20  C when compared to the same extraction protocol using heat inactivation [1]. It is highly convenient for lab settings with high sample numbers or field settings where dry ice is available for storing samples and local laws allow use of organic solvents in nonlaboratory settings. The third method is adapted from a protocol of Shiva et al. [7]) that uses isopropanol at 75  C to first inactivate the lipases, then extracts lipids by addition of additional organic solvents and subsequent shaking at 37  C for up to 24 h [7]. Before analysis of the extract, leaf tissue is removed and the lipids are concentrated by evaporation under nitrogen gas. This method is convenient when extractions cannot be carried out at low temperature. Heating blocks have some advantages in that they are typically cheap, have low energy requirements, and can be run from generators. The samples produced from this method can be kept at room temperature for several days. However, this method is not without caveats, as we noted substantial degradation of pigments. We tested lipid extraction methods in a training environment with an experienced lipid technician and lab novices that were not previously exposed to any of the methods. When our protocols were followed, extractions could be performed equivalently well at each expertise level, and both expert- and novice-collected and analyzed data are included in our results (see Note 1). After the extractions were completed, lipid classes were separated via thin-layer chromatography, derivatized into fatty acid methyl esters and separated via gas chromatography with detection by flame ionization, for which detailed protocols are available elsewhere (see Note 2) (see Chapters 1 and 4) [8, 11]. Lipid analyses of leaves from both A. thaliana and S. bicolor showed that the three extraction methods are similarly efficient at both lipase inactivation and glycerolipid extraction (Figs. 2 and 3). Specifically, the amounts of the most abundant nonplastidic lipid, phosphatidylcholine, and the most abundant plastidic lipid, monogalactosyldiacylglycerol (MGDG), were similar across all methods tested. The lipid classes expected to increase due to lipase activity (DAG, FFA, PA) were of very low abundance in both A. thaliana and sorghum across all

18

Samira Mahboub et al.

a

b 60

80 60

40

40 20

20

0

0

c

Wang et al 60

Total

Browse 40

Shiva et al TAG

30

40

20 20

10

0 60

0 DAG

60

40

40

20

20

0

0

50

PA

40

80

MGDG

60

30

40

20 10

20

0

0

50

FFA

PG

60

PE/PS

40 40

30 20

20

10 0

0

80

80

DGDG

60

60

40

40

20

20

0

0

SQDG/PC/PI

Fig. 2 Three extraction methods of Arabidopsis thaliana lipids followed by TLC separation and quantification of FAMEs by GC show similar lipid abundances and fatty acid profiles. (a) Mole percent of lipid classes

Leaf Glycerolipid Extraction

19

methods tested (Fig. 2a and 3a). The lipases with highest activity likely differ in each extraction method, as there were significant differences between the acyl composition of DAG, FFA, and PA (Figs. 2c and 3c). The most consistent lipase effect observed for the two plant species is on PA, which shows similar and statistically significant changes with regard to the abundances of 18:2. The method by Wang and Benning [8] appears to provide the most efficient extraction of lipids (measured as total fatty acids) of all presented methods (Fig. 2a and 3a). The method showing the highest lipase activity, as judged by a statistically significant increase in the levels of PA extracted from sorghum, may be the Browse method (Fig. 3b). The levels of MGDG showed a corresponding decline, possibly due to the activity of a plastidic lipase. The similarity of the lipid profiles suggests that all three methods reduce lipase activity, though subtle differences have been observed and are likely due to extraction capability and selective lipase inactivation.

2

Materials All solutions containing organic solvents should be prepared in a fume hood with glass measuring implements and stored in glass bottles with screw top Teflon (polytetrafluoroethylene, PTFE)lined lids to prevent evaporation. All solvents should be HPLCgrade, other chemicals should be more than 99% pure, if not stated otherwise. All reagents containing either acid or organic solvents should be used with appropriate personal protective equipment and stored in safety cabinets. All aqueous solvents are prepared with ultrapure water (18 MΩ-cm at 25  C). All waste should be correctly disposed of according to the institution’s safety guidelines.

2.1

Solvents

1. Isopropanol. 2. Formic acid (88%). 3. Phosphoric acid (85%). 4. Isopropanol lipid extraction solvent: chloroform–methanol– water (30:41.5:3.5, v/v/v) (see Note 3).

ä Fig. 2 (continued) normalized to total fatty acid abundance in the entire extract. (b) Abundance of lipid classes was normalized to the sum of all lipid classes. (c) Mole percent of each acyl chain versus total acyl chains per lipid class. cis (C16:1c) and trans (C16:1t) were analyzed separately. Mean  standard deviation, with n  4 for each lipid class. Statistical significance is denoted by (*) for p < 0.05. Asterisks without brackets indicate that each data point is distinct. Statistical analysis was completed using a two-way ANOVA with multiple comparisons after outliers were removed using ROUT analysis with a Q value of 10%. Due to its low abundance, 16:2 was not used for outlier calculations. Note that PE/PS and SQDG/PC/PI comigrate during TLC

20

Samira Mahboub et al.

a

b 50

50 40 30

40 30 20

20 10 0

10 0

c

Wang et al 80

Total

Browse

Shiva et al

50 40

60

30

40

20

20

10

0

0

60

TAG

DAG

40

100 80

FFA

60 40

20

20 0

0

60

100 80

PA

40

MGDG

60 40

20

20 0

0

60

60

PG

40

40

20

20

0

0

80 60

DGDG/PS

60

PE

SQDG/PC/PI

40

40 20 0

20 0

Fig. 3 Three extraction methods of Sorghum bicolor lipid followed by TLC separation and quantification of FAMEs by GC show similar lipid abundances and fatty acid profiles. (a) Mole percent of lipid classes

Leaf Glycerolipid Extraction

21

5. Formic acid lipid extraction solvent: methanol–chloroform– 88% formic acid (2:1:0.1, v/v/v) (see Note 3). 6. Phase separation solvent: 1 M potassium chloride, 0.2 M phosphoric acid (see Note 3). 2.2 Laboratory Equipment

1. Razor blades or other sharp tissue-collection devices (see Note 4). 2. 2 mL screw-cap tubes without skirts or compatible with bead beater. 3. 3.2 mm chrome steel disruption beads. 4. Screw-cap glass tubes with accompanying PTFE-lined lids. 5. Heating block that holds the screw-cap glass tubes, capable of heating to 37  C. 6. Rotating shaker equipped with a with tube rack capable of up to 250 rpm. 7. Bead beater capable of 4000 rpm (see Note 5). 8. 2 mL or similar amber-glass storage vials with PTFE-lined lids.

3

Methods Unless stated otherwise, all steps are carried out at room temperature. For data shown in Figs. 2 and 3, 4-week-old plants (Arabidopsis thaliana and Sorghum bicolor) were used. All steps involving organic solvents should be completed with adequate ventilation for safety.

3.1 Plant Tissue Collection

1. S. bicolor: Start with 0.2 to 0.3 g of fresh plant tissue. We used 4 cm of the second leaf taken from 4-week-old plants grown at 28  C for a 12-h day, and 22  C for a 12-h night. Tissue was collected with a single razor blade cut (see Notes 6 and 7). 2. A. thaliana: Start with 0.3 to 0.45 g of fresh plant tissue. We harvested from the second-oldest leaves on the rosette of 4-week-old plants grown at 22  C, 16-h day and 8-h night. Leaves were removed with a razor blade (see Notes 6 and 7).

ä Fig. 3 (continued) normalized to fatty acid abundance in the entire extract. (b) Lipid class abundance normalized to the sum of all lipid classes. (c) Relative mole percent of each acyl chain versus total acyl chains per lipid class. Mean  S.D., with n  4 for each lipid class. Statistical significance is denoted by (*) for p < 0.05. Asterisks without brackets indicate that each data point is distinct. Statistical analysis was completed using a two-way ANOVA with multiple comparisons; outliers were removed using ROUT analysis with a Q value of 10%. Due to its low abundance, 16:1, 16:2, and 16:3 were not used for outlier calculations. Note that PE/PS and SQDG/PC/PI comigrate during TLC

22

Samira Mahboub et al.

3.2 Extraction Method (Adapted from Wang and Benning, 2011 [8])

In this method, lipases are inactivated by combining an acidified solvent with a quick phase extraction to remove the lipases into the aqueous phase and lipids are separated into the organic phase. It must be completed as quickly as possible and most tube-handling steps are performed on ice. 1. Add one volume of the formic acid lipid extraction solvent (about 1 mL/0.3 mg of tissue) to a 2 mL screw-cap microcentrifuge tube and place it on ice (see Note 8). 2. Gently add leaf tissue to the extraction solvent on ice and move tubes into the mini bead beater. 3. Shake vigorously at 4000 rpm for 30 s, then immediately return the tubes to ice (see Note 9). 4. Initiate phase separation by adding ice-cold phase separation solvent at ½ volume of the solvent added in step 1. 5. Vortex well. 6. Centrifuge at 16,000  g for 3 min at 4  C. This will allow separation of an organic (lower) and an aqueous top layer (see Note 10, Fig. 4). 7. Remove the organic layer (lower) and transfer to an amber 1.5 mL glass vial (see Notes 11 and 12).

3.3 Extraction Method (Adapted from J. A. Browse; See Chapter 1)

This method inactivates lipases by incubating leaf tissue in an acidified solvent mixture for 24 h prior to tissue disruption. Enzyme activity is reduced at low temperatures, thus providing increased time for inactivation at 20  C, and prolonging the time available for lipid extraction. 1. Add one volume of the formic acid lipid extraction solvent (about 1 mL/0.3 mg of tissue) to a screw-cap 2 mL microcentrifuge tube, and place it on ice (see Note 8). 2. Gently add leaf tissue to the extraction solvent on ice and move samples to 20  C for 24 h. 3. After 24 h, place tubes into the mini bead beater and shake vigorously at 4000 rpm for 30 s, then immediately return the tubes to ice (see Note 9). 4. Initiate phase separation by adding ice-cold phase separation solvent at ½ volume of the solvent added in step 1. 5. Vortex well. 6. Centrifuge at 16,000  g for 3 min at 4  C. This will allow separation of an organic (lower) and an aqueous top layer (see Note 10, Fig. 4). 7. Remove the organic layer (lower) and transfer to an amber 1.5 mL glass vial (see Notes 11 and 12).

Leaf Glycerolipid Extraction

23

Fig. 4 Arabidopsis thaliana and Sorghum bicolor phase separation. The methods adapted from Wang and Benning (2011) [8] and J. A. Browse (Chapter 1) both require complete tissue disruption for efficient extraction. After centrifugation, the completeness of tissue disruption can be assessed by the presence of multiple phases. The aqueous phase (clear) containing lipases, the organic phase (dark green) containing the majority of lipids, and the interphase (black arrowhead) containing plant tissue and insoluble material. Examples of complete tissue disruption from (1) A. thaliana or (2) S. bicolor. Note the clearly defined interphase. S. bicolor tissue disruption required the use of disruption beads (white arrowhead). Examples of incomplete tissue disruption of (3) A. thaliana or (4) S. bicolor. When disruption is incomplete, the interphase contains visibly green plant material and samples might be discarded 3.4 Extraction Method (Adapted from Shiva et al., 2018 [7])

In this method lipases are inactivated by boiling the tissue samples in isopropanol. No formic acid is used for the inactivation. 1. Add 1 volume of isopropanol to glass tubes and cap with PTFE-lined screw caps. The volume should be large enough to cover the leaf, which is added in step 2 and may be gently folded (see Note 13). 2. Set dry bath to 75  C. When temperature is reached, place the tubes with isopropanol into the dry bath and bring to boiling. Once boiling, carefully open the tubes and add the leaf to the tube. Close tube with the cap. 3. Boil for 15 min (see Note 14). 4. Remove tubes from the hot block and let them cool down to room temperature. 5. Add 3 volumes of the isopropanol lipid extraction solvent to each tube.

24

Samira Mahboub et al.

6. Move the tubes to a 37  C incubator and shake them at 250 rpm for 24 h (see Note 15). 7. After 24 h, remove the leaf tissue and return the tubes to the 75  C dry bath (see Note 16). 8. Place the samples under a flow of nitrogen gas until the solvent has completely evaporated (see Note 17). 9. Reconstitute the sample with 200 μL of chloroform and move to an amber 1.5 mL glass vials with screw top lids (see Notes 11 and 12).

4

Notes 1. Misinterpretation or lack of emphasis on a crucial step of the extraction protocol can lead to less effective lipase inactivation or lipid extraction. We found that undergraduate students performed the extraction methods well, but struggled with the prolonged series of steps that occurred later in the protocol. We compared results from an experienced technician and an undergraduate student, and the results were sufficiently equivalent (Fig. 3a, b). 2. Detailed procedures for TLC analysis, FAME reaction and lipid quantification using gas chromatography can be found in Chapters 1 and 4 [8, 11]. Here, TLC was performed with silica plates using two resolving solvents. First, the plate was developed with resolving solvent 1, chloroform–methanol–glacial acetic acid–ultrapure water (85:12.5:12.5:3, v/v/v/v) to separate polar lipids. After drying the plate, it was developed again in the same direction with resolving solvent 2, chloroform– acetone–glacial acetic acid (96:4:1, v/v/v) to separate the nonpolar lipids without affecting the polar lipids. With this solvent system, sulfoquinovosyldiacylglycerol, phosphatidylcholine, and phosphatidylinositol are not separated from each other, and are analyzed together. We chose the solvents because our primary focus was to have a good separation of diacylglycerol, free fatty acids and phosphatidic acid. 3. This solvent can be stored for up to 1 month at room temperature in a glass bottle with a PTFE-lined cap. Note that formic acid reacts with methanol over time and fresh solvent should be used for deactivating lipases. 4. It is critical to choose a sharp tissue collection device. 5. Instead of the bead beater, samples can alternatively be crushed by mortar and pestle prechilled with liquid nitrogen. Fresh tissues can be added to the mortar containing chilled extraction solvent and then ground with the pestle. The sample can then be poured into the 2 mL tubes for extraction. This should be done in a ventilation hood.

Leaf Glycerolipid Extraction

25

6. It is imperative to introduce the lowest amount of plant-toplant variation as possible. Plant samples should be at the same age and tissues should be as similar as possible. Samples for replicates should be collected at the same time of the day to prevent introducing noise from the diurnal cycle [12]. 7. Great effort should be taken to reduce the amount of time between tissue harvest and the application of the lipase inactivation solvent. 8. When extracting lipids from robust plant tissue such as grasses (e.g., sorghum leaves) or seeds, 3–4 disruption beads must be added to the bead-beating step to ensure the tissue is sufficiently broken. When using softer tissue such as A. thaliana leaves, the friction of the agitation is sufficient for tissue disruption and the beads are not necessary. 9. Even with the use of disruption beads, a single 30 s round of shaking might not be sufficient for tissue disruption of robust plant tissues. In this case, a second round of 30 s can be used. The duration of continuous time in the bead beater should be limited to 30 s, as heat is generated and can lead to lipid degradation. Samples should be placed on ice between rounds in the bead beater. 10. The ground tissue remnants will form an interphase between the upper aqueous phase and the lower organic phase. The interphase should be uniform and have little to no color. If it has obvious chunks or remains colored, the tissues were not broken sufficiently and samples should be discarded. 11. For analysis of lipids by TLC or GC with flame ionization detector, transfer of the organic layer with a low-retention plastic pipette tip is sufficient. However, if an analytical method more sensitive to plastic contamination is used, the organic layer can be transferred with a stainless-steel syringe or a glass Pasteur pipette. 12. Before capping for storage, overlay the samples with an inert gas such as nitrogen to displace any oxygen in the tube and prevent lipid oxidation. The lipid extracts can be stored under an inert gas at 20  C for 2 weeks, and at 80  C for less than 6 months. 13. The tubes should be able to hold a total volume 4.5 fold larger than the initial solvent volume. We use 13  100 mm screw top vials with a total volume of 9 mL, and one solvent volume is equivalent to 2 mL. 14. It is critical that the isopropanol is already heated at the time tissues are introduced to prevent lipase activity. If the tissue sample extends out of the solvent, the tube should be inverted every few min during the boiling. Alternatively, an additional volume of isopropanol can be added to the tube.

26

Samira Mahboub et al.

Fig. 5 Manifold for drying lipid extractions. The end of the pipette has been heat-sealed and the tip is attached to the tubing for gas delivery

15. This time can be extended to 3 days. Temperature should be between 22  C and 37  C. 16. The tissue will still appear to be intact. We remove it with stainless steel forceps cleaned in acetone or by making a hook out of a glass Pasteur pipette. With some tissues, it is preferable to decant the solvent away from the leaf tissue into a new tube. 17. The drying process can be accelerated by placing the tubes back into a gently heated block (25  C to 30  C). The nitrogen can be distributed to multiple tubes simultaneously through use of a manifold, which can be purchased or made (Fig. 5). We make our own through attachment of 16 gauge, stainless-steel tubing (O.D. 1.65 mm, I.D. 1.41 mm) into a plastic Pasteur pipette into which holes have been melted with a soldering iron. They are held in place by use of epoxy resin or hot glue.

Acknowledgments The authors would like to thank C. Benning, J. Browse, and R. Welti for meaningful conversations about these extraction methods. This project was partially supported by the National Science Foundation (IOS-1845175) and the Nebraska Agricultural Experiment Station with funding from the Hatch Multistate Research capacity funding program (NEB-30-133) from the USDA National Institute of Food and Agriculture. References 1. Bieleski RL (1964) The problem of halting enzyme action when extracting plant tissues. Anal Biochem 9:431–442. https://doi.org/ 10.1016/0003-2697(64)90204-0 2. Slack CR, Campbell LC, Browse JA, Roughan PG (1983) Some evidence for the reversibility of the cholinephosphotransferase catalysed reaction in developing linseed cotyledons in vivo. Biochim Biophys Acta 754:10–20.

https://doi.org/10.1016/0005-2760(83) 90076-0 3. Yang SF, Freer S, Benson AA (1967) Transphosphatidylation by phospholipase D. J Biol Chem 242:477–484 4. Christie WW (1993) Preparation of lipid extracts tissues. In: Christie W (ed) Advances in lipid methodology - two. Oily Press, Dundee, pp 195–213

Leaf Glycerolipid Extraction 5. Folch J, Lees M, Sloane Stanley GH (1957) A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226:497–509 6. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37:911–917. https://doi. org/10.1139/o59-099 7. Shiva S, Enninful R, Roth MR, Tamura P, Jagadish K, Welti R (2018) An efficient modified method for plant leaf lipid extraction results in improved recovery of phosphatidic acid. Plant Methods 14:14. https://doi.org/ 10.1186/s13007-018-0282-y 8. Wang Z, Benning C (2011) Arabidopsis thaliana polar glycerolipid profiling by thin-layer chromatography (TLC) coupled with gas-liquid chromatography (GLC). J Vis Exp 49:2518. https://doi.org/10.3791/2518 9. Matos AR, Pham-Thi A-T (2009) Lipid deacylating enzymes in plants: old activities, new

27

genes. Plant Physiol Biochem 47:491–503. https://doi.org/10.1016/j.plaphy.2009.02. 011 10. Salem M, Bernach M, Bajdzienko K, Giavalisco P (2017) A simple fractionated extraction method for the comprehensive analysis of metabolites, lipids, and proteins from a single sample. J Vis Exp 124:55802. https://doi. org/10.3791/55802 11. Barnes AC, Benning C, Roston RL (2016) Chloroplast membrane remodeling during freezing stress is accompanied by cytoplasmic acidification activating SENSITIVE TO FREEZING2. Plant Physiol 171:2140–2149. https://doi.org/10.1104/pp.16.00286 12. Ekman Å, Bu¨low L, Stymne S (2007) Elevated atmospheric CO2 concentration and diurnal cycle induce changes in lipid composition in Arabidopsis thaliana. New Phytol 174:591–599. https://doi.org/10.1111/j. 1469-8137.2007.02027.x

Chapter 3 Thin-Layer Chromatography Georg Ho¨lzl and Peter Do¨rmann Abstract Lipid extracts from plants represent a mixture of polar membrane lipids and nonpolar lipids. The main constituents of the polar lipid fraction are glycerolipids, that is, galactolipids, sulfolipid, and phospholipids. In addition, betaine lipids are found in pteridophytes, bryophytes, and algae. Nonpolar lipids include the storage lipid triacylglycerol, wax esters, diacylglycerol and free fatty acids. The complex lipid mixtures from plant tissues can be separated by thin-layer chromatography (TLC) into different lipid classes. In most cases glass plates coated with a silica gel are used as stationary phase and an organic solvent as mobile phase. Different solvent systems are required to separate polar membrane lipids or nonpolar lipids by TLC. Depending on the complexity of the lipid mixture, lipids are separated using one- or two-dimensional TLC systems. Different dyes and reagents allow the visualization of all lipid classes, or the selective staining of glycolipids or phospholipids. Lipids can be isolated from the TLC plate for subsequent analysis, provided that nondestructive methods are used for visualization. Key words Polar lipid, Nonpolar lipid, Glycerolipid, TLC, Phospholipid, Glycolipid, Fatty acid, Betaine lipid, Wax ester, Iodine

1

Introduction Plant tissues including leaves, roots, or seeds are characterized by unique lipid compositions. Glycerolipids constitute the majority of lipid classes in leaves, while sterol lipids, sphingolipids or cuticular lipids are less abundant. According to their physicochemical characteristics, lipids are divided into polar and nonpolar lipids, which are mainly represented by membrane and storage lipids, respectively. Monogalactosyldiacylglycerol (MGDG), digalactosyldiacylglycerol (DGDG), and sulfoquinovosyldiacylglycerol (SQDG) are the predominant glycolipids in leaves. Phospholipids are found in green and non-green tissues and are mainly comprised of phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and phosphatidylcholine (PC), while phosphatidylinositol (PI), phosphatidylserine (PS), cardiolipin (CL) and phosphatidic acid (PA) are less abundant. Glucosylceramides (GlcCer), glycosylinositolphosphoryl ceramides (GIPCs), and sterylglucosides

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_3, © Springer Science+Business Media, LLC, part of Springer Nature 2021

29

30

Georg Ho¨lzl and Peter Do¨rmann

(SG) represent further glycerol-free membrane lipids. The betaine lipid diacylglyceryl-trimethylhomoserine (DGTS) is found in pteridophytes, bryophytes and algae [1]. Nonpolar lipids include triacylglycerol (TAG), wax esters (WE), diacylglycerol (DAG), monoacylglycerol (MAG), and free fatty acids (FFA). TLC was one of the first chromatographic techniques established for lipid analysis. It is still used today and offers many advantages over mass spectrometry-based techniques. TLC is very effective because it only requires low-budget equipment, it is a rather simple technique which can be performed in any molecular or biochemical laboratory, and the samples do neither require extensive cleanup procedures nor derivatization. Lipids extracted from plant tissues using conventional chloroform–methanol–water systems can be directly applied to the TLC plate. In addition, lipids can be purified by solid-phase extraction prior to application to the plate (see Chapter 7). Such a purification is, for example, necessary to remove the oil in oil-rich samples (seed lipid extracts) from the polar lipids to prevent overloading of the plate. Phospholipids and glycolipids often occur in small amounts and are therefore separated by solid phase extraction (SPE) before TLC analysis to enable the detection of minor lipids which would not be seen in a TLC analysis of total lipid extracts. During TLC separation, a porous layer, the sorbent (in most cases silica gel), which is bound to an inert support (a carrier, e.g., glass plate) represents the stationary phase. The sorbent and carrier together form the TLC plate. For chromatography, the plate is placed in a separation chamber which can be closed airtight. The bottom of the chamber is filled with the eluent up to a height of ~1 cm, a solvent mixture which establishes the mobile phase. During the development of the TLC plate, the eluent penetrates into the sorbent due to capillary forces and transports the analytes in the direction of flow and thereby separates them into individual components. Different techniques are available to develop the TLC plates. The simplest method is the vertical linear technique. The TLC plate is placed in the separation chamber so that the eluent wets the sorbent below the application point of the samples. With the horizontal linear technique, the solvent is transferred over a bridge onto the horizontal plate. In this case, the solvent does not have to move against gravity. In a two-dimensional linear development, one substance mixture is applied in a plate corner and first developed linearly. After drying, the plate is rotated by 90 and developed again linearly in a second solvent with chemical properties different to the first solvent. Silica gel is used as the sorbent, which is a polycondensate of ortho-silica [Si(OH)4]. Silica gel can be obtained in defined particle sizes with an average diameter of 5–50 μm and with different pore sizes. Since the average pore diameter can vary by three orders of magnitude, the commercial preparations differ greatly in the size of their inner and outer

Thin-Layer Chromatography

31

surface (measured in m2/g) which is an important parameter for the separation properties. The lipids are separated according to the principle of distribution chromatography. Since the eluent itself is subject to the chromatographic process, the composition of the mobile phase changes steadily during development, because it is constantly depleted of polar components during migration. In addition, solvent components evaporate from the liquid to the gas phase of the separation chamber. Because of this, the development of a TLC plate is carried out in a closed separation chamber, the gas space of which is saturated with the solvent vapor after a certain time. It is therefore not advisable to open the chamber during chromatography or to reuse the eluent. However, several TLC plates can be developed in parallel in one tank. After the development of the TLC plate has been completed, the separated substances must be visualized. Most lipids (with the exception of some compounds like carotenoids or chlorophyll) do not absorb sufficiently in the range of the wavelength of the visible light. Lipids that absorb in the UV range can be detected by fluorescence under UV light. Another detection option with UV radiation is local fluorescence quenching: If the silica layer contains a fluorescence indicator (added during production) that can be excited at 254 nm or 366 nm, UV-absorbing substances appear as dark spots on a fluorescent background when the plate is irradiated. Often, however, the separated substances can only be detected after spraying the plate with a suitable detection reagent (in some cases heating is required to trigger a chemical reaction). If the lipids are used for further experiments, non-destructive staining methods should be applied. Staining of lipid spots with iodine vapor in a glass chamber is a very common nondestructive method. Iodine associates with hydrophobic compounds, and therefore all lipids appear in a yellow-brown color. The TLC plate can be sprayed with primuline or with anilino naphthalene sulfonic acid (ANS) to visualize the lipids as fluorescent spots under UV light. The spots can be marked with a pencil and isolated for further applications. Glycolipids (galactolipids, sulfolipid) which are abundant in leaves are stained by spraying the plate with α-naphthol–sulfuric acid and subsequent heating to 137  C. α-Naphthol reacts irreversibly with the sugar head groups, and the glycolipid spots appear in a reddish color. Due to the presence of sulfuric acid in this reagent, all lipids can be charred by heating. The same charring effect can also be achieved by spraying the TLC plate only with sulfuric acid. Radioactively labeled substances can be detected by autoradiography or fluorography (see Chapter 5). TLC techniques coupled with mass spectrometry (MS) are increasingly used for lipid identification [2]. For this purpose, the lipids are removed from the TLC plate and extracted from the silica gel for subsequent analysis by LC-MS. Lipids isolated from TLC plates can be derivatized and quantified by GC (see Chapters 1 and 4).

32

Georg Ho¨lzl and Peter Do¨rmann

A parameter frequently used for the chromatographic characterization of a substance is the retention factor, the Rf value. The Rf value is defined as the ratio of the distance migrated by the substance and the distance migrated by the the solvent, measured in each case from the position of the sample application. The Rf value is therefore a number without a dimension ranging from 0 to 1. In a mixture, the components can often be identified provisionally on the basis of the Rf value, provided that suitable reference substances (standards) are co-chromatographed under the same conditions. Therefore, in addition to the substance mixture to be analyzed, a reference substance is applied for compound identification by co-migration.

2

Materials

2.1 One-Dimensional Thin-Layer Chromatography (1D-TLC)

1. Silica 60 plates with a concentration zone (20 cm  20 cm) (see Note 1). 2. Developing solvent for polar lipids: acetone–toluene–H2O (91:30:8, v/v/v) (see Notes 2 and 3). 3. Developing solvent for non-polar lipids: hexane–diethyl ether– acetic acid (70:30:1, v/v/v) (see Note 4). 4. Separation chamber (glass tank) with airtight lid for TLC. 5. Glass micro pipettes or pipettor with suitable tips.

2.2 Two-Dimensional Thin-Layer Chromatography (2D-TLC)

1. Developing solvent, first dimension: chloroform–methanol– H2O (65:25:4, v/v/v). 2. Developing solvent, second dimension: chloroform–methanol–acetic acid–H2O (90:15:10:3.5, v/v/v/v). 3. Silica 60 plates without concentration zone (20 cm  20 cm). 4. Separation chamber (glass tank) with lid for TLC. 5. Glass micro pipettes or pipettor with suitable tips.

2.3 Reference (Standard) Lipids and Staining and Visualization of Lipids

1. Different reference lipids used as standards for TLC are presented in Table 1. 2. Iodine (crystalline) and a glass tank. 3. Anilino naphthalene sulfonic acid (ANS) reagent: 0.2% (w/v) ANS in methanol. The ANS reagent is light sensitive. Therefore, the vial containing the ANS reagent should be wrapped in aluminum foil. 4. Primuline solution: 0.05% (w/v) primuline in acetone–H2O (8:2, v/v).

Thin-Layer Chromatography

33

Table 1 Reference (standard) lipids used for thin-layer chromatography Lipid

Abbreviation

Supplier

Product code

Specifications

Polar lipids (phospholipids) Phosphatidylglycerol Phosphatidylethanolamine Phosphatidylcholine Phosphatidylinositol Phosphatidylserine Cardiolipin Phosphatidic acid

PG PE PC PI PS CL PA

Sigma Sigma Sigma Sigma Sigma Avanti Sigma

P-8318 P-8193 P-7443 P5766 P-0474 710335P P9511

Egg yolk Soy bean Soy bean Soy bean Soy bean Tetra-18:1 Egg yolk

Polar lipids (nonphospholipids) Monogalactosyldiacylglycerol Digalactosyldiacylglycerol Sulfoquinovosyldiacylglycerol Diacylglyceryl-trimethylhomoserine Glucosylceramide Sterylglucosides

MGDG DGDG SQDG DGTS GlcCer SG

Larodan Larodan Larodan Merck Avanti Matreya

59-1200 59-1210 59-1230-7 857463P 131304P 1117

Spinach Spinach

Nonpolar lipids Wax esters Triacylglycerol Diacylglycerol Monoacylglycerol Oleic acid Linolenic acid

WE TAG DAG MAG 18:1 18:3

Weleda Weleda Avanti Sigma Fluka Fluka

Deuterium (d9) Soy bean From plants Birch oil Birch oil

800815C M7765 75090 62159

5. α-Naphthol–sulfuric acid reagent: dissolve 8 g of α-naphthol in 250 ml methanol. Add 30 ml concentrated H2SO4 dropwise to the α-naphthol solution while stirring on ice (see Note 5). 6. Sulfuric acid reagent: 50% (v/v) H2SO4 in H2O (see Note 6). 7. Glass sprayers. 8. UV transilluminator or handheld lamp (366 nm).

3

Methods (See Note 7)

3.1 Lipid Extraction from Plant Tissues for TLC Separation

Lipids for TLC separation can be isolated using any standard extraction protocol for plant lipids. Care should be taken that lipases are inactivated during tissue grinding and isolation to avoid lipid degradation (see Note 8). 1. The protocol using acidic solvents (chloroform–methanol–formic acid, 2:1:0.1 v/v/v, and 0.2 M H3PO4–1 M KCl) can be employed for lipase inactivation and lipid extraction (see Chapter 1 and 2).

34

Georg Ho¨lzl and Peter Do¨rmann

2. Alternatively, lipases in plant tissues (e.g., Arabidopsis leaf) can be inactivated using boiling water [3], and lipids are then extracted with chloroform–methanol (see Chapter 7) [4]. 3. Crude lipid extracts can be separated into polar and nonpolar fractions via solid-phase extraction (e.g. Chapter 7). 4. Finally, the solvent from the organic phase should be evaporated and polar or nonpolar lipids dissolved in a defined volume (100 μl) of chloroform–methanol (2:1, v/v) or chloroform (100%), respectively. Store lipids at 20  C. 3.2 Separation of Polar Lipids Via 1D-TLC [12] (Fig. 1)

1. Fill a separation chamber with 130 ml of the developing solvent for polar lipids (acetone–toluene–H2O, 91:30:8, v/v/ v) such that the height of the solvent will be ~1 cm which is sufficient to develop up to two plates. Close the tank and wait until the gas phase is equilibrated with solvent vapor (see Note 9). 2. Load the lipid extract as a spot or short line onto the concentration zone of the plate, about 1 cm from the bottom of the plate, using a glass micro pipette or a suitable disposable pipette tip. Several samples can be applied in parallel lanes, about 1–2 cm apart from each other. 3. Reference lipids (galactolipids, phospholipids) can be loaded in separate lanes if required. 4. After drying the lipid spots, the plate is placed in the glass tank and the tank is closed.

Fig. 1 Separation of polar lipids by one-dimensional (1D)-TLC. Polar lipid standards together with a lipid extract from an Arabidopsis thaliana leaf were separated on a TLC plate which was developed with acetone–toluene–H2O (91:30:8). Lipids were stained for 15 min in iodine vapor. CL cardiolipin, DGDG digalactosyldiacylglycerol, DGTS diacylglyceryl-trimethylhomoserine, GlcCer glucosylceramide, MGDG monogalactosyldiacylglycerol, PA phosphatidic acid, PC phosphatidylcholine, PE phosphatidylethanolamine, PG phosphatidylglycerol, PI phosphatidylinositol, PS phosphatidylserine, SG sterylglucoside, SQDG sulfoquinovosyldiacylglycerol

Thin-Layer Chromatography

35

WE TAG

18:1 18:3

DAG

MAG start

Fig. 2 One dimensional (1D) separation of nonpolar lipid standards, after development in hexane–diethyl ether–acetic acid (70:30:1). Lipids were stained with iodine vapor. DAG diacylglycerol, MAG monoacylglycerol, TAG triacylglycerol, WE wax ester, 18:1 oleic acid and 18:3 linolenic acid, representing free fatty acids

5. Develop the plate until the solvent front has reached a distance of about 1 cm from the top of the plate. 6. Remove the TLC plate from the tank and dry it in a fume hood. 3.3 Separation of Non-Polar Lipids Via 1D-TLC (Fig. 2)

1. Fill a glass tank with 100 ml of the developing solvent for nonpolar lipids (hexane–diethyl ether–acetic acid, 70:30:1, v/v/v) such that the height of the solvent will be ~1 cm which is sufficient for developing up to two plates. Close the tank and wait until the gas phase is equilibrated with solvent vapor (see Note 9). 2. Load the lipid extract as a spot or short line onto the concentration zone of the plate, about 1 cm from the bottom of the plate, using a glass micro pipette or a suitable disposable pipette tip. Several samples can be applied in parallel lanes about 1–2 cm apart from each other. 3. Reference lipids (triacylglycerol, diacylglycerol, wax esters, free fatty acids) can be loaded in separate lanes if required. 4. After drying the lipid spots, the plate is placed in the glass tank and the tank is closed. 5. Develop the plate until the solvent front has reached a distance of about 1 cm from the top of the plate. 6. Remove the TLC plate from the tank and dry it in a fume hood.

36

Georg Ho¨lzl and Peter Do¨rmann

MGDG

GlcCer

SG DGTS

DGDG SQDG 1

PI

PE PA

PC PG

CL

PS

Start 2

Fig. 3 Polar lipid separation by two-dimensional (2D) TLC. A mixture of polar lipid standards was separated on a 2D-TLC plate developed in chloroform–methanol–H2O (65:25:4; first dimension, arrow 1) and chloroform–methanol–acetic acid–H2O (90:15:10:3.5; second dimension, arrow 2). Lipids were stained with iodine vapor. For abbreviations of lipid standards, see Fig. 1 3.4 Separation of Polar Lipids Via 2D-TLC [5] (Fig. 3) (See Note 10)

1. Label the upper right corner of the TLC plate (without concentration zone) with a pencil. 2. Load the lipid extract with a distance of 2 cm to the left and bottom edges as a single spot onto the plate with a micro pipette or a suitable disposable pipette tip. Only one sample can be applied, which can be spiked with reference lipids if required. 3. After drying the spot, place the plate (with the label in the upper right corner) into the tank with ~100 ml of chloroform– methanol–H2O (65:25:4, v/v/v) (first dimension). 4. After the TLC development is completed, dry the plate completely in a fume hood. 5. Turn the plate by 90 counterclockwise (now with the label in the upper left corner) and develop the plate in ~120 ml of chloroform–methanol–acetic acid–H2O (90:15:10:3.5, v/v/ v/v) (second dimension). 6. After the TLC separation is completed and the plate is dried, it is ready for further analysis.

Thin-Layer Chromatography

3.5 Staining of Lipids with Iodine [5]

37

1. Prepare a glass tank with a lid containing about 5 g of crystalline iodine that sublimates into the gas phase. Equilibrate overnight so that the iodine vapor fills the entire glass tank. 2. Place the dry TLC plate into the glass tank with iodine vapor for about 5–15 min. 3. Remove plate from the tank. 4. Lipids will be stained with a yellowish to brownish color. If lipids are exposed to iodine for only a short time, the staining is reversible. However, unsaturated fatty acids may be covalently modified, in particular during long exposure to iodine. 5. Mark spots with a pencil, or cover the plate with saran wrap or a glass plate to prevent fading of the staining. 6. The plate can be photographed or scanned.

3.6 Staining of Lipids with Anilino Naphthalene Sulfonic acid (ANS) [6]

1. Spray the dried TLC plate carefully with ANS reagent in a fume hood (see Note 11). 2. Observe the lipid spots using a handheld UV lamp or a UV transilluminator (366 nm wavelength). 3. In the presence of ANS, lipids are visualized by fluorescence light and their positions are marked with a pencil. 4. Lipids can be isolated from the TLC plate because they are not covalently modified.

3.7 Staining of Lipids with Primuline [7, 8]

1. Spray the dried TLC plate carefully with primuline reagent in a fume hood (see Note 11). 2. Observe the lipid spots using a handheld UV lamp or a UV transilluminator (366 nm wavelength). 3. In the presence of primuline, lipids are identified by fluorescence light and the positions are marked with a pencil. 4. Lipids can be isolated from the TLC plate because they are not covalently modified.

3.8 Staining of Glycolipids with α-Naphthol– Sulfuric Acid [9]

1. Spray the dried TLC plate carefully with α-naphthol–sulfuric acid reagent in a fume hood. 2. Heat the TLC plate on a heating plate to 137  C until the staining appears. 3. Glycolipids (MGDG, DGDG, SQDG, GlcCer, and SG) are specifically stained in a red color. This staining method leads to the destruction of the lipids. 4. After longer heat exposure, the sulfuric acid reacts with organic compounds and the color of the lipid bands turns from red to dark grey (see Subheading 3.9).

38

Georg Ho¨lzl and Peter Do¨rmann

3.9 General Lipid Staining by Charring with Sulfuric Acid [10, 11 ]

3.10 Recovery of Lipids from TLC Plates

1. Spray the dried TLC plate carefully with the sulfuric acid reagent in a fume hood. 2. Heat the plate on a heating plate to 120  C for 15 min until the staining appears. All lipid molecules will be charred resulting in a grey or black color. This staining method leads to the destruction of the lipids. Lipids recovered from TLC plates can be analyzed by further techniques, such as mass spectrometry (Chapters 7–11) or nuclear magnetic resonance (NMR) spectroscopy (Chapter 14). Lipids from TLC plates can be converted into fatty acid methyl esters (FAMEs) for lipid quantification by GC. The lipid-containing silica material can be directly used for the FAME reaction (see Chapters 4 and 6) in the presence of an internal standard, without prior lipid extraction from the silica. A simple protocol for lipid extraction from silica material, modified from [4], is presented below. 1. Stain the TLC plate with a nondestructive reagent (iodine, ANS, primuline; see Subheadings 3.5–3.7) and mark the lipid spots with a pencil. 2. Scrape off the silica of the marked spots with a razor blade or a spatula and collect the silica material in a glass vial (see Note 12). 3. Add 3 ml of chloroform–methanol (2:1, v/v) to the silica material (see Note 13) and incubate for 30 min by gentle shaking. Alternatively, the sample can be exposed in an ultrasonic bath for 30 min. 4. Centrifuge for 3 min at 4000  g and transfer the supernatant into a new glass vial. 5. Add 750 μl of 0.9% (w/v) NaCl solution to the organic phase to obtain a ratio of chloroform–methanol–0.9% NaCl of 2:1:0.75, and mix vigorously. Instead of 0.9% NaCl solution, pure water can be used to suppress adduct formation with Na+ during ionization if the lipids shall be analyzed by mass spectrometry. 6. Centrifuge for 3 min at 4000  g to facilitate phase separation. 7. Transfer the lower organic phase with a Pasteur pipette into a new glass vial. 8. Evaporate the solvent from the organic phase under a stream of nitrogen and dissolve polar or nonpolar lipids in a defined volume (100 μl) of chloroform–methanol (2:1, v/v) or chloroform, respectively, store at 20  C.

Thin-Layer Chromatography

4

39

Notes 1. The use of TLC plates with a concentration zone facilitates loading of lipid extracts. The extract can be loaded as a spot, which is concentrated to a thin line during the migration through the concentration zone before entering the separation zone of the plate. 2. Note that the benzene in the original protocol [12] has been replaced by toluene which is less toxic [13]. 3. Silica plates for the analysis of polar plant lipids can be impregnated with ammonium sulfate (submerge plate for ~30 s in 0.15 M (NH4)2SO4, then dry for several days at ambient air and activate for 2.5 h at 120  C (oven) prior to use) [12, 14]. The silica of these plates is acidified leading to the protonation of PG with improved lipid separation (see Chapter 5). 4. Variations in the ratio of hexane to diethyl ether lead to slight changes of the solvent polarity which can be used to adjust the separation of lipids in the nonpolar fraction. 5. The addition of sulfuric acid to the methanolic α-naphthol solution leads to heat development. Therefore, place the solution into an ice bath, and use safety glasses. 6. Due to heat development, add sulfuric acid dropwise into water on an ice bath, and use safety glasses. 7. Solvents and staining reagents are harmful to skin and eyes and bear a common health risk. Therefore, separation and staining of lipids must be done under the fume hood. Wear gloves, a lab coat, and safety glasses. 8. Lipid quantification can be normalized to fresh weight, dry weight or leaf area. Add internal lipid standards to the sample prior to or after extraction. A simple way to determine the dry weight of plant tissues is to collect the extracted plant material (leaf) and dry it before weighing until the weight is constant under the fume hood or in the oven (at 40–50  C). 9. Optionally, a filter paper (~20 cm  20 cm) can be placed into the tank such that it is soaked with solvent for saturation of the gas phase with solvent vapor. 10. In addition to 2D-TLC, separation of lipids can be improved in 1D TLC systems by running the same plate successively in two different solvent mixtures (see Chapters 2, 6 and 17). 11. Primuline and ANS (or derivatives) are frequently used dyes for lipid staining. Primuline was described as being more sensitive in comparison with other nondestructive staining procedures [8, 15, 16].

40

Georg Ho¨lzl and Peter Do¨rmann

12. This method is adapted for the extraction of low amounts of separated lipids (from 1 or 2 Arabidopsis leaves). For lipid extraction from increased amounts of silica material adjust the amount of solvents and the size of the vials. 13. To facilitate the release of lipids from the silica material, add some drops (10–30 μl) of water. Alternatively, after the solvent was added to the silica material, continue directly with step 5 starting with the addition of 750 μl of 0.9% (w/v) NaCl solution. This alternative is used for samples with low amounts of silica. High amounts of silica material would impair the transfer of the lower organic phase with a pipette through the intermediate silica layer, possibly resulting in contamination of the lipids with silica. Therefore, the silica material should be removed first by centrifugation, as described in step 4. References 1. Sato N (1992) Betaine lipids. Bot Mag 105 (1):185–197. https://doi.org/10.1007/ BF02489414 2. Tuzimski T, Sherma J (2000) In: Meyers RA (ed) Thin-layer chromatography, vol 15. Wiley, Chichester, UK, pp 1–26 3. Ho¨lzl G, Leipelt M, Ott C et al (2005) Processive lipid galactosyl/glucosyltransferases from Agrobacterium tumefaciens and Mesorhizobium loti display multiple specificities. Glycobiology 15(9):874–886. https://doi.org/10.1093/ glycob/cwi066 4. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37(8):911–917. https://doi. org/10.1139/o59-099 5. Roughan PG, Slack CR, Holland R (1978) Generation of phospholipid artefacts during extraction of developing soybean seeds with methanolic solvents. Lipids 13(7):497–503. https://doi.org/10.1007/BF02533620 6. Siebertz HP, Heinz E, Linscheid M et al (1979) Characterization of lipids from chloroplast envelopes. Eur J Biochem 101 (2):429–438. https://doi.org/10.1111/j. 1432-1033.1979.tb19736.x 7. White T, Bursten S, Federighi D et al (1998) High-resolution separation and quantification of neutral lipid and phospholipid species in mammalian cells and sera by multi-one-dimensional thin-layer chromatography. Anal Biochem 258(1):109–117. https://doi.org/10. 1006/abio.1997.2545 8. Skipski VP (1975) Thin-layer chromatography of neutral glycosphingolipids. Meth Enzymol

35:396–425. https://doi.org/10.1016/ 0076-6879(75)35178-1 9. Jorasch P, Wolter FP, Z€ahringer U et al (1998) A UDP glucosyltransferase from Bacillus subtilis successively transfers up to four glucose residues to 1,2-diacylglycerol: expression of ypfP in Escherichia coli and structural analysis of its reaction products. Mol Microbiol 29 (2):419–430. https://doi.org/10.1046/j. 1365-2958.1998.00930.x 10. Heinze FJ, Linscheid M, Heinz E (1984) Release of diacylglycerol moieties from various glycosyl diacylglycerols. Anal Biochem 139 (1):126–133. https://doi.org/10.1016/ 0003-2697(84)90397-X 11. Wang Z, Benning C (2011) Arabidopsis thaliana polar glycerolipid profiling by thin layer chromatography (TLC) coupled with gas-liquid chromatography (GLC). J Vis Exp 49. https://doi.org/10.3791/2518 12. Benning C, Somerville CR (1992) Isolation and genetic complementation of a sulfolipiddeficient mutant of Rhodobacter sphaeroides. J Bacteriol 174(7):2352–2360. https://doi. org/10.1128/jb.174.7.2352-2360.1992 13. Do¨rmann P, Hoffmann-Benning S, Balbo I et al (1995) Isolation and characterization of an Arabidopsis mutant deficient in the thylakoid lipid digalactosyl diacylglycerol. Plant Cell 7(11):1801–1810. https://doi.org/10.1105/ tpc.7.11.1801 14. Khan M-U, Williams JP (1977) Improved thin-layer chromatographic method for the separation of major phospholipids and glycolipids from plant lipid extracts and phosphatidyl glycerol and bis(monoacylglycery) phosphate

Thin-Layer Chromatography from animal lipid extracts. J Chromatogr A 140 (2):179–185. https://doi.org/10.1016/ S0021-9673(00)88412-5 15. Heyneman RA, Bernard DM, Vercauteren RE (1972) Direct fluorometric microdetermination of phospholipids on thin-layer chromatograms. J Chromatogr A 68(1):285–288. https://doi.org/10.1016/S0021-9673(00) 88791-9

41

16. Mu¨thing J, Kemminer SE (1996) Nondestructive detection of neutral glycosphingolipids with lipophilic anionic fluorochromes and their employment for preparative highperformance thin-layer chromatography. Anal Biochem 238(2):195–202. https://doi.org/ 10.1006/abio.1996.0275

Chapter 4 Lipid Analysis by Gas Chromatography and Gas Chromatography–Mass Spectrometry Mathias Brands, Philipp Gutbrod, and Peter Do¨rmann Abstract Gas chromatography (GC) and gas chromatography–mass spectrometry (GC-MS) represent powerful tools for the quantitative and structural analysis of plant lipids. Here, we outline protocols for the isolation, separation, and derivatization of plant lipids for subsequent GC and GC-MS analysis. Plant lipids are extracted with organic solvents and separated according to their polarity by thin-layer chromatography or solid phase extraction. As most lipids are not volatile, the analytes are derivatized by transmethylation or trimethylsilylation to enable the transition of the molecules into the gas phase. After separation on the polymer matrix of the GC column, the analytes are detected by flame ionization or mass spectrometry. This chapter includes methods suitable for the analysis of lipid-bound or free fatty acids, long chain alcohols, and monoacylglycerols and for the determination of double bond positions in fatty acids. Key words Thin-layer chromatography, Fatty acids, Methyl esters, Gas chromatography, Mass spectrometry, Silylation, Double bond position, Alcohols, Monoacylglycerols

1

Introduction Since its invention in 1952, gas chromatography (GC) became a powerful tool for the quantitative and structural analysis of plant lipids [1]. The prerequisite for GC analysis is that the analytes can change from the liquid to the gas phase without decomposition. To facilitate this transition, often derivatization of the analytes is required to block functional polar groups (e.g., COOH, OH, NH2). Lipid mixtures can be separated by GC into individual compounds using a glass capillary column whose inner surface is coated with a thin polymer film (the stationary phase). Separation of individual compounds is based on the strength of their interactions with the stationary phase. Depending on the application and analytes, different stationary phases, suitable for the separation of specific compounds, may be used [2]. Typically, polar stationary phases are suitable for the separation of esters, alcohols, acids,

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_4, © Springer Science+Business Media, LLC, part of Springer Nature 2021

43

44

Mathias Brands et al.

ethers, amines, and thiols, while nonpolar stationary phases are used for separation of alkanes. The type of stationary phase has an impact on the separation of closely related molecules, such as isomers of unsaturated fatty acid methyl esters (FAMEs) with cis/trans double bonds or different positions of the double bonds. For example, the Supelco SP2380 column used for routine analysis of FAMEs contains a highly polar stationary phase (poly(90% biscyanopropyl/10% cyanopropylphenyl siloxane)) that enables the separation of cis/trans FAMEs. On a polar phase, FAMEs are eluted in the order of increasing chain lengths and of increasing degree of unsaturation. In contrast, the Agilent HP-5MS with a (5%-phenyl)methylpolysiloxane phase is nonpolar and achieves separations of monounsaturated FAMEs that differ only in the position of the double bond. Nonpolar phases are often more thermostable than polar phases. On nonpolar phases, FAMEs are also eluted in the order of increasing chain lengths, but polyunsaturated FAMEs elute before monounsaturated and saturated ones. Once a compound is eluted from the column, a detector monitors the composition of the gas and generates the input signals for data acquisition [3]. The time required for an analyte to reach the detector (retention time) is an important parameter for the identification of the compound. By recording the signal intensity over time, a chromatogram is generated in which the peak area corresponds to the relative abundance of a compound in the mixture. Several detector types including flame ionization (FID) and mass spectrometric (MSD) are used for GC. FID shows a larger linear detection range for quantification compared with MSD, while MSD allows structural analysis due to the ability to generate a mass spectrum of a compound. Comparison of this mass spectrum to publicly available libraries allows for compound identification [4]. Compound quantification is based on the addition of defined amounts of internal standards, substances with similar properties as the analytes, but absent from the samples analyzed. By correlating the peak area of a given analyte with unknown quantity to that of the internal standard, an absolute quantification is possible. Total fatty acids from a biological sample can be analyzed by GC-FID after transmethylation. Either the entire plant organ can be used for the transmethylation reaction, or alternatively lipids are first isolated and subsequently transmethlyated. It is also possible to first separate the lipid extract into different lipid classes by thin layer chromatography (TLC) or solid phase extraction (SPE) prior to the transmethylation reaction. Analysis of long chain alcohols and monoacylglycerols (MAGs) requires purification from the total lipid extract by solid phase extraction (SPE). Long chain alcohols and MAGs are trimethylsilylated prior to analysis by GC-MS [5, 6]. In this chapter, protocols are provided for the quantification of fatty acids, long chain alcohols and MAGs by GC-MS and GC-FID and for structural determination of double bond positions of fatty acids by GC-MS.

Gas Chromatography and Gas Chromatography-Mass Spectrometry

2

45

Materials Use HPLC-analysis grade solvents and rinse all glassware with organic chloroform–methanol (2:1, v/v) prior to use (see Note 1).

2.1 Materials for Lipid Derivatization

1. Sample tissue, total lipid extract or SPE fraction (see Chapters 1, 2, 7). 2. Glass vials with threads (8 mL, 12
100 mm). 3. Teflon (PTFE) lined screw caps. 4. Disposable glass tubes (12
100 mm). 5. Pasteur glass pipettes (150 mm). 6. Plastic tips for micropipette (yellow). 7. Sample evaporator equipped with N2 gas. 8. GC vials with inserts and Teflon-lined caps. 9. Methanolic HCl, 1 N. Store at 4  C (see Note 2). 10. n-Hexane. 11. 0.9% (w/v) NaCl in H2O. 12. Oven or water bath set to 80  C. 13. Centrifuge capable of centrifugation of the glass vials (12
100 mm) at 4000
g (see Subheading 3.1, step 9). 14. N-Methyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA). 15. Dimethyldisulfide (DMDS). 16. Iodine (crystalline, I2). 17. Diethyl ether. 18. 10% (w/v) sodium thiosulfate (Na2S2O3) in H2O (freshly prepared).

2.2 Internal Standards

2.3 Gas Chromatography

The internal standards should be added to the tissue before lipid extraction and SPE to compensate for sample loss (see Chapter 7). Appropriate internal standards are often commercially available as a pure powder. To prevent errors due to weighing of extremely low amounts of standard, it is recommended to weigh the standard for the preparation of a “superstock” (e.g., 100 or 1000 fold concentrated) with an analytical fine balance. This “superstock” is diluted in organic solvents to create an appropriate “working stock” (Table 1). All standards are stored at 20  C. 1. Gas chromatograph with split/splitless injector and flame ionization detector (FID) (e.g., Agilent GC 7890A Plus, or Shimadzu). 2. Gas chromatograph with split/splitless injector and mass detector (quadrupole) (e.g., Agilent 7890A Plus GC-MS).

46

Mathias Brands et al.

Table 1 Standards used for GC-FID and GC-MS analysis Superstock (Solvent; concentration)

Working stock (Solvent; concentration)

Amount of working stock added to the sample

Standarda

Analytes

Pentadecanoic acid (15:0)

Fatty acids

(Methanol; 19.5 nmol/μL)

(Methanol; 0.195 nmol/μL)

100 μL (19.5 nmol)

Pentadecanoylglycerol (15:0 MAG) (87.7% α-form, 12.3% β-form; SigmaAldrich)

Monoacylglycerol

(Chloroform; 3.4 μmol/μL)

(Chloroform; 3.4 nmol/μL)

10 μL (34 nmol [29.81 nmol α-form, 4.182 nmol β-form])

9-cis-Octadecenol (18:1ol)

Long chain alcohols

(Chloroform; 37 μmol/μL)

(Chloroform; 0.37 nmol/μL)

27 μL (9.99 nmol)

a

The standards are weighed using an analytical balance scale and dissolved in 1–10 mL of the solvent to provide the superstock and then diluted to give the working stock

3. GC column for FAME analysis: e.g., Supelco SP 2380 (30 m
0.25 mm, 0.25 μm film). 4. GC column for long chain alcohol or MAG analysis: for example, Agilent HP-5MS column (30 m
0.25 mm, film thickness 0.25 μm). 5. Gas cylinders: helium for column flow; hydrogen and synthetic air (20.5% oxygen, 79.5% nitrogen) for FID detection. 6. GC autosampler vials (2 mL) with inserts (300 μL) and Teflonlined caps.

3

Methods

3.1 Preparation of FAMEs by Transmethylation and Quantification by GCFID

The analysis of fatty acids by gas chromatography requires conversion into their corresponding FAMEs to enable a transition into the gas phase. FAMEs are prepared by (trans)methylation under acidic conditions and heating. For quantitative analysis of total fatty acids, a defined amount of an internal standard (15:0, free pentadecanoic acid) is added prior to lipid extraction. FAMEs can be used for quantification of total lipids (¼total fatty acids) or a specific lipid class that was isolated, for example, via TLC or SPE from the total lipid extract (see Chapter 3 and 7). In the latter case, the internal free fatty acid standard (e.g., 15:0) needs to be added after TLC or SPE, or an alternative internal standard might be used that can be added prior to lipid extraction (see Note 3). Ideally, fatty acids that are absent from plants are used as internal standards for fatty acid quantification. The transmethylation reaction can be performed directly with whole plant tissues like leaves or seeds or with

Gas Chromatography and Gas Chromatography-Mass Spectrometry

47

extracted lipids [7]. However, homogenization of resilient tissues (e.g., seeds or thick roots) may be necessary (see Note 4). The incubation time required for complete derivatization of whole tissue samples may need to be extended. (Trans)methylation with methanolic HCl under acidic conditions has proven to represent a very robust method to convert free and esterified fatty acids into their FAMEs. Alternative (trans)methylation reagents are available (see Note 5). 1. The lipid extract or SPE fraction is transferred to a glass screw cap vial and the organic solvent is dried under a stream of N2 gas (see Note 6). Plant tissues with thin cell walls, for example, leaves or seeds of A. thaliana, can be placed directly into the 1 N methanolic HCl (see next step) (see Note 4). 2. 1 mL of 1 N methanolic HCl is added to the sample using rinsed plastic pipette tips (see Note 7). 3. If no internal standard has been added during lipid extraction or purification, 100 μL (19.5 nmol) pentadecanoic acid (15:0) is added to the sample using plastic pipette tips (see Note 8). 4. The glass vial is tightly closed with a screw cap with Teflon septum. 5. The glass vial is transferred to an oven or water bath and incubated at 80  C for 20–60 min (see Note 9). 6. The glass vial is removed from the oven or water bath and allowed to cool to room temperature. 7. Add 1 mL n-hexane and 1 mL 0.9% (w/v) NaCl in H2O to the sample, shake and vortex vigorously. 8. Phase separation is achieved by centrifugation for 3 min at 4000
g. 9. The upper organic phase containing the FAMEs is transferred to a disposable glass tube and the solvent evaporated under a stream of N2 gas. 10. For GC analysis, the FAME sample is dissolved in 50–100 μL n-hexane and transferred to a clean GC vial containing an insert using plastic pipette tips. Alternatively, samples may be stored at 20  C or used in other procedures (e.g., determination of double bond position, see Subheading 3.3). 11. For quantification of FAMEs, an Agilent GC 7890A Plus coupled to a flame ionization detector and the following settings are used: injector temperature: 220  C; 3 μL splitless injection; detector temperature: 250  C; hydrogen flow rate 30 mL/min; synthetic air flow rate 400 mL/min; helium makeup flow 19 mL/min; GC column, Supelco SP 2380 (see Note 10); helium flow rate 7 mL/min; Oven ramp: from initial 100  C to 160  C with a 25  C/min increase, to 220  C with a 10  C/min increase and back to 100  C with a decrease of 25  C/min.

48

Mathias Brands et al. 1200 18:3

Signal (pA)

1000 800 600 16:0

400

16:3 18:2

200 15:0 (IS)

16:1

18:1 16:2 18:0

0 3.5

4.0

4.5 5.0 5.5 Retention time (min)

6.0

6.5

Fig. 1 GC-FID chromatogram of fatty acid methyl esters (FAMEs) generated from Arabidopsis thaliana total lipid extracts of leaves. The number above the peaks indicates the number of carbon atoms and the degree of desaturation. IS, intnernal standard. The x-axis indicates the retention time and the y axis the current (pA) of the FID detector

12. Quantification of fatty acids is performed by correlating the peak area of the fatty acids to be quantified, to that of the internal standard (15:0; pentadecanoic acid) (Fig. 1). As FAMEs differ in their carbon chain lengths and degree of desaturation, the detector response of the FAMEs can be different. To account for this discrepancy and allow for accurate quantification, response factors should be determined for all fatty acids [8]. For this, equal amounts (nmol or μg) of a fatty acid (e.g., 16:3) and the internal standard (e.g., 15:0) are simultaneously measured by GC. The two peak areas are used to calculate the response factor (e.g., for 16:3). Note that response factors can be based on nmol or μg quantification and should be around 1.0. Typically, response factors are determined once but may have to be readjusted when GC parameters change (e.g., changing the type of column or detector). The relative response factor has to be experimentally obtained for every fatty acid which is quantified. 3.2 Trimethylsilyation of Long Chain Alcohols and Quantification by GC-MS

The analysis of long chain alcohols from plants (e.g., phytol) by gas chromatography requires the conversion into their corresponding trimethylsilyl (TMS) ethers to enhance transition into the gas phase. Long chain alcohol-TMS ethers are synthesized using MSTFA (see Note 11). Heating accelerates the reaction. For quantitative analysis, an internal standard (octadecenol) of a defined amount is added prior to lipid extraction and SPE (see Chapter 7).

Gas Chromatography and Gas Chromatography-Mass Spectrometry

49

Because of their low abundance, it is generally required to purify (e.g., by SPE) long chain alcohols after extraction from the plant tissue. 1. The SPE fraction containing long chain alcohols is transferred to a glass screw cap vial and dried under a stream of N2 gas (see Note 6). 2. 80 μL MSTFA is added to the sample using plastic pipette tips (see Note 12). 3. The glass vial is tightly closed with a screw cap with Teflon septum. 4. The glass vial is transferred to an oven or water bath set to 80  C and incubated for 30 min. 5. The glass vial is removed from the oven or water bath and allowed to cool to room temperature and the solvent evaporated under a stream of N2 gas. 6. For GC analysis, the sample is dissolved in 50–100 μL n-hexane and transferred with a rinsed plastic pipette tip to a clean GC vial containing an insert. Because of their instability, trimethylsilylated alcohols should be measured directly after derivatization. They can be dissolved in n-hexane stored for a short time (a few days) at 20  C (see Note 13). 7. For GC-MS analysis, 2 μL are injected on a GC-MS (Agilent 7890A Plus GC-MS) in splitless mode and separated on an Agilent HP-5MS column (see Note 14) using the following oven temperature gradient: 70  C—start; increase to 310  C at 5  C/min steps; hold for 1 min; decrease to 70  C at 25  C/ min steps. 8. Quantification is performed by correlating the peak area in the total ion chromatogram (TIC) of the long chain alcohol (e.g., phytol) to that of the internal standard (octadecenol) (Fig. 2) (see Note 15). 3.3 Trimethylsilylation and GC-MS Analysis of Monoacylglycerols

Structural identification and quantification of monoacylglycerols (MAG) is achieved by GC-MS analysis of the TMS derivatives of MAGs as described in [6]. Two regioisomeric forms of MAGs exist, α-MAG and β-MAG, depending on the attachment of the fatty acid to the sn-1/sn-3 (α) or the sn-2 (β) carbon of the glycerol. The two forms of α-MAG (sn1-MAG and sn3-MAG) are enantiomeric and cannot be separated by conventional GC-MS. However, the α-MAG and β-MAG regioisomers can be separated on the HP-5MS GC column (Fig. 3a) and give rise to distinct mass spectra. Characteristic fragment ions serve for the identification of α-MAGs and β-MAGs (see Note 15). For quantification, the ion with m/z [M+-103] is used for α-MAGs (Fig. 3b) and the ion with m/z [M+-161] for β-MAGs (Fig. 3c). Because in most cases, the

Mathias Brands et al.

8e+6 Signal intensity

50

6e+6

18:1ol

Phytol

4e+6

2e+6

0 23.5

24.0

24.5

25.0

25.5

Retention time (min)

Fig. 2 GC-MS chromatogram of trimethylsilylated alcohols. Long chain alcohols were extracted from Arabidopsis thaliana leaves with organic solvents and purified on a silica SPE column (see Chapter 7). The peaks representing the internal standard octadecenol (18:1ol) and phytol are indicated. The x-axis indicates the retention time and the y axis represents the signal intensity of a total ion chromatogram

lipid fraction containing the MAGs (e.g., isolated via SPE; see Chapter 7) includes other unidentified compounds, it is important to employ single ion monitoring for MAG quantification using the characteristic ions described above. 15:0 α-MAG and β-MAG are used as internal standards. 1. Transfer MAG-containing lipid fraction (e.g., SPE fraction) into glass vials (8 mL, 12
100 mm). 2. Evaporate the solvent under a stream of N2 (see Note 6). 3. Add 200 μL MSTFA and vortex. 4. Close the vial with screw cap containing a Teflon septum. 5. Incubate at 80  C for 30 min. 6. Evaporate MSTFA under a stream of N2 (see Note 6). 7. Add 200 μL n-hexane. Transfer n-hexane containing lipids with a rinsed plastic pipette tip to a GC vial containing an insert and then perform GC-MS analysis (see Note 13). 8. For GC-MS analysis, 3 μL is injected in splitless mode on the GC-MS (Agilent 7890A Plus GC-MS) and separated with an HP-5MS column using the following temperature gradient: 120  C for 4 min; increase to 300  C at steps of 10  C/min; hold for 2 min; decrease to 120  C at steps of 15  C/min.

Gas Chromatography and Gas Chromatography-Mass Spectrometry

(a)

51

15:0α (IS) 6e+6

Signal intensity

16:0α 18:0α 18:1+ 18:2α

16:0β

14:0α 2e+6

18:3α

16:1α

15:0β (IS)

4e+6

14:0β

18:3β

16:1β

18:1+ 18:2β 0 16

18 Retention time (min)

(b)

16:0 α-MAG-diTMS O

Si

4e+5

O

371

371 m/z C21H43O3Si

5e+5 Signal intensity

20

O

O 3e+5

Si

2e+5

147

1e+5

M+ 474 m/z C25H54O4Si2

205

239

M+-15 459

313

0 50

(c)

100

129

14e+4

150

200

250 m/z

300

400

450

500

16:0 β-MAG-diTMS 218

M+ 474 m/z C25H54O4Si2

Si

12e+4 Signal intensity

350

O 218 m/z C9H22O2Si2

10e+4 8e+4

O O

O

Si 101

6e+4

313 191

4e+4

239

2e+4

401

M+-15 459

0 50

100

150

200

250 m/z

300

350

400

450

Fig. 3 Analysis of monoacylglycerols (MAGs) by GC-MS. (a) GC-MS total ion chromatogram of bis-trimethylsilylated compounds. MAGs were measured in a lipid extract of Lotus japonicus roots colonized with the mycorrhiza fungus Rhizophagus irregularis after SPE purification (see Chapter 7) α-MAG and β-MAG peaks are indicated with the abbreviation of the acyl groups. IS internal standard. (b) GC-MS electron impact

52

Mathias Brands et al.

3.4 Analysis of Double Bond Position in Unsaturated Fatty Acids by GC-MS

The double bond position in unsaturated fatty acids can be determined after derivatization with dimethyldisulfide (DMDS). Fragmentation in the ion source at the derivatized position gives rise to characteristic fragments which can be used to calculate the position of the double bond in the acyl chain. For DMDS derivatization of unsaturated fatty acids, FAMEs are first prepared from lipid extracts by transmethylation (see Subheading 3.1). The protocol for the determination of the double bond position is based on [9]. 1. Transfer FAMEs into glass vials (8 mL, 12
100 mm). 2. Evaporate n-hexane from the FAME solution under a stream of N2 (see Note 6). 3. Dissolve FAMEs in 100 μL of a solution of iodine (I2) in diethyl ether (60 mg/mL) and add 350 μL dimethyldisulfide (DMDS). Close the vial with a screw cap containing a PTFE septum. 4. Incubate with shaking at 37  C for 3 h (see Note 16). 5. Add 1 mL n-hexane/diethyl ether (1:1) while gently vortexing. 6. Slowly add a few drops of 10% (w/v) Na2S2O3 (sodium thiosulfate in H2O) with a Pasteur pipette under strong vortexing until the solution becomes completely clear again (see Note 17). 7. Transfer upper hexane phase to a GC vial containing an insert and perform GC-MS analysis. 8. The DMDS-treated FAMEs (3 μL) are injected in splitless mode on a GC-MS (Agilent 7890A Plus GC-MS) and separated on a HP-5MS column (see Note 14) using the following temperature gradient: 120  C for 4 min; increase to 300  C at 10  C/min steps; hold for 2 min; decrease to 120  C at 15  C/ min steps. 9. Monounsaturated FAMEs will give rise to bis(methylthio) derivatives that are eluted later in the GC-MS chromatogram, as compared to the corresponding nonderivatized FAMEs. For example, 16:1Δ11 methyl ester is eluted at 12.795 min while the 16:1Δ11-bis(methylthio) methyl ester elutes at 18.528 min (Fig. 4a). Fragmentation between the two methylthioderivatized carbons that formed the original double bond gives rise to intense fragments whose m/z values provide the possibility to calculate the position of the double bond (Fig. 4b, c).

ä Fig. 3 (continued) (EI) mass spectrum (MS) and the structure of the bis-TMS derivative of 16:0 α-MAG. The dotted red line indicates the site of fragmentation giving rise to the 371 m/z fragment ion. (c) GC-MS electron impact (EI) mass spectrum (MS) and the structure of the bis-TMS derivative of 16:0 β-MAG. The dotted red line indicates the site of fragmentation giving rise to the 218 m/z fragment ion. The other diagnostic ion, 313 (m/z) is also indicated in the mass spectrum

Gas Chromatography and Gas Chromatography-Mass Spectrometry

(a) 3.0e+7

11,12-bis (mehtylthio) 16:1 methyl ester

2.0e+7 Signal intensity

53

1.0e+7

9,10-bis (mehtylthio) 16:1-methyl ester

1.0e+6

0.0 18.2

(b)

18.4

9,10-bis(mehtylthio)16:1-methyl ester 217

O 217 m/z C11H21O2S

145

8e+4

S

S

O

1e+5 55

Signal intensity

19.0

18.8

18.6 Time (min)

145 m/z C8H17S

M+ 362 m/z C19H38O2S2

6e+4 97 4e+4

185

137

[M+] 362

109 2e+4

168 236

0 50

(c)

100

150

341

250

200 m/z

300

11,12-bis(mehtylthio)16:1-methyl ester S

O 4e+5 Signal intensity

350

S

O

3e+5 55 61

245

117

245 m/z C13H25O2S

117 m/z C8H13S

M+ 362 m/z C19H38O2S2

69 2e+5

1e+5

81 87 95 147

[M+] 362 165

196

0 50

100

150

200 m/z

236 250

300

350

Fig. 4 Determination of the double bond position of unsaturated fatty acids by GC-MS. A total lipid extract from Lotus japonicus roots colonized with the mycorrhiza fungus Rhizophagus irregularis was used to synthesize FAMEs. The

4

Mathias Brands et al.

Notes 1. Exposure of the samples dissolved in organic solvents to plastic ware such as pipet tips or microfuge tubes should be minimized to avoid leakage of contaminants (e.g., plasticizers) into the sample. Always use vials with Teflon (PFTE) septa, not rubber septa. For derivatization, glass tubes are recommended whose inner surface was rinsed with organic solvents (e.g., chloroform–methanol, 2:1 v/v). Prepare a control vial containing an internal standard but no biological tissue to evaluate the background contamination from solvents, glassware and plasticware, or instruments (“solvent control”). 2. 1 N HCl in methanol can be prepared by diluting 3 N HCl (commercially available) in methanol. Alternatively, 1 N HCl can be obtained adding 7.85 g (¼7.13 mL) of acetylchloride drop by drop into cold methanol. Be careful because the reaction is exothermic! 3. Instead of using a free fatty acid (e.g., 15:0) as internal standard for FAME analysis, lipid-bound fatty acids (e.g., tri-15:0-TAG) can be used to avoid discrepancies in the degree of (trans) methylation between free and esterified fatty acids. 4. Although some tissues can be used directly in transmethylation reactions [7], the water content shall not exceed 5% of the final volume of the reaction to avoid lipid hydrolysis. Soft or very small tissues like leaves or seeds of A. thaliana can be directly transmethylated, but more rigid material such as large seeds (Lotus japonicus, cereals, rapeseed, etc.) or rigid roots need to be cut into small pieces or homogenized and subjected to lipid extraction before transmethylation. 5. Methanol/BF3 can be used as an alternative acidic (trans)methylation reagent for esterification of free and lipid-bound fatty acids; it is potentially even more efficient at catalyzing methylation of less common or unusual fatty acids. However, methanol/BF3 is more expensive [10]. Additionally, methylation with BF3/methanol can result in the formation of methoxy artifacts by addition of methanol to a double bond [11]. Longer reaction times might be necessary for transesterification ä

54

Fig. 4 (continued) FAMEs contain 16:1Δ9 (derived from plant) and 16:1Δ11 (fungal fatty acid). (a) GC-MS chromatogram of DMDS-derivatized FAMEs containing 16:1Δ9 and 16:1Δ11. (b, c) GC-MS electron impact (EI) mass spectra (MS) and the structures of the bis(mehtylthio) derivatives of 16:1Δ9 or 16:1Δ11 FAMEs. The dotted red lines indicate the sites of fragmentation giving rise to the fragment ions containing the two parts of the carbon chain each with a single methylthiol group on the left and the right side of the original double bond. The fragment ions are indicated in red in the mass spectra

Gas Chromatography and Gas Chromatography-Mass Spectrometry

55

with methanol/BF3 compared with methanolic HCl [12]. Another acidic methylation reagent (for free and esterified acyl groups) is 2.5% sulfuric acid (H2SO4) in methanol (see Chapter 1). Basic transesterification protocols include sodium or potassium methoxide in methanol (methanolic NaOCH3 or KOCH3). While esterified acyl groups are rapidly methylated, free fatty acids do not react with NaOCH3 or KOCH3 [13]. Diazomethane can be used for the selective methylation of free fatty acids (esterified acyl groups are not transmethylated) [14]. However, because of its high toxicity and because it is a gas that needs to be generated in the lab, strict precautions have to be taken during methylation reactions with diazomethane. Alternatively, trimethylsilyldiazomethane which is a liquid and easier to handle can be used for methylation of free fatty acids instead of diazomethane (see Chapter 9). 6. FAMEs and TMS derivatives of lipids are volatile. Therefore, the N2 stream used for solvent evaporation should be very gentle, and should be stopped once the vial is dry, to avoid blowing away the analyte. 7. It is sometimes difficult to completely dissolve dried lipid extracts in some organic solvents. A short sonication treatment may help to overcome this problem. 8. Internal standards should be added during the extraction to compensate for matrix effects or sample losses during the handling (see Chapter 7). 9. Extracted, dried, total lipids from up to 100 mg (fresh weight) plant tissue are incubated for 20 min at 80  C. If more tissue is used, the reaction time should be extended to 60 min or until all esterified and free fatty acids are converted into FAMEs. If necessary, this can be experimentally determined by a methylation kinetic using different reaction times with the same lipid extract. For transesterification of lipids in intact A. thaliana seeds, incubate the seeds for 60 min in 1 N HCl at 80  C. If FAMEs are prepared from A. thaliana leaves, 20 min of treatment are sufficient. 10. Columns with polar phases are optimal for the separation of plant-derived FAMEs. 11. Instead of using MSTFA, lipids can be trimethylsilylated with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA). BSTFA and MSTFA have similar reactivity for the introduction of TMS groups. However, the by-products of MSTFA derivatization are more volatile and therefore elute very early during GC analysis, and may prevent detection of early-eluting analytes. 12. The trimethylsilylation reaction is highly sensitive to the presence of water. Therefore, remove any residual amounts of water from the lipid sample prior to adding MSTFA.

56

Mathias Brands et al.

13. TMS ethers are instable and volatile and even though they can be stored, it is recommended to measure them directly after synthesis. Store them airtight excluding humidity, because the TMS ether linkage is prone to hydrolysis. 14. The Agilent HP5-MS column (nonpolar phase) can be used for the separation of nonpolar compounds including TMS derivatives of alcohols and MAGs. It is also suitable for the separation of FAMEs or FAME-DMDS derivatives. The column is operated with helium gas as carrier at a flow rate of 2 mL/min. Injector temperature is set at 250  C and detector operated at 220  C. Ions are produced by electron impact (EI) ionization with 69 eV, and mass spectra are recorded in full-scan mode at 2.91 s intervals. 15. For identification of compounds (e.g., phytol and MAGs) in a chromatogram, GC-MS mass spectra from candidate peaks can be searched at the National Institute of Standards and Technology (NIST) database (www.nist.gov/srd/nist-standard-ref erence-database-1a-v17). 16. To ensure constant mixing, use a rotary platform and tape the tube horizontally but slightly inclined so that the liquid does not come in excessive contact with the septum that seals the lid. 17. In this step, S2O3 (thiosulfate) is oxidized and iodine is reduced to iodide. It is therefore important to proceed until the solution is completely clear, as unreacted iodine will damage the GC column. The total volume of sodium thiosulfate that needs to be added can therefore differ. In the end, the DMDS derivatives will be concentrated in the upper hexane phase. References 1. James AT, Martin AJP (1952) Gas-liquid partition chromatography: the separation and micro-estimation of volatile fatty acids from formic acid to dodecanoic acid. Biochem J 50:679–690. https://doi.org/10.1042/ bj0500679 2. McNair H, Miller J (1997) Basic gas chromatography. Wiley, Hoboken, ISBN 0-471-17261-8 3. Bartle KD, Myers P (2002) History of gas chromatography. Trends Analyt Chem 21 (9–10):547–557. https://doi.org/10.1016/ S0165-9936(02)00806-3 4. Babushok VI, Linstrom PJ, Reed JJ, Zenkevich IG, Brown RL, Mallard WG, Stein SE (2007) Development of a database of gas chromatographic retention properties of organic compounds. J Chromatogr A 1157

(1–2):414–421. https://doi.org/10.1016/j. chroma.2007.05.044 5. Ischebeck T, Zbierzak AM, Kanwischer M, Do¨rmann P (2006) A salvage pathway for phytol metabolism in Arabidopsis. J Biol Chem 281(5):2470–2477. https://doi.org/10. 1074/jbc.M509222200 6. Destaillats F, Cruz-Hernandez C, Nagy K, Dionisi F (2010) Identification of monoacylglycerol regio-isomers by gas chromatographymass spectrometry. J Chromat B 1217:1543–1548. https://doi.org/10.1016/ j.chroma.2010.01.016 7. Browse J, McCourt PJ, Somerville CR (1986) Fatty acid composition of leaf lipids determined after combined digestion and fatty acid methyl ester formation from fresh tissue. Anal Biochem 152:141–145. https://doi.org/10. 1016/0003-2697(86)90132-6

Gas Chromatography and Gas Chromatography-Mass Spectrometry 8. de Saint Laumer J, Leocata S, Tissot E, Baroux L, Kampf DM, Merle P, Boschung A, Seyfried M, Chaintreau A (2015) Prediction of response factors for gas chromatography with flame ionization detection: algorithm improvement, extension to silylated compounds, and application to the quantification of metabolites. J Sep Sci 38(18):3209–3217. https:// doi.org/10.1002/jssc.201500106 9. Francis GW (1981) Alkylthiolation for the determination of double-bond position in unsaturated fatty acid esters. Chem Phys Lipids 29:369–374. https://doi.org/10.1016/ 0009-3084(81)90070-0 10. Weston R, Derner JD, Murrieta CM, Rule CD, Hess BW (2006) Comparison of catalysts for direct transesterification of fatty acids in freeze dried forage samples. Crop Sci 48:1636–1641. https://doi.org/10.2135/cropsci2007.07. 0376sc

57

11. Christie WW (1993) In: Christie WW (ed) Preparation of ester derivatives of fatty acids for chromatographic analysis. Advances in lipid methodology-two. Oily Press, Dundee 12. Morrison WR, Smith LM (1964) Preparation of FAMEs and dimethylacetals from lipids with boron fluoride-methanol. J Lipid Res 5 (4):600–608 13. Luddy FE, Barford RA, Rimenschneider RW (1960) Direct conversion of lipid components to their fatty acid methyl esters. J Am Oil Chem Soc 37:447–451. https://doi.org/10.1007/ BF02631205 14. Schlenk H, Gellermann JL (1960) Esterification of fatty acids with diazomethane on a small scale. Anal Chem 32(11):1412–1414. https:// doi.org/10.1021/ac60167a011

Chapter 5 14

C-Tracing of Lipid Metabolism

Hari Kiran Kotapati and Philip D. Bates Abstract Lipids are produced through a dynamic metabolic network involving branch points, cycles, reversible reactions, parallel reactions in different subcellular compartments, and distinct pools of the same lipid class involved in different parts of the network. For example, diacylglycerol (DAG) is a biosynthetic and catabolic intermediate of many different lipid classes. Triacylglycerol can be synthesized from DAG assembled de novo, or from DAG produced by catabolism of membrane lipids, most commonly phosphatidylcholine. Quantification of lipids provides a snapshot of the lipid abundance at the time they were extracted from the given tissue. However, quantification alone does not provide information on the path of carbon flux through the metabolic network to synthesize each lipid. Understanding lipid metabolic flux requires tracing lipid metabolism with isotopically labeled substrates over time in living tissue. [14C]acetate and [14C]glycerol are commonly utilized substrates to measure the flux of nascent fatty acids and glycerol backbones through the lipid metabolic network in vivo. When combined with mutant or transgenic plants, tracing of lipid metabolism can provide information on the molecular control of lipid metabolic flux. This chapter provides a method for tracing in vivo lipid metabolism in developing Arabidopsis thaliana seeds, including analysis of 14C labeled lipid classes and fatty acid regiochemistry through both thin-layer chromatography (TLC) and high-performance liquid chromatography (HPLC) approaches. Key words Radiolabel, Metabolic flux, Pulse–chase, Flow liquid scintillation counting, Carbon-14

1

Introduction The use of radioisotopes to trace and quantify carbon flux through in vivo lipid metabolism has a long history in plant biology research, see reviews: [1–6]. In the early stages of developing Arabidopsis thaliana as a model plant it quickly became clear that utilizing isotopic tracing of metabolism in combination with genetic mutants was crucial to defining the role of specific genes in plant lipid metabolic pathways [7]. Modern molecular biology approaches allow to easily knockout, knock-down, or overexpress genes of interest. Combining these modified plants with in vivo analysis of acyl fluxes in A. thaliana leaves and seeds has provided key insights into the molecular control of metabolism and metabolic bottlenecks that limit oilseed engineering, for examples see [8–

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_5, © Springer Science+Business Media, LLC, part of Springer Nature 2021

59

60

Hari Kiran Kotapati and Philip D. Bates

13]. This book chapter illustrates a way to analyze oilseed metabolism in mutant and/or transgenic plants with a focus on tracing glycerolipid precursor–product relationships within developing A. thaliana seeds actively producing triacylglycerols (TAG). There are several key factors to consider when designing a lipid metabolic tracing experiment in developing A. thaliana seeds: 1.1 Stage of Seed Development

Tracing of metabolic flux needs to be done during the stage of tissue development that is actively producing the compounds of interest. A. thaliana seeds develop over approximately a three-week period. Small amounts of TAG are produced during the first week, then TAG rapidly accumulates during the second week before TAG levels stabilize and then decrease during seed maturation in the third week [14, 15]. However, the amount of TAG, the TAG fatty acid composition, and the timing of the rapid TAG accumulation phase is highly dependent on the growth conditions, and can vary between lines [11, 16, 17]. Therefore, to choose the proper stage of silique development to collect seeds for a labeling experiment, a developing seed TAG accumulation curve should be produced for the lines and exact growth conditions that will be used to grow the plants for seed labeling. We have observed at least a two-day shift in the start and end of the TAG accumulation phase between different growth chambers programmed for similar growth conditions. Choosing a seed age suitable for metabolic labeling, developing A. thaliana seeds of a narrow age range (e.g., 2–3 days apart) give more consistent results than seeds collected from a wider developmental range.

1.2 Choice of Radiolabeled Substrate

Different substrates (e.g., isotopically labeled acetate or glycerol) can label various parts of different lipid classes to varying degrees. This is dependent on the relative flux through competing metabolic pathways in vivo and the labeling time points, and it affects the relative quantification of total isotope incorporation into each glycerolipid. [14C]acetate incorporated into plant tissues is activated to acetyl-CoA in the plastid and cytosol. Fatty acids are labeled in the first 18 carbons through de novo fatty acid synthesis in the plastid, and fatty acids
20 carbons can also be labeled through cytosolic fatty acid elongation. However, the specific activity of label incorporation into the cytosolic elongated portion of the fatty acid is typically higher than the de novo synthesized portion and varies with time [10, 18]. Therefore, fatty acids with different chain lengths will have nonequivalent molar ratios of total labeling, and this will affect the relative labeling of glycerolipids depending on their fatty acid composition and number of fatty acids in each lipid class. [14C]glycerol is converted to glycerol-3-phosphate and predominantly labels the glycerol backbone of glycerolipids. However, the glycerol headgroup of phosphatidylglycerol (PG) is also labeled leading to twice the amount of 14C in PG. In addition,

14

C-Tracing of Lipid Metabolism

61

approximately 10–20% of the [14C]glycerol is converted to acetylCoA through glycolysis and leads to fatty acid labeling having the same considerations as for [14C]acetate. The relative flux of labeled carbon from [14C]glycerol into the glycerol backbone or the fatty acids of glycerolipids is dependent on the lipid class and time. For example a 10 min [14C]glycerol labeling of Arabidopsis seeds produced TAG containing 60% of the radioactivity in the glycerol backbone and 40% in the acyl chains, whereas the 14C incorporation into phosphatidylcholine at the same time point was 90% in the glycerol backbone and 10% in the acyl chains [9]. Therefore, accurate quantification of glycerol backbone labeling from [14C]glycerol requires the separation of the acyl chains from the backbone prior to measuring the radioactivity by liquid scintillation counting. The 14C label of both [14C]acetate and [14C]glycerol is incorporated into glycerolipids, starting from initial steps of lipid metabolism, and they are preferred substrates to trace de novo biosynthesis of glycerolipids. Sometimes full-length fatty acids can be used for targeted in vivo labeling experiments. However, fatty acids are less easily taken up into plant tissue, can induce feedback inhibition of fatty acid synthesis [19], and it is not exactly clear at which point(s) in the lipid metabolic network they are first incorporated into glycerolipids. Therefore, it cannot be assumed that labeled full length fatty acids will trace the same metabolism as [14C]acetate labeled fatty acids produced by de novo fatty acid synthesis; thus, care must be taken when interpreting the results. 1.3 Type of Labeling Experiment and Time Points

The relative accumulation of isotopic label between any two metabolite pools is dependent on the size and metabolic flux (rates of synthesis/turnover) through all metabolic precursor pools to the given metabolites, and both the size and flux through the metabolite pools of interest [5]. It is important to point out that the size of active metabolite pools cannot be determined from the mass quantification of tissue extracts because the same metabolite can have multiple subcellular pools with different sizes and turnover rates [4, 12]. Metabolic labeling time points that are too long will report mostly on pool size differences between metabolites, not precursor–product relationships within metabolism. While time points that are too short, risk not labeling the metabolite of interest, or only labeling it through one of multiple possible pathways. Therefore, single time point labeling studies are of limited informational value, and time-course studies are preferred. There are two types of time-course based metabolic labeling studies, continuous pulse and pulse–chase studies. During continuous pulse labeling studies, the tissue of interest is incubated with the radioisotope over the entire time-course and the accumulation of radiolabel into metabolite pools can be followed until they reach steady state. Metabolism is traced by following the precursor–product relationships of pool filling and typically requires short time

62

Hari Kiran Kotapati and Philip D. Bates

points (seconds to tens of minutes). In pulse–chase studies, the tissue is incubated with the labeled substrate for a short time (tens of minutes), the substrate is removed and then the incubation of the tissue is continued for a longer time period (hours to days). Pulse–chase time courses follow the turnover of initially labeled intermediate pools and the filling of downstream and endpoint metabolite pools. In studying plant lipid metabolism, short continuous time courses are useful for tracking the initial steps of newly synthesized fatty acid incorporation into glycerolipids [10, 12, 20– 22], whereas pulse–chase labeling experiments are useful for following the turnover of intermediates such as PC prior to TAG or galactolipid synthesis [7, 12, 23, 24]. 1.4 Variations in Structural Labeling Aid in Defining Pathway Flux

Beyond determining the rate of radiolabel incorporation into different lipid classes, additional information on the regiochemistry of labeled fatty acid incorporation into each lipid class, or the specific lipid molecular species that are labeled can be an essential component of deciphering the pathway of metabolic flux [9, 10, 12, 20, 21, 23, 25–27]. For example, in many plant species and various plant tissues, at short time points of [14C]acetate labeling (e.g., less than 10 min) most newly synthesized fatty acids are initially incorporated into mostly the sn-2 position of PC by acyl editing, TAG is mostly labeled at the sn-3 position by acylation of an unlabeled DAG, and the 14C labeled DAG pool contains labeled fatty acids in similar proportions at sn-1 and sn-2 (typically with a slight preference for sn-1) by de novo DAG synthesis through the Kennedy pathway. Regardless of the rate of labeling of each of these lipid classes the different regiochemistry of acyl labeling implies that de novo DAG synthesized by the Kennedy pathway is not the immediate precursor of the initially labeled PC and TAG. Thus, combining metabolic labeling with regiochemical analysis has helped to define lipid biosynthesis as a metabolic network rather than a linear pathway [4, 6]. The above considerations are to help in designing a metabolic labeling experiment tailored to the needs of the specific scientific question. The procedure outlined below is a general method for short time point continuous pulse labeling in developing A. thaliana seeds. We have previously used this protocol to analyze acyl editing in A. thaliana mutants with [14C]acetate [10], and the role of de novo DAG or PC-derived DAG in production of TAG containing unusual fatty acids with [14C]glycerol labeling in transgenic plants [9]. This general procedure can be adapted for various other needs, such as pulse–chase metabolic labeling. The experimental procedure involves four major aspects as outlined below: 1. Developing staging and collection of Arabidopsis thaliana seeds. 2. Continuous pulse labeling of developing seeds.

14

C-Tracing of Lipid Metabolism

63

3. Lipid extraction. 4. Analysis of labeled lipids—thin-layer chromatography (TLC) or high-performance liquid chromatography (HPLC), and lipid class regiochemical analysis.

2

Materials All organic solvents used are either optima grade or reagent grade. Chemicals used are of reagent grade. All organic, aqueous, and radioactive waste must be disposed of properly according to the local regulatory guidelines. While working with radioactive substances, it is ideal to perform the experiment in a designated area to prevent accidental contamination. Any equipment that comes in contact with the radioactive materials must be thoroughly cleaned and tested for any possible contamination afterward. Safety glasses, lab coat and gloves must be worn during all times while handling the volatile organics and any radioactive materials.

2.1 Seed Staging and Collection

1. Dissecting needle, dissecting stainless steel scissors, razor blade, and glass plate. 2. Glass stirring rod. 3. Ice bath. 4. 1 Incubation medium: 5 mM MES buffer pH 5.8, 0.5% (w/v) sucrose, 0.5 Murashige and Skoog (MS) salts (see Note 1).

2.2 Continuous Labeling Experiment

1. Labeling incubator: For the labeling incubator, we utilize a heated/cooled/shaking water bath with a bank of fluorescent white lights on pulleys hanging above the water bath. The light intensity can be adjusted by moving the lights up and down. 2. Sterile glass culture tubes (16  100 mm). 3. Radioactive waste containers for disposing pipette tips and unlabeled media. 4. Substrates: [14C(U)]Glycerol stock purchased as a solution in sterile water, or [1-14C]acetate purchased as its sodium salt in ethanol. 5. Preparation of 14C incubation medium: For [14C]acetate medium, the required amount of the stock solution in ethanol is placed in a 15 ml disposable centrifuge tube using a Pipetman, and ethanol is gently evaporated under the stream of nitrogen (leaving about 0.05 ml liquid behind to limit evaporation of the substrate). The required amount of 1 incubation medium is added to the evaporated [14C]acetate to prepare the 14 C medium at the desired concentration. For preparing the

64

Hari Kiran Kotapati and Philip D. Bates

[14C]glycerol incubation medium, the water in the substrate stock solution must be taken into account. Place the desired volume of [14C]glycerol stock solution in a 15 ml centrifuge tube and add an equal volume of 2 incubation medium to it. Then make up the volume of the [14C]glycerol labeling medium with 1 incubation medium (see Note 2). 6. Isopropanol for quenching of samples from lipid labeling. 7. 3,5-di-tert-4-butylhydroxytoluene (BHT) to protect lipids from oxidation. 2.3

Lipid Extraction

1. Nitrogen evaporator stationed in a laboratory fume hood for drying the lipid extracts. 2. Polytron homogenizer. 3. Spectrophotometer for chlorophyll measurements. 4. Chloroform, methanol, potassium chloride (KCl) solutions (1 M and 0.8% (w/v)).

2.4

Lipid Analysis

1. Silica gel 60 TLC plates (20  20 cm) for neutral lipid analysis. 2. Analtech HL (20  20 cm) TLC plates for polar lipid analysis. 3. Hexane–ethyl ether–acetic acid (70:30:1, v/v/v) for separating neutral lipids. 4. Acetone–toluene–water (91:30:8, v/v/v) for separating polar lipids. 5. Glass TLC tanks. 6. (NH4)2SO4 and hot air oven operated at 120  C for activating the polar lipid plates prior to use. 7. 0.005% Primulin in acetone–water (80:20, v/v) and UV light screen or lamp. 8. Bottle-type sprayer connected to nitrogen gas source for staining the TLC plates to visualize lipid bands. 9. Imager of radioactive TLC plates (e.g., phosphorimager or radio-TLC scanner). 10. High-performance liquid chromatograph (HPLC) equipped with a PVA-Sil column (250  4.6 mm, 5 μm), flow liquid scintillation counter (e.g., LabLogic β-Ram 6 with 500 μl flow cell), and flow scintillation cocktail. 11. Solvents for HPLC separation: A, isopropanol; B, hexanes; C, methanol; and D, aqueous 25 mM formic acid, 25 mM triethylamine at pH 4.2. 12. Gas chromatograph (GC) equipped with flame ionization detector for lipid quantification. 13. Liquid scintillation counter and scintillation cocktail.

14

2.5 14C Labeled Lipid Regiochemical Analysis

C-Tracing of Lipid Metabolism

65

1. Rhizomucor miehei lipase (activity
20,000 U/g; from SigmaAldrich). 2. Phospholipase A2 (PLA2) from honeybee venom salt-free lyophilized powder (activity ¼ 1775 units/mg solid; from Sigma-Aldrich). 3. Borate buffer: 50 mM boric acid, 5 mM calcium chloride, pH 7.8. Weigh 3.09 g of boric acid into a 1 l glass beaker with a magnetic stir bar. Add 800 ml of water to the beaker and stir the contents at room temperature with the help of a stir plate until all the boric acid is dissolved. Adjust the pH of the solution to 7.8 using 1 M sodium hydroxide solution. Weigh out 0.555 g of calcium chloride, add it to the beaker, and make up the volume to 1 l. The pH of the solution should be 7.8. If not, adjust. 4. Tris buffer: 50 mM Tris–HCl, 5 mM calcium chloride, pH 8.7. 5. Honey bee venom phospholipase A2 (PLA2) stock solution: Prepare stock solution of 1 unit/μl in the Tris buffer, and aliquot 50 μl of this stock into 1.5 ml liquid nitrogen compatible microcentrifuge tubes. Snap-freeze the tubes containing enzyme stock in liquid nitrogen and store at 80  C.

3

Methods

3.1 Developing A. thaliana Seed Staging and Collection

1. Seeds are either started on germination plates and seedlings of consistent size transferred to soil (2 per pot), or germinated on soil directly. The growth conditions are as follows: constant fluorescent white light, temperature 21  C, humidity 60–70% and light intensity of 100–150 μmol photons m2 s1 at pot height (see Note 3). 2. Plants chosen for staging must be numbered and staging of the flowers must be started only after each plant has at least four flowers/siliques. Often, the first 2–3 flowers are sterile. 3. Select one main shoot per plant for staging, and the side shoots may be trimmed off carefully. Trimming can be done periodically during the staging process until the week before harvest. 4. Count the siliques and open flowers starting from the base of the shoot and record the number of open flowers and siliques per plant on the first counting day as the baseline. 5. Count again the open flowers and siliques per plant the following day. The difference in the total number of open flowers and siliques between the first two counting days will give the number of new flowers. These new flowers are considered “zero” days after flowering (DAF) corresponding to “baseline day”.

66

Hari Kiran Kotapati and Philip D. Bates

6. Repeat counting open flowers and siliques until the desired range of siliques of different developmental stages needed for the experiment has been reached. For example, if the siliques that are 9, 10, and 11 DAF are to be harvested, then the flower counting must be performed for four continuous days (baseline day, day “zero”, day 1, and day 2). The harvest date would be 11 days after the day zero. On harvest day the day zero flowers will be 11 DAF, the new flowers from day 1 will be 10 DAF and the new flowers from day 2 will be 9 DAF. 7. On the day of harvest, collect the siliques that correspond to the appropriate age range selected (e.g., 9–10 DAF based on our chamber specific growth curve). Typically, 10 siliques are collected per plant line, time point and replicate. Therefore, 150 siliques of the plant line will be needed for a five-point time course in triplicate. Cut the silique tip off with a razor blade, then insert the dissecting needle at the top of the pedicel between the two silique halves and move up toward the cut tip. The silique should split into two halves that can be scraped out with the needle or a very small scoop. Seeds of early–mid TAG accumulation phase are green, slightly sticky, and will stick to a dissecting needle until it is dipped into medium. Take care not to smash or damage the seeds and place them immediately into a 16  100 mm sterile culture tube containing 1.5 ml of 1 incubation medium on ice. Typically, all siliques of the desired age range are dissected from a single plant before collecting siliques from a new plant, all seeds of a single genotype should be collected into one or a few tubes (see Note 4). 8. Each labeling time course will be a “one pot” labeling strategy consisting of a single tube containing all the seeds for every point in the time course. Replicate time courses will consist of a separate tube of seeds (e.g., three tubes for three replicate time courses, each containing enough seeds for five time points). For consistency of labeling replicates with seeds collected from different plants over several hours, mix all seeds collected of a single genotype into a single tube and redistribute into the desired number of aliquots (see Note 5). 3.2 Continuous Labeling Experiment

1. All collected developing seed samples should be placed on the incubator with gentle shaking to equilibrate to light (100 μmol photons m2 s1 at sample level) and temperature (23  C) conditions for at least 30 min prior to the start of labeling (see Note 6). 2. After the equilibration period, remove all the incubation medium from each replicate time course tube and replace with 14C labeling medium (1 medium containing the 14C isotope, typically at 0.1–1.0 mM). Gently mix the seeds with a

14

C-Tracing of Lipid Metabolism

67

pipette and start the timer. The volume of 14C medium added to each replicate tube is 0.2 ml per time point to be collected, and therefore 1 ml of medium for five time points. The seeds will add significant volume (0.1–0.2 ml) and will need to be accounted for when taking time point aliquots (see Note 7). For a pulse–chase experiment the protocol is similar up to this point to perform the “pulse” part of the pulse–chase (see Note 8). 3. At each time point, mix the seeds gently and remove an aliquot (0.22 ml) containing medium and seeds and quench metabolism by placing directly into 1 ml of hot isopropanol at 85  C containing 0.01% (v/v) 3,5-di-tert-4-butylhydroxytoluene (BHT) in an 8 ml (13  100 mm) screw capped test tube. Cap the tube and incubate at 85  C for 15 min. Typical time points for tracing initial steps of A. thaliana TAG metabolism are 3, 6, 10, 30, and 90 min (see Note 9). 4. After quenching all time points, allow the samples to cool to room temperature and store at 20  C until lipid extraction. 3.3

Lipid Extraction

The protocol used for A. thaliana seed lipid extraction is a modified Bligh and Dyer method [28]. Always begin the lipid extraction with shortest time point labeled samples and finish with longest time point labeled samples to prevent cross contamination of the label. The extraction must be performed in a laboratory fume hood as the procedure involves the use of volatile organic solvents such as chloroform and methanol. Glass serological pipets are used to add solvent, and glass Pasteur pipets are used to transfer solvents between tubes. Because each time point aliquot will have a different number of seeds in it, the extracts must be normalized to each other for linear quantification of 14C accumulation over the time course. If the A. thaliana plant lines that are compared have the same amount of oil in the seeds, then the radioactivity can be normalized to the mass of the lipid in the extract measured by gas chromatography of fatty acid methyl esters. If there will be differences in lipid content per seed, then chlorophyll content is suitable for normalization. 1. Allow the 14C labeled Arabidopsis seeds in isopropanol to reach room temperature prior to extraction. 2. A polytron tissue homogenizer is used to grind the A. thaliana seed tissue in the isopropanol used for quenching. First rinse the homogenizer with water, isopropanol, methanol, chloroform and isopropanol in the same order. 10 ml of these solvents are placed in five 15 ml glass tubes for cleaning the homogenizer (see Note 10). 3. Homogenize samples of one-time course at a time using the polytron starting with shortest time points. Rinse the

68

Hari Kiran Kotapati and Philip D. Bates

homogenizer in the same order as in step 2 in between the samples (after step 5) to prevent cross contamination of the 14 C labeled lipids. Replace the wash solvents and give extra washes when you move from a highly labeled long time point to the short-labeled time points of the next time course to prevent contamination with highly labeled lipids. 4. Transfer the homogenized seed/isopropanol sample to a clean 15 ml round bottom threaded glass test tube with a glass Pasteur pipet. Add 3 ml of methanol to the previous 8 ml sample tube, rinse the homogenizer, and combine the methanol portion with the homogenized seeds in isopropanol. Repeat the rinse with 2 ml chloroform and combine with the seed tissue extract. 5. To the 15 ml tube containing the seed extract, add 1.4 ml water, cap the test tube and mix the contents well. Make sure the solution is monophasic. If not, add few drops of methanol until the solution turns monophasic. Rinse the homogenizer as in step 2 and move on to the next sample. 6. Force a phase separation after all samples are ground and have sat for at least 10 min without phase separation: to each sample add 2 ml of chloroform and 2 ml of 0.88% potassium chloride (w/v). Mix well and centrifuge the contents of the test tube to split the phases (2 min at 2000  g in a clinical centrifuge). 7. Remove the lower chloroform layer containing the lipids with a Pasteur pipette and place it in a clean 8 ml screw capped glass test tube. Add 2 ml of chloroform to the remaining aqueous portion in the 15 ml tube, mix well and centrifuge. Remove the lower chloroform layer and combine to the previous chloroform portion in the 8 ml test tube. 8. Wash the chloroform lipid extract to remove nonlipid components and excess unincorporated radioisotope: first add 0.5 ml of 1 M potassium chloride, mix well and centrifuge to cause phase separation. Discard the top aqueous layer appropriately as designated. Second, add 1 ml water to the test tube containing chloroform (lipid extracts), mix well, centrifuge, and remove the top aqueous layer and discard appropriately. 9. Evaporate the solvent under nitrogen stream at 35  C on the nitrogen evaporator equipped with a heated water bath. Vortex the contents in the test tube when the volume is down to approximately 2 ml, to wash the sides of the tube, and then continue to fully evaporate the solvent. Excess unincorporated radioisotope in the lipid extract is one of the main issues for variability in total radioactivity in each extract and nonlinear accumulation of radioactivity over time. Two different approaches are used to remove excess labeled substrate (steps 10 or 11):

14

C-Tracing of Lipid Metabolism

69

10. For the samples from [14C]glycerol labeling experiments, a Folch wash may be required to remove any excess unincorporated [14C]glycerol present in the sample. To perform a Folch wash, dry extract from previous step and add 6 ml of chloroform–methanol (2:1, v/v) followed by 1.5 ml of 0.88% (w/v) potassium chloride. Vortex the contents of the tube, and centrifuge to cause phase separation. Discard the upper aqueous phase and add 1.5 ml of methanol–water (1:1, v/v) to the tube containing organic phase. Vortex and centrifuge the contents of the tube. Remove the lower chloroform phase containing lipids to a new 8 ml tube and remove the solvent under a stream of nitrogen. 11. For the samples from [14C] acetate labeling, resuspend them in 1 ml chloroform and dry again, repeat twice. This helps to evaporate off any excess [14C]acetate. 12. Measure the total chlorophyll content for normalization of different number of seeds extracted per sample (note if normalization by chlorophyll is not used, go to step 13). Add 1 ml of 100% acetone to dried lipid extract and vortex well. Transfer entire sample to an acetone compatible cuvette (see Note 11). Measure chlorophyll content at A644.8 and A661.6. After the measurements, transfer the sample back to its original tube and dry under the stream of nitrogen again. The total μg chlorophyll a and b is calculated as 7.05  A661.6 + 18.09  A644.8 [29]. 13. After the final drying under nitrogen add 0.5 ml toluene to the test tube, vortex the contents and store the sample at 20  C until further analysis. 14. Quantify total radioactivity incorporated into lipids and normalize the radioactivity of the extracts: A small aliquot (0.01 ml) of the total extract should be used for liquid scintillation counting to determine the total disintegrations per minute (DPM). An additional small aliquot (0.05 ml) can be used to quantify total lipid extracted by conversion to fatty acid methyl esters and quantification by gas chromatography (see Chapters 1, 4) [30]. The total radioactivity of extracts from nonequivalent numbers of seeds is normalized by calculating DPM per unit chlorophyll or fatty acid content. 3.4 Analysis of the 14C Labeled Lipid Extract by Thin-Layer Chromatography

To determine the amount of radioactivity in each lipid class, complete separation of the lipid classes of interest must be achieved. Here we discuss two different methods of chromatographic separation of lipid classes from the 14C-labeled seed extracts. The first method involves separation using thin-layer chromatography (TLC). The second method utilizes high-performance liquid chromatography (HPLC) for separating lipid classes and an in-line radio

70

Hari Kiran Kotapati and Philip D. Bates

detector for measuring the 14C label in the lipid classes (Subheading 3.5). The data can be represented as the relative (%) labeling of each lipid at each time point, or as the quantitative accumulation over time as disintegrations per minute (DPM) normalized to chlorophyll or total fatty acid content as indicated above. Since different size aliquots of sample will be used for different analyses, all reported total radioactivity in each lipid should be calculated as for the total sample (e.g., # DPM of TAG (in the whole sample)/μg chlorophyll). Separation of the major A. thaliana seed lipids by one-dimensional TLC requires two separate analyses for neutral and polar lipid classes containing common fatty acids (e.g., 16–20 carbons, 1–3 double bonds). For A. thaliana seeds engineered to produce unusual fatty acids (such as hydroxylated fatty acids), an alternative neutral lipid TLC system will be needed to separate the different hydroxy-fatty-acid-containing TAG species [28]. In this section, we discuss separation of 14C labeled lipid classes in the lipid extract obtained from wild-type A. thaliana. 1. For neutral lipid class separation, Silica gel 60 TLC plates (20  20 cm) are used. 2. The solvent system for the separation of neutral lipids is comprised of hexanes–ethyl ether–acetic acid (70:30:1, v/v/v). The plate is developed to 1 cm from the top (Fig. 1a). 3. For polar lipid class separation, TLC plates (Analtech HL 20  20 cm) are first dipped in 0.15 M ammonium sulfate and dried for 3 days prior to use. TLC plates must be activated on the day of use by heating to 110  C for 3 h. Once cooled to room temperature, use them immediately. 4. The mobile phase for polar lipid class separation consists of acetone–toluene–water (91:30:8, v/v/v). The plate is developed to 1 cm from the top (Fig. 1b). 5. For both TLC systems, load the lipid extracts and unlabeled standards onto the TLC plates in a 1 cm band of up to approximately 0.3 mg total lipid (polar lipid separation) or 0.5 mg (neutral lipid separation). Overloading of the samples will cause the lipids to smear and affect the separation between the lipid classes. 6. After development, let the plate dry for 15 min and then spray it with 0.005% primulin in acetone/water (80:20, v/v) for visualization of the lipid mass from standards or unknowns. Bottle-type sprayer connected to air or a nitrogen line, in the laboratory fume hood, can be used to facilitate the nebulization of primulin stain. Allow the plate to dry for 5 min in the fume hood. Then visualize the lipids under UV light and mark the lipid bands with a pencil.

14

C-Tracing of Lipid Metabolism

a

71

b NL

TAG MGDG PG

FFA PI 1,2-DAG

PE

PL

PC

Fig. 1 TLC-phosphorimaging analysis of 14C labeled neutral lipids and polar lipids. (a) Neutral lipids from 14C labeled A. thaliana seed extract. (b) Polar lipids from 14C labeled A. thaliana leaf extract. DAG diacylglycerol, DGDG digalactosyldiacylglycerol, FFA free fatty acids, MGDG monogalactosyldiacylglycerol, NL Neutral lipids, PC phosphatidylcholine, PE phosphatidylethanolamine, PG phosphatidylglycerol, PI phosphatidylinositol, PL polar lipids, TAG triacylglycerol

7. Radioactivity is determined either by autoradiography methods (such as phosphorimaging or with a radio-TLC scanner) or by removing the lipid mass bands from the TLC plate and liquid scintillation counting the silica. If phosphorimaging is used, add a small radioactive dot to each unlabeled standard for visualization by the phosphorimager. Labeled lipids can be quantified by multiplying the percent of the radioactivity in the lipid band versus the total label in the lane from the phosphorimager by the normalized DPM of the total extract quantified in step 12 of Subheading 3.3. 3.5 Analysis of the 14C Labeled Lipid Extract High-Performance Liquid Chromatography

An alternative to TLC for lipid class separation is normal phase HPLC. Reversed phase HPLC analysis is not ideal because the molecular species in a lipid class separate and tend to overlap with other lipid classes making it impossible to quantify the radioactivity in a single lipid class. In normal phase HPLC, the stationary phase in the column is polar (usually silica based) and the mobile phase is

Hari Kiran Kotapati and Philip D. Bates

relatively nonpolar. The separated lipids can either be collected and quantified by static liquid scintillation counting, or an in-line flow liquid scintillation counter can be utilized. Various normal phase HPLC methods can be found in the literature, see review [31] and Chapter 9. We have reported multiple normal phase HPLC methods with in-line flow liquid scintillation counting to separate neutral lipid classes (including hydroxy fatty acid containing lipids) [32], polar lipid classes [33] and both neutral and polar lipid classes in a single run [33]. The combined HPLC method for the neutral and polar lipid classes reported in [33] separates neutral lipids (TAG, diacylglycerol, free fatty acids, monoacylglycerol including hydroxy-fatty acid containing neutral lipids) and polar lipids (MGDG, DGDG, SQDG, PG, PE, PI, PS, PC, and LPC). Below is the outline for the HPLC analysis of the 14C labeled lipid extract obtained from metabolic labeling of the developing seed tissue with 14C acetate for 30 min in wild type A. thaliana seeds (Fig. 2). 1. Dissolve total lipid extract in toluene. 2. Separate the lipid extract on a YMC-Pack PVA-Sil column (250  4.6 mm, 5 μm). 3. The mobile phase is comprised of four different solvents; A, isopropanol; B, hexanes; C, methanol; and D, aqueous 25 mM formic acid, 25 mM triethylamine at pH 4.2. 4. The HPLC method used for this experiment is method 2 described in [33]. This method involves fifteen steps with linear gradient and the total run time for the method is 58 min 1500

1- TAG

1

2- 1,2-DAG 3- PC

1000 C CPM

3 2

14

72

500

0 0

5

10 15 20 25 30 35 40 45 50 55 Retention time (min)

Fig. 2 HPLC-Radio chromatogram of A. thaliana wild-type seed extract continuously labeled for 30 min with [14C]acetate. CPM counts per minute, DAG diacylglycerol, PC phosphatidylcholine, TAG triacylglycerol

14

C-Tracing of Lipid Metabolism

73

including 8 min for the column equilibration. The mobile phase flows at 1 ml/min, column temperature at 35  C, and sample compartment maintained at 20  C. The solvent gradient for the method is linear starting at 0.5%A/99.5%B to 0.75% A/99.25%B at 1 min; 1%A/99%B at 4 min; 2.5%A/97%B/ 0.5%C at 9 min; 2.5%A/97%B/0.5%C at 12 min; 63.5%A/35% B/1%C/0.5%D at 16 min; 77.75%A/20%B/1.5%C/0.75%D at 19 min; 82%A/15%B/2%C/1%D at 24 min; 87%A/7%B/ 4%C/2%D at 29 min; 63%A/25%C/12%D at 34 min; 55%A/ 30%C/15%D at 36 min; 65%A/35%C at 37 min; 65%A/35%C at 42 min; 100%A at 44 min; 100%A at 48 min and 0.5%A/ 99.5%B at 50 min. The column is equilibrated with 0.5%A/ 99.5%B for 8 min, prior to next injection. 5. An in-line radio detector equipped with a liquid flow cell is used to measure the 14C counts per min in individual lipid classes. The mobile phase flow of the HPLC pump was at 1 ml/min while the scintillation flow was set to 2 ml/min. 6. Regardless of the method of lipid separation, the data can be represented as the relative (%) labeling of each lipid at each time point, or as the quantitative accumulation over time as DPM normalized to chlorophyll or total lipid as indicated above. However, in most cases a quantitative description of the total radioactivity in each lipid at each time point is required for determining the rates of synthesis and/or turnover of each pool essential for tracing the metabolic pathway. 3.6 Elution of Lipids from Silica Gel

The lipid classes TAG, DAG, and PC are separated either by TLC or HPLC as detailed earlier in the Subheadings 3.4 and 3.5 and can be employed for regiochemical analysis (Subheadings 3.7 and 3.8). If fractions are collected by HPLC, the lipids should be dried under N2, resuspended in the solvents indicated below and used for digestion with lipase. If separated by TLC, the lipids should be eluted from the silica. The following protocol works for smaller amounts of silica gel (e.g., lipid bands from 1 to 2 cm lanes). For larger amounts, use proportionately larger volumes of solvent for extraction. 1. After the TLC plates are developed in the appropriate solvents and allowed to dry, use primulin stain to visualize the lipid bands as described in Subheading 3.4. 2. Using a razor blade, scrape the lipid band off the TLC plate and transfer carefully to an 8 ml screw capped tube. Weighing papers can be used to facilitate the transfer of the silica gel with lipids. 3. Add 2.5 ml of freshly prepared chloroform–methanol–water (5:5:1, v/v/v) to the test tube containing silica gel/lipid. Mix the contents of the tube well and centrifuge to pellet the silica.

74

Hari Kiran Kotapati and Philip D. Bates

4. Remove the supernatant liquid (only one phase) to a clean 8 ml screw capped tube. Extract the silica once more and combine the supernatant with the previous. 5. Add 2.25 ml of chloroform and 1.4 ml of 0.88% potassium chloride to the tube containing supernatant with lipids from the previous step. 6. Mix the contents of the test tube well and centrifuge to cause the phase separation. 7. Remove the upper aqueous layer in the test tube and discard it in the designated waste container. 8. Dry the chloroform layer containing lipid under a gentle stream of nitrogen. Redissolve the lipid in a suitable solvent appropriate for the next analysis of the experiment (e.g., the solvents listed in the first step of regiochemical analysis). 3.7 Regiochemical Analysis of 14C Labeled TAG and DAG

In this section, the enzymatic digestion protocols for determining the amount of labeled fatty acids at the sn-1/3 and sn-2 positions of TAG, or the sn-1 and sn-2 position of DAG and PC are discussed. The experimental procedure for both TAG and DAG is similar and utilizes the enzyme Rhizomucor miehei lipase [34]. PC hydrolysis is done with Phospholipase A2 (PLA2) from Honeybee Venom. The following digestion protocol works well for lipids up to 1 mg. However, the time of digestion and amount of enzyme must be optimized for the amount of the respective lipid class and batch of enzyme prior to analysis of the samples. The goal of the experiment is 50–60% digestion within the shortest amount of time possible to limit acyl migration on the glycerol backbone. 1. Dissolve the TAG or DAG isolated from the lipid mixture in 1 ml of freshly obtained diethyl ether, in an 8 ml screw cap tube. 2. Add 0.8 ml of borate buffer, pH 7.8, containing 50 mM boric acid and 5 mM calcium chloride. 3. Using a 1 ml Pipetman, add 0.2 ml of Rhizomucor miehei lipase suspension to the test tube containing lipid in diethyl ether (see Note 12). 4. Mix the contents of the test tube vigorously at room temperature using a vortex mixer or a bench top orbital shaker equipped with a test tube rack (see Note 13). 5. Add 2 ml of freshly prepared chloroform–methanol (1:1, v/v) to the test tube to stop the reaction. Mix the contents of the test tube well using a vortex mixer and then centrifuge to facilitate phase separation. 6. Using a glass Pasteur pipette, remove the lower chloroform phase and transfer to a clean 8 ml screw capped test tube. Add 1 ml of chloroform to the test tube containing the aqueous

14

C-Tracing of Lipid Metabolism

75

phase from the lipase reaction. Mix well, centrifuge, and remove the lower chloroform portion and combine it with the previous chloroform phase. 7. Remove chloroform gently under a stream of nitrogen on an evaporator. When the volume of chloroform is approximately 0.5 ml, vortex the contents of the tube, and then continue with the drying process until all the solvent has been removed. 8. Redissolve the lipids from the enzyme reactions in 0.1 ml of chloroform and load it onto TLC plates (Silica gel 60 TLC plates (20  20 cm)). The developing solvent in the TLC system is hexanes–diethyl ether–acetic acid (35:70:1.5, v/v/ v). The HPLC system mentioned above, in Subheading 3.5, will also work for this analysis. 9. After the development of the TLC plate, the band visualization and further analysis is the same as explained in Subheading 3.4. 3.8 Regiochemical Analysis of 14C Labeled PC

Phosphatidylcholine (PC) is isolated from the lipid mixture either by TLC or HPLC as described in the Subheadings 3.4 and 3.5. This protocol works for lipid amounts up to 0.5 mg. 1. Add 1 ml of fresh diethyl ether to the dried PC sample, in an 8 ml screw capped test tube, isolated from the lipid mixture. 2. Add 50 μl of 50 mM Tris–HCl pH 8.7, 5 mM CaCl2 to the lipid in diethyl ether. 3. Dilute the Honeybee Venom PLA2 stock solution (1 unit/μl) to 0.005 unit/μl with 50 mM Tris–HCl pH 8.7, 5 mM CaCl2. Add 50 μl of the diluted PLA2 enzyme to the lipid in diethyl ether. 4. Mix the contents in the test tube vigorously for 4 min, as detailed in the Subheading 3.7, step 4. 5. Evaporate off the upper diethyl ether under a gentle stream of nitrogen. Add 3.8 ml of freshly prepared chloroform–methanol (2:1, v/v) followed by 1 ml of 0.15 M acetic acid. Mix the contents well and centrifuge the test tubes to obtain phase separation. 6. Remove the lower chloroform layer and transfer to a clean 8 ml screw-capped test tube. Add 2.5 ml of chloroform to the aqueous portion, mix well, centrifuge and combine the chloroform portion with the previous chloroform extract. Repeat the extraction with chloroform once more and combine all the organic portions. 7. Remove chloroform gently under a stream of nitrogen. When the volume in the test tube is approximately 2 ml, vortex the test tube and continue with the drying process. Repeat the vortexing process when the volume in the test tube is down

76

Hari Kiran Kotapati and Philip D. Bates

to about 0.5 ml, and then fully evaporate the residual chloroform. 8. Add 100 μl of chloroform–methanol (9:1, v/v) to the test tube containing dried lipid from the hydrolysis reaction and load the sample on to the TLC plates. 9. The solvent system for the TLC analysis on Silica gel 60 TLC plates (20  20 cm) is chloroform–methanol–acetic acid–water (50:30:8:4, v/v/v/v). After the development of the plate is completed, follow the protocol for mass visualization and subsequent 14C label measurement as in Subheading 3.4.

4

Notes 1. 1 Incubation medium is prepared on the day prior to the labeling experiment, and autoclaved or sterile filtered. Prepare 50–100 ml of the medium. For [14C]glycerol experiments, a 2 incubation medium (double the concentration of all components) is used to equilibrate the substrate into the medium to a final 1 concentration. Therefore, start with producing a 2 medium and dilute to 1 as needed. 2. The 14C labeling medium is best if prepared on the day of experiment. Make enough 14C labeling medium for all samples and replicates such that variations in concentration between batches of labeling medium will not skew results. Extra labeling medium can be stored at 20  C. Place the tube of the labeling medium into the incubator (alongside the collected seeds) for both to equilibrate to the desired temperature prior to starting the labeling. 3. 50–60 plants of each genotype to be analyzed are typically needed to obtain enough developing seeds for a five-point time course in triplicate. A. thaliana seed oil biosynthesis is highly dependent on growth conditions [17]. Therefore, all plants to be compared should be grown in the same growth chamber at the same time. Also, because light intensity can vary by more than 50 μmol photons m2 s1 across a growth chamber, it is important to randomize the different genotypes across the growth area. 4. Developing seed harvesting can take a significant time (1–2 min per silique) and it is best to use the seeds shortly after removing them from the silique. Therefore, it can be very beneficial to have 1–3 people harvesting siliques for each plant line such that all developing seeds can be collected in about 2 h. 5. Developing A. thaliana seeds in medium can be easily pipetted using a 1 ml Pipetman with the tip cut off such that the opening is 2–3 mm wide. The seeds will sink to the bottom

14

C-Tracing of Lipid Metabolism

77

of the medium, and thus gentle pipetting up and down will mix them such that aliquots of similar amounts of seeds can be pipetted. 6. Time of temperature equilibration of seeds may be longer if an open air or air temperature-controlled shaker is used instead of a water bath incubator. In that case moving the seed samples from ice where they were collected to a room temperature water bath will speed up the temperature equilibration. 7. Differences in residual volumes of medium used for seed collection can dilute the labeling medium and cause variations in label uptake between the replicates. The easiest way to remove all the medium is to hold a Pasteur pipette against the bottom of the tube while sucking up the medium so that seeds do not enter the pipet. 8. To perform the chase part of a pulse–chase experiment, remove the 14C-labeled medium from the seed tissue at the end of the pulse period (e.g., 60 min). Wash this seed tissue five times with the 1x incubation medium containing 12C substrate at the same concentration as the 14C labeled medium. Then add the 1x incubation medium containing the 12C substrate and allow the seed tissue to incubate until the end of the chase period. Aliquot the seed tissue at the desired time intervals during the chase period as in steps 3 and 4 for the continuous pulse labeling experiment. Typically, the first time point is at the end of the pulse just before the washes. Some 14C lipids will be continued to be produced from substrate taken up during the washes. Therefore, the second time point should be relatively shortly after the washes to determine the total uptake (e.g., 60 min). Additional time points can be chosen from multiple hours to days. 9. To take aliquots use a 1 ml Pipetman with a filter tip cut to make a 2–3 mm opening. The filter tip limits contamination of the Pipetman when some of the medium is aspirated during pipetting, especially when using [14C]acetate which is slightly volatile. 10. If a polytron is not available, ground glass conical tissue homogenizers with ground glass pestles work well. Make sure to wash the homogenizer between the samples as for the polytron. 11. The whole sample must be used, as the chlorophyll content of seeds is too low to measure from a small aliquot. Many disposable plastic cuvettes are not compatible with acetone, check organic solvent compatibility first (disposable BrandTech® UV cuvettes are acetone compatible). Glass or quartz cuvettes will also work. Other chlorophyll measurement protocols that call for 80% (v/v) acetone or 100% methanol are not suitable for

78

Hari Kiran Kotapati and Philip D. Bates

A. thaliana seed extracts because the high amount of TAG in the sample is not fully soluble in these solvents and will produce a cloudy mixture that will alter the absorbance. 12. Shake the enzyme container well prior to pipetting the enzyme out. If the buffer or enzyme has not been used for a long time, it is always a good idea to do a test run with the lipid of interest for optimizing the reaction parameters such as digestion time and temperature. 13. The shaker is set up in a laboratory fume hood and the test tube rack must be attached to the shaker in an inclined position to ensure proper mixing of the contents in the test tube. For larger samples of TAG (e.g., 0.5–1.0 mg) the ideal reaction time is usually between 30 and 60 min. For DAG and smaller amounts of TAG (e.g., 100  C. Determine the dry weight of your leaf sample using an accurate balance. 3.2 Lipid Fractionation Via SPE

Lipid fractionation using a silica column (100 mg bed volume) is carried out as previously described with modifications [7] (http:// cyberlipid.gerli.com/).

Direct Infusion Mass Spectrometry

107

1. The column is equilibrated with three times 1 ml n-hexane. 2. Lipids isolated from plants are dissolved in 500 μl of n-hexane and loaded onto the equilibrated silica column (see Note 8). 3. The flow through and the elution with 3 ml of n-hexane (containing hydrocarbons and squalene) is collected in a test tube without screw cap (Fraction A). 4. Sterol esters, fatty acid phytyl esters, waxes and fatty acid methyl esters are eluted with 3 ml of n-hexane/diethyl ether (99:1, v/v) (Fraction B) (see Note 9). 5. TAGs and tocopherols are eluted with 3 ml of n-hexane– diethyl ether (95:5, v/v) (Fraction C). Lipid fraction C containing TAGs is dried using nitrogen flow. 6. Free fatty acids and fatty alcohols (e.g., phytol) are eluted with 3 ml of n-hexane/diethyl ether (92:8, v/v) (Fraction D). 7. Sterols and DAGs are eluted with 3 ml of n-hexane–diethyl ether (85:15, v/v) (Fraction E). MAGs and some DAGs are eluted with 3 ml of diethyl ether (Fraction F). Lipid fractions E and F containing DAGs are combined, and dried under a nitrogen flow. 3.3 Direct Infusion MS/MS of Plant Lipids

1. The procedure for direct infusion Q-TOF MS/MS analysis of galactolipids, phospholipids and sulfolipid has been adapted from Welti et al. [4]. Phospholipids and galactolipids are the most abundant lipids in leaves, and therefore only a small aliquot of a crude lipid extract is sufficient for analysis. A crude lipid extract dissolved in 1 ml of chloroform/methanol (2:1, v/v). Ten μl of the leaf extract are mixed with 10 μl of internal standards (Table 1), then 80 μl of direct infusion MS/MS solvent are added and transferred to 1.5 ml autosampler glass vials with glass inserts (see Subheading 2.3). During the first injection, the galactolipids and the phospholipids PC, PE, PG are measured. After acidification of the samples with 8 μl of 100% acetic acid, the acidic lipids PA, PS and SQDG are measured after a second injection (Fig. 2) (see Note 10). 2. Neutral lipids (TAGs and DAGs) are minor lipids in leaves. Therefore, they are purified via SPE before MS/MS analysis. Internal standards for neutral lipids are added at the beginning of the lipid isolation procedure. The lipid fractions eluted from SPE (Subheading 3.2) are dried, dissolved in 100 μl of direct infusion MS/MS solvent (Subheading 2.3) and transferred to 1.5 ml autosampler glass vials with glass inserts (Fig. 3). 3. MS/MS analysis is carried out on an Agilent Series 6530 Accurate Mass Q-TOF LC-MS instrument equipped with a flow infusion chip/ChipCube interface for nanospray ionization. The sample is directly delivered to the ion source

108

Katharina Gutbrod et al.

a

70 Precursor Ion Phosphocholine [F+H]+, m/z=184.0739

60 50 40 30

Phosphatidylcholine di-18:3-PC (36:6) [M+H]+, m/z=778.5382

20 10 100

200

300

400 500 m/z

18:3 Fatty Acid

600

700

800

Precursor Ion Phosphocholine O m/z=184.0739

O O

O O

P O O–

+ N

H

O

18:3 Fatty Acid

b

18 16 14 12 10 8 6 4 2 0

Fragment Ion 18:3-16:3-DAG [F+NH4]+, m/z=585.4752 18:3-16:3-MGDG [M+NH4]+, m/z=778.5382

Fragment Ion-H2O [F+NH4-H2O]+, m/z=567.4646

200

400

600

800

m/z

18:3 Fatty Acid

Neutral Loss of Galactose (179.0556) OH OH

O HO O

O O

16:3 Fatty Acid

O

OH

H

O

Fig. 2 MS/MS spectra of phosphatidylcholine (PC) and monogalactosyldiacylglycerol (MGDG). (a) Fragmentation of the molecular ion [M + H]+ of PC (asterisk) results in the generation of the fragment ion [F + H]+ derived from the phosphocholine head group. (b) Fragmentation of the molecular ion [M + NH4]+ of MGDG (asterisk) results in the generation of two fragments ions, [F + NH4]+ caused by the neutral loss of galactose (179.0556), and [F + NH4-H2O]+ corresponding to the loss of galactose plus one water molecule (197.0661)

Direct Infusion Mass Spectrometry

a

109

12 10

Fragment Ion 16:0-MAG [F+H]+, m/z=313.2743

8 6 4

Neutral Loss of 18:2-NH3 Fragment Ion 18:2-MAG [F+H]+, m/z=337.2743

16:0-18:2DAG [M+NH4]+, m/z=610.5411

Neutral Loss of 16:0-NH3

2 0 100

200

300

400 500 600 700 m/z Neutral Loss of Fatty Acid-NH3 (273.2667) 16:0 Fatty Acid

O O

OH H

O O

18:2 Fatty Acid

Neutral Loss of Fatty Acid-NH3 (297.2667)

b

18 16 14 12 10 8 6 4 2 0

Fragment Ion di-18:3-DAG [F+H]+, m/z=595.4727

tri-18:3-TAG [M+NH4]+, m/z=890.7238

Neutral Loss of 18:3-NH3

200

400

m/z

600

800

1000

Neutral Loss of Fatty Acid-NH3 (295.2511) 18:3 Fatty Acid

O

O

O

O O

18:3 Fatty Acid

H

O

Fig. 3 MS/MS spectra of diacylglycerol (DAG) and triacylglycerol (TAG). (a) Fragmentation of the molecular ion [M + NH4]+ of DAG leads to the generation of a fragment ion [F + H]+ of monoacylglycerol (MAG) ion after the loss of a fatty acid-NH3 adduct. (b) Fragmentation of the molecular ion [M + NH4]+ of TAG results in the generation of a fragment ion [F + H]+ of DAG fragment ion, accompanied by the neutral loss of a fatty acid-NH3 adduct

110

Katharina Gutbrod et al.

(operating in positive mode) with an Agilent nano pump at a flow rate of 1 μl/min. The mobile phase is the direct infusion MS/MS solvent (see Note 11). 4. The collision energy (CE) has to be optimized for each lipid class during the establishment of the method using internal standards. 5. All lipids are measured using the following settings of the Q-TOF MS/MS instrument: drying gas, 8 l/min of nitrogen; fragmentor voltage, 200 V; gas temperature, 300  C, HPLC chip capillary voltage Vcap, 1700 V; scan rate, 1 spectrum/s. Collision energies and adducts for each lipid class are indicated in Tables 3 and 4. Table 3 Molecular ions, fragment ions, and parameters for the analysis of phospholipids and galactolipids from plant leaves by Q-TOF MS/MS.

Lipid class

Molecular species (example)

Molecular Adduct Ion

Molecular adduct ion (m/z)

CE (V)

MGDG

34:6-MGDG

[M + NH4]+

764.5307

DGDG

36:6-DGDG

[M + NH4]+

SQDG

34:3-SQDG

PA

Detection method

Mass difference of neutral loss scana

12

Neutral loss

179.0556 [17], 197.0661

954.6148

17

Neutral loss

341.1084 [18], 359.1189

[M + NH4]+

834.5395

19

Neutral loss

261.0518 [19, 20], 279.0624

34:3-PA

[M + NH4]+

688.4912

20

Neutral loss

115.0034 [4]

PS

34:3-PS

[M + H]+

758.4967

22

Neutral loss

185.0089 [3]

PI

34:2-PI

[M + NH4]+

852.5597

20

Neutral loss

277.0563 [21]

PE

34:3-PE

[M + H]+

714.5069

20

Neutral loss

141.0191 [3]

PG

34:2-PG

[M + NH4]+

764.5436

20

Neutral loss

189.0402 [21]

Pc

36:6-PC

[M + H]+

778.5382

35

Precursor ion

184.0739b [3]

For each lipid class, one characteristic molecular species abundant in plant leaves, its adduct [M + H]+ or [M + NH4]+ and the corresponding mass of the molecular adduct ion (m/z) are presented a Fragmentation of most phospholipids, galactolipids and sulfolipid is characterized by the neutral loss of their head group, which can be detected using a “neutral loss scan” for the mass indicated. Note that two neutral losses are detected for glycolipids, that is, the loss of the sugar head group and of the sugar plus one additional water molecule b PC forms a characteristic head group ion of m/z 184.0739, which can be detected via precursor ion scan

Direct Infusion Mass Spectrometry

111

Table 4 Molecular ions, fragment ions and measuring parameters for abundant TAGs and DAGs in plant leaves Plant Molecular species Molecular lipid (example) adduct ion

Molecular adduct Detection Mass difference of ion (m/z) CE (V) method neutral loss scana

54:9-TAG

[M + NH4]+ 890.7238

20

Neutral loss

295.2511

DAG 34:3-DAG

[M + NH4]+ 608.5254

20

Neutral loss

273.2667, 295.2511

TAG

For each lipid class, one characteristic molecular species abundant in plant leaves, its adduct [M + NH4]+ and the corresponding mass of the molecular adduct ion (m/z) are presented a Fragmentation of TAGs and DAGs is characterized by the neutral loss of their fatty acid-NH3 adduct, which can be detected via neutral loss scan for the mass indicated. Note that two neutral losses are detected for 34:3-DAG corresponding to the losses of the 18:3-NH3 and 16:0-NH3 fatty acyl groups

3.4 Data Analysis of Direct Infusion MS/MS Experiments

1. The Q-TOF instrument records MS/MS spectra (mass spectra after fragmentation) and MS only spectra (mass spectra without fragmentation). 2. The Agilent MassHunter Qualitative Analysis software is used to simulate a “neutral loss scan” or a “precursor ion scan” for characteristic fragments (Tables 3 and 4) in the recorded MS/ MS spectra (see Notes 12–14). 3. The intensities of the characteristic fragment peaks are exported to Microsoft Excel. For MGDG, DGDG and SQDG, intensities of the DAG fragment peak (after the neutral loss of the head group, galactose, 179.0556; galactosylgalactose, 341.1084; or sulfoquinovose, 261.0518, respectively) and the dehydrated DAG fragment peak (after the neutral loss of the head group plus a molecule of water, 197.0662, 359.1190, and 243.0412, respectively) are summed up (Figs. 2 and 3). 4. Trend correction is performed using internal standards. Internal standards of a glycerolipid carrying acyl groups with different chain lengths and/or degree of unsaturation are used to calculate the dependency of the signal intensity on total acyl chain lengths and number of double bonds (see Note 15). 5. Corrections for isotopic overlap for phospholipids, galactolipids, and sulfolipid are performed by calculating the 13C isotope distribution pattern for a given lipid molecular species to exclude the contribution of isotopic overlap from the measured peak intensities. This is particularly important for molecular species of one lipid class differing only by one double bond.

112

4

Katharina Gutbrod et al.

Notes 1. Organic solvents for lipid isolation, fractionation and MS/MS analysis should be “LC-MS grade” or “HPLC grade.” Adequate purity has to be tested prior to use by injecting a solvent control into the mass spectrometer. 2. Internal standards are molecular species of a glycerolipid which are absent from plants. They can be synthetic or isolated from another plant or nonplant organism. Standards for galactolipids and sulfolipid can be generated from plant lipids via hydrogenation [8]. 3. After wounding of the intact plant tissue by cutting, proceed as quickly as possible by placing the leaf in boiling water to avoid the activation of phospholipases, in particular phospholipase Dα, which can lead to accumulation of phosphatidic acid (PA) after hydrolysis of PC [5]. 4. Prior to use, glass tubes and screw caps with Teflon septa should be rinsed several times with 100% chloroform to remove contamination of detergents or other compounds derived from previous cleaning (dish washer detergents) or manufacturing processes. 5. Avoid the transfer of large amounts of water sticking to the leaf material, this may interfere with lipid extraction. 6. At this point, the leaf tissue should look whitish, indicating a complete extraction of lipids. If the tissue remains green, continue with lipid extraction by adding again 1 ml of chloroform to the tissue, vortex and harvest the organic phase. In this case, omit the addition of 1 ml of chloroform in step 11. 7. For an accurate quantification it is important to always transfer the complete organic phase containing the total lipid extract. 8. The silica material should not run dry during an SPE separation. Therefore, always keep some solvent on top of the silica material. 9. Other lipids, such as sterol esters, free sterols, fatty acid phytyl esters, and free fatty acids, can be collected from the column in the indicated fractions and measured with Q-TOF MS/MS (direct infusion MS/MS or LC-MS/MS) as previously described [9–12]. 10. Direct infusion MS/MS of acidic lipids (PA, PS, and SQDG) is hampered by the fact that they can interact with steel surfaces during the delivery to the ion source. To reduce this effect, the needle, needle seat, needle seat capillary, and all tubings should be lined with inert material (e.g., polyetheretherketone, PEEK). In addition, acetic acid is added to the sample for

Direct Infusion Mass Spectrometry

113

measurement of acidic lipids to increase their solubility in the direct infusion MS/MS solvent. 11. The direct infusion MS/MS solvent [4] is used to dissolve the lipids and is used as mobile phase to deliver a constant flow of 1 μl/min. This solvent favors the formation of [M + NH4]+ and [M + H]+ adducts. 12. As the Q-TOF MS/MS instrument lacks a second “scanning” mass analyzer, the neutral loss scanning or precursor ion scanning experiments are performed electronically using the Agilent MassHunter Qualitative Analysis Software. Thus, after Q-TOF MS/MS measurements, the software scans through all MS/MS spectra recorded at different settings of the quadrupole, and extracts the data for the characteristic fragments resulting from CID of molecular species. From the MS/MS spectra of the molecular species of PC, “precursor ion scans” for the m/z 184.0739 (head group fragment) are electronically extracted with the MassHunter software. For all other lipids, the MS/MS spectra are electronically scanned for fragment ions with a defined mass difference to the parental ion (neutral loss scanning) (Figs. 2 and 3). 13. This approach yields information on the total number of carbon atoms and degree of unsaturation of the fatty acyl moieties. For example, 36:6-PC contains 36 carbon atoms and 6 double bonds. To determine the acyl composition of the molecular species, phospholipids, galactolipids, and sulfolipid can be analyzed in the negative ion mode, yielding fragment ions characteristic for fatty acyl residues [13]. 14. Most TAGs and DAGs carry more than one type of fatty acid esterified to glycerol (Fig. 3). In this case, “neutral loss scans” for all present fatty acids are performed using the MassHunter software, the abundances of the resulting fragment peaks are added to determine the amount of each molecular species. For example, the molecular ion for 34:3-DAG gives rise to fragment ions for 16:0-MAG (neutral loss of 18:3) and 18:3-MAG (neutral loss of 16:0) after CID. The peak heights for 16:0MAG and 18:3-MAG are added to calculate the amount of 34:3-DAG. This strategy has been validated by comparison of the amount of TAG and DAG in A. thaliana seeds after quantification with (1) direct infusion Q-TOF MS/MS and (2) transmethylation and quantification of fatty acid methyl esters using GC-FID. 15. Trend correction is performed to correct for deviations in ionization efficiency for different molecular species of one lipid class. Differences in chemical structure, for example chain lengths or the degree of unsaturation of the fatty acyl moiety, can affect the ionization efficiency [3, 14, 15]. For

114

Katharina Gutbrod et al.

phospholipids, galactolipids, and sulfolipid, two internal standards each with different chain lengths of the fatty acyl moiety are used. For DAGs, four internal standards with different chain lengths and different degrees of unsaturation are used. A linear regression line is calculated from two internal standards, respectively, and used to calculate the exact amount of target molecules in the sample.

Acknowledgments We would like to thank the DFG for funding of MS equipment (Forschungsgroßgera¨te nach Art. 91b GG). This work was partially funded by the DFG SFB645. References 1. Slabas T (1997) Galactolipid biosynthesis genes and endosymbiosis. Trends Plant Sci 2 (5):161–162. https://doi.org/10.1016/ S1360-1385(97)01029-7 2. Horn PJ, Korte AR, Neogi PB, Love E, Fuchs J, Strupat K, Borisjuk L, Shulaev V, Lee Y-J, Chapman KD (2012) Spatial mapping of lipids at cellular resolution in embryos of cotton. Plant Cell 24(2):622–636. https://doi. org/10.1105/tpc.111.094581 3. Bru¨gger B, Erben G, Sandhoff R, Wieland FT, Lehmann WD (1997) Quantitative analysis of biological membrane lipids at the low picomole level by nano-electrospray ionization tandem mass spectrometry. Proc Natl Acad Sci U S A 94:2339–2344. https://doi.org/10.1073/ pnas.94.6.2339 4. Welti R, Li W, Li M, Sang Y, Biesiada H, Zhou H-E, Rajashekar CB, Williams TD, Wang X (2002) Profiling membrane lipids in plant stress responses. Role of phospholipase Dα in freezing-induced lipid changes in Arabidopsis. J Biol Chem 277:31994–32002. https://doi. org/10.1074/jbc.M205375200 5. Yang SF, Freer S, Benson AA (1967) Transphosphatidylation by phospholipase D. J Biol Chem 242(3):477–484 6. Roughan PG, Slack CR, Holland R (1978) Generation of phospholipid artefacts during extraction of developing soybean seeds with methanolic solvents. Lipids 13(7):497–503. https://doi.org/10.1007/BF02533620 7. vom Dorp K, Dombrink I, Do¨rmann P (2013) Quantification of diacylglycerol by mass spectrometry. Methods Mol Biol 1009:43–54. https://doi.org/10.1007/978-1-62703-4012_5

8. Buseman CM, Tamura P, Sparks AA, Baughman EJ, Maatta S, Zhao J, Roth MR, Esch SW, Shah J, Williams TD, Welti R (2006) Wounding stimulates the accumulation of glycerolipids containing oxophytodienoic acid and dinor-oxophytodienoic acid in Arabidopsis leaves. Plant Physiol 142(1):28–39. https:// doi.org/10.1104/pp.106.082115 9. Wewer V, Dombrink I, vom Dorp K, Do¨rmann P (2011) Quantification of sterol lipids in plants by quadrupole time-of-flight mass spectrometry. J Lipid Res 52:1–16. https://doi. org/10.1194/jlr.D013987 10. Wewer V, Do¨rmann P (2014) Determination of sterol lipids in plant tissues by gas chromatography and Q-TOF mass spectrometry. Methods Mol Biol 1153:115–133. https:// doi.org/10.1007/978-1-4939-0606-2_8 11. Lippold F, vom Dorp K, Abraham M, Ho¨lzl G, Wewer V, Yilmaz JL, Lager I, Montandon C, Besagni C, Kessler F, Stymne S, Do¨rmann P (2012) Fatty acid phytyl ester synthesis in chloroplasts of Arabidopsis. Plant Cell 24 (5):2001–2014. https://doi.org/10.1105/ tpc.112.095588 12. Brands M, Wewer V, Keymer A, Gutjahr C, Do¨rmann P (2018) The Lotus japonicus acylacyl carrier protein thioesterase FatM is required for mycorrhiza formation and lipid accumulation of Rhizophagus irregularis. Plant J 95(2):219–232. https://doi.org/10. 1111/tpj.13943 13. Devaiah SP, Roth MR, Baughman E, Li M, Tamura P, Jeannotte R, Welti R, Wang X (2006) Quantitative profiling of polar glycerolipid species from organs of wild-type Arabidopsis and a PHOSPHOLIPASE Dα1 knockout

Direct Infusion Mass Spectrometry mutant. Phytochemistry 67:1907–1924. https://doi.org/10.1016/j.phytochem.2006. 06.005 14. Han X, Gross RW (2001) Quantitative analysis and molecular species fingerprinting of triacylglyceride molecular species directly from lipid extracts of biological samples by electrospray ionization tandem mass spectrometry. Anal Biochem 295:88–100. https://doi.org/10. 1006/abio.2001.5178 15. Li X, Evans JJ (2005) Examining the collisioninduced decomposition spectra of ammoniated triglycerides as a function of fatty acid chain length and degree of unsaturation. I. the OXO/YOY series. Rapid Commun Mass Spectrom 19(18):2528–2538. https://doi.org/10. 1002/rcm.2087 16. Gasulla F, vom Dorp K, Dombrink I, Za¨hringer U, Gisch N, Do¨rmann P, Bartels D (2013) The role of lipid metabolism in the acquisition of desiccation tolerance in Craterostigma plantagineum: a comparative approach. Plant J 75(5):726–741. https:// doi.org/10.1111/tpj.12241 17. Moreau P, Bessoule JJ, Mongrand S, Testet E, Vincent P, Cassagne C (1998) Lipid trafficking

115

in plant cells. Prog Lipid Res 37:371–391. https://doi.org/10.1111/tra.12187 18. Moreau RA, Doehlert DC, Welti R, Isaac G, ˜ez A (2008) The idenRoth M, Tamura P, Nun tification of mono-, di-, tri-, and tetragalactosyldiacylglycerols and their natural estolides in oat kernels. Lipids 43(6):533–548. https://doi. org/10.1007/s11745-008-3181-6 19. Gage DA, Huang ZH, Benning C (1992) Comparison of sulfoquinovosyl diacylglycerol from spinach and the purple bacterium Rhodobacter sphaeroides by fast atom bombardment tandem mass spectrometry. Lipids 27 (8):632–636. https://doi.org/10.1007/ BF02536123 20. Welti R, Wang X, Williams TD (2003) Electrospray ionization tandem mass spectrometry scan modes for plant chloroplast lipids. Anal Biochem 314(1):149–152. https://doi.org/ 10.1016/s0003-2697(02)00623-1 21. Taguchi R, Houjou T, Nakanishi H, Yamazaki T, Ishida M, Imagawa M, Shimizu T (2005) Focused lipidomics by tandem mass spectrometry. J Chromatogr B 823(1):26–36. https://doi.org/10.1016/j.jchromb.2005. 06.005

Chapter 8 Fatty Acid Composition by Total Acyl Lipid Collision-Induced Dissociation Time-of-Flight (TAL-CID-TOF) Mass Spectrometry Pamela Tamura, Carl Fruehan, David K. Johnson, Paul Hinkes, Todd D. Williams, and Ruth Welti Abstract Total acyl lipid collision-induced dissociation time-of-flight (TAL-CID-TOF) mass spectrometry uses a quadrupole time-of-flight (QTOF) mass spectrometer to rapidly provide a comprehensive fatty acid composition of a biological lipid extract. Samples are infused into a QTOF instrument, operated in negative mode, and the quadrupole is used to transfer all, or a wide mass range of, precursor ions to the collision cell for fragmentation. Time-of-flight-acquired mass spectra provide mass accuracy and resolution sufficient for chemical formula determination of fatty acids in the complex mixture. Considering the limited number of reasonable CHO variants in fatty acids, one can discern acyl anions with the same nominal mass but different chemical formulas. An online application, LipidomeDB Data Calculation Environment, is employed to process the mass spectral output data and match identified fragments to target fragments at a resolution specified by the user. TAL-CID-TOF methodology is a useful discovery or screening tool to identify and compare fatty acid profiles of biological samples. Key words TAL-CID-TOF, LipidomeDB Data Calculation Environment (DCE), QTOF, Collisioninduced dissociation, Fatty acid composition, Fatty acid analysis, Fatty acyl, Lipidomics, Mass spectrometry

1

Introduction Fatty acids are integral building blocks of phospholipids, galactolipids, triacylglycerols, sphingolipids, cholesteryl esters, and other lipids. They also occur as free fatty acids. Normal-chain fatty acids and their oxidized counterparts (oxylipins) play diverse roles in cellular structures, cell biology, development, signaling, stress responses, and defense mechanisms in plants and animals, both as free fatty acids and as components of complex lipids [1–12]. Characteristic alterations in fatty acid composition occur in many human pathophysiologies, such as inflammation, diabetes, coronary heart disease, and colorectal cancer [13–19].

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_8, © Springer Science+Business Media, LLC, part of Springer Nature 2021

117

118

Pamela Tamura et al.

The fatty acid composition of a lipid extract is commonly determined by gas chromatography (GC), with either flame ionization detection (GC-FID) or mass spectrometry (GC-MS) detection, after derivatization to fatty acid methyl esters. Liquid chromatography and various other separation techniques have also been used, with special handling sometimes required for easily degradable oxylipin species [20–25]. Alternatively, we have developed a simple method utilizing direct infusion of a crude lipid extract into a quadrupole time-of-flight (QTOF) mass spectrometer to provide rapid identification of component fatty acyl species [26]. Because it is performed on a high-mass-resolution QTOF instrument, this technique, “total acyl lipid collision-induced dissociation time-of-flight mass spectrometry” (TAL-CID-TOF MS), can identify fatty acyl anion fragments to the level of the chemical formula. Anions with the same nominal mass (m/z), but different chemical formulas, can be differentiated by small m/z differences. Optionally, internal standards can be added to the samples for approximate quantification. In TAL-CID-TOF MS analysis, a lipid extract is directly infused into a QTOF instrument, operated in negative mode with an electrospray ionization source, and either the quadrupole is turned off or it is adjusted to allow ions of a wide m/z range to enter the collision cell. If there is no selection by the quadrupole, fragments from collision-induced dissociation result from both complex lipids and free fatty acids in the extract. If the QTOF method is designed such that primarily ions of m/z > 400 enter the collision cell, the fatty acyl fragments result mainly from complex lipids; free fatty acids are attenuated or eliminated from the analysis. TAL-CID-TOF MS is a precise and reproducible tool for screening biological samples for fatty acid composition. Differences among samples due to experimental treatments, genetic mutations, pathological conditions, and so on can be easily determined with minimal sample processing. Figure 1 shows results from an experiment investigating the effects of mechanical wounding on Arabidopsis thaliana leaves [8]. Forty-five minutes after wounding, the levels of ten fatty acid species have significantly changed from the levels in unwounded leaves, with most oxylipins increasing and most unoxidized fatty acids decreasing. Due to the mass accuracy of this technique, clear distinctions among anion fragments with the same nominal mass can be made and formulas determined. Figure 2 displays centroided mass spectra from another experiment studying the effect of wounding on the lipid profile of A. thaliana leaves [2]. Left insets in both Panels a and b of Fig. 2 show that a glycerol-galactose anion at m/z 253.0929 (C9H17O8), a fragment from monogalactosyldiacylglycerol (MGDG) molecular species, is distinguishable from the 16:1 fatty acyl anion at m/z 253.2173 (C16H29O2). The right inset in Panel b documents the appearance in the wounded sample of an

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

119

Fig. 1 Levels of fatty acids in Arabidopsis Col-0 leaves, unwounded and 45 min after mechanical wounding, as detected by TAL-CID-TOF methodology. Fatty acid nomenclature is indicated by the number of acyl carbons: number of double bonds or double bond equivalents beyond the acid carbonyl carbon—number of “extra” oxygens in addition to the two in the carboxylic acid group. Levels of lipids in wounded leaves were compared to those in unwounded leaves using a Student’s T-test with correction for false discovery rate (FDR), and levels which were significantly different ( p < 0.05) are labeled with an asterisk (*). Error bars indicate standard deviation, n ¼ 10. This figure is modified from Fig. S4 in Vu et al. [8]

oxylipin fragment at m/z 281.1758 (C16H25O4), which correlates to a 16-carbon ketol fatty acyl anion, 16:3-2O. (The 16:3-2O nomenclature indicates 16 carbons, 3 double bonds or double bond equivalents beyond the acid carbonyl carbon, and 2 “extra” oxygens in addition to the two in the carboxylic acid group.) The peak for 16:3-2O is distinct from the unoxidized 18:1 fatty acyl anion at m/z 281.2486 (C18H33O2). These spectra also clearly show increases in three additional oxylipin species in wounded plants: dinor-oxophytodienoic acid (dnOPDA, 16:4-O, m/z 263.1653, C16H23O3), oxophytodienoic acid (OPDA, 18:4-O, m/z 291.1966, C18H27O3), and an 18-carbon ketol compound (18:3-2O, m/z 309.2071, C18H29O4). Once the TAL-CID-TOF MS data are obtained, they are exported from the mass spectrometer’s acquisition software as peak lists of anion fragments (m/z and intensity), formatted appropriately, and uploaded into an online application, LipidomeDB

120

Pamela Tamura et al.

Fig. 2 Centroided TAL-CID-TOF anion mass spectra from fragmentation of lipids in Arabidopsis Col-0 leaves, unwounded (a) and 5 h after wounding (b). Spectra are expanded 10 from approximately m/z 262–273 and 5 from m/z 290–310. Insets show m/z ranges where there are two peaks at the same nominal mass. Peak intensities are plotted relative to the base (largest) peak in the spectrum. This figure is modified from Fig. 2 in Buseman et al. [2]

Data Calculation Environment (DCE). LipidomeDB DCE is a web-based tool, which was designed to process mass spectral data obtained from direct infusion of lipid-containing biological extracts into a triple quadrupole mass spectrometer [27]. The initial capabilities provided interpretation and quantification of data from multiple precursor and neutral loss scans on multiple samples, and a later update added multiple reaction monitoring (MRM) functionalities [28]. LipidomeDB DCE has now been updated to process CID-TOF data. A list of target fragment ions, e.g., a list of fatty acyl anions, is defined within the application in terms of fragment names and chemical formulas, which indicate m/z. LipidomeDB DCE matches fragments in the uploaded TAL-CID-TOF mass spectral sample data with those in the target list at a mass resolution specified by the user. The output is the intensity of each target fragment for each sample. Users may design their own target lists and further interpret the data based on experimental goals.

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

121

If internal standards are added to samples before analysis, TALCID-TOF MS provides semiquantitative elucidation of fatty acid levels, the limitations of which are detailed in Esch et al. [26]. Confirming earlier studies [29, 30], it was shown that the sn2 position acyl moiety on the glycerol backbone in phospholipids tends to fragment more readily than the sn1 acyl moiety [26]. Fragmentation of MGDG molecular species conversely produces higher intensity peaks from the sn1 fatty acyl group. Additionally, different classes of complex lipids (i.e., those with different head groups) vary in their ability to ionize and/or form fragments in negative mode; changes in collision energy alter the response. Although the fragmentation disparities among different complex lipids render this technique only approximately quantitative, it is useful for making sample-tosample acyl composition comparisons within an experiment, as demonstrated in Fig. 1 [8]. TAL-CID-TOF MS can be used as a “discovery” tool, providing rapid screening of new tissues, genetic variants, or physiological states. Accurate identification of the fatty acyl moieties enables subsequent quantitative probes of the lipidome through triple quadrupole scanning for their intact lipid precursor ions [2, 8]. Utilization of this methodology avoids lengthy purification, derivatization, and potential lipid degradation and effectively provides comprehensive fatty acid composition analysis of lipid extracts from diverse biological samples.

2

Materials 1. Lipid extract from a biological sample (see Note 1). 2. 2 mL glass vials with Teflon-lined lids (Thermo Scientific, B7800-1). 3. Nitrogen gas stream evaporator, in hood. 4. Methanol–chloroform–300 mM ammonium acetate in water (665:300:35, v/v/v). 5. Chloroform–methanol (1:1, v/v), for washing between samples. 6. Methanol–acetic acid (9:1, v/v), for occasional washing between samples. 7. Phosphatidylethanolamine (PE)(46:0) [di23:0], available by transphosphatidylation of phosphatidylcholine (PC)(46:0) [di23:0] from Avanti [31], for use as a mass spectrometer calibrant or internal standard. 8. External syringe pump for infusing sample into the mass spectrometer, capable of delivering 25 μL/min, unless the mass spectrometer has an integrated syringe pump.

122

Pamela Tamura et al.

9. Glass syringe for use in the syringe pump, such as a 1 mL Hamilton gastight syringe (Fisher Scientific, 13-684-94). 10. Quadrupole time-of-flight mass spectrometer (QTOF MS), such as Q-Tof 2, Q-Tof Premier, Xevo QTof, or Synapt (Waters/Micromass, Ltd., Milford, MA), equipped with an electrospray ionization source (see Subheading 3.1). A method for MDS/Sciex QStar Elite (Applied Biosystems, Foster City, CA) is provided in Subheading 3.2. The instrument must be capable of allowing all ions, or ions of a wide mass range, into the collision cell before TOF analysis. 11. Access to the online web-based application LipidomeDB Data Calculation Environment (DCE) at http://lipidome.bcf.ku. edu:8080/Lipidomics/.

3

Methods

3.1 Sample Preparation and Mass Spectrometry (Waters Mass Spectrometers) (See Note 2)

1. To prepare the sample for analysis, transfer a volume equal to approximately 5–50 nmol of lipids from a complex biological lipid extract to a 2 mL glass vial (see Note 3). Add internal standard(s) if desired (see Note 4). Evaporate the solvent with a nitrogen gas stream. Add 1.2 mL of methanol–chloroform– 300 mM ammonium acetate in water (665:300:35, v/v/v) (see Note 5). Prepare an additional “background solvent” vial containing 1.2 mL of the solvent mixture. 2. Calibrate the QTOF MS (see Note 6). 3. Design an acquisition method for the QTOF MS that will either turn off mass selection in the Q1 quadrupole entirely or will allow ions of a specified wide m/z range to enter the collision cell before TOF analysis. Set up the method, with electrospray ionization in negative mode, to continuously infuse sample at 25 μL/min. The TOF analyzer should be tuned for at least 8000 full width at half height (FWHH) resolving power. Masses of fragments should be determined to four decimal places and be accurate to within 2 millimass units (0.002 u). Create methods in continuum (individual scans saved) and multiple channel analyzer (MCA) modes. In MCA mode, delay data collection until adequate time has passed for sample to move into the mass spectrometer and for signal to stabilize (see Note 7). 4. Infuse the “background solvent” into the QTOF MS, acquiring TAL-CID-TOF data with the MCA method developed above. Potential background noise and contamination in the solvent and/or instrument can be visualized in the resulting spectra.

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

123

5. Infuse the lipid extract sample into the QTOF MS. Perform a preliminary test acquisition examining data in individual scans (not MCA mode), and make sure that the sample concentration is appropriate (see Note 8). 6. Continue to infuse the sample, and acquire the TAL-CID-TOF data in MCA mode. 7. In between samples, infuse chloroform–methanol (1:1, v/v) to rinse the tubing and source and remove residue from the ion path. If many samples are being analyzed, periodically infuse methanol–acetic acid (9:1, v/v) to remove residual, “sticky” fatty acyl groups and contaminants. 8. Select background subtraction, smoothing, and centroiding parameters such that each resolved ion is represented as one m/z and intensity pair. Obtain a peak list from the spectrum. Copy the peak mass (m/z) and intensity columns, and paste them into an Excel file (see Note 9). 9. Format the Excel file appropriately for uploading into LipidomeDB DCE. The peak mass and intensity columns should be pasted into Columns A and B, respectively. Two additional rows should be inserted above the data. Cell A1 should be the word “Mass,” and Cell B1 should be “Signal.” Cells A2 and B2 should be the number “0.” Fragment data from all samples in an experiment should be pasted into one Excel workbook, with individual samples on separate sheets. Each sample should have a unique name or number indicated on the sheet tab. There is no limit to the number of sheets (samples) or rows of fragment data for TAL-CID-TOF analysis. 3.2 Mass Spectrometry (MDS/Sciex QStar Elite)

Here we provide alternate instructions for the QStar Elite QTOF MS (MDS/Sciex). 1. Using Analyst QS 2.0 software, design an acquisition method, with electrospray ionization, which allows ions of a specified wide m/z range to enter the collision cell before TOF analysis. Our laboratory uses an installed script developed by Applied Biosystems, “Q1 Resolving Overrides.” When operated in product ion mode with the script activated, the Q1 quadrupole allows only ions larger than the mass filled in at the “Products of ___” prompt, plus slightly smaller ions, to pass through to the collision cell. Experiment to clarify this quadrupole ion selection range for your potential sample components. We have found that “Products of 540” allows only ions of m/ z > 430 to enter the collision cell; smaller ions are filtered out. Fragmentation of lipid extracts in negative mode should thus produce fatty acyl anions from complex lipid precursors; free fatty acids should be eliminated from analysis.

124

Pamela Tamura et al.

2. Develop the acquisition method in negative product ion mode (“Products of 540”), continuously infusing sample at 30 μL/ min with the integrated Harvard syringe pump. The ion spray voltage should be set at 4.5 kV, the source temperature at 150  C, the curtain gas at 25 psi, and the ion source gases at 20 and 30 psi. The declustering potential should be 80 V, the declustering potential 2, 15 V, and the focusing potential, 300 V. The collision gas, nitrogen, should be set at 3 psi, and the collision energy at 45 V. Multiple channel acquisition (MCA) mode (cumulative scans) is best if any of the lipid extract components are in low abundance; continuum mode (individual scans) is adequate if the sample is concentrated. Data should be collected over m/z 100–1200, 60–300 scans over 1–5 min, and accurate masses of fragments should be determined to four decimal places. 3. Calibrate the QTOF MS externally with one or more lipid standards, which when fragmented in TAL-CID-TOF analysis, will produce both low and high mass anions that bracket the fatty acyl range of interest. A good choice for a calibrant is PE (di23:0). Upon TAL-CID-TOF fragmentation, PE(di23:0) produces a glycerol phosphate-related peak at m/z 152.9958 and a 23:0 fatty acyl anion peak at m/z 353.3425. If the calibrant concentration is unknown, determine the concentration of the calibrant by phosphate assay [32], and add approximately 1 nmol to 1.2 mL of methanol–chloroform–300 mM ammonium acetate in water (665:300:35, v/v/v). Infuse the calibrant solution, acquire the data with the TAL-CID-TOF method developed above, and calibrate the QTOF MS on the produced anion fragments. Perform calibration daily or as needed. 4. In between calibrants/samples, infuse chloroform–methanol (1:1, v/v) to rinse the tubing, source, and ion path. If many samples are being analyzed, periodically infuse methanol–acetic acid (9:1, v/v) for more thorough rinsing. 5. Prepare samples as indicated in Subheading 3.1, step 1. 6. Infuse the lipid extract sample into the QStar Elite. Perform a preliminary test acquisition, examining data in individual scans (not MCA mode), to make sure that the concentration is appropriate. The intensity of each peak in a single scan should be 5000 counts to ensure that the detector is not saturated. If the sample is too concentrated, dilute by adding methanol– chloroform–300 mM ammonium acetate in water (665:300:35, v/v/v) until the peaks are on scale. Alternatively, if the peak intensities are weak, prepare a more concentrated sample or adjust the method to accumulate more scans than usual.

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

125

7. Continue to infuse the sample, and acquire the TAL-CID-TOF data. Analyze the “background solvent” sample with the same method, acquisition time, and mode (continuum or MCA) as the lipid extract samples to determine background noise and contamination in the solvent and/or instrument. 8. Utilize Analyst QS 2.0 software to open the data file. If the acquisition was performed in MCA mode, the spectrum will open immediately. If the acquisition was performed in continuum mode, left-click and drag across the total ion current chromatogram to average all scans, right-click, and choose “Show Spectrum.” Smooth with a 0.4–1–0.4 setting (previous point weight-current point weight-next point weight), choose “List Data,” then “Peak List,” then “Save as Text,” and save the table as a “.txt” file. Open the “.txt” file with Excel, and copy the columns “Centroid Mass (Da)” and “Centroid Intensity” to a new Excel file. 9. See Subheading 3.1, step 9, to continue with data set-up and subsequent processing. 3.3 LipidomeDB Data Calculation Environment (DCE)

1. Utilize LipidomeDB DCE (http://lipidome.bcf.ku. edu:8080/Lipidomics/) to process the mass spectral data and match identified fragments in the sample extract to designated target fragments (see Note 10). After logging in, choose “Add CID-TOF Experiment.” The database will assign an experiment number. Enter an experiment name and description. Enter a mass unit (u) value in the “Mass Tolerance” box, which will define the m/z window for peak searching. A mass tolerance of 0.004 u is recommended; this setting will locate sample fragment ion peaks within 0.004 u of the theoretical m/z value of each designated target fragment ion. Select a “Mass to find” option from the three choices in the dropdown menu: “Find the Nearest Mass” (nearest to the theoretical m/z value of the designated target fragment ion), “Find the Largest Signal” (largest within the m/z window designated by the mass tolerance setting), and “Find the Sum of all Signals” (sum within the m/z window designated by the mass tolerance setting). For TAL-CID-TOF data sets, “Find the Nearest Mass” is recommended. Press “Continue.” 2. The next step involves entering a CID-TOF Target Fragment list, which will be used to mine the sample spectral data. Choose an existing, preformulated list (see Note 11), modify and save such a list with a new name (see Note 12), or build a new list (see Note 13). It is possible and sometimes desirable, especially with many modifications, to directly add fragments to the database or to add or edit existing CID-TOF Target Fragment lists independent of an actual experimental workflow (see Notes 14 and 15, respectively).

126

Pamela Tamura et al.

3. After the CID-TOF Target Fragment list has been entered and “Continue” selected, upload the spectral data. These data must be in an Excel file, formatted as detailed above in Subheading 3.1, step 9. Press “Continue” to display the results on screen, which will appear as a summary table of the target fragment intensities for all samples. Press “Export to Excel” at the bottom of the table to download the results. 4. If an internal standard was added to the samples, approximate quantification of fatty acyl fragments is possible. Calculate the amount of each target fragment, in relation to the internal standard (IS) (see Note 16). Alternatively, determine the level of each target fragment, as a percentage of the total fatty acyl fragment intensities (see Note 17).

4

Notes 1. Many extraction methods can be used to isolate lipids from biological tissues. A modified Bligh–Dyer method [33, 34] and a rapid one-step extraction from plant tissue [35] are both excellent for phospho- and galactolipids and have been commonly used in our laboratory for Arabidopsis, sorghum, maize, grasses, and other plants. Methods can be modified for different tissues [36]. Once the extract is obtained, it is best to evaporate to dryness and redissolve in 1 mL chloroform. See also Chapters 1 and 2 in this book. 2. Mass spectrometry directions in Subheadings 3 and 4 are written primarily for Waters QTOF instruments (Subheading 3.1). An alternative method for MDS/Sciex QStar Elite is provided in Subheading 3.2. 3. For A. thaliana leaves, a lipid extract prepared from 0.2 mg dry leaf tissue is used. An equivalent amount is necessary for other Arabidopsis tissues, such as roots and pollen. For more fibrous tissues (stems, grasses, thick leaves, etc.), a larger dry weight equivalent is necessary, perhaps double or more. In terms of protein, an extract from tissue containing 35 μg is suggested. The sample aliquot required for analysis must often be determined by trial and error. 4. If approximate quantification of the fatty acids in the sample is desired, an internal standard can be added. Internal standards are synthetic or purified compounds that are added to the sample before analysis. Their concentration should be accurately determined by traditional analytical methods, such as phosphate assay [32] or gas chromatography of fatty acid methyl esters. Internal standards for TAL-CID-TOF analysis should be lipids which ionize well in negative mode and

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

127

produce non-naturally occurring fatty acyl anion fragments upon collision-induced dissociation. Good choices are complex lipids with a single, uncommon fatty acid, such as PC(24:0) [di12:0], from Avanti, or PE(46:0) [di23:0]. After phosphate assay [32] and preparation of a 0.5 mM solution in chloroform, add 2 μL (1 nmol) to the extract aliquot. 5. Alternatively, instead of evaporating, add chloroform to the aliquot up to a total volume of 360 μL. Add 840 μL methanol–300 mM ammonium acetate in water (95:5, v/v). The resultant volume and solvent composition (1.2 mL of methanol–chloroform–300 mM ammonium acetate in water, 665:300:35, v/v/v) will be the same with either method. 6. On Waters instruments, you may calibrate in V mode using sodium formate and its reference file, over m/z 50–1000 (a range in which negatively charged sodium formate cluster ions are formed [37]), making sure to obtain an ion below m/z 100. Keep ion counts of all utilized calibrant ions below 0.1 ions per push (ipp). Sodium formate (0.5 mM) can be prepared by adding 100 μL of a 10% formic acid solution (in HPLCgrade water) to 25 mL of 0.5 mM sodium hydroxide (in isopropanol–water, 9:1, v/v) (ToF G2-S Standard Kit-2, Waters, 700008892). Sonicate for 5 min. 7. Parameters should be optimized for your instrument and sample concentration. On the Waters QTOF instruments, with MassLynx as the operating and data processing software, needle voltage should be set at 2.6 kV, cone voltage at 35 V, and collision cell offset at 28 V. The quad profile should be set up to start at m/z 400; larger anions will enter the collision cell, but smaller ions will be subject to Q1 transmission attenuation. To set up the quad profile, select the RF Settings or Quad Profile tab (depending on the instrument/software version) on the tune page, choose “Manual Fixed” as the quadrupole option, and fill in “400” as the quadrupole MS profile mass. Fragmentation of lipid extracts in negative mode should produce fatty acyl anions from primarily complex lipid precursor ions; detection of free fatty acids should be attenuated. Data can be acquired with TOF MS function for 5 min in continuum mode over m/z 50–1000, 10 s per cycle, or in MCA mode (additive scans). In either continuum or MCA mode, analyze only scans where the total ion current is high and stable; the time when this occurs depends on the length and diameter of tubing from syringe to the mass spectrometer. 8. If the peak intensities are too high (>0.1 ions per push (ipp) on Waters instruments) and the detector is saturated, the exact mass (m/z) determination and the intensity will not be accurate. Dilute the sample by adding methanol–chloroform–

128

Pamela Tamura et al.

300 mM ammonium acetate in water (665:300:35, v/v/v), and test again. Repeat the dilution/test run as necessary until the peak intensities in a single scan are on scale. This detector saturation level is specific to each mass spectrometer. If the peak intensities in the test run are weak, prepare a more concentrated sample or adjust the method to accumulate more scans than usual. 9. Verify that the centroid parameters identify only one “stick” per resolved ion in the acyl range (m/z 100–400). Lock on a mass of a known acyl anion in the spectrum (e.g., 18:3 at m/z 277.2173). Display the spectrum from m/z 100–400. Consider setting a threshold at 0.01–1% of the base (largest) peak to reduce noise. Switch from spectrum view to peak list. Copy two columns, mass and intensity, and paste into an Excel sheet. 10. A tutorial, which details the function and capabilities of the application and provides example data files to illustrate its use, is available at the LipidomeDB DCE website. 11. Select an existing list from the drop-down menu. The list “fatty1 (95fa)” is a list of 95 unoxidized and oxidized fatty acyl anions suitable for analysis of lipids derived from many plants (Table 1). The second option, “fatty2 (39fa)”, includes only the unoxidized fatty acyl anions from the “fatty1” list. The third list, “fatty3 (34 animal fa)”, is optimized for mammalian lipids. After selection, the boxes will populate; the mode is negative. Press “Continue” at the bottom of the page. 12. To modify an existing list, select it from the drop-down menu. New fragments can be entered by typing in a fragment name already stored in the database, which will retrieve the proper fragment formula. The maximum number of fragments in a target list is 100 (designated as #0–99), and they do not need to be listed in any particular order. If a fragment is not in the database, entering both the fragment name and chemical formula is required. Elements that can be interpreted in the formula are C, H, O, N, P, and S. Chemical formulas should be entered as CmHnOqNrPsSt, where the large letters represent the elements, and the small letters indicate the number of atoms of each element. Leaving an element out of the formula indicates 0 atoms of that element, and no number following an element indicates 1 atom of that element. If a fragment contains atoms due to formation of an adduct ion, these atoms must be included in the fragment formula. Fragments derived from any internal standards in the sample must be included in the target fragment list. The “Mass of Detected Ion” box will calculate automatically when the chemical formula is retrieved or entered. After the modified target fragment list is complete, give it a new name in the “Save current interface” box at the

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

129

Table 1 CID-TOF target fragment list of fatty acyl anions, “fatty1 (95 fa)”, from LipidomeDB DCE Fragment name

Fragment formula

Detected ion, m/z

Fragment name

Fragment formula

Detected ion, m/z

6:0

C6H11O2

115.0765

18:1-O

C18H33O3

297.2435

7:1-O

C7H11O3

143.0714

16:2-3O

C16H27O5

299.1864

8:0

C8H15O2

143.1078

18:0-O

C18H35O3

299.2592

7:1-2O

C7H11O4

159.0663

16:1-3O

C16H29O5

301.2020

9:1-O

C9H15O3

171.1027

16:0-3O

C16H31O5

303.2177

10:0

C10H19O2

171.1391

18:5-2O

C18H25O4

305.1758

9:1-2O

C9H15O4

187.0976

20:3

C20H33O2

305.2486

12:3-O

C12H17O3

209.1183

18:4-2O

C18H27O4

307.1915

13:0

C13H25O2

213.1860

20:2

C20H35O2

307.2643

12:2-2O

C12H19O4

227.1289

16:5-4O

C16H21O6

309.1344

14:0

C14H27O2

227.2017

18:3-2O

C18H29O4

309.2071

15:2

C15H25O2

237.1860

20:1

C20H37O2

309.2799

15:1

C15H27O2

239.2017

16:4-4O

C16H23O6

311.1500

15:0

C15H29O2

241.2173

18:2-2O

C18H31O4

311.2228

16:3

C16H25O2

249.1860

16:3-4O

C16H25O6

313.1657

16:2

C16H27O2

251.2017

18:1-2O

C18H33O4

313.2384

16:1

C16H29O2

253.2173

16:2-4O

C16H27O6

315.1813

16:0

C16H31O2

255.2330

18:0-2O

C18H35O4

315.2541

16:5-O

C16H21O3

261.1496

16:1-4O

C16H29O6

317.1970

16:4-O

C16H23O3

263.1653

16:0-4O

C16H31O6

319.2126

17:3

C17H27O2

263.2017

18:5-3O

C18H25O5

321.1707

16:3-O

C16H25O3

265.1809

18:4-3O

C18H27O5

323.1864

16:2-O

C16H27O3

267.1966

19:3-2O

C19H31O4

323.2228

17:1

C17H31O2

267.2330

21:1

C21H39O2

323.2956

16:1-O

C16H29O3

269.2122

18:3-3O

C18H29O5

325.2020

17:0

C17H33O2

269.2486

21:0

C21H41O2

325.3112

16:0-O

C16H31O3

271.2279

18:2-3O

C18H31O5

327.2177

16:5-2O

C16H21O4

277.1445

18:1-3O

C18H33O5

329.2333

18:3

C18H29O2

277.2173

18:0-3O

C18H35O5

331.2490

16:4-2O

C16H23O4

279.1602

22:3

C22H37O2

333.2799

18:2

C18H31O2

279.2330

22:2

C22H39O2

335.2956 (continued)

130

Pamela Tamura et al.

Table 1 (continued) Fragment name

Fragment formula

Detected ion, m/z

Fragment name

Fragment formula

Detected ion, m/z

16:3-2O

C16H25O4

281.1758

18:5-4O

C18H25O6

337.1657

18:1

C18H33O2

281.2486

22:1

C22H41O2

337.3112

16:2-2O

C16H27O4

283.1915

18:4-4O

C18H27O6

339.1813

18:0

C18H35O2

283.2643

22:0

C22H43O2

339.3269

16:1-2O

C16H29O4

285.2071

18:3-4O

C18H29O6

341.1970

16:0-2O

C16H31O4

287.2228

18:2-4O

C18H31O6

343.2126

18:5-O

C18H25O3

289.1809

18:1-4O

C18H33O6

345.2283

18:4-O

C18H27O3

291.1966

18:0-4O

C18H35O6

347.2439

19:3

C19H31O2

291.2330

23:1

C23H43O2

351.3269

16:5-3O

C16H21O5

293.1394

23:0

C23H45O2

353.3425

18:3-O

C18H29O3

293.2122

24:1

C24H45O2

365.3425

19:2

C19H33O2

293.2486

24:0

C24H47O2

367.3582

16:4-3O

C16H23O5

295.1551

25:1

C25H47O2

379.3582

17:3-2O

C17H27O4

295.1915

25:0

C25H49O2

381.3738

18:2-O

C18H31O3

295.2279

26:1

C26H49O2

393.3738

19:1

C19H35O2

295.2643

26:0

C26H51O2

395.3895

16:3-3O

C16H25O5

297.1707

LipidomeDB DCE is an online application which processes mass spectral data from direct infusion of lipid extracts into a tandem mass spectrometer (http://lipidome.bcf.ku.edu:8080/Lipidomics/). The displayed target list is preloaded within the application and is suitable for TAL-CID-TOF analysis of lipids derived from many plants. LipidomeDB DCE searches uploaded sample data for these target fragments and returns lists of identified targets and their intensities for all samples. Fatty acyl anion fragment nomenclature is indicated by the number of acyl carbons: number of double bonds or double bond equivalents beyond the acid carbonyl carbon—number of “extra” oxygens in addition to the two in the carboxylic acid group. “Detected Ion m/z” is the theoretical exact mass of the fatty acyl anion fragment, calculated from the listed fragment formula

top of the page. It will then be available for immediate and future use and is visible only to the user who created it. Press “Continue” at the bottom of the page. 13. A new CID-TOF Target Fragment List can be designed by following the instructions in Note 12, without selecting an existing list to start. Press “Continue” at the bottom of the page. 14. Fragments may be added to the database by selecting “Add or Edit Fragments” on the page following log-in. Already-entered fragment names and formulas are listed in the drop-down menus. To edit an existing fragment, the “Copy Current

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet

131

Fragment” button can be used. To enter a new fragment, type the name and formula into the boxes. Alternatively, two columns of up to 100 rows at a time (fragment names and formulas) can be copied and pasted from Excel into the box at the bottom of the page; the “Fill Boxes” button will then generate the numbered list. When ready to commit the new fragments to the database, click “Add Fragments.” Further instructions can be found on the website and in the tutorial. 15. CID-TOF Target Fragment lists can be added or edited by selecting “Add or Edit CID-TOF Interfaces” on the page following log-in. The instructions in Notes 12 and 13 above can generally be followed to make these changes. Additionally, a box at the bottom of the page allows for copying and pasting two columns of up to 100 rows (fragment names and formulas) from Excel; “Fill Boxes” will then generate the numbered list. After the new or edited CID-TOF Interface is ready to save to the database, give it a name in the “Interface name” box at the top of the page, and press “Save” at the bottom. The new interface (target fragment list) will be available for use only by the user who created it. 16. To determine the amount of a fatty acyl fragment in comparison to the IS, calculate as follows: nmol of target fragment ¼ðintensity of target fragmentÞ  ðnmol IS target fragmentÞ =ðintensity of IS target fragmentÞ: If the IS has two identical target fragments (e.g., two identical fatty acyl chains), nmol IS target fragment in the above formula should be double the IS concentration added to the sample. 17. To determine the level of a fatty acyl fragment intensity as a percentage of the total intensity of all fragments, calculate as follows: % of target f ragment ¼ ðintensity of target f ragmentÞ  100=ðsum of target f ragment intensitiesÞ: Note that sum of target fragment intensities should not include any intensities of internal standards.

Acknowledgments The authors would like to thank Mary Roth for sharing her expertise in sample preparation methods. This work was supported by the USDA National Institute of Food and Agriculture, Hatch/ Multi-State project 1013013, National Science Foundation MCB

132

Pamela Tamura et al.

1413036, and the Chemical Biology of Infectious Disease, Center of Biomedical Research Excellence (COBRE) of National Institute of Health (P20GM113117). Instrument acquisition and maintenance at KLRC was supported by National Science Foundation (EPS 0236913, DBI 0521587, DBI 1228622, DBI 1726527), K-IDeA Networks of Biomedical Research Excellence (INBRE) of National Institute of Health (P20GM103418), and Kansas State University. Contribution no. 20-230-B from the Kansas Agricultural Experiment Station. References 1. Creelman RA, Mulpuri R (2002) The oxylipin pathway in Arabidopsis. Arabidopsis Book 1: e0012 2. Buseman CM, Tamura P, Sparks AA et al (2006) Wounding stimulates the accumulation of glycerolipids containing oxophytodienoic acid and dinor-oxophytodienoic acid in Arabidopsis leaves. Plant Physiol 142:28–39 3. Kourtchenko O, Andersson MX, Hamberg M et al (2007) Oxo-phytodienoic acid-containing galactolipids in Arabidopsis: jasmonate signaling dependence. Plant Physiol 145:1658–1669 4. Chehab EW, Kaspi R, Savchenko T et al (2008) Distinct roles of jasmonates and aldehydes in plant-defense responses. PLoS One 3:e1904 5. Degenkolbe T, Giavalisco P, Zuther E et al (2012) Differential remodeling of the lipidome during cold acclimation in natural accessions of Arabidopsis thaliana. Plant J 72:972–982 6. Lee H, Park WJ (2014) Unsaturated fatty acids, desaturases, and human health. J Med Food 17:189–197 7. Okazaki Y, Saito K (2014) Roles of lipids as signaling molecules and mitigators during stress response in plants. Plant J 79:584–596 8. Vu HS, Shiva S, Roth MR et al (2014) Lipid changes after leaf wounding in Arabidopsis thaliana: expanded lipidomic data form the basis for lipid co-occurrence analysis. Plant J 80:728–743 9. Li N, Xu C, Li-Beisson Y et al (2016) Fatty acid and lipid transport in plant cells. Trends Plant Sci 21:145–158 10. Martin SA, Brash AR, Murphy RC (2016) The discovery and early structural studies of arachidonic acid. J Lipid Res 57:1126–1132 11. Ho¨lzl G, Do¨rmann P (2019) Chloroplast lipids and their biosynthesis. Annu Rev Plant Biol 70:51–81 12. de Carvalho CCCR, Caramujo MJ (2018) The various roles of fatty acids. Molecules 23:2583

13. Matthan NR, Ooi EM, Van Horn L et al (2014) Plasma phospholipid fatty acid biomarkers of dietary fat quality and endogenous metabolism predict coronary heart disease risk: a nested case-control study within the Women’s Health Initiative observational study. J Am Heart Assoc 3:e000764 14. Philippova M, Resink T, Erne P et al (2014) Oxidised phospholipids as biomarkers in human disease. Swiss Med Wkly 144:w14037 15. Ma W, Wu JH, Wang Q et al (2015) Prospective association of fatty acids in the de novo lipogenesis pathway with risk of type 2 diabetes: the Cardiovascular Health Study. Am J Clin Nutr 101:153–163 16. Zheng JS, Sharp SJ, Imamura F et al (2017) Association between plasma phospholipid saturated fatty acids and metabolic markers of lipid, hepatic, inflammation and glycaemic pathways in eight European countries: a crosssectional analysis in the EPIC-InterAct study. BMC Med 15:203 17. Bandu R, Mok HJ, Kim KP (2018) Phospholipids as cancer biomarkers: mass spectrometrybased analysis. Mass Spectrom Rev 37:107–138 18. Jackson KH, Harris WS (2018) Blood fatty acid profiles: new biomarkers for cardiometabolic disease risk. Curr Atheroscler Rep 20:22 19. Pakiet A, Kobiela J, Stepnowski P et al (2019) Changes in lipids composition and metabolism in colorectal cancer: a review. Lipids Health Dis 18:29 20. Weber H, Vick BA, Farmer EE (1997) Dinoroxo-phytodienoic acid: a new hexadecanoid signal in the jasmonate family. Proc Natl Acad Sci U S A 94:10473–10478 21. Vollenweider S, Weber H, Stolz S et al (2000) Fatty acid ketodienes and fatty acid ketotrienes: Michael addition acceptors that accumulate in wounded and diseased Arabidopsis leaves. Plant J 24:467–476

Fatty Acid Composition by TAL-CID-TOF Mass Spectromet 22. Mueller MJ, Me`ne-Saffrane´ L, Grun C et al (2006) Oxylipin analysis methods. Plant J 45:472–489 23. Schulze B, Lauchli R, Sonwa MM et al (2006) Profiling of structurally labile oxylipins in plants by in situ derivatization with pentafluorobenzyl hydroxylamine. Anal Biochem 348:269–283 24. Ibrahim A, Schu¨tz AL, Galano JM et al (2011) The alphabet of galactolipids in Arabidopsis thaliana. Front Plant Sci 2:95 25. Li-Beisson Y, Shorrosh B, Beisson F et al (2013) Acyl-lipid metabolism. Arabidopsis Book 11:e0161 26. Esch SW, Tamura P, Sparks AA et al (2007) Rapid characterization of the fatty acyl composition of complex lipids by collision-induced dissociation time-of-flight mass spectrometry. J Lipid Res 48:235–241 27. Zhou Z, Marepally SR, Nune DS et al (2011) LipidomeDB data calculation environment: online processing of direct-infusion mass spectral data for lipid profiles. Lipids 46:879–884 28. Fruehan C, Johnson D, Welti R (2018) LipidomeDB data calculation environment has been updated to process direct-infusion multiple reaction monitoring data. Lipids 53:1019–1020 29. Murphy RC (1993) Mass spectrometry of lipids. In: Handbook of lipid research, vol 7. Plenum Press, New York, NY, pp 223–226 30. Guella G, Frassanito R, Mancini I (2003) A new solution for an old problem: the regiochemical distribution of the acyl chains in galactolipids can be established by electrospray

133

ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 17:1982–1994 31. Comfurius P, Zwaal RF (1977) The enzymatic synthesis of phosphatidylserine and purification by CM-cellulose column chromatography. Biochim Biophys Acta 488:36–42 32. Ames BN (1966) Assay of inorganic phosphate, total phosphate and phosphatases. In: Neufeld E, Ginsburg V (eds) Methods in enzymology: complex carbohydrates, vol VIII. Academic, New York, NY, pp 115–118 33. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37:911–917 34. Welti R, Li W, Li M et al (2002) Profiling membrane lipids in plant stress responses. Role of phospholipase Dα in freezing-induced lipid changes in Arabidopsis. J Biol Chem 277:31994–32002 35. Shiva S, Enninful R, Roth MR et al (2018) An efficient modified method for plant leaf lipid extraction results in improved recovery of phosphatidic acid. Plant Methods 14:14 36. Shiva S, Vu HS, Roth MR et al (2013) Lipidomic analysis of plant membrane lipids by direct infusion tandem mass spectrometry. In: Munnik T, Heilmann I (eds) Plant signaling protocols. Methods in molecular biology, vol 1009. Humana Press, Totowa, NJ, pp 79–91 37. Hao C, March RE, Croley TR et al (2001) Electrospray ionization tandem mass spectrometric study of salt cluster ions. Part 1—investigations of alkali metal chloride and sodium salt cluster ions. J Mass Spectrom 36:79–96

Chapter 9 Targeted Analysis of the Plant Lipidome by UPLC-NanoESI-MS/MS Cornelia Herrfurth, Yi-Tse Liu, and Ivo Feussner Abstract The plant lipidome is highly complex and changes dynamically under the influence of various biotic and abiotic stresses. Targeted analyses based on mass spectrometry enable the detection and characterization of the plant lipidome. It can be analyzed in plant tissues of different developmental stages and from isolated cellular organelles and membranes. Here, we describe a sensitive method to establish the relative abundance of molecular lipid species belonging to three lipid categories: glycerolipids, sphingolipids, and sterol lipids. The method is based on a monophasic lipid extraction and includes the derivatization of a few rare and low-abundant lipid classes. The molecular lipid species are resolved by lipid class-specific reverse-phase liquid chromatography and detected by nanoelectrospray ionization coupled with tandem mass spectrometry. The triple quadrupole analyzer is used for detection with multiple reaction monitoring (MRM). Mass transition lists are constructed based on the knowledge of organism-specific lipid building blocks. They are initially determined by classical lipid analytical methods and then used for combinative assembly of all possible lipid structures. The targeted analysis enables detailed and comprehensive profiling of the entire lipid content and composition of plants. Key words Glycerolipids, Sphingolipids, Sterol lipids, Liquid chromatography, Nanoelectrospray ionization, Mass spectrometry, Membrane enrichment, Monophasic extraction, Methylation, Acetylation

1

Introduction Lipids can function as energy source, structural components of cellular membranes, and signaling molecules [1]. Thousands of lipid molecular species exist in different organisms in a concentration range spanning at least six orders of magnitude, and these lipids can be categorized based on their molecular structure [2, 3]. Lipidomics is an analytical discipline to study lipid metabolism on a broad scale with mass spectrometric techniques [4, 5]. Due to the development of mass spectrometry (MS) over the last 20 years, the lipidomic technology has greatly advanced with respect to its analytical range, detection sensitivity, and speed of analysis. Lipidomic studies aim to detect the complete set of

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_9, © Springer Science+Business Media, LLC, part of Springer Nature 2021

135

136

Cornelia Herrfurth et al.

lipids (the lipidome) of a given cell type or tissue for understanding their relation and function within the context of cellular metabolism [6]. Two major analytical approaches exist, namely, global and targeted analyses. These can be distinguished by their analytical coverage and lipidomic applications, as well as by the presence or absence of liquid chromatography (LC) prior to the MS analysis. For global or “shotgun lipidomics,” a lipid extract is directly infused into the MS system and analyzed without prior chromatographic separation [7, 8]. The entire lipidome is recorded at a constant sample concentration, and the identification and quantification of the molecular lipid species is performed without limitation on the acquisition time other than sample volume. The complexity of the sample matrix, however, limits the selectivity for isomeric and isobaric lipid species, as well as the detection sensitivity for trace lipid species (e.g., those with signaling functions). This restriction can be overcome by the use of LC separation prior to tandem MS (MS/MS) analyses. The targeted methodology reported herein has been developed to perform a sensitive and highly resolved analysis, screening molecular species of a total of 36 lipid subclasses from a minimal amount of plant material. The workflow includes a monophasic propan-2-ol–hexane–water extraction. It was originally developed for the efficient extraction of amphiphilic sphingolipids [9] and is applied here to extract a broad range of plant lipids with highly diverse chemical properties. Prior to the MS-based lipidomics analysis, the overall lipid building blocks, namely, acyl chains, polar head groups, and backbones, are determined by classical lipid analytical methods (e.g., GC, TLC). This information is then used to calculate an array including all possible combinations of putative plant molecular lipid species. Based on this array, the putative precursor ions and corresponding lipid subclass-specific fragment ions (Table 2) are derived and converted into mass transition lists for the MS/MS detection. The lipid extract is subjected to an ultraperformance LC (UPLC) system coupled with a chip-based nanoelectrospray ionization (nanoESI) source and a triple quadrupole analyzer. The robust and efficient resolving power of the sequential UPLC separation and targeted MS/MS detection enables the analysis of the distinct acyl composition of the molecular species within most lipid subclasses. For triacylglycerols, however, only the averaged composition corresponding to the total number of carbon atoms and double bonds can be resolved due to the wide range of possibilities for its acyl combinations in a single targeted molecule. The specificity of the UPLC-nanoESI-MS/MS method is additionally increased by incorporating chemical derivatization approaches (methylation and acetylation) after lipid extraction. Thus, distinct functional lipid groups, such as trace phospholipid species, can be detected in plant tissues.

Plant Lipidomics Analysis

2

137

Materials

2.1 Samples and Buffers

1. Flash-frozen (i.e., in liquid nitrogen) plant tissue (various developmental stages, e.g., seeds, seedlings, leaves), keep at80
C until grinding and extraction. 2. Cell cultures from plants, algae, and yeast, keep at-80
C until extraction. 3. Microsomal fractions, keep at 80
C under argon until extraction. 4. Microsome Extraction Buffer (MEB) 1: 0.1 M Tris–HCl, pH 7.5, 0.81 M sucrose, 5% (v/v) glycerol, 10 mM ethylenediaminetetraacetic acid (EDTA), pH 8.0, 10 mM ethyleneglycoltetraacetic acid (EGTA), pH 8.0, 5 mM KCl, 1 mM 1,4-dithiothreitol (DTT), 1 mM phenylmethanesulfonyl fluoride (PMSF). 5. Lipid Extraction Buffer (LEB): propan-2-ol–hexane–water (60:26:14, v/v/v).

2.2 Chemicals and Standards

Analytical standards were purchased from Merck KGaA (Darmstadt, Germany), Avanti Polar Lipids, Inc. (Alabama, AL, USA) and Matreya (State College, PA, USA). 1. Trimethylsilyldiazomethane solution for methylation: 2 M in hexane (Merck KGaA, Darmstadt, Germany). 2. Pyridine, acetic anhydride for acetylation. 3. Tetrahydrofuran–methanol–water (4:4:1, v/v/v) for dissolving lipid samples. 4. Methylamine for glycerolipid hydrolysis: 33% (v/v) methylamine in ethanol. 5. N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA), pure for silylation (Merck KGaA, Darmstadt, Germany).

2.3 Solvents and Solutions for LC-MS

All solvents used (methanol, propan-2-ol, tetrahydrofuran) are LC-MS grade quality, unless indicated otherwise. Ultrapure water is always freshly generated by an Arium pro VF TOC ultrapure water system (Sartorius, Go¨ttingen, Germany). 1. Solvent system for UPLC analyses with the ACQUITY HSS T3 column (Waters Corporation, Milford, MA, USA): solvent A (methanol-20 mM ammonium acetate, 3:7, v/v, containing 0.1%, v/v acetic acid), solvent B (tetrahydrofuran–methanol– 20 mM ammonium acetate, 6:3:1, v/v/v, containing 0.1%, v/v acetic acid). 2. Tuning mixture for QTRAP6500: Standards chemical kit with low/high concentration polypropylene glycols (PPGs) (AB Sciex, Framingham, MA, USA).

138

2.4

Cornelia Herrfurth et al.

LC-MS System

1. For chromatographic separation: ACQUITY UPLC® system (Waters Corporation, Milford, MA, USA) equipped with an ACQUITY HSS T3 column (100 mm  1 mm, 1.8 μm; Waters Corporation, Milford, MA, USA). This is a silica-based, reversed-phase C18 column. 2. For nanoESI: chip ion source TriVersa Nanomate® (Advion, Incorporation, Ithaca, NY, USA) equipped with nanoESI chip with 5 μm internal diameter nozzles. 3. For mass-spectrometric detection: AB Sciex QTRAP6500® tandem mass spectrometer (AB Sciex, Framingham, MA, USA).

2.5

Software

1. Data acquisition: Analyst 1.6.2 (AB Sciex, Framingham, MA, USA). 2. nanoESI control: ChipSoft 8.3.1 (Advion, Incorporation, Ithaca, NY, USA). 3. Data analysis: MultiQuant 3.0.2 (AB Sciex, Framingham, MA, USA). 4. Data processing and statistics: Excel 2016 (Microsoft Corporation, Redmond, WA, USA) and RStudio (RStudio, Incorporation, Boston, MA, USA).

2.6

Other Equipment

1. Kimble extraction tubes (Kimax-51, 13  100 mm) with Teflon-lined screw caps (Gerresheimer Glass Inc., Vineland, NY, USA). 2. Chemically resistant tips for organic solvents (Safe Seal Tips Premium, Biozym, Oldendorf, Germany). 3. Test strips for semiquantitative determination of hydrogen peroxide and peroxides in tetrahydrofuran (Quantofix® Peroxide 100, Macherey-Nagel, Du¨ren, Germany). 4. Glass sample vials for sample storage (1.1 ml, inner cone in the solid glass bottom, ND9, VWR International GmbH, Darmstadt). 5. Glass micro vials for analysis (12 mm, 250 μl, Macherey-Nagel GmbH, Du¨ren, Germany) fixed by a spring in HPLC glass vials (1.5 ml, N9, Macherey-Nagel GmbH, Du¨ren, Germany). 6. Nitrogen evaporator (Organomation Associates, Incorporation, Berlin, MA, USA). 7. Mixer Ball Mill MM200 with stainless steel grinding jars or PTFE-jars (Retsch GmbH, Haan, Germany). 8. Freeze dryer. 9. Argon.

Plant Lipidomics Analysis

3

139

Methods

3.1 Harvesting and Homogenization of Plant Material

1. Complete the harvesting of plant material as quickly as possible and always in the same time range to avoid unspecific lipid degradation by lipolytic activities. 2. Shock freeze the harvested material immediately in liquid nitrogen (see Note 1). 3. Grind the plant material by using a porcelain mortar and pestle or homogenize using the Mixer Ball Mill MM200 (see Note 2). 4. Use stainless steel grinding jars (for large sample amounts) or PTFE-jars for Eppendorf tubes (for small sample amounts) with the corresponding size of stainless steel balls for mill homogenization. 5. Time and repetitions of the homogenization cycles depends on the amount and rigidity of the biological material (see Note 3). 6. Ensure that the biological material always stays completely frozen under liquid nitrogen. 7. For freeze drying, incubate the sample material in the freeze dryer overnight until the pressure stays constant, indicating that the water of the sample has been completely removed.

3.2 Enrichment of Microsomal Membrane Fractions

To concentrate minor lipids located in cellular membranes, microsomal-type membranes are enriched from small amounts of plant material. Microsomal membranes are isolated without the need for ultracentrifugation (modified from [10]): 1. Prior to use, the 2 ml Eppendorf tubes are kept on ice. All solvents are kept at room temperature. 2. Weigh 50 mg of homogenized deep frozen material, or 5 mg of homogenized freeze-dried material, into a 2 ml Eppendorf tube (see Note 4). Immediately add 0.2 ml of the extraction buffer (MEB) 1. Ensure that the biological material is completely covered with the MEB 1. 3. Vortex the sample strongly. 4. Centrifuge the samples for 3 min at 600  g and 4
C. 5. Transfer the supernatant into a new 2 ml Eppendorf tube and put it aside on ice. Reextract the sample with 0.1 ml MEB 2 (dilute MEB 1 to 0.35 in water: mix 35 ml of MEB 2 with 65 ml of water). 6. Vortex the sample strongly. 7. Centrifuge the samples for 3 min at 600  g and 4
C. 8. Add the supernatant of this second extraction to the first supernatant. Reextract the sample with 65 μl MEB 3 (dilute

140

Cornelia Herrfurth et al.

MEB 1 to 0.48 in water: mix 48 ml of MEB1 with 52 ml of water). 9. Vortex the sample strongly. 10. Centrifuge the samples for 30 s at 2000  g and 4
C. 11. Transfer the supernatant to the combined supernatants. Add 0.25 ml water. 12. Vortex the pooled supernatant strongly. Transfer three 0.2 ml aliquots into 2 ml Eppendorf tubes. 13. Centrifuge the samples for 2 h at 20,000  g and 4
C to obtain the membrane pellets. 14. Remove the supernatant carefully. Wash the membrane pellet with 0.15 ml water. 15. Centrifuge the samples for 45 min at 20,000  g and 4
C. 16. Remove the supernatant carefully. 17. Cover the membrane pellet with argon and use it either immediately for monophasic lipid extraction or store it at 20
C until extraction. To avoid autoxidation of lipids, immediately cover the samples with argon after each extraction step, particularly at the end of membrane isolation procedure before storage. 3.3 Extraction of Lipids from Plant Material and Cultured Cells

The monophasic extraction method with propan-2-ol, hexane, and water as described by Markham et al. [9] was slightly modified as follows [11]: 1. Prior to use, Kimble glass tubes and the lipid extraction buffer (LEB) (propan-2-ol–hexane–water (60:26:14, v/v/v)) are warmed to 60
C. 2. Weigh 200 mg of homogenized and deep frozen material, or 20 mg of homogenized and freeze-dried material, into Kimble glass tubes (see Note 4). Immediately add 6 ml of the warmed LEB (see Note 5). Ensure that the biological material is completely covered with the extraction buffer. 3. Vortex the sample strongly. 4. Shake for 30 min at 60
C. During this incubation process, vortex and sonicate the sample every 10 min. 5. Centrifuge the samples for 20 min at 800  g and 20
C. 6. Collect the supernatant with a glass Pasteur pipette and transfer it into a new Kimble glass tube. 7. Dry the supernatant under a nitrogen stream. 8. Dissolve the samples in 0.8 ml of tetrahydrofuran–methanol– water (4:4:1, v/v/v). Ensure that no material is stuck to the wall of the glass tube.

Plant Lipidomics Analysis

141

9. Centrifuge (5 min, 800  g, 20
C) and transfer the supernatant into a glass sample vial. 10. Cover the sample with argon and store it at 20
C until UPLC-nanoESI-MS/MS analysis (see Note 6). 11. Transfer aliquots (30–100 μl) of the sample into glass micro vials directly before starting the analysis. To avoid autoxidation, immediately cover the samples with argon after each extraction step, particularly at the end of the extraction procedure before storage. 3.4 Extraction of Lipids from Microsomal Membrane Fractions

The monophasic extraction method with propan-2-ol, hexane, and water as described by Markham et al. [9] was modified for the extraction of microsomal membrane fractions as follows: 1. Prior to use, Kimble glass tubes and the LEB are warmed to 60
C. Also warm propan-2-ol and hexane, which are added separately to the sample after resuspension of the pellets in water. 2. Combine the three microsomal membrane pellets derived from a single sample in 0.14 ml water and transfer the sample into a Kimble glass tube (see Note 4). Immediately add the warmed 0.6 ml propan-2-ol and the 0.26 ml hexane (see Note 5), and 5 ml of warmed LEB. 3. Vortex the sample strongly. 4. Shake for 30 min at 60
C. During this incubation vortex and sonicate the sample every 10 min. 5. Centrifuge the samples for 20 min at 800  g and 20
C. 6. Collect the supernatant with a glass Pasteur pipette and transfer it into a new Kimble glass tube. 7. Dry the supernatant under a nitrogen stream. 8. Dissolve the samples in 0.2 ml of tetrahydrofuran-methanolwater (4:4:1, v/v/v). Ensure that no material is stuck to the wall of the glass tube. 9. Centrifuge (5 min, 800  g, 20
C) and transfer the supernatant into a glass sample vial. 10. Cover the sample with argon and store it at 20
C until UPLC-nanoESI-MS/MS analysis (see Note 6) or before chemical derivatization. 11. Transfer aliquots (30–100 μl) of the sample into glass micro vials directly before the analysis. To avoid autoxidation, immediately cover the samples with argon after each extraction step, particularly at the end of the extraction procedure before storage.

142

Cornelia Herrfurth et al.

3.5 Chemical Derivatization of Lipids

Phosphate groups and hydroxyl groups of some lipid classes are chemically modified by methylation (lysophosphatidic acid, phosphatidic acid, phosphatidylinositol phosphate, and phosphatidylinositol bisphosphate) or acetylation (ceramide phosphate, free sterol, and long-chain base phosphate) to improve their chromatographic separation and mass spectrometric detection. 1. Transfer an aliquot (30–100 μl) of the lipid sample into a glass sample vial (see Note 4). 2. Dry the aliquot under a nitrogen stream. 3. For methylation (modified from [12]): dissolve the dry lipid in 0.4 ml methanol and add 6.5 μl of trimethylsilyldiazomethane solution (2 M in hexane). Vortex the sample strongly. Incubate for 30 min at room temperature and terminate the reaction by neutralizing with 2 μl of 1 N acetic acid. 4. For acetylation (modified from [13]): dissolve the dry lipid in 100 μl pyridine and 50 μl of acetic anhydride. Vortex the sample strongly. Incubate for 30 min at 50
C. 5. Dry the derivatized sample aliquot under a nitrogen stream. 6. Dissolve the sample aliquot in an equal volume of tetrahydrofuran–methanol–water (4:4:1, v/v/v) as before the chemical derivatization. Ensure that no material is stuck to the wall of the glass tube. 7. Centrifuge (5 min, 800  g, 20
C) and transfer the sample into a glass micro vial. 8. Cover the derivatized sample aliquot with argon and store it at 20
C until UPLC-nanoESI-MS/MS analysis (see Note 6). To avoid autoxidation, immediately cover the sample aliquot with argon after each derivatization step, particularly at the end of the derivatization procedure before storage.

3.6 Methylamine Treatment for Enhanced Sphingolipid Analysis

To improve the detection efficiency for sphingolipids, the lipid extract is treated with methylamine. This mild base hydrolyzes glycerophospholipids, but not sphingolipids, and therefore reduces interference specifically for sphingolipid analysis, for example during the ESI process [14]. 1. Transfer 200 μl of the extracted lipid into a glass sample vial. 2. Dry the aliquot under a nitrogen stream. 3. Add 1.4 ml of 33% (v/v) methylamine in ethanol and 0.6 ml of water (modified from [15]). 4. Vortex the sample strongly. Incubate for 1 h at 50
C. 5. Dry the methylamine-treated lipid under a nitrogen stream. 6. Dissolve the methylamine-treated lipid in 50 μl of tetrahydrofuran-methanol-water (4:4:1, v/v/v). Ensure that no material is stuck to the wall of the glass tube.

Plant Lipidomics Analysis

143

7. Centrifuge (5 min, 800  g, 20
C) and transfer the supernatant into a glass micro vial. 8. Cover the methylamine-treated lipid aliquot with argon and store it at –20
C until UPLC-nanoESI-MS/MS analysis (see Note 6). 3.7 Lipid Analysis by UPLC-NanoESI-Mass Spectrometry

1. Set the temperature of the autosampler (sample manager) of the UPLC system to 18
C and the column oven temperature to 40
C. 2. Set the flow rate to 0.1 ml/min or 0.13 ml/min (depending on the gradient; Table 1) and the injection volume to 2 μl. 3. Use methanol as strong and methanol-water (1:9, v/v) as weak wash solutions. 4. Use the gradients of solvents shown in Table 1 as mobile phase for the chromatographic separation, depending on the lipid classes of interest (Table 2) [11]. The retention time windows for the elution of the corresponding lipid subclasses are shown in Fig. 1. Representative chromatograms illustrating the separation by the different gradients have been published in [16].

Table 1 Solvent gradients for UPLC separation (see Note 6) prior to detection by mass spectrometry Gradient

Time (min)

Flow (ml/min)

Solvent A (%)

Solvent B (%)

1

0 2 4 8 8.5 12

0.13 0.13 0.13 0.13 0.13 0.13

10 10 0 0 10 10

90 90 100 100 90 90

2a

0 2 10 12 12.5 16

0.1 0.1 0.1 0.1 0.1 0.1

20 20 0 0 20 20

80 80 100 100 80 80

2b

0 2 10 12 12.5 16

0.1 0.1 0.1 0.1 0.1 0.1

35 35 0 0 35 35

65 65 100 100 65 65

2c

0 2 10 12 12.5 16

0.1 0.1 0.1 0.1 0.1 0.1

60 60 0 0 60 60

40 40 100 100 40 40

2c

SQMG

2b

2c

2c

2c

2c

LPA

LPC

LPE

LPG

1

2b

SQDG

TAG

2c

MGMG

2a

2b

MGDG

DAG

2c

DGMG

Phospholipids CL

Neutral lipids

2b

Negative

Negative

Negative

Negative

Negative

Positive

Positive

Negative

Negative

Negative

Negative

Negative

Negative

Positive

2b Glycolipids Ara Ara-A, Ara-B Ara-C, Ara-D Ara-E, Ara-G

DGDG

Negative

2b

DGCC

Negative

a

+

+



[MH]

[MH]





[M+OAc]

[M+MeH]

[M2H]

2

[M+NH4]+

[M+NH4]

[MH]

[MH]



[M+OAc]

[M+OAc]





+



[RCOO]

[RCOO]





[RCOO]

[RCOO]

[RCOO]



[MRCOO]+

[MRCOO]

[RCOO]

[RCOOsn1] /[RCOOsn2]

[RCOO]



[RCOOsn1] /[RCOOsn2]







[RCOO]

[M+OAc]

[M(monoGalH)+NH4] [M(diGal-H)+NH4]+d [M(monoGalacylatedH)+NH4]+e [RCOOsn1]/[RCOOsn2]



CXP [V] Reference

10

30

6

[24]

10

10

40

38

6

10

[16]

[16]

200 10 40 11 [16]

200 10 40 11 [16]

200 10 40 11 [16]

200 10 30 11 [12]

100 10 40 10 [25]

140

100

200 10 40 11 [16]

100 10 40 10 [16]

200 10 40 11 [16]

100 10 45 10 [16]

200 10 40 11 [16]

100 10 40 10 [16]

100

180 10 60 13 f +c

CE [V]

[RCOO]

EP [V]

180 10 60 13 f

DP [V]

[RCOO]

Q3

[M+OAc]

[M+NH4]

Negative

[M+OAc]

Solvent Ionization gradient mode Q1

2b

Subclass

DGTA/ DGTS

Betaine lipids

Glycerolipids

Lipid Category Class

Table 2 Mass transitions and optimized MS parameters for detection of molecular species from various lipid classes by mass spectrometry

144 Cornelia Herrfurth et al.

2b

2b

2b

2b

PI

PIP

PIP2

PS

Positive Negative

2b Hex-GlcAIPCb 2b HexN-GlcAIPC HexNAcGlcA-IPC Hex-HexGlcA-IPC Hex-HexNGlcA-IPC HexHexNAcGlcA-IPC

Positive Positive

2b

Positive

Negative

Negative

Negative

Negative

Negative

2b

Cer

2c

2b

PG

LCB

Negative

2b

PE

Glycolipids HexCer

Neutral lipids

Sphingolipids

Negative

2b

PC

a

a

[LCB2H2O+H]+ [LCB2H2O+H]+ [Cer2H2O+H]+ [373]/[RCOC3H6NOH]

[M+H]+ [M+H]+ [M+NH4]+ [M2H]2







[MH2O+H]+/[M2H2O+H]+ / [M3H2O+H]+

[RCOOsn1] /[RCOOsn2]



[RCOOsn1] /[RCOOsn2]





[RCOOsn1] /[RCOOsn2]





[M+H]+

[MH]



[M+MeH]

[M+MeH]







[RCOOsn1]/[RCOOsn2]

[RCOOsn1] /[RCOOsn2]



[RCOOsn1] /[RCOOsn2]



[MH]

[MH]

[MH]



[RCOOsn1] /[RCOOsn2]



[RCOOsn1]/[RCOOsn2]



[M+OAc]



[M+MeH]

Negativea

2b

PA

[RCOO]





[MH]

[RCOO]

[MH]

Negative

2c

LPS

Negative

2c

LPI

10

10

10

55

50

10

10

20/ 10 25 / 30

[16]

[16]

[16]

(continued)

160 10 76 10 [14] 160 10 44 10 [26]

120

100

50

100 10 40 10 [16]

200 10 60 11 f

200 10 60 11 f

100 10 40 10 [16]

100 10 40 10 [16]

100 10 40 10 [16]

100 10 40 10 [16]

200 10 38 11 [12]

200 10 40 11 [16]

200 10 40 11 [16]

Plant Lipidomics Analysis 145

1

2b

SE

FS

2b

2b

[M+NH4] [M+Ac+NH4]+

Positivec

+

+

[SterolOH]+

[SterolOH]

[SterolOH]

+

+

[M+NH4]

[SterolOH]+

[M+NH4]+

Positive

Positive

Positive

[M+3AcH] and [M+2AcH]/[M+AcH2OH] [M+2AcH] and [M+AcH]/[MH2OH]

Negativec

EP [V]

CE [V]

CXP [V] Reference

240

140

100

100

10

10

10

10

23

22

28

22

34

6

10

10

[28]

[16]

[16]

[16]

100 10 31 10 [27]

100 10 50 10 f

DP [V]

Details of the solvent gradient are described in Table 1. The ionization mode depicts the polarity of the nanoESI source. Q1 and Q3 indicate the parent and product ions, respectively. CP, EP, CE, and CXP indicate the declustering potential, entrance potential, collision energy, and cell exit potential for the molecular species of the corresponding lipid classes, respectively Abbreviations for lipid classes and subclasses: Ara Arabidopside, ASG acylated steryl glucoside, Cer ceramide, Cer-P ceramide phosphate, CL cardiolipin, DAG diacylglycerol, DGCC diacyl-carboxyhydroxymethylcholine, DGDG digalactosyldiacylglycerol, DGMG digalactosylmonoacylglycerol, DGTA diacylglyceryl-hydroxymethyltrimethyl-β-alanine, DGTS diacylglyceryl-O-(N,N,N-trimethyl)-homoserine, FS free sterol, GlcA α-glucuronic acid, Hex hexosyl, HexCer hexosylceramide, HexN hexosaminyl, HexNAc Nacetylhexosaminyl, IPC inositol phosphoceramide, LCB long-chain base, LCB-P long-chain base phosphate, LPA lysophosphatidic acid, LPC lysophosphatidylcholine, LPE lysophosphatidylethanolamine, LPG lysophosphatidylglycerol, LPI lysophosphatidylinositol, LPS lysophosphatidylserine, PA phosphatidic acid, PC phosphatidylcholine, PE phosphatidylethanolamine, PG phosphatidylglycerol, PI phosphatidylinositol, PIP phosphatidylinositol phosphate, PIP2 phosphatidylinositol bisphosphate, PS phosphatidylserine, MGDG monogalactosyldiacylglycerol, MGMG monogalactosylmonoacylglycerol, SQDG sulfoquinovosyldiacylglycerol, SE, steryl ester, SG steryl glucoside, SQMG sulfoquinovosylmonoacylglycerol, TAG triacylglycerol a Methylation before UPLC-nanoESI-MS/MS analysis b Nomenclature corresponding to [29] c Acetylation before UPLC-nanoESI-MS/MS analysis d Equals to [MC6H11O6+NH4]+ e Equals to [MC12H21O11+NH4]+ f Equals to [MC6H11O5RCOOGal+NH4]+ g Unpublished data of the authors

Neutral lipids

ASG

Glycolipids SG

Sterol lipids

2c

LCB-P

[M+AcH]/[MH2OH]

[M+2AcH]

Q3

Negativec

Solvent Ionization gradient mode Q1

2b

Subclass

Phospholipids Cer-P

Lipid Category Class

Table 2 (continued)

146 Cornelia Herrfurth et al.

Plant Lipidomics Analysis

147

5. The performance of the UPLC should be controlled regularly by inspecting the back pressure of the system and the retention time stability using lipid extracts with known composition or analytical standards. 6. Set the ionization voltage of the nanoESI system to 1.5 kV in negative mode, or to 1.5 kV in positive mode, when the UPLC flow is started.

Fig. 1 Solvent gradient-depending retention time windows for the elution of the corresponding lipid subclasses by the UPLC-nanoESI-MS/MS analysis. The amount of solvent B, the flow rate, the ionization mode and the retention time window for the elution of lipids are shown for solvent gradient 1 (a), solvent gradient 2a (b), solvent gradient 2b (c) and solvent gradient 2c (d). All data derive from the UPLC-nanoESI-MS/MS analysis of lipid extracts from leaves of Arabidopsis thaliana except Cer-P, HexNAc-GlcA-IPC, Hex-Hex-Glc-IPC, LCB-P (microsomal membrane fractions [27]), CL (isolated mitochondria), DGTA (Physcomitrium patens). Abbreviations are explained in Table 2

148

Cornelia Herrfurth et al.

Fig. 1 (continued)

7. The performance of the nanoESI device has to be controlled regularly by visually inspecting the surface of the chip and routinely calibrating the LC coupler using ChipSoft. During the analysis, the nanoelectrospray current should be monitored constantly. 8. Operate the QTRAP6500® tandem mass spectrometer in MRM mode in either negative or positive mode depending on the lipid class of interest (Table 2). 9. Import lipid subclass-specific mass transition lists of molecular species constructed on the basis of the identified lipid building blocks (see Subheading 3.8). The general calculation of the precursor m/z values, product m/z values, and the optimized MS parameters are shown in Table 1. 10. Set the dwell time to 5 ms for all mass transitions. 11. Adjust the resolution of the mass analyzers to 0.7 amu full width at half-height (FWHH).

Plant Lipidomics Analysis

Fig. 1 (continued)

149

150

Cornelia Herrfurth et al.

Fig. 1 (continued)

12. Set the ion source temperature to 40
C and the curtain gas at 10 (given in arbitrary units). 13. The performance of the QTRAP6500® tandem mass spectrometer has to be controlled regularly. Inspection of the mass sensitivity with lipid extracts with known composition or analytical standards should be performed before running samples. The mass spectrometer has to be cleaned annually following the instruction of the manufacturer. Moreover, the mass accuracy and resolution have to be tuned using the tuning mixture following cleaning.

Plant Lipidomics Analysis

151

14. For identification of precursor ions and fragment ions, use the Q1 MS mode or the product ion mode and vary the declustering potential and collision energy depending on the requirements of the analyte ion.

3.8 Assembly of the Lipid Building Blocks for the Target Lipid List

To construct the mass transition lists, the identities of the following three lipid building blocks are either taken from the literature or determined experimentally by performing lipid measurements by for example GC or TLC using the biological sample. Lipid building blocks: 1. Acyl chains (nonhydroxylated, hydroxylated). 2. Polar head groups (carboxyhydroxymethylcholine, dihexose, hexose, hexosyl glucosamine inositolphosphate (Hex-GlcAIns-P), hexosaminyl (HexN) GlcA-Ins-P, N-acetylhexosaminyl (HexNAc) GlcA-Ins-P, Hex-Hex-GlcA-Ins-P, Hex-HexNGlcA-Ins-P, Hex-HexNAc-GlcA-Ins-P, hydroxymethyltrimethyl-β-alanine, (N,N,N-trimethyl)-homoserine, sulfoquinovose, phosphate, phosphocholine, phosphoethanolamine, phosphoglycerol, phosphoinositol, phosphoserine). 3. Backbones (glycerol, long-chain bases, sterols). Classical lipid analytical methods are applied to analyze the building blocks as follows. Complex lipids are first partially disrupted by methanolysis under acidic or alkaline conditions (see Chapter 4). Additional modifications of hydroxy or amino groups via silylation (e.g., with BSTFA), and of methyl esters via transesterification into nitrogen-containing derivatives (e.g., pyrrolidides or 4,4-dimethyloxazolines) can be performed prior to the separation and detection of the analytes. Separation approaches including thin layer chromatography (TLC), high-performance liquid chromatography (HPLC), or gas chromatography (GC) can be coupled with detection methods ranging from staining procedures, flame ionization (FID) to mass spectrometry after electrospray ionization as well as electron impact ionization (for example: [17, 18], https://www.lipidhome.co.uk/, see Chapters 1, 4, and 15).

3.9 Data Analysis and Processing

1. Before starting data analysis, peak identification is supported by coelution (same retention time and identical MS/MS patterns including head group-specific fragments) with analytical standards, authentic lipids isolated from biological extracts, and/or lipid data from the literature. Chromatographic rules for elution orders of the lipid classes and molecular species in a reversed-phase mode (structure of the head groups, length of the side chains, number of double bonds and hydroxyl groups) need to be considered (Fig. 1) [16]. The identities of molecular

152

Cornelia Herrfurth et al.

species can be confirmed by performing TLC followed by lipid isolation from the plate and accurate mass measurements with a high-resolution mass spectrometer (time-of-flight analyzer, Orbitrap® analyzer, etc.). 2. For peak integration with MultiQuant Software, set Gaussian smooth width to 2 points and minimal peak height to 300 points. Copy the entire output table from MultiQuant to Excel. 3. For data processing and lipid profiling, correct the raw peak area with the naturally occurring proportion of the 13C isotopes. This contribution is calculated in correlation to the total number of carbon atoms from the respective molecular lipid species by multiplying the raw peak area by the correction factor αCN [19]. αCN ¼ 1 + 0.011 n + 0.0112 n (n  1)/2 (n ¼ total number of carbon atoms of the lipid species). The 13C isotope-corrected peak areas are then used to construct category-specific, class-specific, or subclass-specific lipid profiles. 3.10 Absolute Quantification of Lipid Subclasses by TLC Coupled with GC-FID

UPLC-nanoESI-MS/MS-based lipid analysis is a highly efficient method to separate molecular lipid species due to the robust reduction of sample matrix effects. One drawback of this chromatographic separation is the unequal solubility of the analytes along the elution gradient. Therefore, the ionization efficiency and fragmentation kinetics vary according to the target molecule. Different sample types, such as whole tissue extracts or membrane fractions, could limit the applicability of internal standards as well. Therefore, for absolute quantification of individual lipid subclasses, an analytical approach coupling TLC with GC/FID is used [20]. 1. The lipid extract is first separated by lipid class-specific TLC (for example: [17, 18], https://www.lipidhome.co.uk/, see Chapters 3 and 6). 2. Lipid subclasses are stained reversibly with 0.05% [w/v] primuline (Merck KGaA, Darmstadt, Germany) in 80% [v/v] acetone [21]. 3. The lipids are isolated from the TLC plate. 4. Lipids are converted to fatty acid methyl esters (FAMEs) in methanol containing 5% (v/v) sulfuric acid overnight at 110
C [22]. For sterol quantification, the isolated lipids are silylated with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA). 5. Pentadecanoic acid (15:0) is used as internal standard for glycerolipid quantification [20], heptadecanoic acid (17:0) and 2-hydroxy-pentadecanoic acid for sphingolipid quantification [22], and cholestanol for sterol quantification [23].

Plant Lipidomics Analysis

153

6. The resulting lipid subclass-specific FAMEs are analyzed by GC-FID and total amounts of the individual lipid subclasses are calculated based on normalization to internal standards (see Chapter 4).

4

Notes 1. Wear glasses and cold-protective gloves when handling liquid nitrogen. 2. Cool down all equipment (mortar, pestle, mill jars, cups, tubes, spatula) with liquid nitrogen to avoid thawing and clogging of the sample on the equipment. 3. Arabidopsis thaliana rosettes are usually completely homogenized to fine powder using the Mixer Ball Mill MM200 for 1 min at 30 vibrations s1. 4. Wear gloves for sample extraction and sample handling before analysis. 5. Use pipette tips that are chemically resistant to organic solvents, and use glass tubes instead of plastic tubes for lipid extraction. 6. Lipids in organic solution and lipid extracts can be stored at 20
C covered with argon for at least 1 year.

Acknowledgments We are very grateful to Pablo Tarazona for the initial establishment of the lipidomics method and to Tegan Haslam for editing the manuscript. We thank Sabine Freitag and Pia Meyer for their excellent assistance. References 1. Shulaev V, Chapman KD (2017) Plant lipidomics at the crossroads: from technology to biology driven science. Biochim Biophys Acta 1862(8):786–791. https://doi.org/10.1016/ j.bbalip.2017.02.011 2. Liebisch G, Vizcaı´no JA, Ko¨feler H, Tro¨tzmu¨ller M, Griffiths WJ, Schmitz G, Spener F, Wakelam MJO (2013) Shorthand notation for lipid structures derived from mass spectrometry. J Lipid Res 54 (6):1523–1530. https://doi.org/10.1194/jlr. M033506 3. Fahy E, Cotter D, Sud M, Subramaniam S (2011) Lipid classification, structures and

tools. Biochim Biophys Acta 1811 (11):637–647. https://doi.org/10.1016/j. bbalip.2011.06.009 4. Han X, Gross RW (2003) Global analyses of cellular lipidomes directly from crude extracts of biological samples by ESI mass spectrometry: a bridge to lipidomics. J Lipid Res 44 (6):1071–1079. https://doi.org/10.1194/jlr. R300004-JLR200 5. Oldach L (2019) Harmonizing lipidomics. ASBMB Today 18(5):26–39 6. Yang K, Han X (2016) Lipidomics: techniques, applications, and outcomes related to biomedical sciences. Trends Biochem Sci 41

154

Cornelia Herrfurth et al.

(11):954–969. https://doi.org/10.1016/j. tibs.2016.08.010 7. Simons B, Kauhanen D, Sylv€anne T, Tarasov K, Duchoslav E, Ekroos K (2012) Shotgun lipidomics by sequential precursor ion fragmentation on a hybrid quadrupole time-of-flight mass spectrometer. Meta 2(1):195–213. https://doi.org/10.3390/metabo2010195 8. Samarakoon T, Shiva S, Lowe K, Tamura P, Roth MR, Welti R (2012) Arabidopsis thaliana membrane lipid molecular species and their mass spectral analysis. Methods Mol Biol 918:179–268. https://doi.org/10.1007/ 978-1-61779-995-2_13 9. Markham JE, Li J, Cahoon EB, Jaworski JG (2006) Separation and identification of major plant sphingolipid classes from leaves. J Biol Chem 281(32):22684–22694. https://doi. org/10.1074/jbc.M604050200 10. Abas L, Luschnig C (2010) Maximum yields of microsomal-type membranes from small amounts of plant material without requiring ultracentrifugation. Anal Biochem 401 (2):217–227. https://doi.org/10.1016/j.ab. 2010.02.030 11. Grillitsch K, Tarazona P, Klug L, Wriessnegger T, Zellnig G, Leitner E, Feussner I, Daum G (2014) Isolation and characterization of the plasma membrane from the yeast Pichia pastoris. Biochim Biophys Acta 1838(7):1889–1897. https://doi.org/10. 1016/j.bbamem.2014.03.012 12. Lee JW, Nishiumi S, Yoshida M, Fukusaki E, Bamba T (2013) Simultaneous profiling of polar lipids by supercritical fluid chromatography/tandem mass spectrometry with methylation. J Chromatogr A 1279:98–107. https:// doi.org/10.1016/j.chroma.2013.01.020 13. Berdyshev EV, Gorshkova IA, Garcia JGN, Natarajan V, Hubbard WC (2005) Quantitative analysis of sphingoid base-1-phosphates as bisacetylated derivatives by liquid chromatography–tandem mass spectrometry. Anal Biochem 339(1):129–136. https://doi.org/10. 1016/j.ab.2004.12.006 14. Markham JE, Jaworski JG (2007) Rapid measurement of sphingolipids from Arabidopsis thaliana by reversed-phase high-performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 21(7):1304–1314. https://doi.org/10.1002/rcm.2962 15. Markham JE (2013) Detection and quantification of plant sphingolipids by LC-MS. In: Munnik T, Heilmann I (eds) Plant lipid signaling protocols. Humana, Totowa, NJ, pp 93–101. https://doi.org/10.1007/978-162703-401-2_10

16. Tarazona P, Feussner K, Feussner I (2015) An enhanced plant lipidomics method based on multiplexed liquid chromatography–mass spectrometry reveals additional insights into coldand drought-induced membrane remodeling. Plant J 84(3):621–633. https://doi.org/10. 1111/tpj.13013 17. Christie WW, Han X (2010) Lipid analysis – isolation, separation, identification and lipidomic analysis, 4th edn. Oily Press, Bridgwater 18. Schneiter R (2006) Analysis of yeast lipids. Methods Mol Biol 313:75–84. https://doi. org/10.1385/1-59259-958-3:075 19. Iven T, Herrfurth C, Hornung E, Heilmann M, Hofvander P, Stymne S, Zhu L-H, Feussner I (2013) Wax ester profiling of seed oil by nano-electrospray ionization tandem mass spectrometry. Plant Methods 9 (1):24. https://doi.org/10.1186/17464811-9-24 20. Wang Z, Benning C (2011) Arabidopsis thaliana polar glycerolipid profiling by thin layer chromatography (TLC) coupled with gas-liquid chromatography (GLC). J Vis Exp 49:e2518. https://doi.org/10.3791/2518 21. Kelly AA, van Erp H, Quettier A-L, Shaw E, Menard G, Kurup S, Eastmond PJ (2013) The SUGAR-DEPENDENT1 lipase limits triacylglycerol accumulation in vegetative tissues of Arabidopsis. Plant Physiol 162(3):1282–1289. https://doi.org/10.1104/pp.113.219840 22. Cacas J-L, Bure´ C, Grosjean K, GerbeauPissot P, Lherminier J, Rombouts Y, Maes E, Bossard C, Gronnier J, Furt F, Fouillen L, Germain V, Bayer E, Cluzet S, Robert F, Schmitter J-M, Deleu M, Lins L, SimonPlas F, Mongrand S (2016) Revisiting plant plasma membrane lipids in tobacco: a focus on sphingolipids. Plant Physiol 170 (1):367–384. https://doi.org/10.1104/pp. 15.00564 23. Wewer V, Dombrink I, vom Dorp K, Do¨rmann P (2011) Quantification of sterol lipids in plants by quadrupole time-of-flight mass spectrometry. J Lipid Res 52(5):1039–1054. https://doi.org/10.1194/jlr.D013987 24. Ibrahim A, Schu¨tz A-L, Galano J-M, Herrfurth C, Feussner K, Durand T, Brodhun F, Feussner I (2011) The alphabet of galactolipids in Arabidopsis thaliana. Front Plant Sci 2:95. https://doi.org/10.3389/fpls. 2011.00095 25. Zhou Y, Peisker H, Do¨rmann P (2016) Molecular species composition of plant cardiolipin determined by liquid chromatography mass spectrometry. J Lipid Res 57(7):1308–1321. https://doi.org/10.1194/jlr.D068429

Plant Lipidomics Analysis 26. Bure´ C, Cacas JL, Wang F, Gaudin K, Domergue F, Mongrand S, Schmitter JM (2011) Fast screening of highly glycosylated plant sphingolipids by tandem mass spectrometry. Rapid Commun Mass Spectrom 25 (20):3131–3145. https://doi.org/10.1002/ rcm.5206 27. Zienkiewicz A, Go¨mann J, Ko¨nig S, Herrfurth C, Liu Y-T, Meldau D, Feussner I (2020) Disruption of Arabidopsis neutral ceramidases 1 and 2 results in specific sphingolipid imbalances triggering different phytohormone-dependent plant cell death programs. New Phytol 226(1):170–188. https://doi.org/10.1111/nph.16336 28. Liebisch G, Binder M, Schifferer R, Langmann T, Schulz B, Schmitz G (2006)

155

High throughput quantification of cholesterol and cholesteryl ester by electrospray ionization tandem mass spectrometry (ESI-MS/MS). Biochim Biophys Acta 1761(1):121–128. https://doi.org/10.1016/j.bbalip.2005.12. 007 29. Fang L, Ishikawa T, Rennie EA, Murawska GM, Lao J, Yan J, Tsai AY-L, Baidoo EEK, Xu J, Keasling JD, Demura T, Kawai-YamadaM, Scheller HV, Mortimer JC (2016) Loss of inositol phosphorylceramide sphingolipid mannosylation induces plant immune responses and reduces cellulose content in Arabidopsis. Plant Cell 28(12):2991–3004. https://doi.org/10.1105/tpc.16.00186

Chapter 10 Mass Spectrometry-Based Profiling of Plant Sphingolipids from Typical and Aberrant Metabolism Rebecca E. Cahoon, Ariadna Gonzalez Solis, Jennifer E. Markham, and Edgar B. Cahoon Abstract Mass spectrometry has increasingly been used as a tool to complement studies of sphingolipid metabolism and biological functions in plants and other eukaryotes. Mass spectrometry is now essential for comprehensive sphingolipid analytical profiling because of the huge diversity of sphingolipid classes and molecular species in eukaryotes, particularly in plants. This structural diversity arises from large differences in polar head group glycosylation as well as carbon-chain lengths of fatty acids and desaturation and hydroxylation patterns of fatty acids and long-chain bases that together comprise the ceramide hydrophobic backbone of glycosphingolipids. The standard methods for liquid chromatography–mass spectrometry (LC-MS)-based analyses of Arabidopsis thaliana leaf sphingolipids profile >200 molecular species of four sphingolipid classes and free long-chain bases and their phosphorylated forms. While these methods have proven valuable for A. thaliana based sphingolipid research, we have recently adapted them for use with ultraperformance liquid chromatography separations of molecular species and to profile aberrant sphingolipid forms in pollen, transgenic lines, and mutants. This chapter provides updates to standard methods for LC-MS profiling of A. thaliana sphingolipids to expand the utility of mass spectrometry for plant sphingolipid research. Key words Ceramide, Long-chain base, Serine palmitoyltransferase, Deoxysphinganine, Glycosylinositolphosphoceramide, GIPC, Glucosylceramide

1

Introduction Plant sphingolipids have emerged as central lipids in plant cellular functions, including biotic and abiotic stress responses, because of their abundance in plasma membranes and endomembranes and their functions in processes such as programmed cell death (PCD) that are part of the plant pathogen defense [1–3]. The extensive glycosylation, characteristic of glycosylinositolphosphoceramides (GIPCs), the most abundant plant sphingolipids [4, 5], has also been shown to be important for recognition of pathogenic and symbiotic microbes at the plasma membrane surface [5–

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_10, © Springer Science+Business Media, LLC, part of Springer Nature 2021

157

158

Rebecca E. Cahoon et al.

7]. Sphingolipids are built upon the long-chain base (LCB), typically an 18 carbon molecule in plants, that derives from the condensation of the amino acid serine and the fatty acid intermediate palmitoyl-CoA via the activity of serine palmitoyltransferase (SPT), a highly regulated enzyme in sphingolipid biosynthesis [8]. Acylation of LCBs by ceramide synthases with distinct LCB and fatty acyl-CoA specificities, generates ceramides (Cer) [9, 10]. Cer are the hydrophobic backbone of glycosphingolipids, analogous to the diacylgylcerol backbone of glycerolipids. Cer are glycosylated at their C-1 hydroxyl group to form the abundant membrane glycosphingolipids GIPCs and glucosylceramides (GlcCer or glucocerebrosides). LCBs and Cer in their free forms are also linked with PCD induction, which can be mediated by reversible phosphorylation of Cer [11]. LCBs, in their simplest form, have two hydroxyl groups and are referred to as dihydroxy LCBs (designated d18:X, 18 ¼ numbers of carbon atoms, X ¼ numbers of double bonds) [12]. LCBs can also be further hydroxylated to form trihydroxy LCBs (designated t18:X). The fatty acids of Cer typically contain 16–26 carbon atoms, and fatty acids and LCBs can occur in saturated or unsaturated forms in A. thaliana [8]. Mass spectrometry–based profiling applied to mutants and transgenic lines has been a key method for elucidating details of sphingolipid biosynthesis and metabolism and for understanding implications of dysfunctions in SPT-mediated biosynthetic regulation and downstream alterations in Cer and glycosphingolipid biosynthesis. Typical mass spectrometry methods use multiple reaction-monitoring (MRM) with ultraperformance liquid chromatography (UPLC) for separation of molecular species and detection by a triple quadrupole mass spectrometer to rapidly profile 200 different glycosphingolipid, Cer, and LCB molecular species [13, 14]. MRMs involve detection of the M+H+ ion in positive mode (Q1 mass), and a characteristic collision-induced fragment ion (Q3), typically the long-chain base or ceramide portion of the complex sphingolipid molecule. A unique chromatographic retention time establishes the correct identity of the compounds. Sphingolipid species within each class elute from a C18 column with a typical pattern, where complex sphingolipids containing ceramides with t18:1 LCBs elute first, followed sequentially by species containing ceramides with t18:0, d18:1, and d18:0 LCBs. Sphingolipid species within a given class that contain Cer with d18:2 LCBs elute between those with t18:1 and t18:0 LCBs. Glycosylation reduces retention time of sphingolipids on a C18 column. Less hydroxylation increases retention times, as in the case of Cer (harboring nonhydroxylated fatty acids), which elute before Cer that contain fatty acids with α- or C-2 hydroxyl groups, referred to as hydroxyl-Cer (or hCer). Comparison of peak areas of unknown to a known amount of internal standard is used for quantification [13, 14].

Profiling of Plant Sphingolipids

159

The methods described here provide a guide for mass spectrometry profiling of typical sphingolipid classes found in A. thaliana leaves. The methods have been updated to expand profiling for nontypical or aberrant sphingolipids that we have recently profiled in mutants and transgenic lines with alterations in specific biosynthetic steps and in A. thaliana pollen: 1. Variant products of SPT activity. SPT functions as a heterodimer of LCB1 and LCB2 subunits [15, 16]. The SPT complex includes the noncatalytic proteins, the small subunit of SPT (ssSPT), an essential protein for maintaining sufficient SPT activity, and ORMs (or orosomucoid-like proteins), negative regulators of SPT [17–22]. Abnormal regulation of SPT activity allows the use of alanine, in addition to serine, as a substrate [17]. The result is the production of the aberrant LCB structure 1-deoxysphinganine (DoxSA, deoxy-d18:0 or monohydroxy 18:0, m18:0) that lacks the C-1 hydroxyl group that is required for glycosylation of ceramides to form GlcCer and GIPCs (see Fig. 1). DoxSA is a biomarker and believed to be the causal agent of the human disease hereditary sensory neuropathy (HSN) Type 1, which results from mutations in the LCB1 subunit that affect SPT substrate specificity, allowing for condensation of palmitoyl-CoA with serine or alanine [23– 25]. DoxSA can be acylated by ceramide synthases to form DoxSA-containing Cer that are not further used for glycosphingolipid biosynthesis. 1-Deoxysphingosine (DoxSO, or m18:1) ceramide, formed by desaturation of DocSA, is also detectable in small amounts [17]. Methods for profiling of deoxy-LCBs and -ceramides are described in Subheading 3.3.1. 2. Glucosylceramides with Nonhydroxylated Fatty Acids. The fatty acid component of sphingolipid ceramides predominantly contains an α- or C-2 hydroxyl group when present in glycosphingolipids [26]. The hydroxylation is believed to occur at the ceramide stage of sphingolipid biosynthesis, prior to glycosylation [12]. One of the results of unregulated SPT activity in ORM knockout mutants is the increased production of GlcCer species that lack fatty acid α-hydroxylation, likely due to saturation of fatty acid α-hydroxylase activity by ceramide substrates, which are 50- to 100-fold higher than levels in wildtype plants [17]. We have demonstrated the ability of A. thaliana plants to produce GlcCer with nonhydroxylated fatty acids by feeding psychosines or glucosylated LCBs to a GlcCer synthase knockout mutant [27]. In this experiment, psychosines were acylated predominantly by the LOH2 ceramide synthase that has specificity for C16 fatty acids and LCBs with two hydroxyl groups (or dihydroxy LCBs) [27]. Methods for profiling of nonhydroxylated GlcCer (or nhGlcCer) are described in Subheading 3.3.2.

160

Rebecca E. Cahoon et al.

Fig. 1 Formation of 1-deoxysphinganine and 1-deoxyceramides from aberrant serine palmitoyltransferase (SPT) activity. Loss of ORM regulation of SPT, consisting of LCB1 and LCB2 subunits and small subunit of SPT (ssSPT), and LCB1 mutations associated with hereditary sensory neuropathy (HSN) result in the aberrant use of alanine, rather than serine, as a SPT substrate. The product of this reaction 1-deoxysphinganine (DoxSA; deoxy-d18:0 or m18:0) can be acylated by ceramide synthases to form ceramides (Cer). Because these 1-deoxyceramides lack a C-1 hydroxyl group, they cannot be glycosylated to form glycosphingolipids. In typical sphingolipid metabolism, SPT and 3-ketosphinganine reductase generate the long-chain base (LCB) sphinganine from serine and palmitoyl-CoA. Sphinganine can be further metabolized into Cer and glycosphingolipids

3. Variations in GIPC Classes. GIPCs are formed by glycosylation of an inositolphosphoceramide (IPC) that is synthesized from the exchange of the phosphoinositol head group from phosphatidylinositol to Cer [12]. The predominant GIPC in A. thaliana leaves contains a hexose-hexuronic acid moiety bound to the inositol residue of the IPC core (Hex-HexAIPC). Looking more broadly in the plant kingdom, species with up to seven sugar residues bound to IPC have been identified [28]. While our typical GIPC profiling of A. thaliana leaves and rosettes focuses on the hexosehexuronic acid IPC form, we found that GIPCs in A. thaliana pollen are distinct from those in leaves and include an array of glycosylation patterns with up to six sugar residues, most of which contain N-acetylation of one of the hexose moieties [29]. The procedures below were adapted for

Profiling of Plant Sphingolipids

161

profiling of A. thaliana pollen GIPCs. We detected an abundance of IPCs lacking glycosylation in A. thaliana IPUT1 mutants that are deficient in inositol phosphorylceramide glucuronosyltransferase1 activity that links the hexuronic acid (glucuronic acid) to IPC [30]. Elevated concentrations of nonglycosylated IPCs were also detected in an ORM mutant with defective SPT regulation [17]. We provide methods (Subheading 3.3) that were used for profiling highly glycosylated IPCs and IPCs. This chapter provides an update on mass spectrometry protocols, including the use of UPLC for sphingolipid molecular species separation, to broaden the utility of this methodology to study plant sphingolipid biosynthetic and metabolic plasticity.

2

Materials All solvents should be HPLC grade or above, OmniSolv brand solvents (Millipore Sigma, St Louis, MO, USA) are a good starting point. Measure solvents in clean, detergent-free glassware in a fume hood with appropriate personal protective equipment. Always measure solvents before mixing. Small amounts of solvents are usually pipetted with a glass pipette or clean Hamilton syringe of appropriate size.

2.1

Lipid Extraction

1. EXTRACTION SOLVENT. Prepare a mixture of 2-propanol (isopropanol; 55 mL), water (20 mL), and hexane (25 mL), mix well and allow the phases to separate, use only the lower phase. 2. DUALL, All-Glass Tissue grinder, size 21/3 mL capacity (Kimble Chase, Vineland, NJ). Use once and clean with isopropanol, water and methanol after each use. If many samples are to be extracted have up to five available to use in rotation. 3. Conical glass centrifuge tubes, 16  110 mm, screw thread cap. 4. Round bottom glass culture tubes, 16  100 mm, screw thread cap. 5. Nitrogen drying apparatus (e.g., TurboVap, Biotage, Uppsala, Sweden) or centrivap concentrator for 16  100 mm tubes. 6. METHYLAMINE REAGENT. Combine 7 mL of methylamine solution (33%, v/v, in ethanol, Millipore Sigma) with 3 mL of water.

2.2 UPLC and Mass Spectrometry

1. SCIEX 4000 or 6500 QTRAP or equivalent mass spectrometer fitted with Turbo V ion source and Turbo Ion Spray (TIS) probe (see Note 1).

162

Rebecca E. Cahoon et al.

2. UPLC system consisting of two binary pumps, low volume mixer, autoinjector, and column oven (see Note 2). 3. Zorbax Eclipse Plus C18 column, 2.1  100 mm, 1.8μm particle size, fitted with a 2.1 mm  12.5 mm guard column (Agilent Technologies, Santa Clara, CA) (see Note 3). 4. Ammonium formate: 5 mM solution in water (dilute from 50 mM stock). 5. SOLVENT A: Combine 300 mL of tetrahydrofuran with 200 mL of methanol and 500 mL of 5 mM ammonium formate. Add 1 mL of formic acid. 6. SOLVENT B: combine 700 mL of tetrahydrofuran with 200 mL of methanol and 100 mL of 5 mM ammonium formate. Add 1 mL of formic acid (see Note 4). 7. SAMPLE SOLVENT: combine 400 mL of tetrahydrofuran with 200 mL of methanol and 400 mL of water. Add 1 mL of formic acid (see Note 5). 2.3 Sphingolipid Standards

1. LIPID STANDARDS, including deoxysphingolipids, are available from Avanti Polar Lipids, Alabaster, AL (see Table 1). Dissolve all standards at a concentration of 1 mg/mL in chloroform–methanol–water 16:16:5 (v/v/v) and store at 20  C for up to 1 year. Warm to room temperature and ensure all lipids are dissolved before using them. 2. Combine lipid standards as outlined in Table 1 using a Hamilton-syringe. Use deoxysphingolipid standards as required. Dry standards under a stream of nitrogen and redissolve them in 1 mL of SAMPLE SOLVENT.

Table 1 Internal standards use for quantification of different sphingolipid classes

Sphingolipid class

Internal standard

Microliter stock per mL

Amount per 10 μL, nmol

GIPCs/IPC

GM1

312.5

2

(h/nh)GlcCer

Glucosyl-C12-ceramide

64.5

1

(h/nh)Ceramide

C12-ceramide

4.83

0.1

Long-chain base

C17-sphingosine

2.85

0.1

Long-chain base phosphate

C17-sphingosine-1-phosphate

3.65

0.1

Each standard is dissolved individually as a stock solution of 1 mg/mL and then combined in the quantities indicated to produce a standard solution GIPC glycosylinositolphosphoceramide, IPC inositolphosphoceramide, GM1 GM1 ganglioside, GlcCer glucosylceramide, h hydroxylated, nh nonhydroxylated

Profiling of Plant Sphingolipids

3

163

Methods

3.1 Sphingolipid Extraction

1. Freeze-dry samples overnight to ensure complete removal of water. Once freeze dried, sphingolipids are stable at 80  C for years or at room temperature for several days. 2. Weigh 10–30 mg of tissue at room temperature and place in DUALL all-glass tissue grinder. Record the weight. 3. Add 10 μL of the LIPID STANDARD solution and 3 mL of EXTRACTION SOLVENT and grind tissue until all tissue is completely disrupted. 4. Vortex and pour sample into a glass centrifuge tube. Rinse tissue grinder with 3 mL of EXTRACTION SOLVENT and add to sample. 5. Cap the tube and incubate the sample at 60  C for at least 15 min. 6. Centrifuge the sample at 500  g for 10 min. Decant the supernatant into a clean, round bottom tube. 7. Resuspend the pellet in 3 mL of EXTRACTION SOLVENT and incubate at 60  C for an addition 15 min. Centrifuge as before and combine with the supernatant from step 7. 8. Dry the sample under nitrogen gas or vacuum. 9. Add 2 mL of METHYLAMINE REAGENT and incubate at 50  C for 1 h. Methylamine treatment results in hydrolysis of ester bonds thus removes contaminating glycerolipids. 10. Dry the sample under nitrogen gas and dissolve in 1 mL of SAMPLE SOLVENT (gentle sonication and sample heating is usually required). Transfer to a 2 mL autosampler vial and cap tightly. Store at 20  C until required.

3.2 UPLC-MS Detection of Sphingolipids

1. Set up methods for UPLC separation of the different sphingolipid classes (see Note 6). All methods use a flow rate of 0.2 mL/min, oven temperature of 40  C and a post-run equilibration of 1–2 min. The flow for the first minute after sample injection is directed to waste, the gradient program starts immediately following injection. Chromatography gradient parameters are described in Table 2. 2. Set up the source parameters for each method as follows: Curtain Gas (CUR) 20; Collision Gas (CAD), Medium; IonSpray Voltage (IS) 5000; Temperature (TEM) 350–400  C; Ion Source Gas 1 (GS1), 60; Ion Source Gas 2 (GS2), 50. 3. Input MRM parameters for the mass-spectrometer using Tables 3, 4, 5, 6, and 7.

164

Rebecca E. Cahoon et al.

Table 2 Gradient parameters for separation of individual sphingolipid species within a class by UPLC prior to detection by mass-spectrometry (d/h)Cer

(h)GlcCer

(G/GG)IPC

LCB(P)

[B] start

50%

50%

25%

10%

Gradient [B], time

85%, 10 min 100%, 14 min 100%, 17 min

70%, 11 min 100%, 12 min 100%, 14 min

50%, 10 min 100%, 12 min 100%, 14 min

35%, 5 min 100%, 7 min 100% 11 min

Stop

22 min

17 min

18 min

14 min

The table shows concentrations of buffer B (in %) and the times and end points of the linear gradients (d/h)Cer dihydroxy or hydroxy ceramide, GlcCer glucosylceramide, (G/GG)IPC glycosyl or glycosylglycosylinositolphosphorylceramide, LCB(P) long-chain base (phosphate)

Table 3 MRM parameters for detection of LCB and LCBPs by mass spectrometry LCB(P)

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

d17:1

286.3

268.3

45

16

d18:0

302.3

284.3

55

21

d18:1

300.3

282.3

55

18

d18:2

298.3

280.3

55

18

t18:0

318.3

300.4

50

21

t18:1

316.3

298.4

50

18

t20:0

346.3

328.3

55

20

t20:1

344.3

326.3

55

20

d20:0

330.3

312.3

55

20

d20:1

328.3

310.3

55

20

m18:0 (doxSA)

286.3

268.3

55

20

m18:1 (doxSO)

284.3

266.32

55

18

Glc-d18:1 (psychosine)

462.3

282.3

45

20

d17:1P

366.2

250.3

50

23

d18:0P

382.3

266.3

55

19

d18:1P

380.3

264.3

50

25

d18:2P

378.3

262.3

50

25

t18:0P

398.3

300.3

55

22

t18:1P

396.3

298.3

55

25

Q1 and Q3 are the parent and product ion m/z settings and DP and CE indicate the declustering potential and collision energies for each compound

Profiling of Plant Sphingolipids

165

Table 4 MRM parameters for ceramide (Cer) and 1-deoxyceramides Cer and 1-deoxyCer LCB

FA

Standard

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

482.5

264.3

60

35

t18:0

16:0

556.5

300.3

100

35

t18:0

18:0

584.6

300.3

100

35

t18:0

20:0

612.6

300.3

100

37

t18:0

20:1

610.6

300.3

100

37

t18:0

22:0

640.6

300.3

100

43

t18:0

22:1

638.6

300.3

100

43

t18:0

24:0

668.7

300.3

100

43

t18:0

24:1

666.7

300.3

100

43

t18:0

26:0

696.7

300.3

100

43

t18:0

26:1

694.7

300.3

100

43

t18:1

16:0

554.5

298.3

100

38

t18:1

18:0

582.6

298.3

100

38

t18:1

20:0

610.6

298.3

100

40

t18:1

20:1

608.6

298.3

100

40

t18:1

22:0

638.6

298.3

100

42

t18:1

22:1

636.6

298.3

100

42

t18:1

24:0

666.7

298.3

100

42

t18:1

24:1

664.7

298.3

100

44

t18:1

26:0

694.7

298.3

100

44

t18:1

26:1

692.7

298.3

100

44

d18:0

16:0

540.5

266.3

40

42

d18:0

18:0

568.6

266.3

40

43

d18:0

20:0

596.6

266.3

42

43

d18:0

20:1

594.6

266.3

40

48

d18:0

22:0

624.6

266.3

39

48

d18:0

22:1

622.6

266.3

40

48

d18:0

24:0

652.7

266.3

39

44

d18:0

24:1

650.7

266.3

37

43

d18:0

26:0

680.7

266.3

43

48 (continued)

166

Rebecca E. Cahoon et al.

Table 4 (continued) Cer and 1-deoxyCer LCB

FA

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

d18:0

26:1

678.7

266.3

46

48

d18:1

16:0

538.5

264.3

40

39

d18:1

18:0

566.6

264.3

38

39

d18:1

20:0

594.6

264.3

44

39

d18:1

20:1

592.6

264.3

42

42

d18:1

22:0

622.6

264.3

44

46

d18:1

22:1

620.6

264.3

39

44

d18:1

24:0

650.7

264.3

38

49

d18:1

24:1

648.7

264.3

42

43

d18:1

26:0

678.7

264.3

38

46

d18:1

26:1

676.7

264.3

46

48

d18:2

16:0

536.5

262.3

40

39

d18:2

18:0

564.6

262.3

38

39

d18:2

20:0

592.6

262.3

44

39

d18:2

20:1

590.6

262.3

42

42

d18:2

22:0

620.6

262.3

44

46

d18:2

22:1

618.6

262.3

39

44

d18:2

24:0

648.7

262.3

38

49

d18:2

24:1

646.7

262.3

42

43

d18:2

26:0

676.7

262.3

38

46

d18:2

26:1

674.7

262.3

46

48

m18:0

16:0

524.5

268.3

40

42

m18:0

18:0

552.6

268.3

40

43

m18:0

20:0

580.6

268.3

42

43

m18:0

20:1

578.6

268.3

40

48

m18:0

22:0

608.6

268.3

39

48

m18:0

22:1

606.6

268.3

40

48

m18:0

24:0

636.7

268.3

39

44

m18:0

24:1

634.7

268.3

37

43

m18:0

26:0

664.7

268.3

43

48 (continued)

Profiling of Plant Sphingolipids

167

Table 4 (continued) Cer and 1-deoxyCer LCB

FA

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

m18:0

26:1

662.7

268.3

46

48

m18:1

16:0

522.5

266.3

40

39

m18:1

18:0

550.6

266.3

38

39

m18:1

20:0

578.6

266.3

44

39

m18:1

20:1

576.6

266.3

42

42

m18:1

22:0

606.6

266.3

44

46

m18:1

22:1

604.6

266.3

39

44

m18:1

24:0

634.7

266.3

38

49

m18:1

24:1

632.7

266.3

42

43

m18:1

26:0

662.7

266.3

38

46

m18:1

26:1

660.7

266.3

46

48

The LCB and fatty acid (FA) pairings are shown on the left followed by the parent (Q1) and product ion (Q3) m/z, the declustering potential (DP), and collision energy (CE) for each compound

4. Inject 10 μL of lipid extract for each LC-MS method. After injection rinse the needle in SAMPLE SOLVENT. 5. Once the data have been collected, it will be necessary to create quantification methods to integrate the peak areas (see Fig. 2) and determine the amount of each compound. The formula is as follows: Aanalyte A standard

 nmols of standard  R , grams tissue ðdry weightÞ

where Aanalyte and Astandard are the peak areas for each analyte and its respective standard, nmols of standard is the amount of each standard added for the respective sphingolipid class (shown in Table 1), R is a response factorial (see Table 8) to account for different responses between sphingolipids with different chemical structures compared to the internal standard, and grams tissue is the dry weight of the extracted tissue. The formula gives a value or the amount of sphingolipid in a sample in nmol per gram dry weight.

168

Rebecca E. Cahoon et al.

Table 5 MRM parameters for glucosylceramide (GlcCer) with typical hydroxyl fatty acids, and nonhydroxy GlcCer, which lack the 2-hydroxyl group on the fatty acid component of the ceramide backbone hGlcCer LCB

FA

Standard

nhGlcCer

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

644.5

264.3

56

46

644.5

264.3

56

46

t18:0

16:0

734.6

300.3

65

54

718.6

300.3

65

54

t18:0

18:0

762.6

300.3

70

51

746.6

300.3

70

51

t18:0

20:0

790.6

300.3

72

52

774.6

300.3

72

52

t18:0

20:1

788.6

300.3

72

52

772.6

300.3

72

52

t18:0

22:0

818.7

300.3

70

53

802.7

300.3

70

53

t18:0

22:1

816.7

300.3

70

53

800.7

300.3

70

53

t18:0

24:0

846.7

300.3

80

55

830.7

300.3

80

55

t18:0

24:1

844.7

300.3

75

56

828.7

300.3

75

56

t18:0

26:0

874.7

300.3

75

56

858.7

300.3

75

56

t18:0

26:1

872.7

300.3

75

40

856.7

300.3

75

40

t18:1

16:0

732.6

298.3

75

51

716.6

298.3

75

51

t18:1

18:0

760.6

298.3

75

55

744.6

298.3

75

55

t18:1

20:0

788.6

298.3

75

55

772.6

298.3

75

55

t18:1

20:1

786.6

298.3

75

60

770.6

298.3

75

60

t18:1

22:0

816.7

298.3

70

57

800.7

298.3

70

57

t18:1

22:1

814.7

298.3

70

60

798.7

298.3

70

60

t18:1

24:0

844.7

298.3

80

57

828.7

298.3

80

57

t18:1

24:1

842.7

298.3

80

59

826.7

298.3

80

59

t18:1

26:0

872.7

298.3

75

57

856.7

298.3

75

57

t18:1

26:1

870.7

298.3

70

62

854.7

298.3

70

62

d18:0

16:0

718.6

266.3

65

35

702.6

266.3

65

35

d18:0

18:0

746.6

266.3

70

55

730.6

266.3

70

55

d18:0

20:0

774.6

266.3

75

55

758.6

266.3

75

55

d18:0

20:1

772.6

266.3

75

55

756.6

266.3

75

55

d18:0

22:0

802.7

266.3

75

55

786.7

266.3

75

55

d18:0

22:1

800.7

266.3

75

55

784.7

266.3

75

55

d18:0

24:0

830.7

266.3

80

56

814.7

266.3

80

56

d18:0

24:1

828.7

266.3

80

56

812.7

266.3

80

56 (continued)

Profiling of Plant Sphingolipids

169

Table 5 (continued) hGlcCer

nhGlcCer

LCB

FA

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

d18:0

26:0

858.7

266.3

80

56

842.7

266.3

80

56

d18:0

26:1

856.7

266.3

80

56

840.7

266.3

80

56

d18:1

16:0

716.6

264.3

65

45

700.6

264.3

65

45

d18:1

18:0

744.6

264.3

65

45

728.6

264.3

65

45

d18:1

20:0

772.6

264.3

70

55

756.6

264.3

70

55

d18:1

20:1

770.6

264.3

70

55

754.6

264.3

70

55

d18:1

22:0

800.7

264.3

70

56

784.7

264.3

70

56

d18:1

22:1

798.6

264.3

70

56

782.6

264.3

70

56

d18:1

24:0

828.7

264.3

75

57

812.7

264.3

75

57

d18:1

24:1

826.7

264.3

80

57

810.7

264.3

80

57

d18:1

26:0

856.7

264.3

80

57

840.7

264.3

80

57

d18:1

26:1

854.7

264.3

80

57

838.7

264.3

80

57

d18:2

16:0

714.6

262.3

65

45

698.6

262.3

65

45

d18:2

18:0

742.6

262.3

65

45

726.6

262.3

65

45

d18:2

20:0

770.6

262.3

70

55

754.6

262.3

70

55

d18:2

20:1

768.6

262.3

70

55

752.6

262.3

70

55

d18:2

22:0

798.7

262.3

70

56

782.7

262.3

70

56

d18:2

22:1

796.6

262.3

70

56

780.6

262.3

70

56

d18:2

24:0

826.7

262.3

75

57

810.7

262.3

75

57

d18:2

24:1

824.7

262.3

80

57

808.7

262.3

80

57

d18:2

26:0

854.7

262.3

80

57

838.7

262.3

80

57

d18:2

26:1

852.7

262.3

80

57

836.7

262.3

80

57

MRMs for non-hydroxy fatty acid GlcCers are calculated by subtracting 16 m/z from the Q1 mass of the typical GlcCer. The LCB and fatty acid (FA) pairings are shown on the left followed by the parent (Q1) and product ion (Q3) m/z, the declustering potential (DP), and collision energy (CE) for each compound

170

Rebecca E. Cahoon et al.

Table 6 MRM parameters for inositolphosphoceramides (IPCs) and glucuronic acid–IPCs (GlcA-IPCs), the most common core GIPC in plants IPC LCB

FA

Standard

GlcA-IPC

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

1546.9

366.3

50

46

1546.9

366.3

50

0

t18:0

16:0

814.6

554.5

45

45

652.6

554.5

45

45

t18:0

18:0

842.6

582.5

45

45

680.6

582.5

45

45

t18:0

20:0

870.7

610.6

45

45

708.7

610.6

45

45

t18:0

20:1

868.7

608.6

45

45

706.7

608.6

45

45

t18:0

22:0

898.7

638.6

45

45

736.7

638.6

45

45

t18:0

22:1

896.7

636.6

45

45

734.7

636.6

45

45

t18:0

24:0

926.7

666.6

45

45

764.7

666.6

45

45

t18:0

24:1

924.7

664.6

45

45

762.7

664.6

45

45

t18:0

26:0

954.8

694.7

45

45

792.8

694.7

45

45

t18:0

26:1

952.8

692.7

45

45

790.8

692.7

45

45

t18:1

16:0

812.6

552.5

45

45

650.6

552.5

45

45

t18:1

18:0

840.6

580.5

45

45

678.6

580.5

45

45

t18:1

20:0

868.7

608.6

45

45

706.7

608.6

45

45

t18:1

20:1

866.7

606.6

45

45

704.7

606.6

45

45

t18:1

22:0

896.7

636.6

45

45

734.7

636.6

45

45

t18:1

22:1

894.7

634.6

45

45

732.7

634.6

45

45

t18:1

24:0

924.7

664.6

45

45

762.7

664.6

45

45

t18:1

24:1

922.7

662.6

45

45

760.7

662.6

45

45

t18:1

26:0

952.8

692.7

45

45

790.8

692.7

45

45

t18:1

26:1

950.8

690.7

45

45

788.8

690.7

45

45

d18:0

16:0

798.6

538.5

45

45

636.6

538.5

45

45

d18:0

18:0

826.7

566.6

45

45

664.7

566.6

45

45

d18:0

20:0

854.7

594.6

45

45

692.7

594.6

45

45

d18:0

20:1

852.7

592.6

45

45

690.7

592.6

45

45

d18:0

22:0

882.7

622.6

45

45

720.7

622.6

45

45

d18:0

22:1

880.7

620.6

45

45

718.7

620.6

45

45

d18:0

24:0

910.7

650.6

45

45

748.7

650.6

45

45

d18:0

24:1

908.7

648.6

45

45

746.7

648.6

45

45 (continued)

Profiling of Plant Sphingolipids

171

Table 6 (continued) IPC LCB

FA

d18:0

26:0

d18:0

Q1 (m/z)

GlcA-IPC Q3 (m/z)

DP (V)

CE (V)

938.8

678.7

45

45

26:1

936.8

676.7

45

d18:1

16:0

796.6

536.5

d18:1

18:0

824.7

d18:1

20:0

d18:1

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (V)

776.8

678.7

45

45

45

774.8

676.7

45

45

45

45

634.6

536.5

45

45

564.6

45

45

662.7

564.6

45

45

852.7

592.6

45

45

690.7

592.6

45

45

20:1

850.7

590.6

45

45

688.7

590.6

45

45

d18:1

22:0

880.7

620.6

45

45

718.7

620.6

45

45

d18:1

22:1

878.7

618.6

45

45

716.7

618.6

45

45

d18:1

24:0

908.7

648.6

45

45

746.7

648.6

45

45

d18:1

24:1

906.7

646.6

45

45

744.7

646.6

45

45

d18:1

26:0

936.8

676.7

45

45

774.8

676.7

45

45

d18:1

26:1

934.8

674.7

45

45

772.8

674.7

45

45

Inositol is detected by a neutral loss of 180 m/z, and glucuronic acid (GlcA) is detected by a neutral loss of 176 m/z. The LCB and hydroxyl-fatty acid (FA) pairings are shown on the left followed by the parent (Q1) and product ion (Q3) m/z, the declustering potential (DP), and collision energy (CE) for each compound

3.3 Updates for Profiling of Aberrant Sphingolipids 3.3.1 Variant Products of Serine Palmitoyltransferase Activity

The standard lipid extraction protocol is used to extract DoxSA (m18:0) LCB, DoxSO (m18:1) LCB and the Cer1-deoxysphinganine. m18:0 and m18:1 LCBs can be profiled with other LCBs by incorporating MRMs for their detection as shown in Table 3. Gradient conditions for LCB separation are shown in Table 2. m18:0 elutes after d18:0, and m18:1 elutes after d18:1. m18:0 and m18:1 standards are available from Avanti Polar Lipids and are needed to verify the retention times of these compounds. Loss of H2O in the mass spectrometer ion source causes a d18:0 LCB to have the same MRM transition as m18:1 (302.3–18 ¼ 284.3), and the d17:1 LCB internal standard has the same MRM as m18:0 (286.3/268.3). There are multiple peaks produced by these MRM transitions and therefore the m18:0 and m18:1 LCB standards are essential for verification of the proper peak for quantification. 1-DeoxyCer are profiled alongside typical Cer by the addition of MRM transitions shown in Table 4, using the standard lipid extraction method, and the UPLC gradient for Cer and 1-deoxyCer shown in Table 2. Quantification is done by comparing 1-deoxyceramides to the d18:1/12:0 Cer internal standard and incorporating the response factors shown in Table 8. In this method, Cer have the same fatty acid but a different LCB will

172

Rebecca E. Cahoon et al.

Table 7 MRM parameters for Hexose-GlcA-IPC species commonly found in plants Hex-GlcA-IPC

GlcN-GlcA-IPC

GlcNac-GlcA-IPC

Q1 (m/ Q3 (m/ DP z) z) (V)

CE (V)

Q1 (m/ Q3 (m/ DP z) z) (V)

CE (V)

Q1 (m/ Q3 (m/ DP z) z) (V)

CE (V)

1546.9 366.3

50

46

1546.9 366.3

50

46

1546.9 366.3

50

46

t18:0 16:0 1152.6 554.5

45

45

1151.6 554.5

45

45

1193.6 554.5

45

45

t18:0 18:0 1180.6 582.5

45

45

1179.6 582.5

45

45

1221.6 582.5

45

45

t18:0 20:0 1208.7 610.6

45

45

1207.7 610.6

45

45

1249.7 610.6

45

45

t18:0 20:1 1206.7 608.6

45

45

1205.7 608.6

45

45

1247.7 608.6

45

45

t18:0 22:0 1236.7 638.6

45

45

1235.7 638.6

45

45

1277.7 638.6

45

45

t18:0 22:1 1234.7 636.6

45

45

1233.7 636.6

45

45

1275.7 636.6

45

45

t18:0 24:0 1264.7 666.6

45

45

1263.7 666.6

45

45

1305.7 666.6

45

45

t18:0 24:1 1262.7 664.6

45

45

1261.7 664.6

45

45

1303.7 664.6

45

45

t18:0 26:0 1292.8 694.7

45

45

1291.8 694.7

45

45

1333.8 694.7

45

45

t18:0 26:1 1290.8 692.7

45

45

1289.8 692.7

45

45

1331.8 692.7

45

45

t18:1 16:0 1150.6 552.5

45

45

1149.6 552.5

45

45

1191.6 552.5

45

45

t18:1 18:0 1178.6 580.5

45

45

1177.6 580.5

45

45

1219.6 580.5

45

45

t18:1 20:0 1206.7 608.6

45

45

1205.7 608.6

45

45

1247.7 608.6

45

45

t18:1 20:1 1204.7 606.6

45

45

1203.7 606.6

45

45

1245.7 606.6

45

45

t18:1 22:0 1234.7 636.6

45

45

1233.7 636.6

45

45

1275.7 636.6

45

45

t18:1 22:1 1232.7 634.6

45

45

1231.7 634.6

45

45

1273.7 634.6

45

45

t18:1 24:0 1262.7 664.6

45

45

1261.7 664.6

45

45

1303.7 664.6

45

45

t18:1 24:1 1260.7 662.6

45

45

1259.7 662.6

45

45

1301.7 662.6

45

45

t18:1 26:0 1290.8 692.7

45

45

1289.8 692.7

45

45

1331.8 692.7

45

45

t18:1 26:1 1288.8 690.7

45

45

1287.8 690.7

45

45

1329.8 690.7

45

45

d18:0 16:0 1136.6 538.5

45

45

1135.6 538.5

45

45

1177.6 538.5

45

45

d18:0 18:0 1164.7 566.6

45

45

1163.7 566.6

45

45

1205.7 566.6

45

45

d18:0 20:0 1192.7 594.6

45

45

1191.7 594.6

45

45

1233.7 594.6

45

45

d18:0 20:1 1190.7 592.6

45

45

1189.7 592.6

45

45

1231.7 592.6

45

45

d18:0 22:0 1220.7 622.6

45

45

1219.7 622.6

45

45

1261.7 622.6

45

45

d18:0 22:1 1218.7 620.6

45

45

1217.7 620.6

45

45

1259.7 620.6

45

45

d18:0 24:0 1248.7 650.6

45

45

1247.7 650.6

45

45

1289.7 650.6

45

45

d18:0 24:1 1246.7 648.6

45

45

1245.7 648.6

45

45

1287.7 648.6

45

45

LCB

FA

Standard

(continued)

Profiling of Plant Sphingolipids

173

Table 7 (continued) Hex-GlcA-IPC LCB

FA

GlcN-GlcA-IPC

Q1 (m/ Q3 (m/ DP z) z) (V)

GlcNac-GlcA-IPC

CE (V)

Q1 (m/ Q3 (m/ DP z) z) (V)

CE (V)

Q1 (m/ Q3 (m/ DP z) z) (V)

CE (V)

d18:0 26:0 1276.8 678.7

45

45

1275.8 678.7

45

45

1317.8 678.7

45

45

d18:0 26:1 1274.8 676.7

45

45

1273.8 676.7

45

45

1315.8 676.7

45

45

d18:1 16:0 1134.6 536.5

45

45

1133.6 536.5

45

45

1175.6 536.5

45

45

d18:1 18:0 1162.7 564.6

45

45

1161.7 564.6

45

45

1203.7 564.6

45

45

d18:1 20:0 1190.7 592.6

45

45

1189.7 592.6

45

45

1231.7 592.6

45

45

d18:1 20:1 1188.7 590.6

45

45

1187.7 590.6

45

45

1229.7 590.6

45

45

d18:1 22:0 1218.7 620.6

45

45

1217.7 620.6

45

45

1259.7 620.6

45

45

d18:1 22:1 1216.7 618.6

45

45

1215.7 618.6

45

45

1257.7 618.6

45

45

d18:1 24:0 1246.7 648.6

45

45

1245.7 648.6

45

45

1287.7 648.6

45

45

d18:1 24:1 1244.7 646.6

45

45

1243.7 646.6

45

45

1285.7 646.6

45

45

d18:1 26:0 1274.8 676.7

45

45

1273.8 676.7

45

45

1315.8 676.7

45

45

d18:1 26:1 1272.8 674.7

45

45

1271.8 674.7

45

45

1313.8 674.7

45

45

The first hexose sugar can be (1) Hex detected by neutral loss (NL) of 162 m/z, (2) glucosamine (GlcN-GlcA-IPC) detected by NL 161 m/z, or (3) N-acetylglucosamine (GlcNac-GlcA-IPC), detected by NL 203 m/z. Additional sugar units that are present on these backbones are detected by NL of 162 for hexose sugars and 132 m/z for pentose sugars. MRMs for the more complex GIPCs are calculated by adding 162 (hexose) m/z or 132 m/z (pentose) units to the Q1 mass. The LCB and hydroxyl-fatty acid (FA) pairings are shown on the left followed by the parent M+H ion (Q1) and product ion (Q3) m/z, the declustering potential (DP), and collision energy (CE) for each compound. Precursor ion scanning using the Q3 mass of the most abundant ceramide class can reveal glycosylation patterns to aid in MRM design

Table 8 Response factor for each chemical structure compared to internal standard Class/LCB

d18:0 d18:1 t18:0 t18:1 d20:0 d20:1 t20:0 t20:1 m18:0 m18:1 Glc-d18:1

Ceramides

3

4

6

5

3

4

6

5

3

3

–

Hydroxyceramides 3

4

6

5

3

4

6

5

–

–

–

Glucosylceramides 4

6

4

3

4

6

4

3

–

–

–

GIPCs

0.16

0.45

0.12 0.08 0.16

0.45

0.12 0.08 –

–

–

LCBs

2

1

5

4

2

1

5

4

1

1

1

LCB(P)s

1.8

1

1

1

1.8

1

1

1

1

1

–

Response factors for d18:0, d18:1, t18:0, and t18:1 were calculated by comparing the standard curve for the internal standard with that for purified Arabidopsis sphingolipids [14]. Response factors for C20 LCB were estimated based on similarity to C18 forms with the same degree of hydroxylation and unsaturation. 1-deoxy LCBs were estimated based on comparison to a standard curve

174

Rebecca E. Cahoon et al.

Fig. 2 Elution pattern of sphingolipids is based on fatty acid chain length and LCB. The elution pattern of Cer with 16 carbon fatty acids is shown. For Cer backbones with the same fatty acid, peaks elute as follows: IS (d18:1/12:0 Cer internal standard), (1) t18:1/16:0 Cer, (2) t18:0/16:0 Cer, (3) d18:1/16:0 Cer, (4) d18:0/16:0 Cer, (5) m18:1–16:0 Cer, (6) m18:0/16:0 Cer. Asterisk denotes m18:1/16:0 Cer MRM detected by loss of water from the d18:0/16:0Cer peak (depicted with a dashed line). Peaks in red are derived from aberrant deoxy-ceramides. This elution pattern based on the LCB component where the fatty acid component is constant, is observed for every complex sphingolipid class

elute with regular, predictable intervals as shown in Fig. 2. For example, m18:0/16:0 ceramides will elute after d18:0/16:0 Cer, and m18:1/FACer after d18:1/FA ceramides. Comparison to a wild-type sample lacking 1-deoxyCer will help to verify authentic 1-deoxyCer peaks. A small proportion of d18:0 Cer will lose a water molecule in the ion source producing masses that have the same MRM as m18:1 Cer, so careful verification of the peaks is needed, and m18:1Cer peaks that have the same retention time as the d18:0 Cer are excluded as they are formed from d18:0. Some m18:1 Cer standards are available from Avanti Polar Lipids (e.g., m18:1/16:0 Cer and m18:1/24:0 Cer). Because standards are not available for most 1-deoxyCer species, comparison of retention times and exclusion of peaks that come from water loss by other Cer species is essential for correct peak verification. 3.3.2 Glucosylceramides with Nonhydroxylated Fatty Acids

GlcCer and nonhydroxylated GlcCer(nhGlcCer) can be profiled together in the same injection or separately using the standard lipid extraction method, the UPLC gradient for GlcCers shown in Table 2, and MRM transitions shown in Table 5. Quantification is done by comparison to the d18:1/12:0 GlcCer internal standard. nhGlcCer elute slightly later than the corresponding GlcCer. MRM transitions for nhGlcCer and GlcCer are very specific, so that there are no confusing peaks that arise from source-induced transitions.

Profiling of Plant Sphingolipids 3.3.3 Variations in GIPC Classes

4

175

GIPCs, glucuronosylinositolphosphoceramides (GlcA-IPCs) and IPCs can be profiled together in the same injection or separately using the standard lipid extraction method, the UPLC gradient for GIPCs shown in Table 2, and MRM transitions shown in Tables 6 and 7. Quantification is done by comparison to the GM1 internal standard. Gradient conditions should be optimized to ensure clear separation of species. More highly glycosylated species will elute first, followed by GlcA-IPC and then IPC. Precursor ion scanning based on the ceramide (Q3) fragment of the most abundant GIPC type is useful in establishing gradient conditions that will separate these closely related species. In A. thaliana, precursor ion scanning using the mass 664.6 for the Q3 fragment from abundant t18:1– h24:0 GIPCs is useful. Clear establishment of retention times is essential for proper peak validation, since IPC and GlcA-IPC fragments can be generated from more highly glycosylated GIPCs by ion source-induced fragmentation. Precursor ion scanning using the Q3 fragment of the most abundant GIPC is useful in establishing gradient conditions that can separate (1) Hex-GlcA-IPC and N-acetylglucosamine-glucuronosyl-inositolphosphoceramide (GlcNac-GlcA-IPC), (2) glucosamine-glucuronosyl-inositolphosphoceramide (GlcN-GlcA-IPC), and (3) GlcA-IPC and (4) IPC species which elute closely together and in that order but which can be resolved into distinct peaks. Hex-GlcA-IPC and GlcNac-GlcAIPCs are difficult to resolve but can be distinguished based on their unique MRMs.

Notes 1. Any mass-spectrometer capable of MS/MS can be used but the triple-quadrupole arrangement makes for the best sensitivity and flexibility. Triple quadrupole machines from Shimadzu, Agilent, and Waters among others have all been used for this purpose. Models and makes differ in their sensitivity and requirements so methods may need to be adapted as necessary to different mass spectrometers. 2. Due to the low volume of UPLC systems they are much more sensitive to features that cause band-broadening. Inactivation of stainless-steel surfaces in UPLC machines with 1% phosphoric acid is recommended to reduce excessive peak widths of phosphorylated sphingolipids. 3. Other reverse phase columns that have been used successfully include Kinetex C18 (Phenomenex, Torrance, CA), Acclaim Polar Advantage (ThermoFisher Scientific, Waltham, MA), XSelect CSH-C18 (Waters Corp., Milford, MA), and Accucore C30 (ThermoFisher Scientific).

176

Rebecca E. Cahoon et al.

4. The final concentration of ammonium formate in SOLVENT B affects the separation and detection of the phosphorylated sphingolipids and should be optimized for your system. The 0.5 mM final concentration suggested here is a starting point. Concentrations as high as 5 mM may be optimal for the separation and detection of GIPC. 5. Samples should be dissolved and analyzed within 24 h to prevent sample aggregation and precipitation. Older samples may need warming and /or gentle sonication to ensure all lipids are properly dissolved. 6. To keep the number of MRM transitions manageable for each method, tailor the MRM transitions to the species of interest. For example, Cer and 1-deoxyCer can be conveniently measured in one method, and hCer, which uses the same chromatographic gradient but different MRM parameters, can be measured using a different method. Where retention time differences along the same gradient are needed to confirm peaks, as for GlcCer and GlcCer with nonhydroxylated fatty acids (nhGlcCer), MRMs for the major GlcCer species can be combined in one method with appropriate adjustments of dwell times.

Acknowledgments This work was supported by a National Science Foundation grant MCB 1818297 to E.B.C. and J.E.M. A.G.S. acknowledges funding from the Mexican National Council of Science and Technology (CONACyT). References 1. Huby E, Napier JA, Baillieul F, Michaelson LV, Dhondt-Cordelier S (2020) Sphingolipids: towards an integrated view of metabolism during the plant stress response. New Phytol 225:659–670 2. Liang H, Yao N, Song JT, Luo S, Lu H, Greenberg JT (2003) Ceramides modulate programmed cell death in plants. Genes Dev 17:2636–2641 3. Magnin-Robert M, Le Bourse D, Markham J, Dorey S, Clement C, Baillieul F et al (2015) Modifications of sphingolipid content affect tolerance to hemibiotrophic and necrotrophic pathogens by modulating plant defense responses in Arabidopsis. Plant Physiol 169:2255–2274 4. Cacas JL, Bure C, Grosjean K, Gerbeau-PissotP, Lherminier J, Rombouts Y et al (2016)

Revisiting plant plasma membrane lipids in tobacco: a focus on sphingolipids. Plant Physiol 170:367–384 5. Gronnier J, Germain V, Gouguet P, Cacas JL, Mongrand S (2016) GIPC: glycosyl inositol phospho ceramides, the major sphingolipids on earth. Plant Signal Behav 11:e1152438 6. Lenarcic T, Albert I, Bohm H, Hodnik V, Pirc K, Zavec AB et al (2017) Eudicot plantspecific sphingolipids determine host selectivity of microbial NLP cytolysins. Science 358:1431–1434 7. Mortimer JC, Scheller HV (2020) Synthesis and function of complex sphingolipid glycosylation. Trends Plant Sci 25:522–524 8. Chen M, Cahoon EB, Saucedo-Garcı´a M, Plasencia J, Gavilanes-Ruı´z M (2009) Plant sphingolipids: structure, synthesis and

Profiling of Plant Sphingolipids function. In: Wada H, Murata N (eds) Lipids in photosynthesis: essential and regulatory functions. Springer, Dordrecht, pp 77–115 9. Luttgeharm KD, Cahoon EB, Markham JE (2016) Substrate specificity, kinetic properties and inhibition by fumonisin B1 of ceramide synthase isoforms from Arabidopsis. Biochem J 473:593–603 10. Ternes P, Feussner K, Werner S, Lerche J, Iven T, Heilmann I et al (2011) Disruption of the ceramide synthase LOH1 causes spontaneous cell death in Arabidopsis thaliana. New Phytol 192:841–854 11. Alden KP, Dhondt-Cordelier S, McDonald KL, Reape TJ, Ng CK, McCabe PF et al (2011) Sphingolipid long chain base phosphates can regulate apoptotic-like programmed cell death in plants. Biochem Biophys Res Commun 410:574–580 12. Luttgeharm KD, Kimberlin AN, Cahoon EB (2016) Plant sphingolipid metabolism and function. Subcell Biochem 86:249–286 13. Markham JE (2013) Detection and quantification of plant sphingolipids by LC-MS. Methods Mol Biol 1009:93–101 14. Markham JE, Jaworski JG (2007) Rapid measurement of sphingolipids from Arabidopsis thaliana by reversed-phase high-performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 21:1304–1314 15. Chen M, Han G, Dietrich CR, Dunn TM, Cahoon EB (2006) The essential nature of sphingolipids in plants as revealed by the functional identification and characterization of the Arabidopsis LCB1 subunit of serine palmitoyltransferase. Plant Cell 18:3576–3593 16. Dietrich CR, Han G, Chen M, Berg RH, Dunn TM, Cahoon EB (2008) Loss-of-function mutations and inducible RNAi suppression of Arabidopsis LCB2 genes reveal the critical role of sphingolipids in gametophytic and sporophytic cell viability. Plant J 54:284–298 17. Gonzalez Solis A, Han G, Gan L, Liu Y, Markham JE, Cahoon RE et al (2020) Unregulated sphingolipid biosynthesis in gene-edited Arabidopsis ORM mutants results in nonviable seeds with strongly reduced oil content. Plant Cell. https://doi.org/10.1105/tpc.20.00015 18. Kimberlin AN, Han G, Luttgeharm KD, Chen M, Cahoon RE, Stone JM et al (2016) ORM expression alters sphingolipid homeostasis and differentially affects ceramide synthase activity. Plant Physiol 172:889–900 19. Kimberlin AN, Majumder S, Han G, Chen M, Cahoon RE, Stone JM et al (2013) Arabidopsis 56-amino acid serine palmitoyltransferaseinteracting proteins stimulate sphingolipid synthesis, are essential, and affect mycotoxin sensitivity. Plant Cell 25:4627–4639

177

20. Gable K, Slife H, Bacikova D, Monaghan E, Dunn TM (2000) Tsc3p is an 80-amino acid protein associated with serine palmitoyltransferase and required for optimal enzyme activity. J Biol Chem 275:7597–7603 21. Breslow DK, Collins SR, Bodenmiller B, Aebersold R, Simons K, Shevchenko A et al (2010) Orm family proteins mediate sphingolipid homeostasis. Nature 463:1048–1053 22. Li J, Yin J, Rong C, Li KE, Wu JX, Huang LQ et al (2016) Orosomucoid proteins interact with the small subunit of serine palmitoyltransferase and contribute to sphingolipid homeostasis and stress responses in Arabidopsis. Plant Cell 28:3038–3051 23. Bejaoui K, Uchida Y, Yasuda S, Ho M, Nishijima M, Brown RH Jr et al (2002) Hereditary sensory neuropathy type 1 mutations confer dominant negative effects on serine palmitoyltransferase, critical for sphingolipid synthesis. J Clin Invest 110:1301–1308 24. Gable K, Gupta SD, Han G, Niranjanakumari S, Harmon JM, Dunn TM (2010) A disease-causing mutation in the active site of serine palmitoyltransferase causes catalytic promiscuity. J Biol Chem 285:22846–22852 25. McCampbell A, Truong D, Broom DC, Allchorne A, Gable K, Cutler RG et al (2005) Mutant SPTLC1 dominantly inhibits serine palmitoyltransferase activity in vivo and confers an age-dependent neuropathy. Hum Mol Genet 14:3507–3521 26. Konig S, Feussner K, Schwarz M, Kaever A, Iven T, Landesfeind M et al (2012) Arabidopsis mutants of sphingolipid fatty acid α-hydroxylases accumulate ceramides and salicylates. New Phytol 196:1086–1097 27. Msanne J, Chen M, Luttgeharm KD, Bradley AM, Mays ES, Paper JM et al (2015) Glucosylceramides are critical for cell-type differentiation and organogenesis, but not for cell viability in Arabidopsis. Plant J 84:188–201 28. Cacas JL, Bure C, Furt F, Maalouf JP, Badoc A, Cluzet S et al (2013) Biochemical survey of the polar head of plant glycosylinositolphosphoceramides unravels broad diversity. Phytochemistry 96:191–200 29. Luttgeharm KD, Kimberlin AN, Cahoon RE, Cerny RL, Napier JA, Markham JE et al (2015) Sphingolipid metabolism is strikingly different between pollen and leaf in Arabidopsis as revealed by compositional and gene expression profiling. Phytochemistry 115:121–129 30. Rennie EA, Ebert B, Miles GP, Cahoon RE, Christiansen KM, Stonebloom S et al (2014) Identification of a sphingolipid α-glucuronosyltransferase that is essential for pollen function in Arabidopsis. Plant Cell 26:3314–3325

Chapter 11 Analysis of Free and Esterified Sterol Content and Composition in Seeds Using GC and ESI-MS/MS Richard Broughton and Fre´de´ric Beaudoin Abstract Total sterol content and composition in plant tissues can be easily determined by gas chromatography (GC) after saponification of the total lipid extract. However, in oleogenic tissues a significant proportion of the sterol is esterified to fatty acids, with GC methodologies unable to provide information about the proportion and the molecular species composition of intact steryl esters (SEs). Here we describe an electrospray ionization-tandem mass spectrometry (ESI-MS/MS) and Multiple Reaction Monitoring (MRM) method which, in parallel with GC analysis, allows for the accurate determination of both free and esterified sterol content and composition in seeds. After extraction of seed oil with hexane, free sterols are derivatized with undecanoyl chloride, total steryl esters are then purified from triacylglycerol (TAG) by liquid chromatography, infused and ionized as ammonium adducts, with molecular species identified and quantified by fragmentation in the presence of internal standards. Key words Free sterol, Steryl ester, Saponification, Silylation, Undecanoyl chloride, GC-FID, HPLC, ESI-MS/MS, MRM

1

Introduction Generating crops with a metabolite content and composition better suited to specific industrial applications can be achieved by conventional breeding. This requires the development of high-throughput metabolic profiling procedures allowing for the screening of hundreds of plants to characterize the traits of interest and, if sufficient biochemical and genetic variation exists within natural populations or diversity collections, the identification of markers to assist molecular breeding. Phytosterols are one such group of natural compounds that face increased industrial demand, partly due to their ability to lower blood cholesterol levels [1]. Phytosterols occur naturally as a minor component of seed and vegetable oils in the form of free sterols or stanols (hydrogenated/saturated forms of sterols) and their fatty acyl ester derivatives. The steryl

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_11, © Springer Science+Business Media, LLC, part of Springer Nature 2021

179

180

Richard Broughton and Fre´de´ric Beaudoin

esters (SE) are the form predominantly used for functional food applications because of their good solubility in oil and fat [2]. Analysis of sterol composition in plant and animal tissue is commonly carried out after lipid hydrolysis and silylation of free sterols using gas chromatography and a flame ionization detector (GC-FID). This is an accurate and reproducible method for the quantitative and qualitative determination of total sterol composition in biological samples [3] and quantification of sterols is relatively straightforward, with response factors for the phytosterols close to one [4]. However, for this reason structural information about steryl esters and their molecular species composition is often missing. Soft ionization techniques such as ESI can suffer from variable ionization, with more abundant matrix compounds and those capable of ionizing efficiently obscuring low abundant, neutral compounds. This can be mitigated by incorporating hyphenated liquid chromatography (LC) methods, but this may require a more complex instrumental and software setup. Whilst the chromatographic separation of all compounds is not a prerequisite for LC-MS/MS, data acquisition and scan speed is predicated on the scan type (e.g., MRM) and the duty cycle, with narrow peaks containing multiple compounds potentially limiting the number of data points acquired across the peak. Direct infusion methods however, do not suffer due to complex duty cycles, as samples can be infused for as long as there is sample available, allowing for multiple scan types to be acquired per sample, and thus for more accurate quantification. Although ion suppression is more problematic during sample introduction via direct infusion, due to the more complex matrix, LC also suffers from intra-peak suppression effects [5], many of which may not be compensated due to internal standards eluting elsewhere in the chromatogram. Absolute quantification utilizing LC-MS has been found to add additional complications when compared to standard HPLC and GC methods. For reliable quantification, matrix effects and ion suppression need to be taken into account. An improvement to the single internal standard approach is the bracketing method [6], which is generally more amenable to infusion methodologies, and can be used effectively if compounds of a single class elute in one peak, as it is the case with both hydrophilic interaction liquid chromatography (HILIC) and supercritical fluid chromatography (SFC). For each class multiple standards are added, in effect bracketing the compounds of interest, which takes into account the detector response resultant from the acyl chain length, as well as the level of acyl unsaturation. This means that each sample contains a calibration curve for the classes of interest, allowing for an accurate determination of the absolute amount.

Sterol and Steryl Ester Analysis in Seeds

181

Fig. 1 Flow chart for seed oil sterol quantification. The left part shows the derivatization of the free sterols (FS) with undecanoyl chloride, purification of steryl esters by HPLC, and quantification of steryl esters by ESI-MS/ MS (QTRAP). The right part depicts the total sterol measurement after saponification and silylation with BSTFA by GC-FID

The method described here includes the extraction of free and esterified phytosterol from seed oil, and involves the samples being split between both GC-FID/MS and ESI-MS/MS analyses because isomers such as stigmasterol and Δ-5 avenasterol, or β-sitosterol and Δ-7 stigmastenol cannot be distinguished by ESI-MS/MS (Fig. 1). These isomers, however, can be resolved using GC. To reduce sample complexity for ESI-MS/MS, semipreparative HPLC with a fraction collector is used to isolate total phytosterol fractions, prior to direct infusion. Another consideration for ESI-MS/MS is that free sterols are known to ionize poorly, and therefore require derivatization to improve their ionization efficiency [7]. The use of an acyl chloride as a derivatization reagent allows for the simplification of the analytical method by converting free sterols to non–naturally occurring steryl esters. Undecanoyl chloride was chosen for derivatization of free sterols, due to the esters formed being similar in size, and therefore

182

Richard Broughton and Fre´de´ric Beaudoin

displaying similar ionization performance, to naturally occurring compounds. The conversion of free sterols to steryl esters also allows for a single preparative HPLC peak to be collected, in addition to a single MRM method utilizing similar settings and fragmentation characteristics to be created, which simplifies both data acquisition and compound optimization.

2

Materials

2.1 Seed Oil Extraction

1. Precellys 24 tissue homogenizer (Bertin Instruments, France). 2. Precellys 2 ml tubes with 2.8 mm ceramic beads (CK28; Bertin Instruments, France) (see Note 1). 3. Benchtop centrifuge. 4. Hexane solution: Hexane containing 50 μM of each cholestanol and cholesteryl heptadecanoate (17:0-cholesterol; Avanti Polar Lipids, USA). 5. Hexane (n-hexane, HPLC grade). 6. 3.5 ml screw top glass vials (e.g., Trident, SLS UK). 7. Tube rotator or rotary incubator. 8. Centrifugal evaporator (e.g., Genevac, UK).

2.2 Saponification of Lipid Extracts and Sterol Silylation for GC Analysis

1. Ethanol (absolute, >99%). 2. Saponification solution: 1.5 M potassium hydroxide in 95% ethanol. This solution can be prepared by dissolving 8.416 g KOH in 95% (v/v) ethanol in water solution (i.e., 95 ml ethanol plus 5 ml water) and adjusting the final volume to 100 ml (see Note 2). 3. Sodium chloride solution: 1% (w/v) sodium chloride in water. 4. Hexane (n-hexane, HPLC grade). 5. BSTFA (N,O-bis(trimethylsilyl)trifluoroacetamide) + 1% TMCS (trimethylchlorosilane): BSTFA/TMCS 99:1, v/v; Merck, UK). 6. HP-1 MS UI column (30 m
0.25 mm
0.25 μM) (Agilent, USA) (see Note 3). 7. GC-FID (6890, Agilent) and/or GC-MS (6890 N/5975B, Agilent).

2.3 Free Phytosterol Derivatization with Undecanoyl Chloride

1. Oven for drying glassware. 2. Anhydrous dichloromethane (see Notes 4 and 5). 3. Prepare 4 mM stocks of the molecular standards in anhydrous dichloromethane.

Sterol and Steryl Ester Analysis in Seeds

183

4. Internal standard mixture containing 23.4 μM of each cholesteryl ester (16:0, 18:0, 18:1, 18:2, 22:0, 22:1) and cholesterol (Merck) in anhydrous dichloromethane. To prepare the 23.4 μM mixture add 585 μl of each standard in a dry glass bottle and make the volume up to 100 ml with anhydrous dichloromethane. With 1 ml added to each sample (Subheading 3.3, step 2) this equates to 5 pmol/μl of each standard in the final volume of 300 μl prior to direct infusion (Subheading 3.5) if a split ratio of 4:1 is used during preparative HPLC (Subheading 3.4). 5. External standard mixture containing 25 μM free sterols (β-sitosterol, stigmasterol, cholesterol) (Merck, UK) and cholestanol (Avanti Polar Lipids, USA) in anhydrous dichloromethane. This is made in bulk; to prepare the 25 μM mixture add 62.5 μl of each standard in a dry glass vial and make the volume up to 10 ml with anhydrous dichloromethane. With 1 ml derivatised (Subheading 3.3, step 2), it equates to 5 pmol/μl (5 μM) of each standard in the final volume of 5 ml infusion buffer (Subheading 3.5). 6. Undecanoyl chloride (Tokyo Chemical Industries, UK). 7. Chloroform/methanol solution (1:1, v/v; Fisher MS grade). 2.4 Semi-preparative Purification of Steryl Esters

1. Solvent A: 100% heptane HPLC grade. 2. Solvent B: 100% ethyl acetate HPLC grade. 3. Lichrospher Si60-5 column (250 mm
7.75 mm) (Hichrom, UK). 4. HPLC: Agilent 1260 infinity with evaporative light scattering detector (ELSD) and fraction collector.

2.5 Direct Infusion ESI-MS/MS of Steryl Ester Fraction

1. Infusion buffer: 24:24:2 v/v/v chloroform–methanol–aqueous 250 mM ammonium acetate buffer. Use MS grade solvents. 2. QTRAP 4000 triple quadrupole ESI-MS/MS (Sciex, USA). 3. Autosampler PAL-HTS-xt (CTC, Switzerland). 4. LipidView and Analyst software (Sciex, USA).

3

Methods

3.1 Seed Oil Extraction

1. Weigh 80 mg of seed material into grinding tubes. 2. Seeds are ground with ceramic beads (already present in CK28 tubes) in the Precellys homogenizer at 5500 rpm for three times of 25 s each, with a 30 s break between each grinding step.

184

Richard Broughton and Fre´de´ric Beaudoin

3. Add 1 ml of the hexane solution containing the internal standards (50 μM cholesteryl heptadecanoate (17:0-cholesterol) and cholestanol) to each sample. 4. Vortex samples and incubate at 30  C in a rotary incubator for 1 h. 5. Centrifuge at 20,000
g for 1 min to sediment seed material and transfer the hexane supernatant to 3.5 ml glass vials using glass Pasteur pipettes. 6. Repeat the extraction twice with 1 ml of pure hexane omitting the incubation step (steps 4 and 5) and pool the three hexane extracts in 3.5 ml glass vials. 7. Dry the lipid extracts using a centrifugal evaporator keeping the temperature below 40  C, or alternatively under a stream of nitrogen (see Note 6). 8. Resuspend lipids in 1 ml of hexane. Samples can be stored at 20  C for several days if necessary. 3.2 Saponification of Lipid Extracts and Sterol Silylation for GC-FID/MS Analysis

1. Take a 250 μl aliquot of the lipid extract (see step 8, in Subheading 3.1), place it into a screw top glass tube and dry as described above (Fig. 1) (see step 7, in Subheading 3.1). 2. Add 1.5 ml of 1.5 M potassium hydroxide in 95% (v/v) ethanol, close the tubes tightly with screw caps and incubate in a heating block or water bath at 80  C for 1 h. 3. Allow the samples to cool and add 1 ml of a 1% (w/v) NaCl solution, followed by 1 ml of hexane. 4. Vortex and centrifuge at 1000
g for 1 min to accelerate phase partitioning. Transfer the upper hexane phase to a glass vial (see Note 7). 5. Repeat the extraction with 1 ml of hexane and pool both extracts. 6. Dry samples as described above (see step 7, Subheading 3.1). 7. Sterols are derivatized by adding 100 μl of BSTFA–TMCS (99:1, v/v) and incubated for 1 h at 85  C. 8. Once samples are cool, add 100 μl of heptane and inject 1 μl for GC-FID and/or GC-MS analysis as described in (Fig. 2) [8]. Typically, GC settings are the same for both GC-FID and GC-MS with splitless injection, helium used as the carrier gas at 2 ml/min and the inlet temperature set to 325  C. Set the oven initial temperature to 200  C, apply a ramp of 6.5  C/min until 325  C and hold for 4 min. Set the FID and/or the MSD transfer line temperature to 325  C, the source and quadrupole to 230 and 150  C, respectively, and the scan range between 42 and 520 m/z.

Sterol and Steryl Ester Analysis in Seeds

185

Fig. 2 GC chromatogram of total free sterols (after silylation) isolated from sunflower seed oil showing the 4-desmethylsterols. 1: campesterol; 2: stigmasterol; 3: Δ-7 ergostenol; 4: β-sitosterol; 5: Δ-5 avenasterol; 6: Δ-7stigmastenol; 7: Δ-7 avenasterol 3.3 Free Phytosterol Derivatization with Undecanoyl Chloride

Glassware in which the acylation takes place must be baked at 140  C for 2 h to remove water adsorbed onto the silica (see Note 5). 1. Place a 500 μl aliquot of the lipid extract (see step 8 in Subheading 3.1) into a baked 3.5 ml screw top glass vial and dry for 45 min at 45  C in a centrifugal evaporator to achieve complete dryness. 2. Add 1 ml of the internal standard mixture containing cholesterol and cholesteryl esters (see item 3 in Subheading 2.3) followed by 12 μl of undecanoyl chloride. The free sterol external standards also need to be derivatized. Place 1 ml of the external standard mixture (see item 4 in Subheading 2.3) into a baked 3.5 ml screw top glass vial and add 12 μl of undecanoyl chloride. 3. Close vials tightly with screw caps, vortex and incubate all samples overnight at 42  C. 4. Following incubation, add 1 ml of chloroform–methanol (1:1, v/v) and 500 μl of water to phase partition. The addition of water also quenches the reaction (see Note 4). 5. Centrifuge at 1000
g for 1 min and transfer the lower chloroform/dichloromethane phase to new 3.5 ml screw top glass vials. 6. Repeat the extraction by adding 1 ml of chloroform–methanol (1:1, v/v) and 500 μl of water to the previous aqueous top phase. Centrifuge at 1000
g for 1 min, remove the bottom organic phase and pool it with the first extract. 7. Dry samples as described above (see step 7 in Subheading 3.1), close vials tightly with screw caps and store at 20  C.

186

Richard Broughton and Fre´de´ric Beaudoin

3.4 Semipreparative Purification of Steryl-Esters

This HPLC step allows for high-throughput separation of the steryl ester fraction from the much more abundant triacylglycerol (TAG) and was adapted from [9] (Fig. 3). If smaller numbers of samples are to be analyzed, this can also be achieved rapidly by solid phase extraction as described in [8]. The derivatized external standard mixture does not need purification and can be analyzed directly by ESI-MS/MS (Subheading 3.5). 1. Resuspend dry derivatized seed lipid samples (see step 7 in Subheading 3.3) in 500 μl heptane and centrifuge at 12,000
g for 1 min to sediment insoluble material. 2. Inject 40 μl onto the semi-preparative HPLC column with the following instrumental setup: A flow rate of 2.2 ml/min is used, with a post-column split set to divert 80% to the fraction collector and 20% to the ELSD, keeping post-split valve tubing to a minimum (see Note 8). An initial solvent composition of 97% A is used, held for 0.5 min, followed by a decrease in

Fig. 3 Preparative HPLC chromatogram illustrating the separation of steryl esters from triacylglycerols in C11-derivatised rapeseed oil (bottom panel). Dashed lines indicate the fraction collected. The top panel shows the separation of a standard mixture: SE: steryl ester; TAG: triacylglycerol; DAG: diacylglycerol; FS: free sterol

Sterol and Steryl Ester Analysis in Seeds

187

solvent A to 50% over 8.5 min, which is held for 1 min. Finally, solvent A is decreased to 20% over 3 min, followed by a return to the starting conditions over two min and held for 1.25 min. The column is kept at 25  C and the ELSD at 40  C, with a gain factor of 4. One sample selected as a quality control (QC) can be injected up to 12 times to produce enough steryl ester aliquots for multiple infusion in the ESI-MS/MS analytical sequence. 3. Under these conditions the steryl ester fraction is eluted from 4.3 to 6.8 min and collected using the fraction collector in 6 ml glass collection vials (Fig. 3). 4. Dry samples as described above (Subheading 3.1, step 7). 3.5 Direct Infusion ESI-MS/MS Analysis of Steryl Ester Fractions

1. Resuspend HPLC purified steryl ester samples (see step 4 in Subheading 3.4) in 300 μl of infusion buffer. Resuspend the derivatized external standard mixture in 5 ml of infusion buffer and aliquot it into 300 μl samples for multiple infusions. 2. Infuse samples into the mass spectrometer at a rate of 20 μl/ min using a PAL-HTS-xt autosampler. Flush tubing with infusion buffer after each sample and purge the line with sample for 2.5 min prior to data acquisition (see Note 9). 3. Run the MRM scan for 3.5 min, with an initial 15 s equilibration period prior to acquisition in positive mode, and average peak intensities over the entire data acquisition period. A list of targeted steryl esters with associated MRM transitions for rapeseed oil is given in Table 1. A 20 s negative mode run should follow each sample to maintain signal intensity and reduce quadrupole charging effects (see Note 10). 4. A probe voltage of 5500 V is used, with a nebulizing gas pressure of 20 psig, a curtain gas pressure of 15 psig and a collision cell pressure of 4 psig. The turbo gas pressure and temperature are set to “off,” and the channel electron multiplier (CEM) voltage to 2300 V. The signal intensities of compounds can be automatically improved using the Analyst software, which results in optimized declustering potential, entrance potential, collision energy and collision exit potential for individual MRM transitions (Table 1) (see Note 11). 5. Monitor sample ionization over time by running the following standard mixtures and control samples every 20–40 samples: (a) the cholesteryl ester internal standard mixture (see item 3 in Subheading 2.3) with the addition of cholesteryl heptadecanoate at 5 pmol/μl in infusion buffer, (b) the undecanoyl chloride–derivatized free sterol external standard mixture at a concentration of 5 pmol/μl (see item 4 in Subheading 2.3), (c) an infusion buffer blank and (d) a QC sample (see step 2 in Subheading 3.4) to measure variability in sample quantification.

188

Richard Broughton and Fre´de´ric Beaudoin

Table 1 MRM transitions used to profile steryl ester molecular species in rapeseed oil

Molecular ion [M + NH4]+, (m/z)

Product ion [M-(RCOOH+ NH3)]+ (m/z)

Steryl ester

Scan time (ms)

DP (V)

CE (V)

574.6

371.4

11:0-Cholestanol (IS)

50

51

13

572.5

369.4

11:0-Cholesterol (IS)

50

51

17

642.6

369.4

16:0-Cholesterol (IS)

50

61

17

656.6

369.4

17:0-Cholesterol (IS)

50

51

19

666.6

369.4

18:2-Cholesterol (IS)

50

101

21

668.6

369.4

18:1-Cholesterol (IS)

50

61

21

670.6

369.4

18:0-Cholesterol (IS)

50

66

21

724.7

369.4

22:1-Cholesterol (IS)

50

71

21

726.7

369.4

22:0-Cholesterol (IS)

50

71

19

587.0

383.4

11:0-Campesterol

50

51

15

650.6

383.4

16:3-Campesterol

50

71

19

652.6

383.4

16:2-Campesterol

50

71

19

654.6

383.4

16:1-Campesterol

50

71

19

656.6

383.4

16:0-Campesterol

50

71

19

678.6

383.4

18:3-Campesterol

50

176

21

680.6

383.4

18:2-Campesterol

50

76

21

682.6

383.4

18:1-Campesterol

50

101

21

684.7

383.4

18:0-Campesterol

50

101

21

706.6

383.4

20:3-Campesterol

50

71

21

708.7

383.4

20:2-Campesterol

50

71

21

710.7

383.4

20:1-Campesterol

50

71

21

712.7

383.4

20:0-Campesterol

50

71

19

734.5

383.4

22:3-Campesterol

50

71

21

736.7

383.4

22:2-Campesterol

50

71

21

738.7

383.4

22:1-Campesterol

50

71

21

740.7

383.4

22:0-Campesterol

50

71

19

766.7

383.4

24:1-Campesterol

50

71

21

768.8

383.4

24:0-Campesterol

50

71

19

600.6

397.4

11:0-Sitosterol

50

46

17

664.6

397.4

16:3-Sitosterol

50

71

19 (continued)

Sterol and Steryl Ester Analysis in Seeds

189

Table 1 (continued)

Molecular ion [M + NH4]+, (m/z)

Product ion [M-(RCOOH+ NH3)]+ (m/z)

Steryl ester

Scan time (ms)

DP (V)

CE (V)

666.6

397.4

16:2-Sitosterol

50

71

19

668.6

397.4

16:1-Sitosterol

50

71

19

670.6

397.4

16:0-Sitosterol

50

76

19

692.6

397.4

18:3-Sitosterol

50

166

21

694.6

397.4

18:2-Sitosterol

50

176

21

696.7

397.4

18:1-Sitosterol

50

181

21

698.7

397.4

18:0-Sitosterol

50

96

23

720.7

397.4

20:3-Sitosterol

50

71

21

722.7

397.4

20:2-Sitosterol

50

71

21

724.7

397.4

20:1-Sitosterol

50

71

21

726.7

397.4

20:0-Sitosterol

50

71

19

748.6

397.4

22:3-Sitosterol

50

71

21

750.7

397.4

22:2-Sitosterol

50

71

21

752.7

397.4

22:1-Sitosterol

50

71

21

754.7

397.4

22:0-Sitosterol

50

71

19

780.8

397.4

24:1-Sitosterol

50

71

21

782.8

397.4

24:0-Sitosterol

50

71

19

598.6

395.4

11:0-Stigmasterol

50

61

15

662.6

395.4

16:3-Stigmasterol

50

71

19

664.6

395.4

16:2-Stigmasterol

50

71

19

666.6

395.4

16:1-Stigmasterol

50

81

17

668.6

395.4

16:0-Stigmasterol

50

71

19

690.6

395.4

18:3-Stigmasterol

50

76

21

692.6

395.4

18:2-Stigmasterol

50

96

19

694.6

395.4

18:1-Stigmasterol

50

86

19

696.7

395.4

18:0-Stigmasterol

50

71

19

718.6

395.4

20:3-Stigmasterol

50

71

21

720.7

395.4

20:2-Stigmasterol

50

71

21

722.7

395.4

20:1-Stigmasterol

50

71

21

724.7

395.4

20:0-Stigmasterol

50

71

19 (continued)

190

Richard Broughton and Fre´de´ric Beaudoin

Table 1 (continued)

Molecular ion [M + NH4]+, (m/z)

Product ion [M-(RCOOH+ NH3)]+ (m/z)

Steryl ester

Scan time (ms)

DP (V)

CE (V)

746.5

395.4

22:3-Stigmasterol

50

71

21

748.7

395.4

22:2-Stigmasterol

50

71

21

750.7

395.4

22:1-Stigmasterol

50

71

21

752.7

395.4

22:0-Stigmasterol

50

71

19

778.7

395.4

24:1-Stigmasterol

50

71

21

780.8

395.4

24:0-Stigmasterol

50

71

19

584.5

381.4

11:0-Brassicasterol

50

61

15

654.6

381.4

16:0-Brassicasterol

50

71

17

652.6

381.4

16:1-Brassicasterol

50

71

19

650.6

381.4

16:2-Brassicasterol

50

71

19

648.6

381.4

16:3-Brassicasterol

50

71

19

682.6

381.4

18:0-Brassicasterol

50

91

17

680.6

381.4

18:1-Brassicasterol

50

86

17

678.6

381.4

18:2-Brassicasterol

50

176

19

676.6

381.4

18:3-Brassicasterol

50

96

19

710.7

381.4

20:0-Brassicasterol

50

71

17

708.7

381.4

20:1-Brassicasterol

50

71

19

706.6

381.4

20:2-Brassicasterol

50

71

19

704.6

381.4

20:3-Brassicasterol

50

71

19

738.7

381.4

22:0-Brassicasterol

50

71

17

736.7

381.4

22:1-Brassicasterol

50

71

19

734.7

381.4

22:2-Brassicasterol

50

71

19

732.7

381.4

22:3-Brassicasterol

50

71

19

764.7

381.4

24:1-Brassicasterol

50

71

21

766.7

381.4

24:0-Brassicasterol

50

71

19

612.6

409.4

11:0-Cycloartenol

50

66

17

682.6

409.4

16:0-Cycloartenol

50

71

19

680.6

409.4

16:1-Cycloartenol

50

91

19

678.6

409.4

16:2-Cycloartenol

50

71

19

676.6

409.4

16:3-Cycloartenol

50

71

17 (continued)

Sterol and Steryl Ester Analysis in Seeds

191

Table 1 (continued)

Molecular ion [M + NH4]+, (m/z)

Product ion [M-(RCOOH+ NH3)]+ (m/z)

Steryl ester

Scan time (ms)

DP (V)

CE (V)

710.7

409.4

18:0-Cycloartenol

50

71

19

708.7

409.4

18:1-Cycloartenol

50

66

19

706.6

409.4

18:2-Cycloartenol

50

76

19

704.6

409.4

18:3-Cycloartenol

50

66

19

738.7

409.4

20:0-Cycloartenol

50

71

19

736.7

409.4

20:1-Cycloartenol

50

71

21

734.7

409.4

20:2-Cycloartenol

50

71

21

732.7

409.4

20:3-Cycloartenol

50

71

21

766.7

409.4

22:0-Cycloartenol

50

71

19

764.7

409.4

22:1-Cycloartenol

50

71

21

762.7

409.4

22:2-Cycloartenol

50

71

21

760.7

409.4

22:3-Cycloartenol

50

71

21

792.8

409.4

24:1-Cycloartenol

50

71

21

794.8

409.4

24:0-Cycloartenol

50

71

19

3.6

Data Processing

1. The internal standards cholesteryl heptadecanoate and cholestanol added during seed oil extraction (see step 4 in Subheading 3.1) are used for quantification of both GC-FID and ESI-MS/MS (QTRAP) data. Total sterol content as determined by GC-FID is quantified using cholesterol released from the cholesteryl heptadecanoate internal standard after saponification and silylation. Phytosterol quantification by GC-FID has been described in detail, for a review see [3]. Response factors should be calculated for each sterol compared to the standard to normalize the FID response, and responses in the linear range should be used when quantifying sterols with GC-FID. 2. ESI-MS/MS (QTRAP) data is analyzed initially using the Lipid View software, to identify MRM transitions, and subsequently exported to Microsoft Excel for further processing. 3. Data are analyzed in a sequential fashion to normalize the signal, and to remove the response associated with compound structural features. Type 2 isotopic correction is applied first: Precursor ions with an m/z difference of 2, which are

192

Richard Broughton and Fre´de´ric Beaudoin

predominantly the result of the presence of one additional double bond, are targeted. For example, Table 2 illustrates the calculated type 2 correction factors used, and how they were subtracted sequentially. Product ion isotopic distributions are calculated using the IsoPatrn software [10] (see Note 12). 4. Type 1 isotopic correction is then applied, with calculations made using the formula as stated by Yang and Han [5]: Itotal(n) ¼ In(1 + 0.0109n + 0.01092n(n  1)/2), where Itotal(n) is the total corrected ion intensity, In is the monoisotopic peak intensity of the ion containing n carbons, 0.0109 is the natural abundance of 13C and n the number of carbons within the compound. 5. Signal intensities of cholesteryl ester internal standards (see item 3 in Subheading 2.3) are used to correct the steryl ester signal by removing the response associated with compound ionization. The variations in compound response caused by different ionization efficiencies due to acyl chain length and acyl chain unsaturation are removed by plotting linear regression curves for each sample. This allows for the estimation and subsequent normalization of the acyl chain length and unsaturation response (Fig. 4). In practice, using acyl chain length as an example, the internal standards 16:0, 18:0, and 22:0cholesterol are used to plot a linear response curve relative to 16:0-cholesterol for each sample, generating a three-point calibration curve with which the 11:0, 17:0, 20:0, and 24:0cholesterol response factors relative to 16:0-cholesterol can be calculated (Fig. 5). These response factors are then used to normalize the corresponding acyl chain on different sterol moieties by dividing averaged signal intensities (see step 3, Subheading 3.5) by the response factors. The same process is used for calculating the unsaturation response using 18:0-cholesterol, 18:1-cholesterol, and 18:2cholesterol. 6. The variations of signal intensity derived from sterol carbon number and unsaturation are removed by averaging the response of the undecanoyl chloride-derivatized external free sterol mixture (see item 4 in Subheading 2.3), which is run several times throughout the sample sequence. Sterol carbon number normalization is calculated using a similar process as for the acyl response. However, a two-point linear curve is generated with 11:0-cholesterol (27 carbons within the sterol) and 11:0-sitosterol (29 carbons). The response for campesterol and brassicasterol (28 carbons), stigmasterol (29 carbons), and cycloartenol (30 carbons) can be derived from the linear response curve, and signal intensities adjusted as stated in step 5, Subheading 3.6. Sterol unsaturation is calculated once the carbon number response has been removed. The ratio of

Sterol and Steryl Ester Analysis in Seeds

193

Table 2 Type two correction factors calculated using the IsoPatrn software and deducted in the order of most unsaturated to the most saturated steryl esters Deduct signal intensity of this . . .

From signal intensity of this . . .

Type II correction

11:0-Cholesterol

11:0-Cholestanol

0.040

18:2-Cholesterol

18:1-Cholesterol

0.022

18:1-Cholesterol

18:0-Cholesterol

0.022

22:1-Cholesterol

22:0-Cholesterol

0.031

11:0-Brassicasterol

11:0-Campesterol

0.043

11:0-Stigmasterol

11:0-Sitosterol

0.046

16:3-Brassicasterol

16:2-Brassicasterol

0.018

16:3-Brassicasterol

16:3-Campesterol

0.043

16:2-Brassicasterol

16:1-Brassicasterol

0.018

16:2-Brassicasterol

16:2-Campesterol

0.043

16:3-Campesterol

16:2-Campesterol

0.018

16:1-Brassicasterol

16:0-Brassicasterol

0.018

16:1-Brassicasterol

16:1-Campesterol

0.043

16:2-Campesterol

16:1-Campesterol

0.018

16:0-Brassicasterol

16:0-Campesterol

0.043

16:1-Campesterol

16:0-Campesterol

0.018

16:3-Stigmasterol

16:2-Stigmasterol

0.018

16:3-Stigmasterol

16:3-Sitosterol

0.047

16:2-Stigmasterol

16:1-Stigmasterol

0.018

16:2-Stigmasterol

16:2-Sitosterol

0.046

16:3-Sitosterol

16:2-Sitosterol

0.018

16:1-Stigmasterol

16:0-Stigmasterol

0.018

16:1-Stigmasterol

16:1-Sitosterol

0.046

16:2-Sitosterol

16:1-Sitosterol

0.018

16:0-Stigmasterol

16:0-Sitosterol

0.046

16:1-Sitosterol

16:0-Sitosterol

0.018

18:3-Brassicasterol

18:2-Brassicasterol

0.022

18:3-Brassicasterol

18:3-Campesterol

0.043

18:2-Brassicasterol

18:1-Brassicasterol

0.022

18:3-Campesterol

18:2-Campesterol

0.022

18:1-Brassicasterol

18:0-Brassicasterol

0.022 (continued)

194

Richard Broughton and Fre´de´ric Beaudoin

Table 2 (continued) Deduct signal intensity of this . . .

From signal intensity of this . . .

Type II correction

18:1-Brassicasterol

18:1-Campesterol

0.043

18:2-Campesterol

18:1-Campesterol

0.022

18:0-Brassicasterol

18:0-Campesterol

0.043

18:1-Campesterol

18:0-Campesterol

0.022

18:3-Stigmasterol

18:2-Stigmasterol

0.022

18:3-Stigmasterol

18:3-Sitosterol

0.047

18:2-Stigmasterol

18:1-Stigmasterol

0.022

18:2-Stigmasterol

18:2-Sitosterol

0.047

18:3-Sitosterol

18:2-Sitosterol

0.022

18:1-Stigmasterol

18:0-Stigmasterol

0.022

18:1-Stigmasterol

18:1-Sitosterol

0.047

18:2-Sitosterol

18:1-Sitosterol

0.022

18:0-Stigmasterol

18:0-Sitosterol

0.047

18:1-Sitosterol

18:0-Sitosterol

0.022

20:3-Brassicasterol

20:2-Brassicasterol

0.026

20:3-Brassicasterol

20:3-Campesterol

0.043

20:2-Brassicasterol

20:1-Brassicasterol

0.026

20:2-Brassicasterol

20:2-Campesterol

0.043

20:3-Campesterol

20:2-Campesterol

0.026

20:1-Brassicasterol

20:0-Brassicasterol

0.026

20:1-Brassicasterol

20:1-Campesterol

0.043

20:2-Campesterol

20:1-Campesterol

0.026

20:0-Brassicasterol

20:0-Campesterol

0.043

20:1-Campesterol

20:0-Campesterol

0.026

20:3-Stigmasterol

20:2-Stigmasterol

0.026

20:3-Stigmasterol

20:3-Sitosterol

0.047

20:2-Stigmasterol

20:1-Stigmasterol

0.026

20:2-Stigmasterol

20:2-Sitosterol

0.047

20:3-Sitosterol

20:2-Sitosterol

0.026

20:1-Stigmasterol

20:0-Stigmasterol

0.026

20:1-Stigmasterol

20:1-Sitosterol

0.047 (continued)

Sterol and Steryl Ester Analysis in Seeds

195

Table 2 (continued) Deduct signal intensity of this . . .

From signal intensity of this . . .

Type II correction

20:2-Sitosterol

20:1-Sitosterol

0.026

20:0-Stigmasterol

20:0-Sitosterol

0.047

20:1-Sitosterol

20:0-Sitosterol

0.026

22:3-Brassicasterol

22:2-Brassicasterol

0.031

22:3-Brassicasterol

22:3-Campesterol

0.044

22:2-Brassicasterol

22:1-Brassicasterol

0.031

22:2-Brassicasterol

22:2-Campesterol

0.044

22:3-Campesterol

22:2-Campesterol

0.031

22:1-Brassicasterol

22:0-Brassicasterol

0.031

22:1-Brassicasterol

22:1-Campesterol

0.044

22:2-Campesterol

22:1-Campesterol

0.031

22:0-Brassicasterol

22:0-Campesterol

0.044

22:1-Campesterol

22:0-Campesterol

0.031

22:3-Stigmasterol

22:2-Stigmasterol

0.031

22:3-Stigmasterol

22:3-Sitosterol

0.047

22:2-Stigmasterol

22:1-Stigmasterol

0.031

22:2-Stigmasterol

22:2-Sitosterol

0.047

22:3-Sitosterol

22:2-Sitosterol

0.031

22:1-Stigmasterol

22:0-Stigmasterol

0.031

22:1-Stigmasterol

22:1-Sitosterol

0.047

22:2-Sitosterol

22:1-Sitosterol

0.031

22:0-Stigmasterol

22:0-Sitosterol

0.047

22:1-Sitosterol

22:0-Sitosterol

0.031

24:1-Brassicasterol

24:0-Brassicasterol

0.036

24:1-Brassicasterol

24:1-Campesterol

0.044

24:0-Brassicasterol

24:0-Campesterol

0.044

24:1-Campesterol

24:0-Campesterol

0.036

24:1-Stigmasterol

24:0-Stigmasterol

0.036

24:1-Stigmasterol

24:1-Sitosterol

0.047

24:0-Stigmasterol

24:0-Sitosterol

0.047

24:1-Sitosterol

24:0-Sitosterol

0.036 (continued)

196

Richard Broughton and Fre´de´ric Beaudoin

Table 2 (continued) Deduct signal intensity of this . . .

From signal intensity of this . . .

Type II correction

16:3-Cycloartenol

16:2-Cycloartenol

0.018

16:2-Cycloartenol

16:1-Cycloartenol

0.018

16:1-Cycloartenol

16:0-Cycloartenol

0.018

18:3-Cycloartenol

18:2-Cycloartenol

0.022

18:2-Cycloartenol

18:1-Cycloartenol

0.022

18:1-Cycloartenol

18:0-Cycloartenol

0.022

20:3-Cycloartenol

20:2-Cycloartenol

0.026

20:2-Cycloartenol

20:1-Cycloartenol

0.026

20:1-Cycloartenol

20:0-Cycloartenol

0.026

22:3-Cycloartenol

22:2-Cycloartenol

0.031

22:2-Cycloartenol

22:1-Cycloartenol

0.031

22:1-Cycloartenol

22:0-Cycloartenol

0.031

24:1-Cycloartenol

24:0-Cycloartenol

0.036

The type two correction is calculated for the sterol fragment because the acyl chain is lost as neutral loss, and the sterol fragment is the only fragment left after CID

11:0-cholestanol (0 double bonds) to 11:0-cholesterol (1 double bond) as well as to 11:0-stigmasterol (2 double bonds) is calculated, with the response factors applied to the signal intensities as in step 5, Subheading 3.6. 7. Following the removal of responses due to compound structural features and isotopic correction, natural steryl esters are quantified using cholesteryl heptadecanoate (17:0-cholesterol) whilst undecanoyl chloride-derivatized free sterols are quantified using cholestanyl undecanoate (11:0-cholestanol).

4

Notes 1. These tubes are designed for hard tissue homogenization and we found that they work very well for Brassica seeds with gentle grinding. Reinforced tubes (CK28-R) are available for bead beating applications, allowing grinding of harder samples. 2. Due to the low solubility of KOH at high ethanol concentrations, crushing the KOH pellets or stirring with a magnetic bar is recommended. Alternatively, because this solution is not stable and degrades quickly at room temperature, if large volumes are required, it can be advantageous to prepare it

Sterol and Steryl Ester Analysis in Seeds

197

Fig. 4 The averaged response factors for individual steryl ester structural features observed across several batch runs of steryl esters isolated from rapeseed oil using direct infusion ESI-MS/MS. Each structural feature of a steryl ester contributes to its response factor, and hence needs to be normalized by removing each contributing factor (see Subheading 3.6, step 5) for accurate quantification using ESI-MS/MS. (a) acyl chain lengths; (b) acyl chain unsaturation (number of double bonds); (c) sterol carbon number; (d) sterol unsaturation (number of double bonds)

fresh using a more stable concentrated KOH stock solution in water and pure ethanol. Mixing 10 ml 13.5 M KOH (15.1 g KOH in 13.5 ml water for 20 ml stock solution) with 80 ml ethanol generates a 1.5 M KOH solution in ca. 92.5% ethanol which works well with lipid extracts using the method described in Subheading 3.2. 3. Other GC columns and conditions can be used for phytosterol analysis, for review see [3]. Columns containing nonpolar stationary phases like the Agilent HP-1MS (100% dimethylpolysiloxane) allow for the base line separation of Δ5-desaturated sterols such as β-sitosterol or Δ5-avenasterol, and Δ7-desaturated sterols such as Δ7-stigmastenol and Δ7-avenasterol (Fig. 2). Higher polarity columns may afford greater separation between related stanol and sterols, and between those of Δ5 and Δ7 sterols.

198

Richard Broughton and Fre´de´ric Beaudoin

Fig. 5 Calibration curves illustrating the linear range, which for the majority of compounds is 1000-fold. The calibration curves generated within each sample are however, single point calibration response curves, as only one concentration of standard (16:0-phytol) is present in each sample. The diagrams show the log10 of the intensity ratios of the steryl ester against 16:0-cholesterol, plotted versus the log10 of the amounts of the steryl ester (17:0-cholesterol, 18:0-cholesterol, 18:2-cholesterol, 22:1-cholesterol, 18:1-cholesterol, and 22:0-cholesterol) against a fixed concentration of 16:0-cholesterol in pmol/μl

4. The dichloromethane used for derivatization must be anhydrous because acyl chlorides react very strongly with water, effectively quenching the reaction by producing the corresponding carboxylic acid. For analytical purposes, purchasing anhydrous solvents may be preferred for preparing

Sterol and Steryl Ester Analysis in Seeds

199

in-house dry solvents. Care must be taken to maintain a dry atmosphere in anhydrous solvent stock bottles when purchased. This can be done by filling a balloon with dry nitrogen and affixing a needle to the end via a syringe barrel. This can be inserted into the septum of anhydrous bottles to maintain a nitrogen atmosphere when dichloromethane is removed using another syringe. Solvent mixtures prepared in the laboratory (e.g., molecular standard stocks and mixtures in dichloromethane) can then be kept dry by adding molecular sieve beads which are generally sufficient for keeping the moisture content low. 5. Glassware is inherently wet, in that moisture in the air binds readily to the polar surface. Baking glassware is sufficient to remove the majority of the water, as long as the reaction is carried out soon after with dry solvents. 6. It is generally advised that storing lipid samples, if not under vacuum, is preferential in aprotic solvents such as chloroform or hexane. This reduces oxidation, which is particularly important for polyunsaturated lipids. If the sample volume is critical, less volatile solvents such as heptane may be preferred; however, the addition of internal standards and drying steps makes this less problematic. Whenever volatile solvents are stored in a fridge, it is strongly advised to check the suitability of the fridge, which must be spark proof. 7. Care must be taken during the saponification step when partitioning the phases, as small amounts of the lower alkali phase may be transferred through the process, eventually being injected into the GC. This results in a high and noisy baseline which takes several runs to clear and can affect subsequent retention times. It is therefore advisable to perform an additional hexane extraction to minimize extraction losses, if removing all the hexane upper phase proves difficult. 8. The split ratio between detector and fraction collector may need fine tuning depending on the sensitivity of the detector used. It is also possible to inject more sample onto the column, depending on the size of the sample loop and being aware of any carry over observed, especially from triacylglycerols. Also be aware of the solvent used to redissolve samples prior to chromatography, as when using large injection volumes, mistakenly using ethyl acetate may result in the steryl ester coeluting with the triacylglycerol. 9. Care must be taken to use HPLC or LC-MS grade solvents, as well as to centrifuge or filter samples, as samples are injected directly into the ESI capillary, which can easily block if particulate matter is present. Infusion methodologies are also more likely to result in dirty source components, so these should be

200

Richard Broughton and Fre´de´ric Beaudoin

cleaned regularly to reduce sources of contamination. Flushing tubing after each sample is also important, otherwise sample cross contamination is likely to occur. When using an autosampler like a PAL-CTC, this process can be automated. 10. The method described here pertains predominantly to triple quad mass spectrometers however, quadrupole time-of-flight (QTof) and Orbitrap mass spectrometers can to some extent also utilize MRM transitions. Therefore, additional or alternative issues may arise with different machines. It was observed that the signal would decrease an order of magnitude or more when running a large sample batch, with the loss in sensitivity being attributed to the charging of the quadrupoles. This was remedied to a large extent by introducing a short negative polarity switch after each run to restore instrument sensitivity. 11. These machine settings are likely to vary with different instruments and manufacturers. It is also likely that certain parameters will be named differently, as well as being given in different units. It is always advised to optimize signal intensities of compounds for the machine to be used, as opposed to just copying settings. 12. Type 1 isotopic correction relates to the isotopic distribution. In the case of lipids where carbon is the predominant atom, it is due to the presence of different numbers of the 13C isotope, which has a natural occurrence of approximately 1%. The larger the number of carbons in lipid molecules the greater the probability of the compound containing a 13C isotope, and hence the greater the associated m/z peak and subsequent reduction in the monoisotopic (12C) peak. Type 2 isotopic correction relates to the interference of these heavier isotope-containing molecular species with neighbouring compounds, specifically those that are 2 mass units away. The (2*13C) isotopic peak of a monounsaturated fatty acid containing compound will have the same unit mass as the monoisotopic mass of the saturated form of the same compound, and will subsequently hinder quantification, especially if abundant (unsaturated) masses are located adjacent to less abundant (more saturated) masses.

Acknowledgments The development of the ESI-MS/MS method for analysis of steryl esters was supported by the BBSRC sLoLa “Renewable Industrial Products from Rapeseeds (RIPR)” grant (BB/L002124/1). F.B. receives grant aided support from the BBSRC as part of RRes “Tailoring Plant Metabolism” Institute Strategic Program grant (BB/P012663/1). The authors thank Dr. Richard Haslam for critical reading of the manuscript.

Sterol and Steryl Ester Analysis in Seeds

201

References 1. Hallikainen MA, Sarkkinen ES, Uusitupa MIJ (2000) Plant stanol esters affect serum cholesterol concentrations of hypercholesterolemic men and women in a dose-dependent manner. J Nutr 130:767–776. https://doi.org/10. 1093/jn/130.4.767 2. Nestel P, Cehun M, Pomeroy S, Abbey M, Weldon G (2001) Cholesterol-lowering effects of plant sterol esters and non-esterified stanols in margarine, butter and low-fat foods. Eur J Clin Nutr 55:1084–1090. https://doi.org/ 10.1038/sj.ejcn.1601837 3. Winkler-Moser J (2011) Gas chromatographic analysis of plant sterols. https://doi.org/10. 21748/lipidlibrary.40384 4. Costin CD, Hansen SL, Chambers DP (2009) Using theoretical correction factors for quantitative analysis of sterols and sterol concentrates. J Am Oil Chem Soc 86:111–118. https://doi. org/10.1007/s11746-008-1332-9 5. Yang K, Han X (2011) Accurate quantification of lipid species by electrospray ionization mass spectrometry – meets a key challenge in lipidomics. Meta 1:21–40. https://doi.org/10. 3390/metabo1010021 6. Bru¨gger B, Erben G, Sandhoff R, Wieland FT, Lehmann WD (1997) Quantitative analysis of

biological membrane lipids at the low picomole level by nano-electrospray ionization tandem mass spectrometry. Proc Natl Acad Sci U S A 94:2339–2344. https://doi.org/10.1073/ pnas.94.6.2339 7. Wewer V, Dombrink I, vom Dorp K, Do¨rmann P (2011) Quantification of sterol lipids in plants by quadrupole time-of-flight mass spectrometry. J Lip Res 52:1039–1054. https:// doi.org/10.1194/jlr.D013987 8. Broughton R, Ruı´z-Lopez N, Hassall KL, Martı´nez-Force E, Garce´s R, Salas JJ, Beaudoin F (2018) New insights in the composition of wax and sterol esters in common and mutant sunflower oils revealed by ESI-MS/MS. Food Chem 269:70–79. https://doi.org/10.1016/ j.foodchem.2018.06.135 9. Ruiz-Lopez N, Broughton R, Usher S, Haslam R, Napier JA, Beaudoin F (2017) Tailoring the composition of wax esters in the seeds of transgenic Camelina sativa through systematic metabolic engineering. Plant Biotechnol J 15:837–849. https://doi.org/10. 1111/pbi.12679 10. Ramaley L (2006) Isopatrn. Dalhousie University, Halifax. http://tarc.chemistry.dal.ca/ IsoPatrn_down.htm

Chapter 12 Techniques for the Measurement of Molecular Species of Acyl-CoA in Plants and Microalgae Richard P. Haslam and Tony R. Larson Abstract The acyl-CoA pool is pivotal in cellular metabolism. The ability to provide reliable estimates of acyl-CoA abundance and distribution between molecular species in plant tissues and microalgae is essential to our understanding of lipid metabolism and acyl exchange. Acyl-CoAs are typically found in low abundance and require specific methods for extraction, separation and detection. Here we describe methods for acyl-CoA extraction and measurement in plant tissues and microalgae, with a focus on liquid chromatography hyphenated to detection techniques including ultraviolet (UV), fluorescence and mass spectrometry (MS). We address the resolution of isobaric species and the selection of columns needed to achieve this, including the analysis of branched chain acyl-CoA thioesters. For MS analyses, we describe diagnostic ions for the identification of acyl-CoA species and how these can be used for both discovery of new species (data dependent acquisition) and routine quantitation (triple quadrupole MS with multiple reaction monitoring). Key words Acyl-CoAs; Lipid metabolism, Liquid chromatography, LC-MS/MS, Mass-spectrometry, Multiple reaction monitoring, Camelina sativa, Microalgae

1

Introduction Acyl-CoA thioesters contribute to multiple cellular activities including energy metabolism, the biosynthesis and recycling of complex lipids, posttranslational modifications of proteins and regulation of gene expression. They are derived from the condensation of free fatty acids and CoA by the action of acyl-CoA synthetase (ACS). In plants and microalgae, acyl-CoAs are available for acylation reactions resulting in lipid synthesis via the Kennedy and acyl editing pathways [1]. During lipid degradation, acyl-CoA thioesters are supplied to peroxisomes for β-oxidation. The central role of acyl-CoA thioesters in cellular metabolism has long been accepted, however, techniques for the accurate determination of acyl-CoA pool composition and quantification in plants and microalgae have been limited. Historically, many techniques were

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_12, © Springer Science+Business Media, LLC, part of Springer Nature 2021

203

204

Richard P. Haslam and Tony R. Larson

developed for the measurement of acyl-CoAs in mammalian tissues and cells. These methods relied on detection by either enzyme assays of Coenzyme A from hydrolyzed acyl-CoA thioesters (which provide no information on acyl chain length), or chromatographic assays of released fatty acids (insensitive and prone to contamination by fatty acids derived from other, often more abundant lipids), or the separation and UV detection of acyl-CoA thioesters by reversed-phase high-performance liquid chromatography (HPLC). Latterly, the availability and uptake of bench top mass spectrometers in laboratories has resulted in a number of analytical protocols based on selected reaction monitoring (SRM) or multiple reaction monitoring (MRM) scan modes to monitor cleavage of the 30 -phosphate-adenosine-50 -diphosphate CoA subunit [2]. Neutral loss and precursor ion scanning have also been utilized, enabling a nontargeted profiling approach. Further modifications have incorporated the use of solid-phase extraction to improve recovery rates [3] and hydrophilic interaction liquid chromatography to reduce peak tailing [4]. Moreover, for relatively pure standard mixtures, the complete structural identification of individual acylCoAs, including the location of unsaturated bond(s) in the fatty acid chain has now been reported using a matrix assisted laser desorption ionization/time-of-flight (MALDI/TOF) mass spectrometer [5]. Compared with animal samples, acyl-CoA analyses from plant tissues and microalgae have proved more challenging, in that they are difficult to grind (fibrous or silica-based material), quantities available may only be in the milligram range, and extracts are complicated by the presence of high concentrations of pigments and phenolic compounds, which may interfere with enzyme assays and chromatographic separations. Moreover, any analytical method must have the capacity to capture and discriminate between the full range of plant and microalgal acyl-CoA esters, which extend from short (C10) to very long chain (C30) species and can include multiple desaturations. Over recent years, methods have been developed for the rapid extraction of acyl-CoAs from multiple plant tissues and the accurate determination of tissue acyl-CoA thioester pools. We provide details of methods for acyl-CoA measurement applicable to plants and microalgae that are sensitive and provide the resolution necessary for structural characterization.

2

Materials

2.1 Acyl-CoA Standards

Acyl-CoA standards may be synthesized from commercially available free fatty acids using an enzymatic procedure for acyl chain lengths greater than approximately 12 carbons, as described by Taylor et al. [6], or for shorter chain lengths, using the chemical synthesis procedures described by Kawaguchi et al. [7]. If synthesis

Mass Spectrometry for Measurement of Acyl-CoA

205

is performed, preparative LC is recommended to separate residual free Coenzyme A (CoASH) from acyl-CoAs, before following the steps outlined below in Subheading 3.1. Otherwise, commercial standards can be used as described below. 1. pH Meter and calibration solutions. 2. Potassium phosphate monobasic (KH2PO4) and potassium phosphate dibasic (K2HPO4). 3. 70% (v/v) ethanol (70 mL ethanol and 30 mL Elga analytical grade water). 4. Acyl-CoA, 5 mg standards (from Avanti Polar Lipids Inc., USA); internal standard (ISTD) 870,717—17:0 Coenzyme A; and external standard mix (EXSTD) consisting of 870,712—12:0 Coenzyme A; 870,714—14:0 Coenzyme A; 870,716—16:0 Coenzyme A; 870,717—17:0 Coenzyme A; 870,718—18:0 Coenzyme A; 870,720—20:0 Coenzyme A; 870,722—22:0 Coenzyme A; 870,724—24:0 Coenzyme A; 870,726—26:0 Coenzyme A; 870,743—16:1(n7) Coenzyme A; 870,719—18:1(n9) Coenzyme A; 870,736—18:2 (n6) Coenzyme A; 870,733—18:3(n6) Coenzyme A; 870,744—20:5 Coenzyme A and 870,728—22:6 Coenzyme A. 5. Freshly made spectrophotometer buffer (use at room temperature): 100 mL 100 mM KH2PO4 (store stock at 4  C) and 100 mL 100 mM K2HPO4 (store stock at 4  C). Combine to give 100 mM KH2PO4–K2HPO4 buffer; pH 7.0. 6. UV transparent 1 mL 1 cm cuvettes and spectrophotometer. 7. 2 mL snap-lid microfuge tubes; 0.5 mL microfuge tubes. 8. 1000, 200, 50, and 10 μL pipettors and tips. 9. 50 mL glass volumetric flask with polypropylene stopper and 50 mL Falcon tube with lid. 2.2 Preparation of Chloroacetaldehyde Derivitization Solution

Derivatization of acyl-CoA thioesters with chloroacetaldehyde is only required if the intention is to analyze acyl-CoAs using HPLC with fluorescence detection. Derivatization will partially decrease the UV response at 260 nm. However, in our experience it has no deleterious effect on the MS response, with similar or better detection limits achievable with MRM techniques (depending on instrument). MS has the advantage of being able to resolve coeluting, nonisobaric acyl-CoA species, but with the caveat that ionization efficiency and thus quantitative response varies depending on acyl chain structure. Both UV and fluorescence measurement techniques effectively read out the molar concentration of the adenine group, so may be the better choices for the quantification of acylCoAs with unknown acyl chain structure or for which standards are not available.

206

Richard P. Haslam and Tony R. Larson

1. Chloroacetaldehyde ~7.87 M. 2. 300 mM citric acid. 3. 300 mM trisodium citrate. 4. Sodium dodecyl sulfate (SDS). 5. 250 mL glass beaker with stirrer. 2.3 Acyl-CoA Extraction

1. Heating water bath (maximum temperature required: 60  C). 2. Acetic acid (less than 1% water content). 3. Mini centrifuge. 4. Oven (maximum temperature required: 85  C). 5. Sample concentrator (Techne) and nitrogen gas supply. 6. Motor-driven pestle grinder and pestles. 7. Ball mill (or similar as available) and 500 μm glass beads. 8. Vial crimper (Fisher); Chromacol 1.1 mL Crimp Cap Tapered Vials (1.1-CTVG) and 11 mm Autosampler Vial Crimp Caps (Thermo Scientific; 11-ApC-TST1). 9. 50 mM KH2PO4; pH 7.2: Prepare 1 M stock solutions of K2HPO4 and KH2PO4 (store at 4  C). Combine 30.75 mL of K2HPO4 with 19.25 mL of KH2PO4 to give pH 7.0. Dilute the 1 M stock to 100 mM with analytical grade water (store at 4  C). Dilute 100 mM phosphate buffer; pH 7.0; to 50 mM and alter pH to 7.2 using 1 M K2HPO4 stock (store at 4  C). 10. 50 mg/mL bovine serum albumin (BSA): combine 10 mL analytical grade water with 0.5 g fatty acid free BSA; aliquot 500 μL into vials and store at 20  C. 11. Petroleum ether saturated with isopropanol–water (analytical grade) (1:1; v/v): Prepare 80 mL of isopropanol–water (1:1; v/v); that is, 40 mL of each; then add ~15 mL of petroleum ether (heat to 50  C before use in a water bath). 12. Saturated ammonium sulfate (NH4)2SO4): 20 mL analytical grade water plus ammonium sulfate (ground to powder) until no more dissolves. 13. Methanol–chloroform (2:1, v/v): 40 mL methanol (CH3OH) plus 20 mL chloroform (CHCl3).

2.4 Separation of Acyl-CoA Species Using Liquid Chromatography and Ultraviolet (UV) Detection

1. Binary HPLC system; autoinjector; column oven and UV detector. 2. LUNA 150  4.6 mm column with phenyl-hexyl coated 5 μm silica particles; together with a 4  3 mm phenyl-propyl guard column. 3. Mobile phases: (A) 10 mM KH2PO4; pH 7.0 and (B) 70% (v/v) acetonitrile (Rathburn)/10 mM KH2PO4; pH 7.0.

Mass Spectrometry for Measurement of Acyl-CoA

207

2.5 Separation of Derivatized Acyl-CoA Species Using Liquid Chromatography and Fluorescence Detection

1. Quaternary HPLC system; autoinjector; column oven and fluorescence detector (Agilent 1200).

2.6 Separation of Isobaric Short Chain Acyl-CoA Species (Derivatized or Not) Using Liquid Chromatography

This method resolves C5 branched chain acyl-CoAs and can be used with fluorescence, UV or MS detection as described in Ishizaki et al. [8]. Acyl-CoAs greater than approximately C10 will not elute from the column with this method.

2. Phenomenex C18(2) 150  2 mm; 5 μm; and 4  2 mm guard cartridge (Phenomenex Security Guard C18). 3. Mobile phases (A) 0.25% (v/v) triethylamine (99.5%) (TEA): Add 2.5 mL triethylamine to 1000 mL analytical grade water; (B) 90% (v/v) acetonitrile made with analytical grade water; (C) 90% (v/v) acetonitrile and 1% (v/v) acetic acid; and (D) 1% (v/v) acetic acid made with analytical grade water.

1. HPLC system as in Subheading 2.5 (step 1) and mobile phases as in Subheading 2.5 (step 3). 2. Thermo Hypersil PGC (porous graphitic carbon) column; 100  3 mm and 4  2 mm guard cartridge (Phenomenex Security Guard C18).

2.7 Separation of Acyl-CoA Species Using Liquid Chromatography and Analysis by Mass Spectrometry with Multiple Reaction Monitoring (MRM)

3

1. Quaternary HPLC system; autoinjector; column oven and fluorescence detector (Agilent 1200 system). 2. QTRAP 4000 hybrid triple quadrupole LIT (linear ion trap) mass spectrometer (SCIEX). 3. Agilent Eclipse XDB-C18 column; 3.0  100 mm; 3.5 μm particles. 4. Mobile phases (A) 100% acetonitrile, 15 mM ammonium hydroxide (add concentrated aqueous ammonium hydroxide to acetonitrile to achieve 15 mM); (B) 90% analytical grade water, 10% acetonitrile, 15 mM ammonium hydroxide (add concentrated aqueous ammonium hydroxide to water–acetonitrile (9:1; v/v) to achieve 15 mM); and (C) 30% analytical grade water; 70% acetonitrile; 0.1% formic acid.

Methods

3.1 Acyl-CoA Standards

1. To make primary stocks: Using a 1 mL pipettor, add 100 μL of analytic grade water for every 1 mg powder into the vials in which the stocks were supplied (i.e., 500 μL to 5 mg or 1000 μL to 10 mg). Close vials, vortex briefly (5–10 s) to avoid excessive foaming and transfer entire vial contents to 2 mL snap-lid microfuge tubes labelled with the acyl-CoA formula and the approximate concentration of approximately 10 mM (this is only approximate, because it assumes a constant molecular weight of 1000 g/mol and no hydration for all acylCoAs). Keep tubes on ice.

208

Richard P. Haslam and Tony R. Larson

2. To make secondary stocks: Using pipettors, combine 50 μL of each primary stock with 950 μL of analytical grade water using a 1 mL pipettor. Cap and vortex (2 s). This gives a secondary stock concentration of approximately 0.5 mM (500 pmol/μL). 3. Calculate the concentration of the secondary stocks. Using a 1 mL pipettor, transfer forty-two 900 μL aliquots of 100 mM KH2PO4–K2HPO4 buffer (pH 7.0 measured at RT) to labelled 2 mL microfuge tubes. Using a 200 μL pipettor, add 100 μL of the appropriate secondary stock to each tube (in triplicate; 14  3 ¼ 42 tubes). Cap and vortex (2 s). This gives a nominal concentration of approximately 50 pmol/μL. Using a 1 mL pipettor, add 900 μL 100 mM KH2PO4–K2HPO4 buffer (pH 7.0 at RT) to a clean, dry cuvette. Insert the cuvette into the spectrophotometer and zero at 260 nm. Using a 1 mL pipettor, in turn transfer 900 μL of each diluted secondary standard into a clean cuvette. Insert the cuvette into the spectrophotometer and measure the absorption at 260 nm (A260 nm). Take the average of the three absorbance values and use this to calculate the actual concentrations of the secondary stocks as follows (see Note 1): Concentration of secondary stock ðmMÞ ¼ 10  1000  A ð260 nmÞ=16, 400: Write the exact concentration on the secondary stock microfuge tubes in mM to three decimal places. 4. To make tertiary stocks at concentrations of exactly 0.05 mM (50 pmol/μL): Add (1000–(100  0.5/exact concentration of secondary stock in mM)) μL analytical grade water to a 2 mL snap-lid microfuge tube. Add (100  0.5/exact concentration of secondary stock in mM) μL secondary stock; vortex; cap. Write the exact concentration on the tertiary stock microfuge tubes as 50 pmol/μL. Store remaining secondary stock solutions together with the primary stocks in a plastic box at 80  C. 5. Make stock for 17:0-CoA ISTDs at concentrations of 0.2 pmol/μL in the same tubes: Combine 200 μL tertiary stock for 17:0 in a clean 50 mL volumetric flask; make up to 50 mL with 70% (v/v) ethanol; cap and mix. Pour the contents into a 50 mL Falcon tube. Using a 1 mL pipettor, transfer approximately two hundred 250 μL aliquots into prepared 0.5 mL microfuge tubes and cap. Each tube then contains approximately 20  10 μL aliquots (each 2 pmol). This gives 1 pmol·per 20 μL injection from 40 μL derivatized sample extract (which contain 10 μL or 2 pmol of 17:0-CoA ISTD). Store the 17:0-CoA ISTD tubes and remaining tertiary stock solutions, together with the primary and secondary stocks in a plastic freezer box at 80  C.

Mass Spectrometry for Measurement of Acyl-CoA

209

6. Make quaternary stock for EXSTD at a concentration of 0.5 pmol/μL: Using a 50 μL pipettor, add 20 μL of each of the 16 tertiary stocks for all acyl-CoAs to a 2 mL snap-lid microfuge tube; then add 1720 μL of 70% (v/v) ethanol; cap and mix. Using a 10 μL pipettor, transfer approximately 200  10 μL (5 pmol) aliquots to tapered vials and cap; label as “CoA EXSTD”. When used, each vial is derivatized with 90 μL reagent. This gives 1 pmol·per 20 μL injection from 100 μL (containing 10 μL or 5 pmol of EXSTD). 3.2 Preparation of Chloroacetaldehyde Derivatization Solution

Reaction of acyl-CoA thioesters from standard solutions or plant extracts with chloroacetaldehyde reagent results in the formation of higher-molecular-weight fluorescent derivatives, where the adenine residue of CoA is modified by the addition of an etheno group. For example; the molecular ion for heptadecanoyl-CoA has a mass-tocharge ratio (m/z) of 1020 compared to a value of 1044 for its etheno derivative [9]. 1. Make the following derivatization solutions as outlined below: (A) 100 mL of 1 M chloroacetaldehyde solution buffered to pH 4.0 with 300 mM citrate buffer with 1% (w/v) SDS as a surfactant for 1:1 liquid derivatization of standards or samples; and (B) 100 mL of 0.5 M chloroacetaldehyde solution buffered to pH 4.0 with 150 mM citrate buffer with 0.5% (w/v) sodium dodecyl sulfate (SDS) as a surfactant for direct derivatizations of dried samples. 2. Add 20 mL chloroacetaldehyde to a 250 mL beaker with stirrer. 3. Add 1.5 g SDS while continuously stirring to dissolve. 4. Add 50 mL 300 mM trisodium citrate. 5. Add 50 mL 300 mM citric acid. 6. Adjust pH to 4.0 and to a final volume of 150 mL with 300 mM citric acid–trisodium citrate; mix with a stirrer and take care to maintain the pH at 4.0. This is solution A. 7. To make solution B; dilute 50 mL A to 100 mL with analytical grade water and check pH. Store both solutions in the dark at room temperature ~ 25  C (do not refrigerate). Discard after 3 months.

3.3 Acyl-CoA Extraction

This procedure is essentially derived from that described by Mancha et al. [10]. 1. Freshly prepare Extraction Buffer and store on ice; scale up/down as required. Extraction buffer: 1 mL isopropanol; 1 mL 50 mM KH2PO4, pH 7.2 (stored at 4  C); 25 μL acetic acid (Fisher); and 40 μL 50 mg/ml bovine serum albumin (see Notes 2 and 3).

210

Richard P. Haslam and Tony R. Larson

2. Accurately weigh each sample to determine fresh weight before processing further. 3. Remove plant samples from freezer; add ~1 mL of liquid N2 and grind to powder using a pestle. Add 200 μL freshly prepared ice-cold extraction buffer. Grind (three times) 1 min with Eppendorf grinder or similar device; cool on ice between each grind. Use a fresh pestle for each sample (see Notes 4 and 5 for extraction of acyl-CoA thioesters from microalgae). 4. Add 20 μL 17:0-CoA ISTD to each extraction (ISTD stored at 80  C in 200 μL aliquots). 5. Remove lipids with three or four (depending on the oil content of the tissue) 200 μL washes of petroleum ether (saturated with isopropanol/water); vortex for 2–3 s to mix. Keep the petroleum ether (saturated with isopropanol/water) at 50  C in a water bath. Separate phases with low-speed (500  g) centrifugation. Transfer upper phase to waste. 6. Add 5 μL saturated (NH4) 2SO4 to aqueous phase. 7. Add 600 μL methanol–chloroform (2:1). Cap; vortex; leave at room temperature for 20 min. 8. Spin down in bench top centrifuge at 3500  g for two min. 9. Transfer supernatant to tapered glass HPLC vials. 10. Evaporate using sample concentrator (Techne) at 36  C with flow of nitrogen gas. In a separate vial; 90 μL of pre-prepared EXSTD can be dried down at the same time. 11. For LC-MS/MS with MRM analysis, resuspend samples in 40 μL (EXTSD in 90 μL) of H2O–acetonitrile (9:1), 15 mM ammonium hydroxide (mobile phase B; see Subheading 2.7), cap and vortex for samples. 12. For fluorescence detection, derivatization is required; resuspend samples in 40 μL (EXTSD, 90 μL) of 0.5 M chloroacetaldehyde (solution B; see Subheading 3.2) and heat in oven at 85  C for 20 min; cool to room temperature and inject 35 μL on to LC. 3.4 Analysis of Nonderivatized AcylCoA Thioesters Using Liquid Chromatography with Ultraviolet Detection

1. Acyl-CoAs are separated on a reverse-phase column using HPLC equipment with a gradient elution system of phosphate buffer and acetonitrile. 2. A binary pump supplied vacuum-degassed mobile phase with a flow at 1 mL/min. 3. The column was maintained in an oven at 40  C. 4. Mobile phases were (A) 10 mM KH2PO4, pH 7.0; and (B) 70% acetonitrile–30% 10 mM KH2PO4, pH 7. The gradient used is as follows: 0–23 min, 10–100% B; 23–25 min, 100–10% B; 25–30 min, 10% B; a concave gradient curve with a slope of 1.

Mass Spectrometry for Measurement of Acyl-CoA

211

5. Samples were maintained at 4  C in the autosampler prior to injection of 35 μL. Eluted acyl-CoA thioesters are detected by UV absorbance at 260 nm. 3.5 Separation and Analysis of Derivatized Acyl-CoA Species Using Liquid Chromatography and Fluorescence Detection

1. Sample and EXSTD acyl-etheno-CoA thioesters are separated by reverse-phase HPLC using a quaternary gradient system. 2. The degassed mobile phase is supplied to a C18 (2) 150  2 mm column (Phenomenex) with 5 μm particles and a 4  2 mm guard cartridge (Phenomenex Security Guard C18) at a column oven temperature of 40  C. 3. For elution of the fluorescent acyl-etheno-CoA thioesters, a gradient of triethylamine and acetonitrile is pumped through the column following impurity elution with the acidic mobile phases. For buffers A, B, C, and D, see Subheading 2.5. A linear quaternary gradient program was run as follows: 0 min; flow 0.75 mL/min; 10% C and 90% D; 0–5 min; to 80% C and 20% D; 5–5.1 min; to 80% A and 20% D; 5.1–7 min; to 97% A and 3% B; 7–10 min; to 95% A and 5% B; 10–10.1 min; flow decreased to 0.5 mL/min; 10.1–50 min; to 60% A and 40% B; 50.1–52 min; to 100% B; 52–52.1 min; flow increased to 0.75 mL/min; 52.1–55 min; maintain 100% B; 55–55.1 min; to 10% C and 90% D; 55.1–60 min; maintain 10% C and 90% D (see Note 6 for detail about column life and peak shape). 4. Samples and EXSTD standards are maintained at 15  C (see Note 7) in the autosampler prior to the injection of 35 μL. 5. Peaks are detected using the fluorescence detector with a lamp flash rate of 100 Hz and photomultiplier tube voltage set to 600 V. The excitation wavelength was set to 230 nm and the emission wavelength to 420 nm.

3.6 Separation of Acyl-CoA Species Using Liquid Chromatography and Analysis by Tandem Mass Spectrometry with MRM

The fatty acyl-CoAs are quantified by liquid chromatography–electrospray ionization tandem mass spectrometry (LC-ESI MS/MS) in positive ion mode using multiple reaction monitoring (MRM) pairs (see Table 1). Fragmentation of [M+H]+ precursor ions (selected with Q1) give rise to the structure-specific product ions (selected with Q3) resulting from a neutral loss of 507.0 Da (Fig. 1) [11] (see Notes 8–10). 1. Sample and EXSTD acyl-CoA thioesters are separated by reverse-phase liquid chromatography-tandem MS/MS with MRM, operated in positive mode. 2. The degassed mobile phase is supplied to an Eclipse XDB-C18 column (3.0  100 mm; 3.5 μm particles; Agilent) at a column oven temperature of 25  C. 3. Samples and EXSTD standards are maintained at 15  C in the autosampler prior to the injection of 35 μL.

212

Richard P. Haslam and Tony R. Larson

Table 1 Mass spectrometer (QTRAP 4000, SCIEX) settings for MRM analysis of acyl-CoA species Acyl-CoAa

Q1 Mass (m/z)

Q3 Mass (m/z)

Time (ms)

DP (V)

CE (V)

CXP (V)

4:0

838.3

301.3

10

180

50

13.3

6:0

866.3

329.3

10

180

50

13.3

8:0

894.3

387.3

10

180

50

13.3

10:0

922.3

415.3

10

180

50

13.3

12:0

950.3

443.3

10

180

50

13.3

14:0

978.3

471.3

10

180

50

13.3

14:1

977.3

470.3

10

180

50

13.3

15:0

992.4

485.4

10

180

50

13.3

16:0

1006.4

499.4

10

180

50

13.3

16:1

1004.4

497.4

10

180

50

13.3

16:3

1000.4

493.4

10

180

50

13.3

16:4

998.4

491.4

10

180

50

13.3

17:0

1020.4

513.4

10

180

53

13.5

18:0

1034.4

527.4

10

190

52

14.3

18:1

1032.4

525.4

10

185

52

13.9

18:2

1030.4

523.4

10

180

52

14

18:3

1028.4

521.4

10

180

52

14

18:4

1026.4

519.4

10

180

52

14

18:5

1024.4

517.4

10

180

52

14

19:0

1048.4

541.4

10

190

52

15.3

20:0

1062.4

555.4

10

190

52

15.3

20:1

1060.4

553.4

10

190

52

16

20:2

1058.4

551.4

10

190

52

16

20:3

1056.4

549.4

10

190

53

16

20:4

1054.4

547.4

10

190

53

16

20:5

1052.4

545.4

10

190

53

16

21:0

1076.4

569.4

10

190

53

16

22:0

1090.4

583.4

10

190

53

16

22:1

1088.4

581.4

10

190

53

16

22:4

1082.4

575.4

10

190

53

16

22:5

1080.4

573.4

10

190

53

16 (continued)

Mass Spectrometry for Measurement of Acyl-CoA

213

Table 1 (continued) Acyl-CoAa

Q1 Mass (m/z)

Q3 Mass (m/z)

Time (ms)

DP (V)

CE (V)

CXP (V)

22:6

1078.4

571.4

10

190

53

16

23:0

1104.5

597.5

10

210

57

17

24:0

1118.5

611.5

10

210

57

17

24:1

1116.5

609.5

10

208

57

17

25:0

1132.4

625.5

10

220

58

17.5

26:0

1146.4

639.5

10

220

58

17.5

26:1

1144.4

637.5

10

218

58

17.5

28:0

1174.4

667.5

10

220

58

17.5

28:1

1172.4

665.5

10

220

58

17.5

30:0

1202.4

695.5

10

220

58

17.5

30:1

1200.4

693.5

10

220

58

17.5

32:0

1230.4

723.4

10

220

58

17.5

MRM parameters (Declustering Potential, DP; Collision Energy, CE; and Cell Exit Potential, CXP) are presented for LC-MS analysis coupled to a SCIEX 4000 QTRAP mass spectrometer operating with ESI in the positive mode Acyl-CoAs are shown as number of linear carbon atoms:number of double bonds; Q1 mass and Q3 mass, mass selection (m/z) of the quadrupoles Q1 and Q3; Time, dwell time; DP, declustering potential; CE, collision energy; CXP, collision cell exit potential

4. For elution of the fluorescent acyl-CoA thioesters, a gradient of 100% acetonitrile and 15 mM ammonium hydroxide is pumped through the column following impurity elution with the acidic mobile phases. Buffers A, B, and C are described in Subheading 2.7. A linear tertiary gradient program was run as follows: flow 1 mL/min. Gradient: 0 min; 100% B; 5 min; 75% B and 25% A; 11 min; 100% A; 13 min; 100% A; 15 min; 100% C; 18 min; 100% C; 20 min; 100% B and maintain 100% B to 25 min. 5. Tandem MS is performed using a SCIEX 4000 QTRAP mass spectrometer with a turbo electrospray source; ion spray voltage set to 5.5 kV; source temperature 750  C; nebulizing gas at 40 (arbitrary units); focusing gas at 40 (arbitrary units) and curtain gas at 20 (arbitrary units). Declustering potential and collision energy were optimized on a compound-dependent basis (see Table 1). 3.7 Diagnostic Ions for the Identification of Acyl-CoA Species and Data Processing

Targeted MS/MS or Data Dependent Acquisition techniques can be used on any MS system capable of tandem MS/MS, where candidate [M+H]+ ions generated by ESI are isolated for fragmentation (see Note 11). Isolation windows and collision energies

214

Richard P. Haslam and Tony R. Larson

Fig. 1 LC-MS chromatogram and mass spectra of acyl-CoA standards using positive ionization with a triple quadrupole mass spectrometer. (a) LC-MS chromatogram showing MRM of selected acyl-CoA standards; (b) Enhanced resolution (ER) scan of 20:0-CoA at 1062.6 m/z; and (c) Enhanced product ion (EPI) scan of 20:0-CoA (1062.6 m/z) showing fragmentation including the release of the 507.1 fragment (neutral loss). Spectra of (b) and (c) were collected during the infusion (50 μL/min) of standards at 1 pmol/μL

Mass Spectrometry for Measurement of Acyl-CoA

215

required will be instrument specific and should be adjusted appropriately by infusing acyl-CoA standards. Diagnostic ions typically greater than 40% of the relative abundance in the mass spectrum are useful for identifying acyl-CoAs and are described in [9]. Specifically, the presence of an adenosine diphosphate fragment ion at m/z 428 (452 for etheno derivatives) diagnoses the presence of a CoA moiety, and in the same spectrum a variable acylpantethenoic acid fragment is observed that indicates the carbon and double bond number (m/z 513 for 17:0). This value will shift by 14 units for every CH2 group addition or loss, with additional compounded shifts of two units indicating changes in the degree of saturation between carbons. For example, in Fig. 1c, 20:0 has an adenosine diphosphate fragment ion of m/z 428 and a specific fragment of m/z 558 (513 (from 17:0) + 3  14 (from 3  CH2) ¼ 558 (C20)). 1. Peak areas are integrated using SCIEX Analyst Quantitate software (see Note 12) and exported to Microsoft Excel for data handling. A typical seed profile for engineered Camelina sativa is shown in Fig. 2. 2. Peak areas are normalized by reference to response factors calculated from separately injected EXSTD mixtures. 3. The impact of ion suppression on peak area is normalized via corrections to the heptadecanoyl-CoA internal standard (17:0CoA ISTD). 4. Quantification of acyl-CoA thioesters (pmola) is achieved by the following equation:

Fig. 2 Acyl-CoA MRM profile of Camelina sativa developing seed tissue (23 days after flowering) engineered for the production of eicosapentaenoic acid (20:5) (line RRes_EPA) with the introduction of genes encoding elongase and desaturase activities [12]

216

Richard P. Haslam and Tony R. Larson

Acyl‐CoA ester pmola ¼ ðA a =A is Þ  ðpmolis Þ=ðsample massÞ, where Aa is the analyte peak area; Ais is the internal standard peak area; pmolis is spiked pmol of internal standard and sample mass is the mass (amount) of the tissue extracted (mg). Isotopic correction is not typically applied as the analyte and internal standards are similar. 5. The limit of detection (LOD) and the limit of quantification (LOQ) are then defined at a signal-to-noise (S/N) of three or ten, respectively. 6. To correct for potential differences among samples, each acylCoA thioester is normalized to per mg of tissue (fresh or dry weight).

4

Notes 1. This calculation is based on the absorption coefficient for the CoA group of 16,400 mol/cm/L at pH 7. A 50 pmol/μL solution ¼ 0.05 mM ¼ 5  105 M is equivalent to an absorbance of 5  105  16,400 ¼ 0.82 at 260 nm at pH 7. Absorbance values should therefore be close to 0.82. AcylCoA thioesters are prone to hydrolysis and are unstable in strongly acidic or alkaline solutions. 2. Biological samples should be harvested fresh and snap-frozen in liquid nitrogen. For example, harvest two optical density units (2 mL of OD ¼ 1) of yeast cells for analysis. Snap-freeze immediately in liquid nitrogen and store at 80  C. Place leaf material or microalgae pellet (~15–20 mg fresh weight) in pre-weighed 2 mL microfuge tube. Attempts to improve sensitivity by increasing the amount of material extracted should be avoided; in our experience, acyl-CoA recoveries rapidly decrease when extracting greater than 50 mg FW. 3. The extraction method is designed to minimize loss of acylCoA thioesters from the samples by minimizing sample handling volumes and washing steps. It is essential not to discard the interface when removing lipids (Subheading 3.3, step 5), as acyl-CoAs are amphipathic and will reside across this boundary. 4. Pellet microalgae in a 2.0 mL microfuge tube with centrifugation; remove supernatant and resuspend in 400 μL freshly prepared ice-cold extraction buffer. Add 500 μm glass beads (to one fifth volume). Secure in a ball mill as per manufacturer’s instructions and operate for 6 min in total (three bursts of two min). Continue to step 4; volumes should be adjusted, reflecting the starting 400 μL of extraction buffer in the protocol to maintain ratios. The glass beads can be recovered during the centrifugation of step 8 of Subheading 3.3.

Mass Spectrometry for Measurement of Acyl-CoA

217

5. Other methods of cell disruption may be appropriate for and may vary with different microalgae, for example cycles of rapid freeze–thaw, ultrasonication, or mechanical disruption. The efficacy of these approaches must be determined for each species via a comparison of acyl-CoA recovery. 6. Monitor peak shapes for indications of reducing column performance. The alkaline mobile phase A results in breakdown of the column silica. The acid wash is essential for the removal of coeluting and interfering impurities from the chromatography system. 7. Cooler autosampler temperatures should not be used as this will lead to precipitation of SDS out of the sample buffer. 8. Elution is in order of increasing acyl chain length; acyl-CoA thioesters containing unsaturated acyl chains elute before their saturated versions. If the acidic wash procedure is not used the background fluorescence from contaminants in the sample may be too high to allow acyl-CoA peak recognition and quantification. 9. Dilutions of EXSTD are run to prepare standard curves for quantification and determination of response factors. 10. The identities of different acyl-CoA thioesters can be confirmed by electrospray ionization (ESI) mass spectrometry. Individual standard acyl-CoA thioesters and peak fractions are collected during HPLC runs and infused at 50 μL/min into the ESI interface of the 4000 QTRAP mass spectrometer. 11. Higher resolution spectra can enable better precision and direct calculation of molecular formulae; however, the double bond position cannot be assigned using standard MS/MS techniques. 12. It is recommended that manual integration is performed for each sample run as retention times for individual acyl-CoA thioesters vary slightly with column decay over time. Signal deterioration and poor detection limits are often observed in LC–MS analysis of later eluting acyl-CoA thioesters. Depending on chain length, acyl-CoA thioesters vary in polarity; therefore, it is important to use the full range (chain length) of acylCoA standards for response curves to enable accurate quantifications.

218

Richard P. Haslam and Tony R. Larson

Acknowledgments R.P.H. and T.R.L. thank the editors for their comments and support during the drafting of this chapter. R.P.H. thanks BBSRC (UK) for financial support under Institute Strategic Programme Grants BBS/E/C/000I0420 and BBS/E/C/00005207. T.R.L. thanks the Leverhulme Trust for supporting the original method development work. References 1. Bates PD (2016) Understanding the control of acyl flux through the lipid metabolic network of plant oil biosynthesis. Biochim Biophys Acta 1861(9 Pt B):1214–1225 2. Haynes C (2011) Analysis of mammalian fatty acyl-coenzyme A species by mass spectrometry and tandem mass spectrometry. Biochim Biophys Acta 1811(11):663–668 3. Golovko MY, Murphy EJ (2004) An improved method for tissue long-chain acyl-CoA extraction and analysis. J Lipid Res 45:1777–1782 4. Abranko´ L, Williamson G, Gardner S, Kerimi A (2018) Comprehensive quantitative analysis of fatty-acyl-coenzyme A species in biological samples by ultra-high performance liquid chromatography–tandem mass spectrometry harmonizing hydrophilic interaction and reversed phase chromatography. J Chromatogr A 1534:111–122 5. Wang HJ, Hsu FF (2020) Revelation of acyl double bond positions on fatty acyl coenzyme A esters by MALDI/TOF mass spectrometry. J Am Soc Mass Spectrom. https://doi.org/10. 1021/jasms.9b00139 6. Taylor DC, Weber N, Hogge LR, Underhill EW (1990) A simple enzymatic method for the preparation of radiolabeled erucoyl-CoA and other long-chain fatty acyl-CoAs and their characterisation by mass spectrometry. Anal Biochem 184:311–316

7. Kawaguchi A, Yoshimura T, Okuda S (1981) A new method for the preparation on acyl-CoA thioesters. J Biochem 89:337–339 8. Ishizaki K, Larson TR, Schauer N, Fernie AR, Graham IA, Leaver CJ (2005) The critical role of Arabidopsis electron-transfer flavoprotein: ubiquinone oxidoreductase during darkinduced starvation. Plant Cell 17:2587–2600 9. Larson TR, Graham IA (2001) Technical advance: a novel technique for the sensitive quantification of acyl CoA esters from plant tissues. Plant J 25:115–125 10. Mancha M, Stokes GB, Stumpf PK (1975) Fat metabolism in higher plants. The determination of acyl-acyl carrier protein and acyl coenzyme A in a complex lipid mixture. Anal Biochem 68:600–608 11. Haynes CA, Allegood JC, Sims K, Wang EW, Sullards MC, Merrill AH (2008) Quantitation of fatty acyl-coenzyme As in mammalian cells by liquid chromatography-electrospray ionization tandem mass spectrometry. J Lipid Res 49:1113–1125 12. Ruiz-Lopez N, Haslam RP, Napier JA, Sayanova O (2014) Successful high-level accumulation of fish oil omega-3 long-chain polyunsaturated fatty acids in a transgenic oilseed crop. Plant J 77:198–208

Chapter 13 Quantification of Acyl-Acyl Carrier Proteins for Fatty Acid Synthesis Using LC-MS/MS Lauren M. Jenkins, Jeong-Won Nam, Bradley S. Evans, and Doug K. Allen Abstract The fatty acid biosynthetic cycle is predicated on an acyl carrier protein (ACP) scaffold where two carbon acetyl groups are added in a chain elongation process through a series of repeated enzymatic steps. The chain extension is terminated by hydrolysis with a thioesterase or direct transfer of the acyl group to a glycerophospholipid by an acyltransferase. Methods for analysis of the concentrations of acyl chains attached to ACPs are lacking but would be informative for studies in lipid metabolism. We describe a method to profile and quantify the levels of acyl-ACPs in plants, bacteria and mitochondria of animals and fungi that represent Type II fatty acid biosynthetic systems. ACPs of Type II systems have a highly conserved Asp-Ser-Leu-Asp (DSLD) amino acid sequence at the attachment site for 40 -phosphopantetheinyl arm carrying the acyl chain. Three amino acids of the conserved sequence can be cleaved away from the remainder of the protein using an aspartyl protease. Thus, partially purified protein can be enzymatically hydrolyzed to produce an acyl chain linked to a tripeptide via the 40 -phosphopantetheinyl group. After ionization and fragmentation, the corresponding fragment ion is detected by a triple quadrupole mass spectrometer using a multiple reaction monitoring method. 15N isotopically labeled acyl-ACPs generated in high amounts are used with an isotope dilution strategy to quantify the absolute levels of each acyl group attached to the acyl carrier protein scaffold. Key words Acyl-carrier protein, Aspartyl protease, Fatty acid synthesis, Isotope dilution, Lipid metabolism, Liquid chromatography, Mass spectrometry, Multiple reaction monitoring

1

Introduction Fatty acids are energy dense molecules important for many biological processes. As an energy reserve, fatty acids comprise most of the carbon in the storage lipid, triacylglycerol, and as components of phospho- and galactolipids, the fatty acids establish the molecular species of membrane lipids and affect fluidity [1]. Acyl lipids can also act as signaling molecules and regulate gene expression and protein activity [2, 3]. In plants, fatty acids are synthesized predominantly in the chloroplast and to a lesser extent in the mitochondria [4–6]. Acyl chains are elongated by two carbons at a time through a series of reactions starting with condensation of a

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_13, © Springer Science+Business Media, LLC, part of Springer Nature 2021

219

220

Lauren M. Jenkins et al.

two-carbon acetyl group to produce the ketoacyl derivative that is subsequently reduced, dehydrated, and reduced again to produce a fully saturated acyl chain (Fig. 1) [7]. In plants, these reactions are repeated until the fatty acid is 16–18 carbons in length [8]. The elongating acyl chains are shuttled through the fatty acid synthesis reactions by a scaffold protein called acyl-carrier protein (ACP). A certain proportion of the 18:0-ACP is desaturated by stearoyl-ACP desaturase to produce oleoyl-ACP (18:1-ACP). Chain elongation is terminated by a thioesterase or acyltransferase. A majority of the fatty acids are cleaved by a thioesterase and exported from the chloroplast and reactivated to acyl-CoAs to be used in lipid production at the ER [1, 9, 10]. Alternatively, an acyltransferase can directly

Fig. 1 Fatty acid metabolism in a plant cell. Fatty acid synthesis connects central carbon metabolism with the production of lipids. Acyl-chains are produced by elongating an acyl-ACP backbone with two carbons at a time, involving a cycle of four reactions per acetyl addition. The primary products of fatty acid synthesis are fully saturated 16 and 18 carbon acyl-ACPs (16:0- and 18:0-ACP) and monounsaturated 18 carbon acyl-ACP (18:1-ACP). Most of these acyl chains are cleaved from the ACP protein via acyl-ACP thioesterases (FATA/FATB) and the free fatty acids (i.e., nonesterified fatty acids. NEFA) produced are exported from the chloroplast and used to synthesize lipids in the endoplasmic reticulum (ER). Alternatively, the acyl chains can be directly transferred from ACP to glycerol-3-phosphate (G3P) and lysophosphatidic acid (LPA) by G3P and LPA acyltransferases (GPAT and LPAAT) to form phosphatidic acid (PA)

Acyl-ACP Quantification by LC-MS/MS

221

transfer fatty acyl chains from acyl-ACPs to glycerol-3-phosphate (G3P) and lysophosphatidic acid (LPA) in the chloroplast as part of the prokaryotic Kennedy pathway [1, 11–18]. Fatty acids as major components of lipid molecules are a significant sink for carbon and therefore represent the crucial node connecting primary and lipid metabolism. The current methods of analyzing acyl-ACPs are confounded by low abundance, multiple isoforms of ACPs, and other sources of acyl-chains (i.e., acyl-CoA and nonesterified fatty acids) [6, 19, 20]. We provide a protocol for analyzing fatty acid synthesis intermediates (acyl-ACPs) by using synthesized isotopically labeled standards, an aspartyl protease, and liquid chromatography-tandem mass spectrometry (LC-MS/MS) based on a recent publication [21]. Plants and bacteria carry out fatty acid synthesis using a discrete acyl-carrier protein (9–15 kDa) (Type II FAS), in contrast to yeast and animals that use a large multidomain enzyme (Type I FAS) [22–25]. The ACP involved in Type II FAS contains highly conserved amino acid residues (Asp, Ser, Leu, Asp) at the acyl-chain attachment site. By using an aspartyl protease (Asp-N), the acyl-ACP molecule is effectively reduced in size to a three amino acid peptide with the acyl-chain attached at the serine through a 40 -phosphopantatheine prosthetic group (Fig. 2a). This molecule is sensitively detected and each acyl chain length is well resolved by LC-MS/MS (Fig. 2b) [21]. Acyl-ACPs can be accurately quantified through isotope dilution techniques with enzymatically synthesized acyl-ACPs [21]. The strategy described in this chapter allows for a deeper insight into the steps of fatty acid synthesis. The method is capable of quantifying all acyl-ACPs commonly found in plant tissues and was developed with saturated standards for carbon lengths of 2–18 and monounsaturated 16 and 18 carbon acyl chains [21]. Presumably, other acyl chains could be quantified in a similar way if standards were made and if the basic chemical structure remained the same thus resulting in an equivalent mass loss and fragmentation pattern in the mass spectrometer. Additionally, the 3-hydroxyacyl and 2,3-enoyl fatty acid synthesis intermediates (Fig. 1) can be measured, allowing for precise studies on regulation of the fatty acid synthesis cycle. The method can also be used as a discovery tool to probe for novel acyl-ACP species, such as polyunsaturated hexadecatrienoyl-ACP (16:3-ACP) which was observed in seeds and leaves of Camelina sativa (false flax) and could potentially be involved in novel lipid remodeling events occurring within the chloroplast [21].

222

Lauren M. Jenkins et al.

Fig. 2 Acyl-ACP detection via LC-MS/MS. (a) Acyl-ACP structure with Asp-N protease cleavage sites and the precursor–product fragmentation indicated. The numbers 315.1 and 413.1 are equivalent to the neutral losses after fragmentation on the protein or pantetheine sides, respectively, of the phosphate group from the acylACP. (b) Representative chromatograph of acyl-ACP standards following LC-MS/MS analysis. For each ACP analyte, the MRM signal as shown in Table 1 is plotted. Reproduced from [21] with permission from The Plant Cell (www.plantcell.org; copyright American Society of Plant Biologists). (c) Verification of peak identities. 15N labeled standards (lower panel) are used to validate sample peaks by retention time. 16:1-Δ9cis-ACP and the

Acyl-ACP Quantification by LC-MS/MS

2

223

Materials For all solutions, use ultrapure water (18 MΩ H2O). Use analytical grade chemicals. Clean and autoclave all glassware used for growth media and buffer storage.

2.1 General Material and Equipment

1. UV and visible spectrum compatible cuvettes. 2. Sterile filters for buffer sterilization. 3. 2 mL lined screw cap tubes. 4. Shaking incubator(s) (17 and 37  C). 5. Adjustable wavelength spectrophotometer (UV and visible). 6. Ultrasonic microtip probe (6–12 mm). 7. Heating block and/or water bath for 1.5 and 2 mL tubes. 8. Vacuum centrifugal concentrator. 9. Bead mill. 10. Laminar flow hood.

2.2 Overexpression of apo-ACP and Sfp Transferase in E. coli and Purification

1. 1  2 L baffled shake flask per culture medium, autoclaved. 2. LB Agar growth medium for selection plates: 10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl, 15 g agar, H2O to adjust volume, autoclaved. 3. Sterile petri dishes containing a selective medium for single colony isolation. 4. Sterile inoculation loop(s). 5. 50 mg/mL filter-sterilized kanamycin, store at 20  C. 6. LB Growth Medium: 10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl, H2O to volume, autoclave, store at 4  C. 7. 500 mL flask for starter culture medium preparation. 8. Sterile culture tubes for starter cultures. 9. 1  2 L bottle per buffer (cleaned and autoclaved) for buffer storage. 10. Ni2+ Immobilized Metal Affinity Chromatography (IMAC) Binding Buffer: 10 mM imidazole, 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-NaOH, pH 7.5, 150 mM NaCl, 1 mM dithiothreitol (DTT), 10% (v/v) glycerol, filter-sterilize, store at 4  C.

ä Fig. 2 (continued) 16:1-Δ2trans-enoyl-ACP elongation intermediate (both in orange) are isomers which can be separated by chromatography. Labeling and LC-MS/MS analysis of 15N 16:1-Δ9cis-ACP internal standard identifies 16:1-Δ9cis-ACP as eluting earlier than C16-Δ2trans-enoyl-ACP. The same result is observed for 18:1-Δ9cis-ACP and 18:1-Δ2trans-enoyl-ACP (both in blue). Not drawn to scale

224

Lauren M. Jenkins et al.

11. Ni2+ IMAC elution buffer: 250 mM imidazole, 20 mM HEPES-NaOH, pH 7.5, 150 mM NaCl, 1 mM DTT, 10% (v/v) glycerol, filter-sterilize, store at 4  C. 12. Desalting/Storage Buffer: 20 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM DTT, 10% (v/v) glycerol, filter-sterilize, store at 4  C. 13. 10 3-Morpholinopropane-1-sulfonic acid (MOPS) base: In the order listed, 0.4 mM MOPS, 0.04 mM Tricine–KOH, pH 7.4, 0.5 mM NaCl, 2.76 mM K2SO4, 0.005 mM CaCl2, 5.25 mM MgCl2, 10 Hunter’s Trace Element Solution (Chlamydomonas Resource Center), 10 Gibco® Minimum Essential Medium (MEM), Vitamin solution (10 mg/L choline chloride, 10 mg/L calcium D-pantothenate, 10 mg/L folic acid, 10 mg/L nicotinamide, 10 mg/L pyridoxal hydrochloride, 1 mg/L riboflavin, 10 mg/L thiamine hydrochloride, 20 mg/L myoinositol, 850 mg/L sodium chloride) in water, filter-sterilize, store at 20  C (see Subheading 3.1, step 12). 14. Modified MOPS minimal medium (for isotopic labeling): 10 MOPS base, 0.528 g/L 15NH4Cl, 1.32 mM monopotassium phosphate (KH2PO4), and 0.2% glucose (sterile) in water, filter-sterilize. 15. 1 M Isopropyl β-D-1-thiogalactopyranoside (IPTG), filtersterilize, store at 20  C. 16. 1  250 mL centrifuge bottle per culture, preweigh. 17. 1  50 mL round bottom centrifuge tube per culture, preweigh. 18. Lysozyme, store at 20  C. 19. cOmplete™ Mini EDTA-free Protease Inhibitor Tablets (1 tablet per 10 mL), store at 4  C. ¨ KTA FPLC system (GE Healthcare), or equivalent, with 20. A 2 and 20 mL sample loops. 21. 5 mL HisTrap™ High Performance column. 22. 2  5 mL HiTrap™ Desalting column. 23. 3 and 20 mL Luer lock syringes. 24. Amicon® Ultra Centrifugal Filters (3000 and 10,000 molecular weight cutoff, MWCO). 2.3 Acyl-ACP Standard Synthesis

1. Dimethyl sulfoxide (DMSO). 2. 500 mM MOPS–NH4OH stock solution (pH 6.5). 3. 50 mM Tris(2-caroxyethyl)phosphine hydrochloride (TCEP) solution in water. 4. 1 mM apo-ACP protein (isolated from E. coli BL21 [New England Biolabs] overproducing apo-ACP; see Subheading 3.1) in storage buffer, store at 80  C.

Acyl-ACP Quantification by LC-MS/MS

225

5. 20% (w/v) Polysorbate 20 (synonymous: Tween 20). 6. 5 mM acyl-CoA (see Note 1). 7. 100 mM Magnesium chloride (MgCl2). 8. 100 mM Manganese chloride (MnCl2). 9. 0.2 mM Sfp transferase (isolated from E. coli BL21 [DE3] [Novagen] overproducing Sfp transferase; see Subheading 3.1) in storage buffer, store at 80  C. 2.4 Analysis of Acyl-ACPs by SDS-PAGE

1. 16.5% Mini-PROTEAN® Tris-Tricine Gel (Bio-Rad) or equivalent. 2. 10 Tris/Tricine/SDS Running Buffer: 100 mM Tris–NaOH, 100 mM Tricine, 0.1% (w/v) SDS, pH 8.3 after dilution to 1 in water. 3. Tricine Sample Buffer for Protein Gels: 200 mM Tris–HCl, pH 6.8, 40% (v/v) glycerol, 2% (w/v) SDS, 0.04% (w/v) Coomassie Brilliant Blue G-250. 4. 0.5 M TCEP. 5. Prestained Protein Standard marker proteins. 6. Colloidal Coomassie Brilliant Blue G-250 Stain: 0.05% (w/v) Coomassie Brilliant Blue G-250, 5% (w/v) aluminum sulfate(14–18)-hydrate [Al2(SO4)3  (14–18) H2O] (or 10% [w/v] ammonium sulfate), 10% (v/v) ethanol (96%), and 2% (v/v) orthophosphoric acid (85%) (see Note 2). 7. Destaining Solution: 10% (v/v) ethanol (96%) and 2% (v/v) orthophosphoric acid (85%).

2.5 Extraction of Acyl-ACP Standards

1. 20% (w/v) trichloroacetic acid (TCA). 2. 1% (w/v) TCA. 3. 50 mM MOPS–NH4OH (pH 7.5) with 5% (v/v) DMSO.

2.6 Extraction of Acyl-ACP from Plant Tissues

1. Liquid nitrogen for sample collection. 2. 2 mL lined screw cap tubes (2 per sample). 3. Metal grinding beads. 4. 100% (w/v) TCA solution (minimum 0.3 μL per sample) (for later use to make 5% and 1% solutions). 5. 5% (w/v) TCA (minimum 1.5 mL per sample). 6. 1% (w/v) TCA (minimum 2.5 mL per sample). 7. 50 mM MOPS–NH4OH (pH 7.5) with 5% (v/v) DMSO (minimum 1.6 mL per sample). 8. pH strips. 9. Gel loading pipette tips. 10. 0.8 μm polyethersulfone (PES) centrifuge filters.

226

Lauren M. Jenkins et al.

2.7 Endoproteinase Asp-N Digestion

1. Endoproteinase Aspartate-N (Asp-N) (Sigma-Aldrich, from Pseudomonas fragi, P3303-1VL; lyophilized powder) in chilled ddH2O, prepare immediately before use, keep on ice. 2. Methanol. 3. 0.5 M TCEP solution.

2.8

LC-MS/MS

1. HPLC vials (with inserts). 2. Shimadzu HPLC, or equivalent. 3. AB SCIEX QTRAP® 6500 mass spectrometer, or equivalent. 4. Column oven. 5. Discovery® BIO Wide Pore C18 reversed phase column (10 cm  2.1 mm; 3 μm) (Millipore Sigma), or equivalent. 6. Buffer A: acetonitrile/10 mM ammonium formate and formic acid, pH 3.5 (10:90, v/v). 7. Buffer B: acetonitrile/10 mM ammonium formate and formic acid, pH 3.5 (90:10, v/v).

2.9

Software

1. Fast Protein Liquid Chromatography (FPLC) control and evaluation software: Unicorn™ (GE Healthcare), or equivalent. 2. LC-MS/MS Integration software: Analyst (SCIEX), or equivalent. 3. Image processing software: ImageJ, or equivalent. 4. Microsoft® Office Excel, or equivalent.

3

Methods

3.1 Overexpression of apo-ACP and Sfp Transferase in E. coli and Harvesting of Cells

An isotope dilution-based quantitation relies on individual standards for each analyte being measured (see Subheading 3.9, steps 6–10). However, no acyl-ACP standards are commercially available. This approach uses recombinant proteins: Sfp transferase (Sfp, WP_003234549) from Bacillus subtilis and apo-ACP (AcpP, NP_415612) from Escherichia coli expressed in E. coli expression strains to enzymatically synthesize acyl-ACP standards (Fig. 3). The Sfp transferase and apo-ACP described in this method are expressed as C-terminal hexahistidine (His6) fusion proteins. The plasmids contain ColE1 origins, kanamycin resistance cassettes, and inducible lac promoters. To produce the unlabeled apo-ACP substrate and the Sfp transferase required for acyl-ACP standard synthesis, LB growth medium is used. However, the 15N labeled apo-ACP substrate is obtained using a minimal medium with 15NH4Cl as nitrogen source. Growth of E. coli within this medium results in 15N labeled proteins. For the following steps, perform all inoculations under sterile conditions in a laminar flow hood and use flasks that are at least two times the volume of the media being prepared.

Acyl-ACP Quantification by LC-MS/MS

227

Fig. 3 Acyl-ACP standard synthesis using Sfp transferase. Sfp transferase covalently transfers the 40 -phosphopantetheine group with an acyl chain attached (highlighted in blue) from acyl-CoA to apo-ACP at a conserved serine residue. Reproduced from [21] with permission from The Plant Cell (www.plantcell.org; copyright American Society of Plant Biologists)

1. Generate LB agar medium for single colony isolation and autoclave. 2. Cool the medium to approximately 60  C and then add kanamycin to 50 μg/mL final concentration. 3. Pour medium into plates (~30 mL per plate) in a laminar flow hood and allow to cool for approximately 30 min with the lids slightly off to avoid condensation. 4. Use plates immediately or seal with Parafilm and store at 4  C. 5. Plate E. coli expression strains (for expression of apo-ACP or Sfp transferase) for single colony isolation using a streak plate method. 6. Incubate plates, inverted, at 37  C for 24 h.

228

Lauren M. Jenkins et al.

7. Generate 1 L of LB growth medium in a 2 L flask to express the unlabeled proteins and 250 mL in a 500 mL flask to aliquot to starter cultures (see Note 3). 8. Autoclave the LB medium, baffled flasks, and buffer storage bottles. 9. Store the 1 L of LB medium at 4  C until needed. Allow the 250 mL LB medium to reach room temperature before using. 10. For 15N labeled apo-ACP: Make 500 mL 10 MOPS medium by adding components in the order listed (see Subheading 2.2, item 13 for concentrations): add MOPS, Tricine, NaCl, and water to 150 mL, adjust pH to 7.4 with 10 M KOH as necessary. Add K2SO4, CaCl2, MgCl2, Hunters Trace Element Solution, MEM vitamin solution and water to a final volume of 500 mL. 11. Filter-sterilize 10 MOPS base medium, aliquot 100 mL aliquots, and store at 20  C. 12. Start the production process by aliquoting 5 mL of the LB medium for starter cultures to sterile culture tubes and add 5 μL of 50 mg/mL kanamycin and inoculate with a single colony from a selection plate to serve as starter culture (starter cultures for 15N labeling of apo-ACP are also made with LB medium). Make a separate starter culture for each overexpression condition (i.e., one each for unlabeled apo-ACP and Sfp transferase, and for 15N culturing of apo-ACP expressing cells). 13. Incubate starter cultures overnight at 37  C with shaking at 100–250 rpm. 14. The following day, warm 1 L LB growth medium to room temperature before inoculation. 15. Make 1 L of 1 Modified MOPS Minimal medium (for 15N isotopic labeling of apo-ACP expressing cells), filter-sterilize, and transfer to an autoclaved baffled flask (see Note 4). 16. Add 1 mL of 50 mg/mL kanamycin and 5 mL starter cultures to each 1 L growth medium (LB medium or 1 Modified MOPS Minimal medium). 17. Incubate the cultures at 37  C with shaking at 100–250 rpm (see Note 5). 18. Monitor the O.D. at 600 nm until it reaches approximately 0.6 (between 0.4 and 0.8 is acceptable) (see Note 6). 19. Cool the cultures by placing them on ice or at 4  C. 20. Add 1 mL of 1 M IPTG to 1 L cultures to induce expression. 21. Reduce the incubation temperature to 16–18  C and continue shaking at 100–250 rpm for expression overnight.

Acyl-ACP Quantification by LC-MS/MS

229

22. The next morning, place the cultures on ice (or at 4  C) and keep on ice from this point onward. 23. Preweigh 50 mL round bottom and/or 250 mL centrifuge tubes (one per culture). 24. For each culture, spin down the cells at 4  C in 250 mL centrifuge tubes for 5 min at 15,344  g, dump out supernatant (preserve cell pellet), add more culture to bottle and repeat until the cells of the entire culture are concentrated in one pellet. 25. Consider transferring the cell pellets to preweighed 50 mL round bottom centrifuge tubes or maintain in the 250 mL centrifuge tubes (see Note 7). 3.2 Purification of apo-ACP and Sfp Transferase by Ni2+ IMAC Chromatography

Apo-ACP and Sfp transferase that have been expressed with C-terminal His6 fusions are isolated by Ni2+ IMAC affinity purification. The proteins are retained on 5 mL HisTrap™ High Performance Columns (GE Healthcare) and eluted by increasing the concentration of imidazole over time. 1. Make Ni2+ IMAC Binding Buffer, Ni2+ IMAC Elution Buffer, and Desalting/Storage Buffer. 2. Filter-sterilize buffers and store in autoclaved bottles at 4  C. 3. Reweigh the tubes with the E. coli cells expressing apo-ACP or Sfp transferase to obtain a pellet mass. 4. Add 3 mL of Ni2+ IMAC Binding Buffer per 1 g of cells, then add 10 mg/1 mL of lysozyme and 1 cOmplete™ Mini EDTAfree Protease Inhibitor Tablet per 10 mL. 5. Resuspend cells by pipetting up and down, vortexing, or shaking for 30 min (see Notes 8 and 9). 6. Return the tubes to ice. 7. Sonicate for 1 min at 50% duty cycle at the microtip limit followed by 1 min of cooling on ice, repeat six times. 8. Spin down cell debris by centrifuging at 4  C at 48,384  g for 30–45 min and collect the soluble fraction for protein purification. ¨ KTA FPLC system, equil9. For Ni2+ IMAC purification on the A ibrate the pumps and the 5 mL HisTrap™ High Performance Column with Ni2+ IMAC Binding Buffer. 10. Attach 20 mL sample loop. 11. Elution buffer gradient: 0–200 mL, 0% Ni2+ IMAC Elution buffer; 200–210 mL, 0–100% Ni2+ IMAC Elution buffer; 210–220 mL, 100% Ni2+ IMAC Elution buffer; 220–255 mL, 0% Ni2+ IMAC Elution buffer; 255 mL, stop.

230

Lauren M. Jenkins et al.

12. Flow rate profile: 0–10 mL, 5 mL/min; 10–30 mL, 1 mL/ min; 20–200 mL, 5 mL/min; 200–220 mL, 1 mL/min; 220–255, 5 mL/min; 255 mL, stop. 13. Set valve position to inject at 10 mL, set valve position to load at 200 mL. 14. Collect 1 mL fractions from 200 to 220 mL. 15. Measure the absorbance at 280 nm for each fraction. 16. Combine the protein containing fractions, indicated by an increase in absorbance at 280 nm. 3.3 Concentration and Desalting of Recombinant apo-ACP and Sfp Transferase

1. Concentrate combined protein fractions of Ni2+ IMAC column purification (see Subheading 3.2) to less than 2 mL using Amicon® Ultra Centrifugal Filters (3000 and 10,000 MWCO for apo-ACP and Sfp transferase, respectively). 2. For desalting, use 2  5 mL HiTrap™ Desalting Columns in ¨ KTA FPLC system: tandem connected to the A 3. Equilibrate the pumps and column to Desalting/Storage Buffer. 4. Connect 2 mL sample loop. 5. Elute proteins in Desalting/Storage Buffer. 6. Flow rate profile: 0–25 mL, 5 mL/min; 25–37 mL, 1 mL/min; 37–62 mL, 5 mL/min; 62 mL, stop. 7. Set valve position to inject at 25 mL. 8. Collect 1 mL fractions from 25 to 37 mL. 9. Measure the absorbance at 280 nm for each fraction. 10. Combine the protein containing fractions and concentrate to less than 2 mL using Amicon® Ultra Centrifugal Filters (3000 and 10,000 MWCO for apo-ACP and Sfp transferase, respectively). 11. Measure the absorbance at 280 nm (A280) using UV-transparent cuvettes with the storage buffer as blank. 12. Calculate the molar concentration (c) of the protein by dividing A280 by the molar extinction coefficient (ε) and the pathlength of the cuvette (b) (see Note 10). c¼

A 280 : εb

13. Run a small amount of protein with several dilutions on an SDS-PAGE gel (see Subheading 3.6) to validate the purification of the protein and determine purity. 14. Dilute the apo-ACP to 1 mM and the Sfp transferase to 200 μM with storage buffer. 15. Aliquot (e.g., 500 μL) and store at 80  C.

Acyl-ACP Quantification by LC-MS/MS

3.4 Acyl-ACP Standard Synthesis

231

1. Per 100 μL of reaction add 14 μL (for standards up to ten carbons in chain length) or 10 μL (for standards 12 or more carbons in chain length) of water, 10 μL of DMSO (10%, v/v final concentration), 10 μL of 500 mM MOPS–NH4OH (pH 6.5) (50 mM final) 8 μL of 50 mM TCEP (4 mM final), 20 μL of 1 mM recombinant apo-ACP (200 μM final), 4 μL of 20% (w/v) Polysorbate 20 (1%, v/v, final) to reactions for standards of 12 carbon chain lengths or longer only (omit Polysorbate 20 from reactions up to ten carbons in chain length),and 10 μL of the respective 5 mM acyl-CoA (500 μM final). Mix thoroughly (see Note 11). Add 8 μL of 100 mM MgCl2 (10 mM final) and 8 μL of 100 mM MnCl2 (10 mM final). Mix thoroughly. Add 12 μL of recombinant Sfp transferase (24 mM final). Mix well and quickly spin the reaction tube to collect liquid from the sides of the tube. 2. React at 37  C for 3 h with vortexing every hour. 3. Quench by precipitating the acyl-ACPs from the reaction by adding TCA and preforming Clean Up of Acyl-ACP Standards (see Subheading 3.7). 4. Determine reaction efficiency by SDS-PAGE analysis (see Subheading 3.6). 5. Calculate the concentration of acyl-ACP (Cacyl) in each reaction by multiplying the fraction of acyl-ACP (determined from densitometry of the corresponding lane in the SDS gel) by the concentration of the ACP in the reactions. C acyl ¼

A acyl  C f ðACPÞ : Aholo þ A acyl

Example : Cacyl ¼ 0.80  200 μM ¼ 160 μM. 3.5 Making the Calibration Curve for Quantification of Acyl-ACPs 3.5.1 Single Point Quantification for Estimation

Before making the calibration curve and performing isotope dilution-based quantification, perform a single point quantification analysis on your samples to estimate the amount of internal standard necessary. Though a multipoint isotope dilution-based quantification will measure across a range of peak area ratios of analyte to internal standard, it is good practice to operate at a ratio as close to 1 as possible. 1. Isolate acyl-ACPs from plant tissues (see Subheading 3.8). At step 2, add a range of concentrations of 15N labeled acyl-ACP standards for each acyl-ACP to be quantified that spans 1–2 orders of magnitude (e.g., 0.01–1 μM final concentration) to 30–40 mg fresh weight (FW) of replicate biomass samples. 2. Carry out LC-MS/MS Analysis (see Subheading 3.10). 3. Estimate the amount of 15N standard needed for each acyl-ACP species for isotope dilution (see Subheading 3.11, steps 1–5).

232

Lauren M. Jenkins et al.

3.5.2 15N Labeled Standard Mix/15N Standard Diluent

1. Make two separate stock mixes of 15N labeled acyl-ACP standards at 10 times the concentration determined for quantification (the calculated estimate from the single point quantification in Subheading 3.5.1). One stock will contain each 15N labeled standard with acyl chains up to ten carbons in length (10 short 15N standard mix, usually contains: apo form and C2, C4, C6, C8, C10 saturated acyl chains), and a second stock will contain each 15N labeled standard with acyl chains of 12 carbons or longer (10 long 15N standard mix, usually contains: C12, C14, C16, C18 saturated and C16:1 and C18:1 acyl chains). These two acyl-ACP standard mixes will serve as the 15N labeled stock mixes used for the calibration curve as well as for internal standards in plant tissue samples. (Fig. 4a, step 1). 2. Calculate the total volume of 1 15N standard needed for serial dilution of the calibration curve (see Subheading 3.5.4) and carry out steps 3–5 (this section, below) to achieve this volume in step 5. 3. Digest the 10 short and long 15N labeled acyl-ACP mixes separately as outlined in Subheading 3.9, steps 1–3. For step 2 of Subheading 3.9, digest an aliquot of the long and short 15N acyl-ACP standard mixes at 4 concentrations with ratios of 1:50 (short mix) or 1:20 (long mix) of Asp-N enzyme and acyl-ACPs (w/w) in 50 mM MOPS–NH4OH (pH 7.5) with 5% (v/v) DMSO to volume (fourth of the volume calculated in step 2 of this section) (Fig. 4a, step 2). 4. For each mix, add methanol to a final concentration of 50% (v/v). This will bring the two standard mix concentrations to 2 each (Fig. 4a, step 3). 5. Mix equal volumes of the 2 short 15N standard mix and 2 long 15N standard mix together to produce the 1 15N standard diluent (Fig. 4a, step 4). This is called “1 15N standard diluent” because it is later used to dilute the unlabeled acylACP standard mix to generate the standard curve (Fig. 4b, c).

3.5.3 Unlabeled Standard Mix

1. Calculate the volume of each individual acyl-ACP standard (VInd Std) necessary for your most concentrated point of the standard curve (dilution point 1) by multiplying the final concentration (Cf) by the final volume (Vf) and dividing by the concentration of acyl-ACP (Cacyl; calculated in Subheading 3.4, step 5). Include an unlabeled standard for each acyl-ACP to be quantified in the biological sample. V Indv Std ¼

Cf  V f : C acyl

For example, for a dilution point 1 with 5 μM of each standard in a final volume of 400 μL:

Acyl-ACP Quantification by LC-MS/MS

233

Fig. 4 Generating calibration curves for quantification. (a) Steps 1–4 (in red), making the 15N labeled standard mix/diluent (Subheading 3.5.2). Steps 5 and 6, preparing the Unlabeled Standard Mix (Subheading 3.5.3). Steps 7 and 9, serial dilution of the unlabeled standards into the 15N standard diluent (Subheading 3.5.4). Repeat step 9 until the full standard curve is made. MeOH, methanol; L, long chain acyl-ACPs (acyl chains of 12 carbons or longer); S, short chain acyl-ACPs (acyl chains up to 10 carbons); NA, not applicable; Unlb, unlabeled. (b) Table showing fold dilution of each dilution point. (c) Graphical representation of standard concentrations

234

Lauren M. Jenkins et al.

V Indv Std ¼

5 μM  400 μL ¼ 12:5 μL each standard: 160 μM

2. Add the calculated amount of each unlabeled acyl-ACP standard into two separate mixes, one for standards with acyl chains up to ten carbons in length (short unlabeled standard mix), and a second for standards with acyl chains of 12 carbons or longer (long unlabeled standard mix) (Fig. 4a, step 5). 3. Digest each mix as outlined in Subheading 3.9, steps 1–3 (Fig. 4a, step 5). 4. Validate Asp-N protease digest via SDS-PAGE. Include a negative control (undigested acyl-ACP). A successful digestion results in the absence of the acyl-ACP band (Fig. 5).

Fig. 5 SDS-PAGE analysis of acyl-ACP standards. Acyl-ACP separation on a 16.5% Tris-Tricine SDS-PAGE gel with lane by lane densitometry analysis using ImageJ software. The ACP’s acidity and acyl-chain lengths both affect migration patterns. Acylated forms of ACP with chains of four carbons or more migrate faster on the gel than apo-ACP, though holo-, acetyl-, and malonyl-ACP migrate slower and cannot be distinguished from each other by migration pattern. Acyl chains of 10 to 18 carbons in length are also indistinguishable from each other based on migration patterns. After scanning the gel and densitometric image analysis, integrate the signal in the lane to obtain the peak areas for each band (e.g., of 10:0-ACP reaction analysis in blowout pane). Use the peak areas to determine the fraction of contaminating holo-ACP (0.20, i.e., 20%) and validate reaction efficiency (0.80, i.e., 80% 10:0-ACP). Reaction efficiency is observed by the complete disappearance of the apo-ACP substrate that has been turned into acyl-ACP

Acyl-ACP Quantification by LC-MS/MS

235

5. For each mix, add methanol to a final concentration of 50% (v/v) and dry down using a vacuum centrifugal concentrator (Fig. 4a. step 6). 3.5.4 Serial Dilution

1. Resuspend both the short unlabeled standard mix and the long unlabeled standard mix that were dried down (from Subheading 3.5.3, step 5) to a 2 final concentration (e.g., 10 μM) in the 1 15N standard diluent produced in Subheading 3.5.2, step 5 (Fig. 4a, step 7). 2. Mix equal volumes of the 2 short and long unlabeled standard mixes together to obtain a 1 total unlabeled standard mix in 1 15N standard diluent (this is dilution point 1) (Fig. 4a, step 8). 3. Serially dilute this 1 total unlabeled standard mix from step 2 into the 1 15N standard diluent (Fig. 4a, step 9) to obtain 10–12 dilutions that span six orders of magnitude (Fig. 4b, c) (see Notes 12 and 13).

3.5.5 Example Standard Curve Generation

1. Individual unlabeled and 15N labeled acyl-ACP standards are synthesized from 200 μM purified E. coli ACP and commercial 500 μM acyl-CoA substrates (Subheading 3.4). 2. SDS-PAGE analysis (Fig. 5) determines that all apo-ACP (80% of total; remaining 20% is holo-form) is converted to acyl-ACP form (160 μM) (i.e., each acyl-ACP as an individual stock at 160 μM). 3. A single point quantification with 0.05, 0.5 and 5 μM 15N acylACP standards per 30 mg biomass in a final volume of 100 μL determines each sample analyte to be close to 0.1 μM (i.e., biomass contains ~10 pmol per analyte per 30 mg FW). 4. Make the 10 short and long 15N labeled acyl-ACP standard stock mixes, combining the individual 15N acyl-ACP standards: Two separate 15N acyl-ACP standard mixes are made at 10 the concentration needed for the plant sample, one containing 25 μL of each individual 15N acyl-ACP up to ten carbons (10 short 15N standard mix) in length and one containing 25 μL of each individual 15N acyl-ACP 12 carbons or longer (10 long 15 N standard mix) and bringing both mixes to 4 mL with 50 mM MOPS–NH4OH (pH 7.5) and with 5% (v/v) DMSO (Fig. 4a, step 1). C f ¼ 10  0:1 μM ¼ 1 μM each: Vol indv std ¼

C f  V f 1 μM  4 mL ¼ 0:025 mL ¼ 25 μL each: ¼ 160 μM C acyl

5. An aliquot of each 15N acyl-ACP standard mix is used to make the 1 15N standard diluent for the standard curve (see step 7

236

Lauren M. Jenkins et al.

below), the rest is put aside for use as internal standards for quantification in biological samples. 6. 2.8 mL of 1 (here, 0.1 μM as determined in step 3) 15N standard diluent is needed for the standard curve in Fig. 4b, c. 7. Approximately 3 mL of 1 (0.1 μM) 15N standard diluent is made by first separately diluting 300 μL of each 10 (1 μM) short and long 15N labeled acyl-ACP mixes to 750 μL by adding Asp-N at 1:50 and 1:20 (w/w) to short and long mixes, respectively and bring to volume with 50 mM MOPS– NH4OH (pH 7.5) containing 5% (v/v) DMSO (Fig.4a, step 2). The mixes are then at 4 (0.4 μM) concentration and are reacted. Then 750 μL of methanol is added to quench each reaction (Fig. 4a, step 3)—the two standard mixes are then at 2 (0.2 μM) concentration. 1.48 mL of both the short and the long 15N labeled acyl-ACP mixes are combined to make 1 (0.1 μM) 15N standard diluent (2.96 mL, i.e., ~3 mL) (Fig. 4a, step 4). This is used for the serial dilution in the standard curve. 8. Now, make the unlabeled standard for the standard curve: The amount of each individual unlabeled standard (160 μM) necessary for a final concentration of 5 μM in a final volume of 400 μL for the highest dilution point on the standard curve (dilution point 1) is calculated to be 12.5 μL. V Indv Std ¼

C f  V f 5 μM  400 μL ¼ ¼ 12:5 μL: C acyl 160 μM

9. Make short unlabeled standard mix by adding 12.5 μL of each individual unlabeled acyl-ACP with up to ten carbons in length. Make long unlabeled standard mix by adding 12.5 μL of each individual unlabeled acyl-ACP with 12 carbons or longer are added to another tube. 1:50 and 1:20 (w/w) Asp-N is added to the short and long mixes, respectively, and 50 mM MOPS–NH4OH (pH 7.5) with 5% (v/v) DMSO is added to a final volume of 500 μL and both tubes reacted (Fig. 4a, step 5). Then 500 μL of methanol is added to quench each reaction and the samples are dried using a vacuum centrifugal concentrator (Fig. 4a, step 6). 10. Dilution point 1 (5 μM) is made by resuspending each of the two digested, dried, unlabeled standard mixes in 200 μL of 1 (0.1 μM) 15N standard diluent (Fig. 4a, step 7)—the unlabeled standards are then at 2 (10 μM) concentration. 190 μL of each are combined to a final concentration of 5 μM unlabeled standards in 1 (0.1 μM) 15N standard diluent (Fig. 4a, step 8). This is dilution point 1. 11. The standard curve is generated by serial diluting dilution point 1) using 1 (0.1 μM) 15N standard diluent (Fig. 4a, step 9, b, c).

Acyl-ACP Quantification by LC-MS/MS

3.6 SDS-PAGE Analysis of Synthesized Acyl-ACP Standards

237

1. Dilute 9 parts of acyl-ACP sample (synthesized standard), 1 part 0.5 M TCEP, and 10 parts Tricine Sample Buffer (see Notes 14 and 15). 2. Heat at 60  C for 10 min. 3. Set up the gel electrophoresis chamber with16.5% Mini-PROTEAN® Tris-Tricine Gel (Bio-Rad) or equivalent and 1 Tris/ Tricine/SDS Running Buffer. 4. Load 10 μL (1–3 μg) of acyl-ACP in each lane of the gel and 7 μL Precision Plus Protein™ Dual Xtra Prestained Protein Standard, or equivalent. 5. Run the gel at 90–100 V until the dye front has completely cleared (see Note 16). 6. Clean the gel by placing it in ~200 mL deionized water and microwave for 1 min or alternatively shake for 15 min at room temperature then replace the water and repeat three more times. 7. Add Colloidal Coomassie Brilliant Blue G-250 Stain and microwave for 15 s then shake overnight. 8. De-stain by rinsing several times with deionized water, then add destaining solution and shake gently until the protein bands are clearly visible. 9. Rinse the gel with deionized water. 10. Image the gel using a gel imager with white light. 11. Analyze the image and perform densitometry using image analysis software (Fig. 5 for an example). 12. Obtain the peak areas of densitometric scans for each band and use to assess the fractions of holo-ACP apo-ACP or acyl-ACP and reaction efficiency (Fig. 5).

3.7 Clean-Up of Acyl-ACP Standards by TCA Extraction

Trichloroacetic acid (TCA) precipitation is used for protein concentration and partial purification prior to proteomic or other analysis [26] and effectively cleans up protein by removing contaminating metabolites and other components of biomass. 1. Add 20% (w/v) TCA to the individual acyl-ACP standard reactions (from Subheading 3.4, steps 1 and 2) to a final concentration of 5% (w/v) and vortex. V 20%TCA ¼ V sample  0:333: 2. Centrifuge samples at 14,000 rpm (~15,000–20,000  g depending on rotor) for 10 min at 4  C. 3. Discard the supernatant and add 1 mL of 1% (w/v) TCA to wash the pellet. 4. Centrifuge samples at 14,000 rpm for 10 min at 4  C.

238

Lauren M. Jenkins et al.

5. Discard the supernatant and resuspend the pellet in 50 mM MOPS–NH4OH (pH 7.5) with 5% (v/v) DMSO to the original reaction volume from Subheading 3.4, steps 1 and 2, to achieve a final concentration of (Cf (ACP)) of 200 μM ACP. 3.8 Acyl-ACP Extraction from Plant Tissues

Throughout the extraction process, keep samples on ice and perform all steps (centrifugation, bead milling, and mixing) at 4  C. Minimize disruption of the pellet while handling at all times to reduce losses; it is recommended that gel loading pipette tips are used to remove/collect the supernatant close to the pellet to minimize disruption of the pellet and remove/collect as much of the supernatant as possible. At least three biological replicates are recommended for this analysis. 1. Harvest and homogenize fresh tissue to a fine powder in the presence of liquid nitrogen (see Notes 17 and 18). 2. Add isotopically labeled acyl-ACP standards (see Subheading 3.5.1 for details on amounts), 1.5 mL of 5% (w/v) TCA, and two steel beads to 30–40 mg of frozen, homogenized tissue and then bead mill the samples for 5 min (see Note 19). 3. Remove the beads and centrifuge at 14,000 rpm for 10 min at 4  C. 4. Discard the supernatant and preserve the biomass pellet (see Notes 20 and 21). Then add 1.5 mL 1% (w/v) TCA and two steel beads and vortex (or bead mill if necessary) to break up the pellet. 5. Remove the beads and centrifuge at 14,000 rpm for 10 min at 4  C. 6. Discard the supernatant (see Notes 20 and 22) and suspend the pellet in 1.5 mL 50 mM MOPS–NH4OH (pH 7.6) with 5% (v/v) DMSO, add two steel beads, and vortex well (or bead mill if necessary). 7. Check that the pH is above 6.5 with 1 μL for a few samples and incubate the samples on a mixer for 1 h in a cold room. Use diluted NH4OH to raise the pH as required. 8. Remove the beads and centrifuge at 14,000 rpm for 10 min at 4  C. 9. Collect the supernatant in a new tube (see Note 23) and add 100% (w/v) TCA to achieve a final concentration of 10% (w/v) and gently invert the samples 2–3 times after this addition (do not vortex). V 100%TCA ¼ V 100%TCA ¼

V sample : 9

1:5 mL ¼ 0:167 mL: 9

Acyl-ACP Quantification by LC-MS/MS

239

10. Chill samples to 80  C overnight (or for a minimum of 1 h if continuing extraction the same day). 11. Thaw the samples on ice without mixing and centrifuge at 14,000 rpm for 10 min at 4  C. 12. Discard the supernatant and wash the pellet carefully with 1 mL 1% TCA. Do not vortex. Centrifuge again and return to ice. 13. Discard the supernatant and dissolve the pellet in a minimal volume of 50 mM MOPS–NH4OH (pH 7.6) with 5% (v/v) DMSO. Record the volume added (see Note 24). 14. Check that the final pH is at or above 7.5 but below 9 to allow for optimal protease activity in the following steps. 15. Proceed to endoproteinase Asp-N digestion of acyl-ACPs and sample cleanup (see Subheading 3.9). 3.9 Endoproteinase Asp-N Digestion of Acyl-ACPs and Sample Clean-Up

1. Resuspend endoproteinase Asp-N (lyophilized powder) in chilled water to 0.04 μg/μL just before use (2–5 μL per sample needed). Keep on ice. 2. Add endoproteinase Asp-N to samples at a 1:50–1:20 enzyme– protein (w/w) ratio (see Note 25). 3. Vortex, quickly spin the reaction contents to the bottom of the tube, and digest the samples in a water bath at 37  C overnight (~16 h) with periodic mixing. 4. Quench the reactions by adding 0.5 M TCEP to a final concentration of 1–5 mM and methanol to 50% (v/v). Record the final volume of the sample. 5. Store at 80  C until LC-MS/MS analysis. 6. The day of LC-MS/MS analysis, thaw samples and remove any insoluble materials using 0.8 μm PES centrifuge filters. 7. Transfer the filtrate to LC autosampler vials for analysis.

3.10 LC-MS/MS Analysis

After digestion and sample clean-up, acyl-ACPs are separated on a C18 column with a Shimadzu HPLC or equivalent and detected by LC-MS/MS using an AB Sciex QTRAP 6500 instrument with Analyst software or equivalent. 1. Set up LC-MS/MS with C18 column and appropriate buffers (Subheading 2.8). 2. Set flow rate to 0.2 mL/min, 100% buffer A. 3. LC buffer gradient: 0–4 min, 0–10% B; 4–12 min, 10–100% B, 12–17 min, hold at 100% B; 17–18 min, 100–0% B; 18–27 hold at 0% B, stop at 27 min. 4. MS parameters: positive ion and multiple reaction monitoring (MRM) mode; curtain gas, 30 psi; ion spray voltage, 4.5 kV;

240

Lauren M. Jenkins et al.

source temperature, 400  C; nebulizing gas (GS1), 30 psi; focusing gas (GS2), 30 psi. Routine declustering potential (DP), collision energy, (CE), cell exit potential (CXP), and the m/z parameters for the mass filters (Q1 and Q3) are presented in Table 1. 5. Inject 10 μL of sample. 6. Run three technical replicates (replicate injections of the same sample) for each standard concentration of the calibration curve. 7. During the calibration curve, make a twofold dilution of the lowest detectable concentration (above background) and run to determine true limits of detection (LOD) and quantification (LOQ) values. 8. For subsequent analyses, only a single dilution point that lies within the linear range is necessary to validate the standard curve. Table 1 LC-MS/MS Parameters for acyl-ACPs and elongation intermediates on the AB Sciex QTRAP 6500 mass spectrometer

Acyl-ACP

Acyl Chain

Q1 (m/z)

Q3 (m/z)

DP, CE, CXP (V, eV, V)

RRT (min)

apo

NA

334.2

203.1

61, 15, 24

2.2

Holo

NA

674.2

261.1

80, 40, 10

5.9

Malonyl

C3H3O3

760.2

303.2

90, 50, 10

a

Acetyl

C2H3O

716.3

303.2

80, 40, 10

7.0

3-Ketobutyryl

C4H5O2

758.3

345.2

90, 50, 10

a

3-Hydroxybutyryl

C4H7O2

760.3

347.2

90, 50, 10

7.1

2,3-trans-Butenoyl

C4H5O

742.3

329.2

90, 50, 10

8.0

Butyryl (4:0)

C4H7O

744.3

331.2

90, 40, 10

8.2

3-Ketohexanoyl

C6H9O2

786.3

373.2

100, 50, 10

a

3-Hydroxyhexanoyl

C6H11O2

788.3

375.2

100, 50, 10

8.1

2,3-trans-Hexenoyl

C6H9O

770.3

357.2

100, 50, 10

8.9

Hexanoyl (6:0)

C6H11O

772.3

359.2

100, 40, 10

9.1

3-Ketooctanoyl

C8H13O2

814.3

401.2

100, 50, 10

a

3-Hydroxyoctanoyl

C8H15O2

816.3

403.2

100, 45, 10

8.9

2,3-trans-Octenoyl

C8H13O

798.3

385.2

100, 40, 10

9.8

Octanoyl (8:0)

C8H15O

800.4

387.3

100, 45, 10

10.0

100, 40, 10

a

3-Ketodecanoyl

C10H17O2

842.4

429.3

(continued)

Acyl-ACP Quantification by LC-MS/MS

241

Table 1 (continued)

Acyl-ACP

Acyl Chain

Q1 (m/z)

Q3 (m/z)

DP, CE, CXP (V, eV, V)

RRT (min)

3-Hydroxydecanoyl

C10H19O2

844.4

431.3

100, 40, 10

9.8

2,3-trans-Decenoyl

C10H17O

826.4

413.3

100, 40, 10

10.7

Decanoyl (10:0)

C10H19O

828.4

415.3

100, 45, 10

11

3-Ketododecanoyl

C12H21O2

870.4

457.3

100, 40, 10

a

3-Hydroxydodecanoyl

C12H23O2

872.4

459.3

100, 40, 10

10.7

2,3-trans-Dodecenoyl

C12H21O

854.4

441.3

100, 40, 10

11.7

Dodecanoyl (12:0)

C12H23O

856.4

443.3

100, 45, 10

12.0

3-Ketotetradecanoyl

C14H25O2

898.4

485.3

100, 45, 10

a

3-Hydroxytetradecanoyl

C14H27O2

900.4

487.3

100, 45, 10

11.7

2,3-trans-Tetradecenoyl

C14H25O

882.4

469.3

100, 45, 10

12.8

Tetradecanoyl (14:0)

C14H27O

884.4

471.3

100, 45, 10

13.1

3-Ketohexadecanoyl

C16H29O2

926.5

513.4

100, 45, 10

a

3-Hydroxyhexadecanoyl

C16H31O2

928.5

515.4

100, 45, 10

12.8

2,3-trans-Hexadecenoyl

C16H29O

910.5

497.4

100, 45, 10

14.0

Hexadecanoyl (16:0)

C16H31O

912.5

499.4

100, 50, 10

14.3

cis-9-Hexadecanoyl (16:1)

C16H29O

910.5

497.4

100, 50, 10

13.3

cis, cis-7, 10-Hexadecanoyl (16:2)

C16H27O

908.4

495.3

100, 50, 10

a

cis, cis, cis-7,10,13-Hexadecanoyl (16:3)

C16H25O

906.4

493.3

85, 46, 14

12.0

3-Ketooctadecanoyl

C18H33O2

954.5

541.4

100, 45, 10

a

3-Hydroxyoctadecanoyl

C18H35O2

956.5

543.4

100, 45, 10

14.0

2,3-trans-Octadecenoyl

C18H33O

938.5

525.4

100, 45, 10

15.1

Octadecanoyl (18:0)

C18H35O

940.5

527.4

100, 50, 10

15.5

cis-9-Octodecanoyl (18:1)

C18H33O

938.5

525.4

100, 50, 10

14.4

cis, cis-9, 12-Octadecanoyl (18:2)

C18H31O

936.5

523.4

100, 50, 10

a

cis, cis, cis-9,12,15-Octodecanoyl (18:3)

C18H29O

934.5

521.3

100, 50, 10

a

Reproduced from [21] with permission from The Plant Cell (www.plantcell.org; copyright American Society of Plant Biologists) CE collision energy, CXP collision cell exit potential, DP declustering potential, NA not applicable, Q1 quadrupole 1 mass filter, Q3 quadrupole 3 mass filter, RRT relative retention time a Unvalidated

242

Lauren M. Jenkins et al.

3.11 LC-MS/MS Data Analysis and Quantification

Peaks are integrated using the compatible integration software. Analytes can be estimated by single point quantification or quantified by isotope dilution based quantification, both methods using 15 N-labeled acyl-ACP internal standards. 1. Verify peak identities by internal standards and retention times (Fig. 2b, c). 2. Integrate the peak areas for each analyte and standard using compatible integration software (Fig. 6a, b). 3. For single point quantification (for estimation), obtain the peak area ratios of sample analyte to internal standard. 4. For each analyte use the sample with a ratio closest to 1 to calculate the amount of sample analyte: C analyte ¼

A analyte  C standard : A standard

5. Normalize the calculated concentration to the sample weight:   pmol  sample vol ðμLÞ C analyte μL pmol : ¼ Example : mg sample weight ðmgÞ 6. For isotope dilution-based quantification, plot the ratios of the peak areas of the unlabeled to labeled standards versus the concentrations of the unlabeled standards for each analyte (Fig. 6c). 7. Perform a linear regression analysis on each analyte to generate standard curves, determine linear ranges, and calculate limits of detection (LOD) and quantification (LOQ) (Fig. 6d) (see Notes 26 and 27): 8. LOD and LOQ: Calculate the standard deviation (SD) of the ratio of peak areas for the technical replicates of the lowest concentration of standard with peak areas above background. Multiply the standard deviation by 3.3 (for LOD) and 10 (for LOQ) and divide by the slope (m) of the linear regression line. 3:3  SD : m 10  SD LOQ ¼ : m

LOD ¼

9. Calculate the ratio of peak areas of the sample analyte (unknown) to internal standard (15N labeled) (Fig. 6e). 10. Calculate the sample analyte (unknown) concentration using the linear regression line obtained from the standard curve (Fig. 6f). 11. Normalize the calculated concentration to the sample weight:   pmol  sample vol ðμLÞ C analyte μL pmol : Example : ¼ mg sample weight ðmgÞ

Acyl-ACP Quantification by LC-MS/MS

243

Fig. 6 Isotope dilution-based quantification. (a) Peak area integration of standards from the standard curve. A and B, one unlabeled standard at different concentrations; C, 15N standard. (b) Plots of unlabeled (blue) and 15 N (orange) standard peak areas across the standard curve. A, B and C, peak areas from panel (a); R1 and R2, link between peak areas to calculate ratios of peak areas for unlabeled to 15N standard. (c) Plots of the ratios of peak areas (unlabeled to 15N standard) over the concentration of the unlabeled standard across the standard curve. (d) Plots of the ratios of peak areas (unlabeled to 15N standard) over the concentration of the unlabeled standard across the standard curve with replicate injections. Peak areas of the lowest concentration above background (black box) are used to calculate LOD and LOQ. The linear range is the range in which the peak area increases linearly in response to changes in concentration. (e) Peak area integration of biological sample analyte with internal standard. C, 15N standard; D, sample analyte (concentration unknown). (f) The ratio of sample analyte D (unknown) to internal standard C is used to calculate the concentration of the sample analyte (unknown concentration) using the standard curve

244

4

Lauren M. Jenkins et al.

Notes 1. Acyl-CoA may be dissolved in chilled ddH2O or 50% (v/v) DMSO solution which may increase solubility of longer acylchains. Aliquot and store at 20  C. 2. The order in which each component is added must be maintained to retain stain sensitivity. First dissolve aluminum sulfate in water, then add ethanol and mix, then mix in Coomassie Brilliant Blue G-250. Once the Coomassie Brilliant Blue G-250 is dissolved, add phosphoric acid and add water to reach the final volume. The staining solution is a colloidal solution so particulates should not be filtered out. 3. LB Medium can be made the day before inoculation and stored at 4  C. 4. 1 Modified MOPS Minimal medium should be made the day of inoculation. 5. The temperature can be lowered as far as 30  C to slow the growth process if necessary. Lower than 30  C will result in too little growth. 6. E. coli cells double every 30–60 min at 37  C (depending on conditions). LB medium will result in faster growth than modified MOPS minimal medium. The O.D. can be checked infrequently until the culture looks turbid. Once the culture is near the target O.D., frequent O.D. monitoring is necessary. Beyond an O.D. of 0.8, the cells enter the stationary phase which results in reduced protein expression. An O.D. below 0.4 will not be sufficient to produce high yields of recombinant protein. 7. Transferring the cells breaks up the pellet, speeding up the resuspension process; however, this method results in minute losses. 8. The shaker method works well for stable proteins and allows time for the lysozyme to take effect. 9. If there is not enough time to carry out the protein purification, resuspend the cells in Ni2+ IMAC Binding Buffer and store them at 80  C. Add lysozyme and protease inhibitor upon thawing to help with cell lysis. 10. Do not attempt to quantify ACP using Bradford assay-based methods due to differences in dye binding capacities between ACP and bovine serum albumin (BSA). 11. Polysorbate 20 (Tween 20) is crucial for solubilization of the medium- and long-chain acyl-CoA (12 and longer) in the presence of MgCl2 and MnCl2 [27] and omitted from reactions where solubilization is not an issue (10 and shorter) to

Acyl-ACP Quantification by LC-MS/MS

245

decrease loading of polysorbate 20 onto the reverse phase column during LC-MS/MS analysis. 12. It is recommended to increase the fold dilution for the lower concentrations. 13. During LC-MS/MS analysis, make a twofold dilution of the lowest detectable concentration (above background) to determine true LOD and LOQ values. 14. Prepare samples at approximately 1–3 μg/10 μL final concentration for best detection and resolution with Coomassie stain. 15. TCEP is included to prevent disulfide bonding between the holo-forms of ACP (ACP-40 -phosphopantetheine-SH). 16. Run the gel in a cold-room to avoid overheating and prevent curvatures in the band pattern. This will enhance the resolution when performing densitometry. 17. An increasing amount of tissue during isolation does not always result in higher yields of acyl-ACPs. This case is observed in leaf tissue where 30 mg FW yields higher amounts of acyl-ACPs than >60 mg FW. It is generally better to extract less tissue and concentrate the sample either by resuspending in less volume at step 13 (Subheading 3.8) or by vacuum centrifugal concentration. Always note the final volumes for later calculations. 18. A fine powder must be obtained after homogenization of the tissue. Anything not well homogenized will result in losses and poor sample handling. 19. It is recommended that you estimate the amount of each acylACP standard necessary for your samples by adding differing amounts to replicate samples and performing a single point quantification before performing isotope dilution-based quantification for best results. 20. Use a 1 mL pipet tip first, then follow using a gel loading tip when close to the bottom to avoid sucking up any of the pellet. 21. It is more important to maintain the integrity of the pellet than to remove every last drop of TCA solution at this step. 22. It is important to remove as much of the TCA solution as possible to achieve a pH above 6.5 in the next step. 23. If the supernatant is cloudy or discolored, filtering may be helpful for downstream steps. 24. Resuspension in 30–50 μL is ideal for achieving adequate mass spectrometry detection (~100 μL final volume in autosampler vial for MS after Asp-N endoproteinase digestion and quenching). 25. Short and medium chain acyl-ACPs (up to ten carbons) can be digested efficiently using 1:50 (w/w) and medium to long chain acyl-ACPs (12 carbons or longer) require 1:20 (w/w) for efficient digestion.

246

Lauren M. Jenkins et al.

26. Generally, the evaluation of six data points within the linear range is desirable for quantification. 27. A weighted linear regression can be performed to account for variance in data points that range across multiple orders of magnitude. An integration software (like Analyst) or other available online worksheets [28] can do this.

Acknowledgments The authors acknowledge support relevant to this project from: The National Science Foundation (IOS-1829365 and MCB-1616820), The United States Department of Agriculture National Institute of Food and Agriculture (USDA-NIFA 201767013-26156), and the United States Department of Agriculture—Agriculture Research Service. Grants that enabled the acquisition of mass spectrometers were obtained from the National Science Foundation (DBI-1427621 and DBI-0521250). References 1. Li-Beisson Y, Shorrosh B, Beisson F, Andersson MX, Arondel V, Bates PD, Baud S, Bird D, Debono A, Durrett TP, Franke RB, Graham IA, Katayama K, Kelly AA, Larson T, Markham JE, Miquel M, Molina I, Nishida I, Rowland O, Samuels L, Schmid KM, Wada H, Welti R, Xu C, Zallot R, Ohlrogge J (2013) Acyl-lipid metabolism. Arabidopsis Book 11:e0161. https://doi.org/10.1199/ tab.0161 2. Kim SC, Nusinow DA, Sorkin ML, PrunedaPaz J, Wang X (2019) Interaction and regulation between lipid mediator phosphatidic acid and circadian clock regulators. Plant Cell 31 (2):399–416. https://doi.org/10.1105/tpc. 18.00675 3. Loewen CJ, Gaspar ML, Jesch SA, Delon C, Ktistakis NT, Henry SA, Levine TP (2004) Phospholipid metabolism regulated by a transcription factor sensing phosphatidic acid. Science 304(5677):1644–1647. https://doi.org/ 10.1126/science.1096083 4. Ohlrogge JB, Kuhn DN, Stumpf PK (1979) Subcellular localization of acyl carrier protein in leaf protoplasts of Spinacia oleracea. Proc Natl Acad Sci U S A 76(3):1194–1198. https://doi.org/10.1073/pnas.76.3.1194 5. Chuman L, Brody S (1989) Acyl carrier protein is present in the mitochondria of plants and eucaryotic micro-organisms. Eur J Biochem 184(3):643–649. https://doi.org/10.1111/j. 1432-1033.1989.tb15061.x

6. Fu X, Guan X, Garlock R, Nikolau BJ (2020) Mitochondrial fatty acid synthase utilizes multiple acyl carrier protein isoforms. Plant Physiol. https://doi.org/10.1104/pp.19. 01468 7. Goldman P, Alberts AW, Vagelos PR (1963) The condensation reaction of fatty acid synthesis: III. Identification of the protein-bound product of the reaction and its conversion to long chain fatty acids. J Biol Chem 238 (11):3579–3583 8. Ohlrogge J, Browse J (1995) Lipid biosynthesis. Plant Cell 7(7):957–970. https://doi.org/ 10.1105/tpc.7.7.957 9. Allen DK (2016) Assessing compartmentalized flux in lipid metabolism with isotopes. Biochim Biophys Acta 1861(9 Pt B):1226–1242. https://doi.org/10.1016/j.bbalip.2016.03. 017 10. Kalinger RS, Pulsifer IP, Hepworth SR, Rowland O (2020) Fatty acyl synthetases and thioesterases in plant lipid metabolism: diverse functions and biotechnological applications. Lipids. https://doi.org/10.1002/lipd.12226 11. Kunst L, Browse J, Somerville C (1988) Altered regulation of lipid biosynthesis in a mutant of Arabidopsis deficient in chloroplast glycerol-3-phosphate acyltransferase activity. Proc Natl Acad Sci U S A 85(12):4143–4147. https://doi.org/10.1073/pnas.85.12.4143

Acyl-ACP Quantification by LC-MS/MS 12. Xu C, Yu B, Cornish AJ, Froehlich JE, Benning C (2006) Phosphatidylglycerol biosynthesis in chloroplasts of Arabidopsis mutants deficient in acyl-ACP glycerol-3- phosphate acyltransferase. Plant J 47(2):296–309. https://doi.org/ 10.1111/j.1365-313X.2006.02790.x 13. Kim HU, Huang AH (2004) Plastid lysophosphatidyl acyltransferase is essential for embryo development in Arabidopsis. Plant Physiol 134 (3):1206–1216. https://doi.org/10.1104/ pp.103.035832 14. Yu B, Wakao S, Fan J, Benning C (2004) Loss of plastidic lysophosphatidic acid acyltransferase causes embryo-lethality in Arabidopsis. Plant Cell Physiol 45(5):503–510. https:// doi.org/10.1093/pcp/pch064 15. Allen DK, Bates PD, Tjellstro¨m H (2015) Tracking the metabolic pulse of plant lipid production with isotopic labeling and flux analyses: past, present and future. Progr Lipid Res 58:97–120. https://doi.org/10.1016/j. plipres.2015.02.002 16. Heinz E, Roughan PG (1983) Similarities and differences in lipid metabolism of chloroplasts isolated from 18:3 and 16:3 plants. Plant Physiol 72(2):273–279. https://doi.org/10. 1104/pp.72.2.273 17. Browse J, Warwick N, Somerville CR, Slack CR (1986) Fluxes through the prokaryotic and eukaryotic pathways of lipid synthesis in the ‘16:3’ plant Arabidopsis thaliana. Biochem J 235(1):25–31. https://doi.org/10.1042/ bj2350025 18. Ho¨lzl G, Do¨rmann P (2019) Chloroplast lipids and their biosynthesis. Annu Rev Plant Biol 70:51–81. https://doi.org/10.1146/ annurev-arplant-050718-100202 19. Post-Beittenmiller D, Jaworski JG, Ohlrogge JB (1991) In vivo pools of free and acylated acyl carrier proteins in spinach. Evidence for sites of regulation of fatty acid biosynthesis. J Biol Chem 266(3):1858–1865

247

20. Ohlrogge JB, Kuo TM (1985) Plants have isoforms for acyl carrier protein that are expressed differently in different tissues. J Biol Chem 260 (13):8032–8037 21. Nam JW, Jenkins LM, Li J, Evans B, Jaworski JG, Allen DK (2020) A general method for quantification and discovery of acyl groups attached to acyl carrier proteins in fatty acid metabolism using LC-MS/MS. Plant Cell. https://doi.org/10.1105/tpc.19.00954 22. Alberts AW, Goldman P, Vagelos PR (1963) The condensation reaction of fatty acid synthesis. I. Separation and properties of the enzymes. J Biol Chem 238:557–565 23. White SW, Zheng J, Zhang YM, Rock (2005) The structural biology of type II fatty acid biosynthesis. Annu Rev Biochem 74:791–831. https://doi.org/10.1146/ annurev.biochem.74.082803.133524 24. Bressler R, Wakil SJ (1961) Studies on the mechanism of fatty acid synthesis: IX. The conversion of malonyl coenzyme A to long chain fatty acids. J Biol Chem 236(6):1643–1651 25. Hsu RY, Wasson G, Porter JW (1965) The purification and properties of the fatty acid synthetase of pigeon liver. J Biol Chem 240 (10):3736–3746 26. Niu L, Zhang H, Wu Z, Wang Y, Liu H, Wu X, Wang W (2018) Modified TCA/acetone precipitation of plant proteins for proteomic analysis. PLoS One 13(12):e0202238. https://doi. org/10.1371/journal.pone.0202238 27. Constantinides PP, Steim JM (1986) Solubility of palmitoyl-coenzyme A in acyltransferase assay buffers containing magnesium ions. Arch Biochem Biophys 250(1):267–270. https://doi.org/10.1016/0003-9861(86) 90726-5 28. O’Haver T (2017) Worksheets for analytical calibration curves. https://terpconnect.umd. edu/~toh/models/CalibrationCurve.html

Chapter 14 Structural Analysis of Glycosylglycerolipids Using NMR Spectroscopy Wiebke Knaack, Georg Ho¨lzl, and Nicolas Gisch Abstract Glycosylglycerolipids are essential components of plant and bacterial membranes. These lipids exert central roles in physiological processes such as photosynthesis in plants or to maintain membrane stability in bacteria. They are composed of a glycerol backbone esterified with two fatty acids at the sn-1 and sn-2 positions, and carbohydrate moieties connected via a glycosidic bond at the sn-3 position. Nuclear magnetic resonance (NMR) spectroscopy is a state-of-the-art technique to determine the nature of the bound carbohydrates as well as their anomeric configurations. Here we describe the analysis of intact glycosylglycerolipids by NMR spectroscopy to determine structural details of their sugar head groups without the need of chemical derivatization. Key words Glycosylglycerolipids, Galactolipid, NMR spectroscopy, Carbohydrates, Anomeric configuration, Fatty acids

1

Introduction Chloroplasts of plants are characterized by their high abundance of glycosylglycerolipids like the two galactolipids mono- and digalactosyl-diacylglycerol (MGDG, DGDG), and the sulfolipid sulfoquinovosyl-diacylglycerol (SQDG). These lipids play crucial roles for example in maintaining an optimal efficiency of photosynthesis and in chloroplast development and morphology [1]. The bulk of MGDG (1,2-di-O-acyl-3-O-β-D-galactopyranosyl-sn-glycerol; βGal-DAG, 1) in plants like Arabidopsis thaliana is synthesized by MGD1, which transfers a galactose moiety from UDP-galactose (UDP-Gal) onto the diacylglycerol precursor [2]. The galactosyl residue is bound in β-anomeric configuration to the glycerol at position sn-3. Formation of DGDG (1,2-di-Oacyl-3-O-(α-D-galactopyranosyl-(1,6)-β-D-galactopyranosyl)-snglycerol; αGal-(1,6)-βGal-DAG, 2) is mainly mediated by DGD1 by galactosylation of MGDG with UDP-Gal as sugar donor [3]. The second galactose in DGDG is characterized by an

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_14, © Springer Science+Business Media, LLC, part of Springer Nature 2021

249

250

Wiebke Knaack et al.

α-anomeric configuration. Under stress conditions, plants can activate an alternative pathway for the synthesis of oligo-galactolipids which involves the galactolipid–galactolipid galactosyltransferase SFR2. This enzyme forms DGDG and higher glycosylated galactolipids in a processive manner by transferring a galactosyl residue from one MGDG molecule as sugar donor onto another MGDG as primary acceptor [4]. One of the resulting products of this pathway is 1,2-di-O-acyl-3-O-(β-D-galactopyranosyl-(1,6)-β-D-galactopyranosyl)-sn-glycerol (βGal-(1,6)-βGal-DAG, 3). Trigalactosyl(TGDG) and tetragalactosyl-diacylglycerol (TeGDG) are further products synthesized through this progressive galactosylation. Therefore, the knowledge of the specific anomeric configuration of galactose moieties in such galactolipids allows the determination of the biosynthesis pathway [5]. In addition to the abundant galactolipids, glucose has been found as carbohydrate moiety in plant glycolipids as well. 1,2-Di-O-acyl-3-O-β-D-glucopyranosyl-snglycerol (βGlc-DAG, 4) is present in for example rice bran, where it represents 36% of the total monoglycosyl-diacylglycerol fraction [6]. This lipid is also found in cyanobacteria synthesized by the glucosyltransferase MgdA, and it is the precursor for the synthesis of MGDG catalyzed by the epimerase MgdE [7]. The diversity of glycosylglycerolipids in the bacterial kingdom is much broader due to a higher variation in their carbohydrate moieties, including differences in glycosidic linkages and anomeric configurations, often found in a species- and sometimes even strain-specific manner [8, 9]. A sensitive and powerful way to analyze the structure of such glycosylglycerolipids can be achieved by mass spectrometry (MS) approaches, which allow the elucidation of single molecular species even in complex mixtures (see Chapters 7–9). However, carbohydrate stereochemistry can only be investigated using LC-MS with separation by advanced ion mobility techniques [10]. Nuclear magnetic resonance (NMR) spectroscopy is the gold standard for a correct and full structural interpretation of stereochemistry and connectivity in complex sugars. General concepts of carbohydrate structural determination by NMR spectroscopy have been discussed elsewhere [11–13]. One common way of analyzing glycosylglycerolipids by NMR spectroscopy is the per-O-acetylation of the molecules to enhance their solubility in CDCl3 (i.e., deuterated chloroform, chloroform-d1), leading to NMR spectra with a good resolution [14, 15]. In this chapter, we focus on the analysis of glycosylglycerolipids by NMR spectroscopy without prior derivatization, thus enabling a fast and practical analysis. Due to the nondestructive nature of the NMR method, one advantage of this approach is that the glycosylglycerolipids remain intact and, hence, can be further analyzed or used in biological assays after completing the NMR analysis.

NMR Spectroscopy of Glycosylglycerolipids

2

251

Materials All NMR spectroscopic measurements were performed on a Bruker AvanceIII 700 MHz NMR spectrometer (BrukerBioSpin GmbH, Rheinstetten, Germany), equipped with an inverse 5 mm quadruple-resonance Z-grad cryoprobe. Spectrometer frequencies were 700.43 MHz (1H) and 176.12 MHz (13C), respectively.

2.1 NMR Tubes and Solvents

1. NMR tubes were purchased from Deutero GmbH (Kastellaun, Germany): 3 mm diameter: Norrell Tube S-3-900-7; 5 mm diameter: Norrell Tube S-5-800-7. 2. Deuterated solvents were purchased from Deutero GmbH (Kastellaun, Germany). For optimal results, we generally use solvents of highest purity packaged in closed brown-glass vials: deuterium oxide (D2O) 99.98% (00509–075 mL), methyl alcohol-d4 (MeOH-d4) 99.96% (01108–075 mL), and chloroform-d1 (CDCl3) 99.95% (00407–075 mL).

2.2

Lipids Analyzed

The glycosylglycerolipids used for NMR spectroscopy were obtained from different sources. MGDG and DGDG were obtained from commercial suppliers. Additional lipid samples from nonplant sources were included to illustrate the characteristics regarding anomeric and epimeric configurations of sugar head groups. These lipids were isolated from Escherichia coli expressing glycosylglycerolipid synthesis genes or directly from Streptococcus pneumoniae. The structures of the glycosylglycerolipids chosen differ in their anomeric (α, β) or epimeric (glucose, galactose) configuration as well as in the number of sugars in their head group (mono- or diglycosyl-diacylglycerols) and in the specific glycosidic linkages of monosaccharides in the diglycosyldiacylglycerols. The analysis of this set of lipids exemplifies how structural details of plant glycolipids, including glycosylglycerolipids, glycosterols, and glycosphingolipids can be determined. 1. MGDG (βGal-DAG, 1) was purchased from Larodan (MGDG from spinach, 59-1200) or Avanti Polar Lipids (MGDG from parsley, 840523P). 2. DGDG (αGal-(1,6)-βGal-DAG, 2) was obtained from Avanti Polar Lipids (DGDG from parsley, 840524P). 3. βGal-(1,6)-βGal-DAG (3) was isolated from an E. coli culture expressing the processive glycosyltransferase Pgt from Agrobacterium fabrum (formerly: A. tumefaciens C58) [14]. Total lipids were extracted from this expression culture and separated by one-dimensional TLC. The digalactosyl lipid band was isolated from the plate and purified (for isolation of lipids from TLC plates, see Chapter 3).

252

Wiebke Knaack et al.

4. βGlc-DAG (4) was isolated from an E. coli culture expressing the cyanobacterial glucosyltransferase MgdA [7]. Cells were harvested from an 1 L culture and crude lipids were extracted. Lipids were dissolved in 300 μL CHCl3–MeOH (90:10, v/v) and 40% of the crude lipid purified by column chromatography. For this, a 25
1 cm glass column (P3) was filled half-height with silica gel 60 (0.040–0.063 mm; Merck, 1.09385.1000) slurried up in CHCl3. The lipid was loaded in CHCl3–MeOH (90:10, v/v) and lipid fractions sequentially eluted in 10 mL portions of CHCl3, CHCl3–MeOH (95:5, v/v) and CHCl3– MeOH (90:10, v/v). Fractions of approximately 2 mL were collected. The lipids in the fractions were analyzed by TLC using silica gel 60 F254 plates (aluminum sheets, Merck, 1.05554.0001; solvent: CHCl3–MeOH (90:10, v/v)) and visualized by Hanessian’s stain [16]. βGlc-DAG 4 eluted within the CHCl3/MeOH (90:10, v/v)-fraction (Rf value: 0.40; final yield approximately 900 μg). 5. 1,2-Di-O-acyl-3-O-α-D-glucopyranosyl-sn-glycerol (αGlcDAG, 5) and αGal-(1,2)-αGlc-DAG (6; see Subheading 2.2, item 6) are the major glycosylglycerolipids in the Grampositive lung pathogen Streptococcus pneumoniae [17]. For extraction of 5 and 6, S. pneumoniae strain D39ΔcpsΔlgt was grown in THY broth medium as described [18]. Two portions of 140–150 mg dry bacterial pellet were individually suspended in 2 mL Millipore-water and 7.5 mL CHCl3–MeOH (1:2, v/v) and thoroughly mixed (in 50 mL Nalgene Oak Ridge Centrifuge Tubes (FEP), Thermo Scientific). After 1 h shaking at RT, 2.5 mL Millipore-water and 2.5 mL CHCl3 were added, mixed and shaken for another hour at RT. In parallel, for each of the two isolation batches, 2 mL Millipore-water was treated in the same way to generate “equilibrated water.” The isolation mixtures were centrifuged at 4000
g for 10 min at 4  C and the organic phases were transferred to fresh tubes. The remaining water phase/interphase was extracted again with the organic phase of the “equilibrated water”-generation (10 s mixing; centrifugation as above). Resulting organic phases were combined with those of the first extractions. These combined phases were washed with 4 mL of the “equilibrated water” (10 s mixing; centrifugation as above). The resulting organic phases were combined and dried under a stream of nitrogen gas (yielding 1.9 and 2.4 mg crude glycosylglycerolipid per pellet portion). These extracts were dissolved in CHCl3–MeOH (2:1, v/v) (10 μg/μL) and applied to HPTLC plates (glass-baked 10
10 cm silica gel 60 F254; Merck, 1.05635.0001). Glycosylglycerolipids were separated with CHCl3–MeOH (80:20, v/v), and a small sidebar of the plate was stained with Hanessian’s stain [16]. Glycosylglycerolipid bands were isolated from

NMR Spectroscopy of Glycosylglycerolipids

253

the silica and recovered by extraction with CHCl3–MeOH (2:1, v/v) for 30 s with thorough mixing and subsequent centrifugation (1080
g, 10 min, 4  C; three rounds of collecting and replacing the organic supernatant). After drying under a stream of nitrogen gas, the residual silica was removed by filtration in CHCl3–MeOH (2:1, v/v) using an Acrodisc CR 13 mm syringe filter (0.2 μm PTFE membrane, PALL Life Sciences; washed with 2
2 mL CHCl3–MeOH (2:1, v/v) prior to use). The yield was 0.416 mg of 5 (Rf range of the band: 0.56–0.72) and 1.34 mg of 6 (Rf range of the band: 0.15–0.34). 6. 1,2-Di-O-acyl-3-O-(α-D-galactopyranosyl-(1,2)-α-D-glucopyranosyl)-sn-glycerol (αGal-(1,2)-αGlc-DAG, 6) was isolated from the Gram-positive lung pathogen Streptococcus pneumoniae as described in Subheading 2.2, item 5.

3

Methods

3.1 Sample Preparation

1. The purified glycosylglycerolipid is transferred into a 1.5-mL screw cap glass vial, dried under a stream of nitrogen gas, and weighed. 2. Hydrogen–deuterium exchange: The exchangeable protons of the hydroxyl groups are exchanged with deuterium by dissolving the glycolipid in 250 μL MeOH-d4 and subsequent drying under a stream of nitrogen gas. This procedure is performed twice (see Note 1). 3. Depending on the amount of the sample, the glycosylglycerolipid is dissolved in an appropriate volume of the appropriate NMR solvent and the solution is transferred into an NMR tube. As a rough guide, if less than 0.3 μmol of glycosylglyc-erolipid are available, it is dissolved in 150 μL solvent and a 3-mm NMR tube is used. If more material is available, the glycosylglycerolipid is dissolved in 400 μL solvent and a 5-mm NMR tube is used. 4. In order to avoid decomposition of the glycosylglycerolipid (s) during long-term NMR analysis, the gas-filled compartment above the solution is shortly filled with nitrogen gas and subsequently the NMR tube is closed by melting, for example using a culinary torch (e.g., Ku¨chenbrenner CB 90, CFH Lo¨tund Gasgera¨te GmbH, Offenbach, Germany) (see Note 2).

3.2 NMR Experiments

To enable a full elucidation of the glycosylglycerolipid structure, a set of one-dimensional (1D) and two-dimensional (2D) NMR experiments is used. As a basis, 1D 1H NMR (with and without water presaturation pulse) and 13C NMR spectra are recorded. 1H assignments were confirmed by 2D 1H,1H-correlation spectroscopy (COSY) and total correlation spectroscopy (TOCSY)

254

Wiebke Knaack et al.

experiments. 13C NMR assignments were indicated by 2D 1 H,13C-heteronuclear single quantum correlation (HSQC), based on the 1H NMR assignments. Interresidue connectivity and further evidence for 13C assignments were obtained from 2D 1H,13C-heteronuclear multiple bond correlation (HMBC) and 1H,13C-heteronuclear single quantum coherence-total correlation spectroscopy (HSQC-TOCSY) experiments. Unless otherwise indicated, all measurements were performed at a temperature of 300 K. All data were acquired and processed using the software Bruker TOPSPIN V 3.1 or higher. The parameter sets used were adapted starting from respective Bruker standard parameter sets, which are all included in this software. The Bruker standard pulse programs used are indicated below in italics; the parameters that are listed are given for the analysis of approximately 800 μg of 1 dissolved in 400 μL MeOH-d4 (Fig. 1 (top), Figs. 2, 3, and 4; see Note 3). 1. 1H NMR: zg (without water presaturation) and zgpr (with water presaturation); parameters: dummy scans (DS): 8, normal scans (NS): 8, Dwell time (DW): 64.800 μs, Relaxation delay (D1): 1.0 s, size of fid (TD): 32768, spectral width (SW): 11.0161 ppm, acquisition time (AQ): 1.9464192 s (see Note 4). The 1H NMR spectra of the monoglycosyl-diacylglycerols 1, 4, and 5 are depicted in Fig. 1, the 1H NMR spectra of the diglycosyl-diacylglycerols 2, 3, and 6 in Fig. 5. 2.

C NMR: zgpg; parameters: DS: 8, NS: 98 k, DW: 64.800 μs, D1: 1.0 s; TD: 32768, SW: 202.7608 ppm, AQ: 0.4587520 s; The 13C spectrum of 1 is shown on the left axis of Fig. 4.

13

3. 1H,1H-COSY: cosygpprqf; parameters: DS: 8, NS: 4, DW: 64.800 μs, D1: 1.0 s; F2: TD: 8192, SW: 11.0161 ppm, AQ: 0.5308416 s; F1: TD: 512, SW: 11.0050 ppm, AQ: 0.033211 s; The COSY spectrum of 1 is shown in Fig. 2. 4. 1H,1H-TOCSY: mlevphpr; parameters: DS: 8, NS: 8, DW: 64.800 μs, D1: 1.0 s, mixing time: 120 ms; F2: TD: 8192, SW: 11.0161 ppm, AQ: 0.5308416 s; F1: TD: 512, SW: 11.0050 ppm, AQ: 0.033211 s. The TOCSY spectrum of 1 is shown in Fig. 3. 5. 1H,13C-HSQCdept: hsqcedetgp; parameters: DS: 16, NS: 48, DW: 64.800 μs, D1: 1.0 s; F2: TD: 8192, SW: 11.0161 ppm, AQ: 0.5308416 s; F1: TD: 512, SW: 185.0845 ppm, AQ: 0.0078527 s. The HSQCdept spectrum of 1 is shown in Fig. 4. 6. 1H,13C-HSQC-TOCSY: hsqcetgpml; parameters: DS: 16, NS: 104, DW: 64.800 μs, D1: 1.0 s, mixing time: 120 ms; F2: TD: 8192, SW: 11.0161 ppm, AQ: 0.5308416 s; F1: TD: 512, SW: 185.0845 ppm, AQ: 0.0078527 s.

NMR Spectroscopy of Glycosylglycerolipids

255

Fig. 1 1H NMR spectra of monoglycosyl-diacylglycerols 1, 4, and 5. Shown are 1H NMR spectra (δH 6.0– (0.5)) of monoglycosyl-diacylglycerols βGalI-DAG (1, top), βGlcI-DAG (4, middle), and αGlcI-DAG (5, bottom) recorded in MeOH-d4 at 300 K with water presaturation (suppressed HOD signal marked with asterisk), the respective chemical structures of the monoglycosyl-diacylglycerols are depicted on the right side. Signals resulting from anomeric protons as well as constantly shifted signals of the glycerol (Gro; only H-1a/b and H-2; H-3a/b are differentially shifted in glycosylglycerolipids depending on the connected carbohydrate moiety) and of characteristic functional groups (olefinic and cyclopropane (cp)) of the fatty acids (R1, R2 ¼ alkyl or alkenyl residues of fatty acid chains) are assigned. The corresponding H atoms in the structures are shown in italics. Complete chemical shift data for 1, 4, and 5 are listed in Table 1

256

Wiebke Knaack et al.

Fig. 2 2D 1H,1H-COSY spectrum of βGalI-DAG (1) recorded with water presaturation in MeOH-d4 at 300 K (depicted section: δH 6.0–0.5 in both dimensions). At the top and on the left, the corresponding 1D 1H NMR

NMR Spectroscopy of Glycosylglycerolipids

257

7. 1H,13C-HMBC: hmbcgpl2ndqf; parameters: DS: 16, NS: 112, DW: 64.800 μs, D1: 1.0 s, CNST6 (¼1J(XH)min): 120.00, CNST7 (¼1J(XH)max): 150.00, CNST13 (¼J (XH) long range): 10.00; F2: TD: 8192, SW: 11.0161 ppm, AQ: 0.5308416 s; F1: TD: 512, SW: 210.0959 ppm, AQ: 0.0069177 s. Spectral processing for 2D experiments was done with a cosine window function in each dimension, line broadening (LB) 1.00 Hz (F2), 0.30 Hz (F1). All spectra were referenced to the residual methanol signal (δH 3.31; δC 49.0 [19]). This set of experiments has been measured for each of the six glycosylglycerolipids analyzed here and was used for the interpretation of the respective structures. The workflow of interpretation is described primarily for 1 in detail below (see Subheading 3.3), the 1H and 13C chemical shift data for all glycosylglycerolipids are summarized in Table 1 (monoglycosyl-diacylglycerols) and Table 2 (diglycosyldiacylglycerols), respectively. 3.3 Interpretation Workflow

1. Assignment of the NMR signals of the glycerol backbone. As a starting point, the spin system of the glycerol backbone, which is esterified with two fatty acids at the sn-1 and sn-2 positions and substituted with a carbohydrate moiety at sn-3, can be identified. The proton chemical shift of the glycerol (Gro) H-2 is found in all the glycosylglycerolipids at an identical position (δH 5.29–5.23) with only marginal deviations in signal width (Tables 1 and 2). Since the C-2 of the glycerol is a chiral center, protons of the two methylene groups H-1 and H-3, respectively, are diastereotopic, thus chemically not equivalent, and therefore result in separate signals for each single proton (H-1a/b and H-3a/b). In the COSY spectrum (Fig. 2) this spin system can be identified by the four cross correlation peaks to the Gro H-2. The proton chemical shift of H-1a/b is identical throughout the glycosylglycerolipids, whereas signals for H-3a/b are differentially shifted depending on the carbohydrate attached to sn-3 (Tables 1 and 2). 2. Identification of the anomeric linkage(s). A second starting point for the interpretation of glycosylglycerolipid NMR spectra is the chemical shift of the characteristic anomeric proton(s). Identification of the anomeric resonances

ä Fig. 2 (continued) spectrum (compare Fig. 1, top) is depicted including assignment of peaks. (a) An overview section (depicted section: δH 6.0–0.5 in both dimensions) with all relevant cross correlations is depicted. The typical cross correlation pattern of the glycerol spin system is present and characteristic for every glycosylglycerolipid as indicated by the dotted lines. (b) The region comprising the cross correlations of the carbohydrate moiety (depicted section: δH 4.6–3.2 in both dimensions) is enlarged. Correlation peaks of protons of the βGalI enabling the sequential elucidation of the spin system are labeled with red numbers

258

Wiebke Knaack et al.

Fig. 3 2D 1H,1H-TOCSY spectrum of βGalI-DAG (1) recorded with water presaturation in MeOH-d4 at 300 K. At the top and on the left, the corresponding 1D 1H NMR spectrum (compare Fig. 1, top) is depicted including assignment of peaks. (a) An overview section (depicted section: δH 6.0–0.5 in both dimensions) with all

NMR Spectroscopy of Glycosylglycerolipids

259

Fig. 4 2D 1H,13C-HSQCdept spectrum of βGalI-DAG (1) recorded in MeOH-d4 at 300 K. At the top, the corresponding 1D 1H NMR spectrum (compare Fig. 1, top) and on the left, the corresponding 1D 13C NMR spectrum is depicted (shown section: δH 6.0–0.5; δC 140–10). CH and CH3 signals are phased up (black) and CH2 signals are phased down (red). The two aliphatic signals indicated with dotted circles are only weakly detected in this experiment, but can be determined in the 1H,13C-HSQC-TOCSY experiment

allows an initial determination of the number of different monosaccharide residues present. In this regard, the glycosylglycerolipids described here only contain galactopyranoside (Galp) and/or glucopyranoside (Glcp) residues. The anomeric proton resonances of common glycopyranosides are usually found in the range of δH 5.5–4.2, depending on the anomeric configuration (α, β) and further substitutions in the neighborhood [13, 20]. In glycopyranosides, the six-membered ring generally forms a chair of fixed conformation providing a classification of protons as axial (ax) or equatorial (eq). Normally the anomeric proton of the α-anomer resonates downfield (at higher δH) compared to the one of the β-anomer in D-glycopyranoses which usually are in the chair (4C1) conformation. The multiplicity of an NMR peak ä Fig. 3 (continued) relevant cross correlations is depicted. (b) The region comprising the cross correlations of the carbohydrate moiety (depicted section: δH 4.6–3.2 in both dimensions) is enlarged. Correlation peaks of protons of the βGalI enabling the elucidation of the complete spin system are labeled with red numbers

260

Wiebke Knaack et al.

depends on the number of protons on its neighboring carbons. Therefore, the signal of the anomeric proton in glycopyranosides always appears as a doublet (d), due to the only neighboring proton H-2. The Karplus equation [21] describes how the coupling constant (J) between two protons is affected by the dihedral angle between them. The highest coupling constants occur

Fig. 5 1H NMR spectra of diglycosyl-diacylglycerols 3, 2, and 6. Shown are the 1H NMR spectra (δH 6.0– (0.5)) of diglycosyl-diacylglycerols βGalII-(1,6)-βGalI-DAG (3, top), αGalII-(1,6)-βGalI-DAG (2, middle), and αGalII-(1,2)-αGlcI-DAG (6, bottom) recorded in MeOH-d4 at 300 K with water presaturation (suppressed HOD signal marked with asterisk), the respective chemical structures of the diglycosyl-diacylglycerols are depicted on the right side. Signals resulting from anomeric protons as well as constantly shifted signals of the glycerol (Gro; only H-1a/b and H-2; H-3a/b are differentially shifted in glycosylglycerolipids depending on the connected carbohydrate moiety) and of characteristic functional (olefinic and cyclopropane (cp)) groups of the fatty acids (R1, R2 ¼ alkyl or alkenyl residues of fatty acid chains) are assigned. The corresponding H atoms in the structures are shown in italics. Complete chemical shift data for 3, 2, and 6 are listed in Table 2

3.53–3.50 (m) 72.4

3.46 (dd) 74.9

3.84–3.82 (m) 70.2

3.52–3.49 (m) 76.8

3.72 (dd) 62.5

3.76 (dd)

H-2 C-2

H-3 C-3

H-4 C-4

H-5 C-5

H-6a C-6

H-6b

4.22 (dd) 64.0

4.44 (dd)

5.29–5.24 (m) 71.8

3.77–3.73 (m) 68.7

3.99 (dd)

H-1a C-1

H-1b

H-2 C-2

H-3a C-3

H-3b

→3)-Gro

4.23 (d) 105.4

βGalI-(1→

11.0, 5.5

12.1, 3.0

12.1, 6.6

11.4, 6.9

11.4, 5.4

9.7, 3.3

7.5

3.99 (dd)

3.74 (dd) 68.8

5.29–5.25 (m) 71.8

4.45 (dd)

4.21 (dd) 64.0

3.87 (dd)

3.67 (dd) 62.8

3.29–3.25 (m) 78.1

3.31–3.27 (m) 71.6

3.35 (dd) 78.0

3.18 (dd) 75.0

4.28 (d) 104.8

βGlcI-(1→

10.9, 5.4

10.9, 5.7

12.1, 2.8

12.1, 6.8

11.9, 2.0

11.9, 5.4

9.1, 8.8

9.1, 7.8

7.8

3.87 (dd)

3.65–3.62 (m) 67.1

5.28–5.24 (m) 71.6

4.48 (dd)

4.20 (dd) 63.8

3.79 (dd)

3.68 (dd) 62.6

3.58 (ddd) 74.0

3.32–3.28 (m) 71.7

3.62 (dd) 75.0

3.39 (dd) 73.5

4.80 (d) 100.7

αGlcI-(1→

δH/δC [ppm]

J [Hz]

δH/δC [ppm]

δH/δC [ppm] J [Hz]

αGlc-DAG (5)

βGlc-DAG (4)

C NMR (176.12 MHz) chemical shift data (δ, ppm) (J, Hz) of monoglycosyl-diacylglycerols 1, 4, and 5

βGal-DAG (1)

13

H-1 C-1

Table 1 1 H (700.43 MHz) and

(continued)

10.8, 5.5

12.0, 3.0

12.0, 6.5

11.9, 2.3

11.9, 5.4

10.0, 5.4, 2.3

9.8, 9.1

9.8, 3.8

3.8

J [Hz]

NMR Spectroscopy of Glycosylglycerolipids 261

J [Hz]

J [Hz]

1.38–1.27 (m)/30.9–30.1 1.31–1.26 (m)/33.1, 33.0 1.34–1.29 (m)/23.7 0.91 (t, 7.0), 0.90 (t, 7.1)/14.4

– at O-1: 2.34–2.30 (m)/35.0 [175.1] at O-2: 2.35–2.31 (m)/35.2 [174.8] 1.65–1.59 (m)/26.0 2.06–2.02 (m)/28.2 1.44–1.27 (m)/31.3–29.9 1.31–1.26 (m)/33.1, 33.0 1.34–1.29 (m)/23.8 0.93–0.88 (m)/14.5, 14.4

2.84–2.79 (m)/26.5, 26.4

at O-1: 2.33 (t, 7.7)/35.0 [175.0] at O-2: 2.36–2.31 (m)/35.1 [174.7]

1.66–1.57 (m)/26.0, 25.9

2.12–2.05 (m)/28.2, 28.1

1.41–1.28 (m)/30.7–29.8

–

–

–

2.12–2.08 (m)/21.5

0.98 (t, 7.5)/14.7

–

–

–

CH2 (α)

CH2 (β)

¼CH–CH2–CH2

Other CH2

–CH2–CH2–CH3

–CH2–CH2–CH3

–CH2–CH2–CH3

¼CH–CH2–CH3

¼CH–CH2–CH3

cp-CH2 (cis-H)

cp-CH2 (trans-H)

cp-CH

0.72–0.65 (m)/16.9, 16.8

–

–

–

0.30–(0.34) (m)/11.6 0.62–0.57 (m)/11.6

–

–

–

–

2.06–2.02 (m)/28.2

1.65–1.58 (m)/26.1, 26.0

at O-1: 2.35–2.31 (m)/35.0 [175.0] at O-2: 2.35–2.31 (m)/35.2 [174.8]

–

–

¼CH–CH2–CH¼

–

5.38–5.28 (m)/129.3, 129.2, 129.1, 128.9, 128.3

¼CH–CH2–CH¼

5.38–5.32 (m)/130.9, 130.8

δH/δC [ppm]

αGlc-DAG (5)

5.41–5.34 (m)/132.8, 131.1, 130.9

5.38–5.32 (m)/130.9, 130.8

δH/δC [ppm]

δH/δC [ppm] J [Hz]

βGlc-DAG (4)

βGal-DAG (1)

¼CH–CH2–CH2

Fatty acid

Table 1 (continued)

262 Wiebke Knaack et al.

3.75–3.71 (m) →6)-βGal -(1→

3.54–3.50 (m) 76.7

3.73–3.70 (m) 62.6

3.79–3.75 (m)

→6)-βGal -(1→

H-5 C-5

H-6a C-6

H-6b

4.24 (d) 105.2

3.52–3.48 (m) 72.4

3.46 (dd) 74.7

3.90–3.88 (m) 70.0

H-1 C-1

H-2 C-2

H-3 C-3

H-4 C-4

I

3.83–3.81 (m) 70.2

H-4 C-4

9.6, 3.3

7.6

9.4, 3.3

3.47 (dd) 74.9

H-3 C-3

I

3.89–3.86 (m) 70.1

3.48 (dd) 74.7

3.51 (dd) 72.4

4.24 (d) 105.3

3.70 (dd) 62.9

3.87–3.83 (m) 72.6

3.91–3.88 (m) 71.1

3.74–3.71 (m) 71.5

3.78 (dd) 70.2

4.87 (d) 100.6

3.53–3.49 (m) 72.6

7.8

αGalII-(1→

H-2 C-2

βGalII-(1→

9.8, 3.3

9.8, 7.5

7.5

11.4, 5.4

10.1, 3.8

3.8

J [Hz]

δH/δC [ppm]

δH/δC [ppm] J [Hz]

αGalII-(1,6)-βGalI-DAG (2)

βGalII-(1,6)-βGalI-DAG (3)

I

3.36 (dd) 71.4

3.78–3.74 (m) 73.4

3.58 (dd) 77.7

5.05 (d) 97.7

→2)-αGlc -(1→

3.73 (m)

3.69 (dd) 62.9

4.11–4.08 (m) 72.6

3.91–3.89 (m) 71.1

3.83–3.79 (m) 71.4

3.79–3.76 (m) 70.3

5.02 (d) 98.2

αGalII-(1→

δH/δC [ppm]

(continued)

9.5, 9.5

9.6, 3.5

3.5

11.3, 7.2

11.3, 5.3

3.8

J [Hz]

αGalII-(1,2)-αGlcI-DAG (6)

C NMR (176.12 MHz) chemical shift data (δ, ppm) (J, Hz) of diglycosyl-diacylglycerols 2, 3, and 6

4.32 (d) 105.4

13

H-1 C-1

Table 2 1 H (700.43 MHz) and

NMR Spectroscopy of Glycosylglycerolipids 263

3.84 (dd) 69.4

4.00 (dd)

H-6a C-6

H-6b

4.44 (dd)

5.29–5.24 (m) 71.8

3.77–3.73 (m) 68.8

3.98 (dd)

H-1b

H-2 C-2

H-3a C-3

H-3b

10.9, 5.4

12.1, 3.0

12.1, 6.7

10.0, 6.3

2.83–2.80 (m)/26.6, 26.4 At O-1: 2.32 (t, 7.8)/35.0 [175.1] At O-2: 2.33 (t, 7.7)/35.1 [174.7]

–

At O-1: 2.32 (t, 7.3)/35.0 [175.1] At O-2: 2.35–2.32 (m)/35.2 [174.8]

¼CH–CH2–CH¼

CH2 (α)

5.38–5.28 (m)/129.2, 128.9, 128.3

–

¼CH–CH2–CH¼

5.41–5.34 (m)/132.8, 131.1

3.94 (dd)

3.76–3.71 (m) 68.8

5.28–5.23 (m) 71.8

4.44 (dd)

4.23 (dd) 64.0

3.92–3.88 (m)

3.67 (dd) 67.8

5.38–5.33 (m)/130.9, 130.8

11.0, 5.4

12.1, 2.8

12.1, 6.8

10.9, 5.9

10.9, 6.6

3.75–3.72 (m) 74.6

¼CH–CH2–CH2

Fatty acid

4.23 (dd) 64.1

H-1a C-1

→3)-Gro

3.74–3.70 (m) 75.4

J [Hz]

δH/δC [ppm]

δH/δC [ppm] J [Hz]

αGalII-(1,6)-βGalI-DAG (2)

βGalII-(1,6)-βGalI-DAG (3)

H-5 C-5

Table 2 (continued)

10.7, 5.2

12.1, 2.8

12.1, 6.5

J [Hz]

At O-1: 2.32 (t, 7.5)/35.0 [175.1] At O-2: 2.35 (t, 7.5)/35.2 [174.8]

–

–

5.38–5.32 (m)/130.9, 130.8

3.88 (dd)

3.68–3.65 (m) 67.1

5.27–5.23 (m) 71.5

4.48 (dd)

4.23 (dd) 64.0

3.82–3.78 (m)

3.70–3.66 (m) 62.6

3.60–3.56 (m) 73.9

δH/δC [ppm]

αGalII-(1,2)-αGlcI-DAG (6)

264 Wiebke Knaack et al.

1.40–1.27 (m)/30.8–29.8 1.30–1.25 (m)/33.1

0.98 (t, 7.5)/14.7 –

1.43–1.25 (m)/31.3–29.9

1.31–1.26 (m)/33.1, 33.0

1.34–1.29 (m)/23.7

0.93–0.88 (m)/14.5

–

–

0.30–(0.34) (m)/11.6

0.62–0.57 (m)/11.6

0.72–0.66 (m)/16.9, 16.8

Other CH2

–CH2–CH2–CH3

–CH2–CH2–CH3

–CH2–CH2–CH3

¼CH–CH2–CH3

¼CH–CH2–CH3

cp-CH2 (cis-H)

cp-CH2 (trans-H)

cp-CH –

–

2.11–2.06 (m)/21.5

0.92 (t, 7.2), 0.90 (t, 7.0)/14.4

1.33–1.27 (m)/23.7

2.12–2.05 (m)/28.2, 28.1

2.06–2.01 (m)/28.2

¼CH–CH2–CH2

1.65–1.57 (m)/26.0, 25.9

1.66–1.56 (m)/26.1, 26.0

CH2 (β)

–

–

–

–

–

0.91 (t, 6.9), 0.90 (t, 7.1)/14.4

1.35–1.29 (m)/23.7

1.32–1.25 (m)/33.1

1.39–1.26 (m)/30.8–30.2

2.07–2.02 (m)/28.2, 28.0

1.66–1.57 (m)/26.1, 26.0

NMR Spectroscopy of Glycosylglycerolipids 265

266

Wiebke Knaack et al.

between protons that have a dihedral angle of either 0 or 180 , and the lowest coupling constants occur at 90 . Therefore, the vicinal coupling constant between the anomeric H-1 and H-2 indicates the relative orientation of these two protons. If both are in ax configuration, a large coupling constant (3J1,2 ~ 7–8 Hz) is observed, whereas if they are in eq-ax configuration, the coupling constant is smaller (3J1,2 ~ 4 Hz), and for ax-eq or eq-eq oriented protons, even smaller coupling constants are observed (3J1,2 < 2 Hz). This holds also true for the other 3JH,H couplings in the pyranosidic sugar ring. Consequently, each glycopyranoside displays a characteristic coupling pattern throughout the sugar ring molecule, which helps to identify its specific nature [11, 20]. 3. Identification of the monosaccharide spin systems. In the 1H NMR spectrum of 1 (Fig. 1, top) the anomeric proton (H-1 βGal) can be found at δH 4.23 (d, 3J1,2 ~ 7.5 Hz). Starting from this, the spin system of the βGal can be evaluated using the 2D 1H,1H-correlated NMR spectra. A COSY spectrum (Fig. 2) has off-diagonal correlations only between coupled spins, whereas a TOCSY spectrum (Fig. 3) has off-diagonal correlations between most/all spins in a spin system. Small values of 3JH,H in monosaccharide rings impose a serious impediment to the standard procedure for making sequential assignments from TOCSY spectra. This becomes evident in the blocking of magnetization transfer from the anomeric proton in a galactose residue at H-4 (Fig. 3b), but not in a glucose residue [12]. Magnetization that is transferred from H-3 to H-4 (J3,4 ¼ 3.3 Hz) in a galactose is not further transferred to H-5 (J4,5 ¼ 1.0 Hz [22]). The more monosaccharide units are present in a molecule, the more informative the TOCSY spectrum becomes. 4. Assignment of 13C Resonances and Interresidue Connectivities The 2D 1H,13C-HSQC NMR experiment permits to obtain a 2D heteronuclear chemical shift correlation map between directly bonded 1H and 13C nuclei, which allows the assignment of a 13C peak observed in the 1D 13C NMR to its corresponding proton(s) (see Note 5). We preferentially use the phase-sensitive dept (distortionless enhancement by polarization transfer)-version for this experiment, because CH and CH3 signals can be differentiated from CH2 signals caused by opposing phases (Fig. 4). The 1H,13C-HSQC-TOCSY is a 2D TOCSY that has been resolved into the carbon dimension. This is especially useful in case of a huge overlap in the proton spectrum or for the assignment of signals only weakly detected in the HSQCdept experiment (like signals indicated by dotted circles in Fig. 4). The 1H,13C-HMBC shows the correlations between protons and carbons that are separated by multiple

NMR Spectroscopy of Glycosylglycerolipids

267

bonds. This information is particularly valuable to confirm bonding through heteroatoms. During glycosylglycerolipid analysis, this data is necessary to substantiate the presence of glycosidic linkages. As an example, the correlation patterns for the anomeric proton of the outermost (αGalII) in glycosylglycerolipids 2 and 6 are depicted in Fig. 6. The correlation to C-2 to C-5 of αGalII is very similar in both molecules, whereas the correlation to C-6 of the innermost (βGalI) in 2 (δC 67.8) is completely different from that to C-2 of αGlcI in 6 (δC 77.7). The HMBC experiment, in contrast to the HSQC experiment, allows to assign carbons that have no protons attached to, in case of the glycosylglycerolipids these are the carbons of the fatty acid carbonyl groups. 5. Temperature dependent shifts in 1H NMR. Sometimes anomeric proton signals are overlapped by the residual signal of HOD (i.e., a water molecule carrying one H

Fig. 6 Section of the 2D 1H,13C-HMBC spectrum of 2 and 6 recorded in MeOH-d4 at 300 K (δH 4.94–4.80 (2), 5.12–4.95 (6); δC 80–60), focussing on the correlation pattern of αGalII H-1 in the respective molecules. In 6 the respective pattern observed for αGlcI H-1 appears in the same region and is shown therefore as well. Cross peaks into carbons of the own spin system are labeled in black, whereas cross peaks to carbons of neighboring spin systems are labeled in red. The latter indicate the respective glycosidic linkages. The baseline of the 1H spectrum of compound 2 along the F2-axis (top) appears as disperse line due to HOD suppression

268

Wiebke Knaack et al.

and one D) and are suppressed as well, if water presaturation is applied, as it is the case for αGalII H-1 in 2 at 300 K (Fig. 7, top). This issue can be solved by changing the measuring temperature, which in general can lead to shifts of some signals relative to the others. In Fig. 7 1H NMR spectra of 2 recorded at 300 and 310 K, respectively, are depicted. Both spectra are referenced to the residual methanol signal (δH 3.31). The increased temperature now enables the detection of the signal for H-1 αGalII without loss of intensity due to a sufficient separation from the suppressed HOD signal, which is now

Fig. 7 Temperature dependent shift of the suppressed HOD signal in 1H NMR spectra of αGalII-(1,6)-βGlcI-DAG (2). Shown are the 1H NMR spectra (δH 6.0–0.5) of αGalII-(1,6)-βGlcI-DAG (2) recorded in MeOH-d4 at 300 K (top; same spectrum as in Fig. 5, middle) or 310 K (bottom), respectively, with water presaturation (suppressed HOD signal is labeled with HODsupp). Both spectra are referenced to the residual methanol signal (δH 3.31). The increased temperature enables the detection of the signal for H-1 αGalII without loss of intensity due to a sufficient separation from the suppressed HOD signal

NMR Spectroscopy of Glycosylglycerolipids

269

shifted. All other signals are not shifted in this example (see Note 6). 6. Assignment of fatty acid-derived signals. Highly indicative signals in the 1H NMR spectra of the investigated glycosylglycerolipids (Figs. 1 and 5) are derived from the methylene groups next to the carbonyl groups in fatty acids, the so-called α-methylene groups (αCH2). They are usually found at δH 2.40–2.30 and occur, if they are fully resolved, as a triplet (t) due to the two chemically identical neighboring protons from the adjacent methylene group (βCH2). The carbon signals of these two groups are generally found around δC 35.0 (αCH2) and 26.0 (βCH2) (see Note 7). Double bonds and their adjacent methylene groups lead to distinct chemical shifts (Tables 1 and 2). Especially the methylene groups between two double bonds (¼CH–CH2–CH¼) as present for example in α-linolenic acid (18:3 Δ9,12,15), one of the major constituents of the plant glycosylglycerolipids 1 and 2, cause distinct signals around δH 2.8 and δC 26.5. Such signals are not observed in the glycosylglycerolipids isolated from E. coli expressing glycosylglycerolipid genes or from S. pneumoniae (Figs. 1 and 5). In turn, the glycosylglycerolipids 3 and 4 produced in and isolated from transformed E. coli contain cyclopropane fatty acids [23] indicated by distinctive signals in the shift range of δH 1.0–(1.0) (labeled with cp in Figs. 1 and 5).

4

Notes 1. An alternative procedure for the hydrogen–deuterium exchange, for example, if nitrogen gas is not available, could be dissolving the glycosylglycerolipid in 250 μL of a 1:1 mixture of D2O and MeOH-d4 and subsequent freeze-drying (requires a laboratory freeze dryer suitable for organic solvents). This procedure has to be performed twice. 2. It is important to fill the gas compartment above the solution in the NMR tube with nitrogen, especially for the analysis of monoglycosyl-diacylglycerols, since otherwise partial decomposition of the sample during long-term measurements can be observed. In the case that decomposition cannot be completely avoided this way, the glycosylglycerolipids can be analyzed using a CDCl3–MeOH–d4–D2O (40:53:7, v/v/v) mixture as an alternative option. This leads to comparable spectral resolutions and almost identical 1H NMR chemical shift data. However, 13C NMR chemical shift data are slightly shifted. This solvent has been used already for example for the analysis of ornithine lipids [24, 25]. Sealing the NMR tube by melting the

270

Wiebke Knaack et al.

top of the tube is highly important in long-term measurements (>12 h at 300 K), since otherwise partial evaporation of the CDCl3 leads to a constant shift in the solvent composition. 3. Some parameters like the receiver gain, 90 high power pulse, and power levels for water presaturation or TOCSY-spinlock have to be determined and adapted individually for each sample. Please note: the numbers of normal scans (NS) used here are excessively high, especially for the 1D 13C NMR, but also for the 2D 1H,13C-HMBC and 1H,13C-HSQC-TOCSY experiments. The focus here was to achieve an optimal representation, not the optimization of measuring time. 4. We noticed that for these glycosylglycerolipids samples, spinning of tubes during measurement of 1D 1H NMR spectra enhances the spectral quality. 5. If the quality of the 1D 13C NMR spectrum is insufficient due to a very low amount of sample, 13C chemical shift data can be determined directly from the 2D 1H,13C-correlated spectra. Review articles regarding 13C NMR chemical shift data of carbohydrates [13, 26] as well as 13C NMR databases (such as CASPER [27]) available in the internet might be helpful in investigating unknown glycolipids. 6. NMR chemical shifts of various conformations are different and the populations of conformations change with temperature. The observed chemical shift is the weighted average of all shifts of the individual conformations. Moreover, temperature can affect the degree of intermolecular hydrogen bonding or other types of aggregation, possibly leading to additional shifts. Therefore, if larger glycolipids and/or glycolipids containing carbohydrates with charged groups are analyzed, signals of the analyzed molecule can shift as well. 7. Note that this is only the case for saturated fatty acids or for unsaturated ones with the double bond(s) not located in the proximity of the αCH2. References 1. Kalisch B, Do¨rmann P, Ho¨lzl G (2016) DGDG and glycolipids in plants and algae. Subcell Biochem 86:51–83. https://doi.org/10. 1007/978-3-319-25979-6_3 2. Awai K, Mare´chal E, Block MA, Brun D, Masuda T, Shimada H, Takamiya K, Ohta H, Joyard J (2001) Two types of MGDG synthase genes, found widely in both 16:3 and 18:3 plants, differentially mediate galactolipid syntheses in photosynthetic and nonphotosynthetic tissues in Arabidopsis thaliana. Proc Natl

Acad Sci U S A 98(19):10960–10965. https:// doi.org/10.1073/pnas.181331498 3. Kelly AA, Do¨rmann P (2002) DGD2, an Arabidopsis gene encoding a UDP-galactosedependent digalactosyldiacylglycerol synthase is expressed during growth under phosphatelimiting conditions. J Biol Chem 277 (2):1166–1173. https://doi.org/10.1074/ jbc.M110066200 4. Moellering ER, Muthan B, Benning C (2010) Freezing tolerance in plants requires lipid remodeling at the outer chloroplast

NMR Spectroscopy of Glycosylglycerolipids membrane. Science 330(6001):226–228. https://doi.org/10.1126/science.1191803 5. Gasulla F, Vom Dorp K, Dombrink I, Za¨hringer U, Gisch N, Do¨rmann P, Bartels D (2013) The role of lipid metabolism in the acquisition of desiccation tolerance in Craterostigma plantagineum: a comparative approach. Plant J 75(5):726–741. https:// doi.org/10.1111/tpj.12241 6. Fujino Y, Miyazawa T (1979) Chemical structures of mono-, di-, tri-, and tetraglycosyl glycerides in rice bran. Biochim Biophys Acta 572 (3):442–451 7. Awai K, Kakimoto T, Awai C, Kaneko T, Nakamura Y, Takamiya K, Wada H, Ohta H (2006) Comparative genomic analysis revealed a gene for monoglucosyldiacylglycerol synthase, an enzyme for photosynthetic membrane lipid synthesis in cyanobacteria. Plant Physiol 141(3):1120–1127. https://doi.org/ 10.1104/pp.106.082859 8. Reichmann NT, Gru¨ndling A (2011) Location, synthesis and function of glycolipids and polyglycerolphosphate lipoteichoic acid in Grampositive bacteria of the phylum Firmicutes. FEMS Microbiol Lett 319(2):97–105. https://doi.org/10.1111/j.1574-6968.2011. 02260.x 9. Brundish DE, Shaw N, Baddiley J (1966) Bacterial glycolipids. Glycosyl diglycerides in Gram-positive bacteria. Biochem J 99 (3):546–549. https://doi.org/10.1042/ bj0990546 10. Chen Z, Glover MS, Li L (2018) Recent advances in ion mobility-mass spectrometry for improved structural characterization of glycans and glycoconjugates. Curr Opin Chem Biol 42:1–8. https://doi.org/10.1016/j. cbpa.2017.10.007 11. Duus JØ, Gotfredsen CH, Bock K (2000) Carbohydrate structural determination by NMR spectroscopy: modern methods and limitations. Chem Rev 100(12):4589–4614. https://doi.org/10.1021/cr990302n 12. Bubb WA (2003) NMR spectroscopy in the study of carbohydrates: characterizing the structural complexity. Concepts Magn Reson A 19A(1):1–19. https://doi.org/10.1002/ cmr.a.10080 13. Agrawal PK (1992) NMR spectroscopy in the structural elucidation of oligosaccharides and glycosides. Phytochemistry 31 (10):3307–3330. https://doi.org/10.1016/ 0031-9422(92)83678-r 14. Ho¨lzl G, Leipelt M, Ott C, Za¨hringer U, Lindner B, Warnecke D, Heinz E (2005) Processive lipid galactosyl/glucosyltransferases from Agrobacterium tumefaciens and Mesorhizobium loti display multiple specificities.

271

Glycobiology 15(9):874–886. https://doi. org/10.1093/glycob/cwi066 15. Ho¨lzl G, Witt S, Kelly AA, Za¨hringer U, Warnecke D, Do¨rmann P, Heinz E (2006) Functional differences between galactolipids and glucolipids revealed in photosynthesis of higher plants. Proc Natl Acad Sci U S A 103 (19):7512–7517. https://doi.org/10.1073/ pnas.0600525103 16. Pirson C, Engel R, Jones GJ, Holder T, Holst O, Vordermeier HM (2015) Highly purified mycobacterial phosphatidylinositol mannosides drive cell-mediated responses and activate NKT cells in cattle. Clin Vaccine Immunol 22(2):178–184. https://doi.org/ 10.1128/cvi.00638-14 17. Meiers M, Volz C, Eisel J, Maurer P, Henrich B, Hakenbeck R (2014) Altered lipid composition in Streptococcus pneumoniae cpoA mutants. BMC Microbiol 14:12. https://doi. org/10.1186/1471-2180-14-12 18. Gisch N, Kohler T, Ulmer AJ, Mu¨thing J, Pribyl T, Fischer K, Lindner B, Hammerschmidt S, Za¨hringer U (2013) Structural reevaluation of Streptococcus pneumoniae lipoteichoic acid and new insights into its immunostimulatory potency. J Biol Chem 288(22):15654–15667. https://doi.org/10. 1074/jbc.M112.446963 19. Gottlieb HE, Kotlyar V, Nudelman A (1997) NMR chemical shifts of common laboratory solvents as trace impurities. J Org Chem 62 (21):7512–7515 20. Hounsell EF (1995) 1H NMR in the structural and conformational analysis of oligosaccharides and glycoconjugates. Progr Nucl Magn Reson Spectrosc 27(5):445–474. https://doi.org/ 10.1016/0079-6565(95)01012-2 21. Karplus M (1959) Contact electron-spin coupling of nuclear magnetic moments. J Chem Phys 30(1):11–15. https://doi.org/10.1063/ 1.1729860 22. Zwahlen C, Vincent SJ (2002) Determination of 1H homonuclear scalar couplings in unlabeled carbohydrates. J Am Chem Soc 124 (24):7235–7239. https://doi.org/10.1021/ ja017358v 23. Raetz CRH (1978) Enzymology, genetics, and regulation of membrane phospholipid synthesis in Escherichia coli. Microbiol Rev 42 (3):614–659 24. Diercks H, Semeniuk A, Gisch N, Moll H, Duda KA, Ho¨lzl G (2015) Accumulation of novel glycolipids and ornithine lipids in Mesorhizobium loti under phosphate deprivation. J Bacteriol 197(3):497–509. https://doi.org/ 10.1128/jb.02004-14

272

Wiebke Knaack et al.

25. Ho¨lzl G, Sohlenkamp C, Vences-Guzmán MA, Gisch N (2018) Headgroup hydroxylation by OlsE occurs at the C4 position of ornithine lipid and is widespread in proteobacteria and bacteroidetes. Chem Phys Lipids 213:32–38. https://doi.org/10.1016/j.chemphyslip. 2018.03.002 26. Bock K, Pedersen C (1983) Carbon-13 nuclear magnetic resonance spectroscopy of monosaccharides. In: Tipson RS, Horton D (eds)

Advances in carbohydrate chemistry and biochemistry, vol 41. Academic, New York, NY, pp 27–66. https://doi.org/10.1016/S00652318(08)60055-4 27. Lundborg M, Widmalm G (2015) NMR chemical shift prediction of glycans: application of the computer program CASPER in structural analysis. Methods Mol Biol 1273:29–40. https://doi.org/10.1007/978-1-4939-23434_3

Part III Lipid Isolation and Analysis from Plant Tissues Cell Compartments and Organelles

Chapter 15 Analysis of Extracellular Cell Wall Lipids: Wax, Cutin, and Suberin in Leaves, Roots, Fruits, and Seeds Johanna Baales, Viktoria V. Zeisler-Diehl, and Lukas Schreiber Abstract Extracellular lipids of plants can be analyzed using gas chromatography and mass spectrometry. Soluble waxes are extracted with chloroform and thus separated from the extracellular polymers cutin and suberin. Cutin and suberin have to be depolymerized using boron trifluoride–methanol or methanolic HCl before analysis. The released monomeric hydroxylated fatty acids are then extracted with chloroform or hexane. Prior to gas chromatography, all free polar functional groups (alcohols and carboxylic acids) are derivatized by trimethylsilylation. Internal standards, that is, long chain alkanes, are used for the quantification of wax molecules and cutin or suberin monomers. Lipids are quantified using gas chromatography coupled to flame ionization detection. Qualitative analysis is carried out by gas chromatography coupled to mass spectrometry. Thus, all wax molecules of chain lengths from C16 to C60 and different substance classes (fatty acids, alcohols, esters, aldehydes, alkanes, etc.) or all cutin or suberin monomers of chain lengths from C16 to C32 and different substance classes (hydroxylated fatty acids, diacids, etc.) can be analyzed from one sample. Key words Plant cuticle, Extracellular lipids, Suberin, Cutin, Wax, GC-MS, GC-FID, Lipid extraction, Qualitative analysis, Quantitative analysis

1

Introduction Plant lipids derived from waxes [1], cutin [2], and suberin [3] can be chemically analyzed using gas chromatography coupled to flame ionization detection (GC-FID) and to mass spectrometry (GC-MS). Since GC is a very sensitive technique with a detection limit in the ng range or lower, it is absolutely essential to work accurately and to avoid any contaminations. Waxes represent a mixture of different long-chain aliphatic molecules (alkanes, acids, aldehydes, esters, etc.) and cyclic triterpenoids, which can be extracted in chloroform. From different chemical and spectroscopical investigations it is known that waxes are heterogeneously distributed in the cuticle [4]. Long-chain aliphatics are located on the surface of the cuticle and within the cuticle, whereas

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_15, © Springer Science+Business Media, LLC, part of Springer Nature 2021

275

276

Johanna Baales et al.

triterpenoids are located in deeper regions of the cuticle, the cutin polymer [1, 5]. Chloroform is the best solvent for wax extraction, since it dissolves the more polar components (e.g., acids and triterpenoids) of the waxes as well as highly unpolar constituents (e.g., alkanes). Different protocols have been described for wax extraction [6]. Total waxes are best extracted by dipping the plant organ (leaf, stem, flower etc.) into chloroform. The correct dipping time for total wax extraction should be determined by performing extraction kinetics, and the volume of chloroform should be optimized to prevent saturation of the solvent with the wax molecules. A specific wax extraction protocol is based on the selective removal of epicuticular wax from the leaf surface by stripping the wax using Collodion (nitrocellulose, 4–8% [5]). Cutin and suberin need to be transesterified using either boron trifluoride–methanol or methanolic HCl to release the single monomers from the polymers [7, 8]. Prior to transesterification it is essential to extract free membrane lipids (delipidation) by intense washing of the tissue in chloroform–methanol (1:1, v/v). All samples need to be internally standardized prior to analysis allowing the quantification of each monomer via GC-FID. Since the FID signal in the chromatograms provides a linear response to the amount of burned CH2 groups, alkanes are suited best as internal standard. To increase the volatility of the monomers, to obtain symmetrical peaks for accurate quantification and to reduce the interaction between the polar compounds and the apolar column, all samples (wax; cutin and suberin monomers) need to be derivatized by trimethylsilylation. The identification of the individual lipid monomers is achieved by running identical samples on the GC-MS and by comparing fragmentation patterns with mass spectra collected in a home-made database or in the literature (e.g., https://lipidlibrary. aocs.org/ or https://www.nist.gov/). Cutin, suberin, and wax amounts need to be calculated individually for each sample and should preferentially be expressed in μmol or μg per surface area and not per dry weight.

2

Materials All materials used for extractions, such as pipettes, vials, syringes, or any other tools need to be solvent resistant (see Subheading 2.1). Only glass ware or tools made of stainless steel should be used and rinsed thoroughly with chloroform prior to use during all steps of sample collection, preparation and extraction. Septa or lids should be lined with Teflon (PTFE) which is resistant to organic solvents. For sample transfer and pipetting, Hamilton syringes should be used, since they enable the accurate measurements of volumes. Chemicals and solvents for chemical analysis need to be of high analytical reagent grade (see Subheading 2.2) and can be obtained from common chemical companies.

Gas Chromatographic Analysis of Wax, Cutin and Suberin

2.1 Glassware, Syringes, Laboratory Tools

277

1. Glass vials with different volumes and Teflon-lined caps. 2. Small reaction glass vials (1 ml) with Teflon-lined screw cap (e.g., Reacti vials, Thermo Fisher Scientific). 3. GC autosampler vials and septa. 4. Hamilton syringes 100–1000μl. 5. Sample evaporator with nitrogen gas stream (99.99% purity). 6. Semi-micro or micro balance. 7. Heating block. 8. Stone mortar and pestle (e.g., agate). Do not use porcelain mortar.

2.2 Solvents for Lipid Extraction and Sample Preparation

1. Acetone. 2. Chloroform. 3. Chloroform–methanol (1:1, v/v). 4. Ethanol (not denatured). 5. Methanol. 6. Hexane. 7. Enzyme solution for cell wall digestion for cutin and suberin isolation: 2% (v/v) cellulase and 2% (v/v) pectinase (v/v) prepared from Beerzym Amber95 (highly concentrated cellulase solution; Erbslo¨h Geisenheim Germany) and Fructozym Color (highly concentrated pectinase solution; Erbslo¨h Geisenheim Germany dissolved in 10 mM citric acid (pH 3.0). NaN3 must be added to 1 mM final concentration to prevent microbial growth. 8. 10 mM sodium borate (Na2B4O7) buffer (pH 9.0) for washing isolated polymer samples. 9. Collodion (nitrocellulose, 4–8%) dissolved in diethyl ether– ethanol (1:1, v/v).

2.3 Reagents for Depolymerization, Transesterification and Derivatization

1. Anhydrous Na2SO4. 2. Methanolic HCl (1 N). 3. 30% (v/v) boron trifluoride–methanol (BF3–methanol). 4. HPLC-grade water. 5. Saturated NaCl in HPLC-grade water. 6. Saturated NaHCO3 in HPLC-grade water. 7. N,O-Bis-(trimethylsilyl)-trifluoroacetamide (BSTFA). 8. Pyridine.

2.4 Standards for GC-FID and GC-MS Analysis

1. Acid standard mixture for GC maintenance: tetracosane (C24 alkane), nonacosanoic acid (C29 acid), triacontanoic acid (C30 acid), hentriacontanoic acid (C31 acid); all components, purity 99.9%, dissolved in chloroform, each at 0.025% (w/v).

278

Johanna Baales et al.

2. Internal standard for suberin and cutin analyses: dotriacontane (C32 alkane), purity 99.9%, dissolved in chloroform at 0.025% (w/v). 3. Internal standard for wax analysis: tetracosane (C24 alkane), purity 99.9%, dissolved in chloroform at 0.025% (w/v). 2.5 GC-FID and GC-MS Instruments

1. GC-FID: Agilent 5980 coupled with on-column injector and flame ionization detector. Column: DB-1 (capillary column with dimethylpolysiloxane coating), length: 30 m, inner diameter: 0.32 mm, film thickness: 0.1μm (J&W Scientific, Folsom, CA, USA). Carrier gas: hydrogen (use high purity gas, 99.99%). Additional gases needed for the detector: nitrogen (use high purity gas, 99.99%) and synthetic air (79.5% nitrogen, 20.5% oxygen). 2. GC-MS: Agilent 6890N with on-column injector and Agilent 5973N mass selective detector. Column: DB-1-MS (capillary column with dimethylpolysiloxane coating), length: 30 m, inner diameter: 0.32 mm, film thickness: 0.1μm. Carrier gas: helium (use high purity gas, 99.99%).

3

Methods

3.1 Collection and Preparation of Plant Samples

1. Fresh leaf and fruit material can be used for wax extraction and analysis, if cuticular membranes cannot be isolated. If possible, cuticular membranes should first be isolated for wax extraction (see Notes 1 and 2). It is crucial to always harvest the same ontogenetic stage of the plant organ, that is, the leaves should be numbered and harvested according to their developmental stage. Leaves, stems, fruits, and flowers which show necrosis, fissures from insects or fungal infection should be avoided. If the plant material is covered with dust, particles can be removed by gently rinsing in water and letting it dry prior to processing. Drying by rubbing with tissue paper should be avoided since waxes might be removed. 2. Roots can easily be harvested from hydroponically grown plants. When roots are harvested from soil grown plants, soil particles should be carefully removed using a fine brush after placing the root system in water. 3. Fresh or dried seeds can be used for seed coat suberin analysis. 4. Lipids from different leaves can be pooled to minimize the biological variability between the samples. Intact leaves should not be cut in pieces since chloroform used for wax extraction diffuses into the leaf via cut edges and dissolves lipids from the leaf interior.

Gas Chromatographic Analysis of Wax, Cutin and Suberin

279

5. For suberin analysis of roots, different root zones, representing different developmental stages can be separated [9]. 6. Cutin and suberin can be isolated from fresh material (leaf discs; roots or seeds) (Subheading 3.2) or from frozen plant material that is first ground to a fine powder under liquid nitrogen. 3.2 Enzymatic Isolation of Extracellular Lipid Polymers

Enzymatic isolation of extracellular lipid polymers often involves the removal of polymeric carbohydrate cell wall material from the lipid polymers using cellulases and pectinases (see Subheading 2.2). Pectinase and cellulase used in food industries for fruit juice or wine production (Erbslo¨h, Germany) are commonly used for cutin and suberin isolation, since they are cheaper than the pure enzymes derived from Aspergillus niger. 1. For cutin isolation from cuticles, leaf discs are punched out from leaves avoiding the veins. 2. For suberin isolation, root segments or discs of for example potato periderm or seeds are harvested. 3. Plant samples are incubated in an Erlenmeyer flask at room temperature containing the enzyme solution made of cellulase and pectinase (see Subheading 2.2). The enzyme solutions are buffered at pH 3.0 (optimal pH for cellulase and pectinase) using 0.01 M citric acid. One millimolar NaN3 is added to prevent the growth of microorganisms and fungi. 4. To accelerate the cell wall digestion, plant samples can be vacuum infiltrated with the enzyme solution. 5. The Erlenmeyer flasks should not be shaken vigorously, since this could cause damage or fragmentation of the samples. 6. During tissue digestion, phenolic compounds are released which inhibit enzyme activity. Therefore, the enzyme solution should be changed every day. Depending on the samples, cutin and suberin polymers are isolated after several days or even up to 2–3 weeks. 7. During digestion at pH 3.0, some compounds (e.g., free fatty acids from the leaf interior and phenolic compounds) are protonated. Thus they become very lipophilic and tend to be absorbed in the lipophilic polymers. Therefore, after enzyme treatment, plant samples should be washed with 0.01 M sodium borate buffer (pH 9.0) to extract these compounds from the lipophilic polymers. Otherwise they might erroneously be identified as wax molecules. 8. Finally, samples are thoroughly washed in deionized water, which is changed daily during 3–7 days.

280

Johanna Baales et al.

3.3 Wax Extraction from Plant Organs and Cutin or Suberin Polymers and Preparation of Wax-Free Cuticular Membranes

Plant cuticles and sometimes suberin contain waxes (linear longchain aliphatic molecules). Waxes can be extracted with chloroform from isolated cutin and suberin samples (extracellular lipid + polymers after enzymatic isolation; Subheading 3.2), leaving the polymer matrix unaffected. Similarly, waxes can be isolated from intact leaves. However, care must be taken that internal lipids are not extracted (see Note 3). Chloroform is the best solvent for extraction of waxes, since both wax fractions, that is, the polar (e.g., acids, alcohols) and nonpolar (e.g., alkanes) molecules, are equally well dissolved in chloroform [5] (see Note 4). Other solvents are also used for wax extraction (see Subheading 2.2), but they do not dissolve all wax molecules with similar efficiency. For example, hexane dissolves alkanes and esters very well, but not fatty acids, alcohols, and triterpenoids. 1. Detach one or several leaves from plants with forceps and scissors. Alternatively, use extracellular lipid polymers after enzymatic isolation (see Subheading 3.2). 2. Determine the leaf area by scanning. 3. Submerge the leaf or extracellular lipid polymers in chloroform in a screw top glass vial for 10 s, and then remove the plant material. 4. Extracted waxes are best stored in organic solvents at 4  C. Solvent evaporation should be avoided since dried wax extracts are sometimes difficult to dissolve. 5. The enzymatically isolated and delipidated extracellular lipid polymers (cutin and suberin) should be dried on Teflon disks, since cell wall samples stick to the glass surfaces of the vials. Samples are best stored at room temperature in a desiccator with activated silica gel reducing relative humidity to 2%. Weights of the samples should be checked until the weights are constant. This can take several days until all solvents (chloroform, methanol) are evaporated.

3.4 Mechanical Isolation of Epicuticular Wax Using Collodion

1. Epicuticular waxes are mechanically removed from leaf surfaces using a small droplet of about 15μl of Collodion (see Subheading 2.3). The Collodion solution is applied to the surface of the plant organ using a soft brush. After about 30 s, a thin polymer film appears as the solvent evaporates. The epicuticular wax layer sticks to the nitrocellulose-strip which is removed from the surface with tweezers (see Note 5). 2. Before further analysis, epicuticular waxes trapped in the polymer film, are extracted from the polymer film by incubating the film in 4 ml chloroform overnight. 3. The internal standard for wax quantification (10μg of C24 alkane) should be directly added to each sample.

Gas Chromatographic Analysis of Wax, Cutin and Suberin

281

4. After wax extraction in chloroform overnight, the Collodion film can be removed and the chloroform is evaporated under a gentle stream of nitrogen at 60  C in a heating block. One and the same leaf area can be treated consecutively with Collodion solution for up to five times to collect all epicuticular lipids, depending on the plant species [5]. 5. Standardized wax extracts in chloroform, already containing the internal standard, can be transferred to clean Reacti vials and stored for several weeks at 4  C. 6. After trimethylsilylation (Subheading 3.8) extracted waxes should be analyzed by on-column injection using GC-FID or GC-MS. 3.5 Wax Extraction from One Leaf Side Using Glass Vials with Rolled Edges

Selective epicuticular wax removal using Collodion is not possible with every plant species, especially with species containing a high number of trichomes and very thin and fragile leaves. Application of Collodion on such leaves often causes damage of the living tissue since trichomes stick to the Collodion film and break which leads to extraction of lipids from inside of the leaf interfering with the quantification of epicuticular waxes. Species with a high number of trichomes require a different method for selective wax removal (see Notes 1 and 3). 1. The intact leaf is carefully placed on a clean Teflon disk. Glass vials with broad rims and a central opening with a defined area are filled with chloroform (2–5 ml). The leaf side of interest is gently pressed on the opening of the glass vial and it is turned upside down for 10 s to allow wax extraction by the chloroform in the vial. 2. If the extracted leaf area is very small, several areas can be treated with one and the same glass vial to ensure that enough waxes are extracted to increase signal intensity during GC analyses. 3. Extractions near leaf veins should be avoided since chloroform might leak. 4. The wax extract is directly spiked with the C24 alkane standard and the volume is reduced to 200μl under a gentle stream of nitrogen at 60  C. 5. The wax extract is transferred to Reacti vials. 6. After trimethylsilylation (Subheading 3.8) extracted waxes should be analyzed by on-column injection using GC-FID or GC-MS.

282

Johanna Baales et al.

3.6 Depolymerization and Transesterification of Cutin and Suberin Using Methanolic HCl

1. Every glass vial, forceps, syringe and spatula has to be cleaned with chloroform prior to use. 2. For each sample, at least 1–3 mg of dried material should be used. 3. The sample is cut in small pieces using clean scissors, a razor blade, or alternatively ground in a mortar. However, only stone mortars should be used (e.g., agate) and not porcelain mortars since the latter show abrasion of mineral material contributing to final dry weights determined prior to analysis. 4. Samples are incubated in glass vials containing 2 ml 1 N methanolic HCl (see Subheading 2.4) at 80  C for 2 h. 5. During heating of the samples, the pressure rises and needs to be released by carefully open and closing the glass vials after 30 min. Wear gloves and safety glasses. 6. After 2 h of incubation, the vials with transesterified cutin and suberin monomers are cooled down to room temperature under the fume hood. 7. 10–20μg of the internal standard C32 alkane (stock solution, 10 mg in 50 ml chloroform) are added to each sample enabling quantification of the individual suberin and cutin monomers during GC analyses. 8. To stop the depolymerization and prepare the sample for monomer extraction, 2 ml of saturated NaCl solution is added to each vial. 9. The methylated fatty acids and other monomers released during depolymerization are extracted with hexane. After addition of hexane (2 ml), samples are vigorously vortexed and subsequently allowed to stand still for 1–2 min for optimal phase separation. The upper hexane phase is carefully separated from the lower aqueous phase using a Hamilton syringe and collected in precleaned glass vials (see Note 6). 10. Extraction with hexane is repeated three times with 2 ml for each extraction. 11. The hexane phases are combined and the volume reduced to 200μl at 60  C under a gentle stream of nitrogen. 12. This final volume of 200μl is transferred to Reacti vials and kept at 4  C until the sample is processed and analyzed. 13. After trimethylsilylation of hydroxylic groups (see Subheading 3.8), transmethylated cutin and suberin monomers can be analyzed by GC-FID or GC-MS using on-column injection or split/splitless injection (see Note 10).

Gas Chromatographic Analysis of Wax, Cutin and Suberin

3.7 Depolymerization and Transesterification of Suberin and Cutin Using Boron Trifluoride–Methanol

283

1. Dry samples are placed in a glass vial, covered with 1.5–3 ml boron trifluoride–methanol solution (BF3–methanol; see Subheading 2.4) and transesterified for 16–18 h at 70  C. 2. After cooling, 10–20μg of the internal C32 alkane standard are added to each sample. 3. The depolymerization reaction is stopped by adding 2 ml saturated NaHCO3 to each sample. 4. Monomeric methylated fatty acids are extracted three times with 2 ml chloroform. After adding chloroform, the samples must be vortexed. Let stand still for phase separation (see Note 6). If phase separation is not sufficient, it may be accelerated by adding more saturated NaHCO3 or chloroform to the sample. 5. The lower chloroform phase with the transesterified monomers is transferred to new clean glass vials using a Hamilton syringe. 6. For further cleaning of the collected chloroform phases, 1–2 ml HPLC-grade water is added, samples are thoroughly vortexed and the upper aqueous phase is discarded. 7. Finally, remaining traces of water in chloroform are dried by adding 1 g or more of anhydrous Na2SO4 to each sample. The Na2SO4 should finally be visible as a precipitate at the bottom of the vial and not as a cloudy substance within the supernatant. This indicates that all traces of water have been removed. 8. The chloroform solution is removed from the Na2SO4 precipitate and the volume reduced to 200μl under a gentle nitrogenstream at 60  C. Then the chloroform extract is transferred to clean Reacti vials and stored at 4  C until further usage. 9. After trimethylsilylation of hydroxylic groups (Subheading 3.8), transmethylated suberin and cutin monomers can be analyzed by on-column injection or split/splitless injection by GC-FID or GC-MS (see Note 10).

3.8 Derivatization with BSTFA

1. Prior to GC, all free hydroxylic and carboxylic groups in the wax components, and all free hydroxylic groups in the depolymerized cutin and suberin samples need to be derivatized. This enhances volatility of the compounds, reduces the interaction with the column and leads to symmetrical peak shapes. 2. Samples are best silylated using BSTFA (N,O-bis-(trimethylsilyl)-trifluoroacetamide; see Subheading 2.5). Polar functional groups like carboxylic groups in wax acids and alcoholic groups in wax, cutin, and suberin are capped by TMS (trimethylsilyl; Fig. 1). 3. For derivatization, 20μl BSTFA and 20μl pyridine as a catalyst are added to the samples dissolved in 200μl solvent (chloroform or hexane) and the samples are heated to 70  C for 40 min in a heating block.

284

Johanna Baales et al.

A

BSTFA + Pyridine 70°C for 40 min

B

BSTFA + Pyridine 70°C for 40 min

Fig. 1 Reaction scheme for the derivatization (trimethylsilylation) of (a) hydroxylic and (b) carboxylic groups occurring in wax lipids, and cutin and suberin monomers

4. Derivatized samples are transferred to GC vials and should best be analyzed immediately, since the derivatization is not stable for longer time periods of days or weeks (see Note 7). 3.9 Preparation of Internal Standards

1. The internal standard is used for quantification (see Subheading 2.6). Therefore, preparation of stock solutions of internal standards and addition of defined amounts of standard to the samples must be precise (see Note 8). 2. A semimicro balance (resolution 10μg) or a micro balance (resolution 100 ng) should be used for weighing the exact amounts of standard in calibrated volumetric flasks. 3. Only material resistant to organic solvent (e.g., aluminum foil, glassware, and metal spoons) can be used. 4. Prepare internal standards with a concentration of 0.025% (w/v) in chloroform. 5. For the quantification of extracellular plant lipids with GC-FID, C24 or C32 alkanes are used. They are best detected by the flame ionization detector and similar to the lipid components of wax, cutin, or suberin, which are largely composed of CH2 chains.

3.10 GC-FID and GC-MS

The derivatized extracellular lipids are analyzed by GC via distribution between the two phases of the GC, the mobile (carrier gas, hydrogen or helium) and a stationary liquid phase (film coating of the capillary column). The lipids are separated according to their

285

ISTD

C28 AL

a

C50 ES

C48 ES

C46 ES

C44 ES

C40 ES

C42 ES

C38 ES C36 ES

C30 OL

C24 OL

C30 AL

C22 OL

200 100

C26 AL C29 AN C26 OL

300

C25 AN C24 AL

Signal (pA)

400

C27 AN

500

C28 OL

Gas Chromatographic Analysis of Wax, Cutin and Suberin

0 10

30

20

50

40

Time (min)

b

16

wax amount [µg/cm2]

14 12 10 8 6 4 2 0 0

10

20

30

40

50

60

70

80

90

100

time [s]

Fig. 2 (a) Chromatogram (GC-FID) of poplar (Populus  canescens) wax extracted from isolated cuticles with chloroform and derivatized with BSTFA. (ISTD: C24 alkane, internal standard; AN: alkanes with chain lengths of C25–C29; OL, primary alcohols, C22–C30; AL, aldehydes, C24–C30; ES, esters, C36–C50). (b) Kinetic of wax extraction with chloroform from intact barley leaves. After 10 s, a plateau of the wax amount is reached indicating complete wax extraction. Data show mean  SD of three replicas

size, polarity and interaction with the stationary phase and are eluted according to their characteristic retention times (for separation of waxes, see Fig. 2a; for cutin monomers, see Fig. 3) (see Note 9). For GC-FID analyses of plant lipids, a DB-1 column is used (see Subheading 2.5). Compounds are ionized in the hydrogen flame of the FID. During combustion, electrons are released, and detected and quantified by measuring the current. The intensity of the

Johanna Baales et al.

C16-di- OH

286

1200

400

ISTD

600

C18-di-OH

600

C16- di- AC C16- ω-OH C16-OH-di- AC

cis-Coumaric-AC trans --Coumaric-AC

Signal (pA)

1000

200

0 10

20

30

40

50

60

Time (min) Fig. 3 Chromatogram (GC-FID) of tomato fruit (Solanum lycopersicum) cutin monomers after transesterification using BF3–methanol and derivatization with BSTFA. cis-Coumaric-AC, cis-coumaric acid; trans-coumaricAC, trans-coumaric acid; C16-di-AC, C16-dicarboxylic fatty acid; C16-ω-OH, C16-ω-hydroxy fatty acid; C16-OH-di-AC, C16-hydroxy-dicarboxylic fatty acid; C16-di-OH, C16-dihydroxy fatty acid; C18-di-AC, C18dicarboxylic fatty acid; ISTD, C32 alkane, internal standard

current is proportional to the amount of the organic compound passing through the hydrogen flame. In the final chromatogram, the current signal is plotted vs. time. Identical samples can be run on GC-MS for identification of the respective compounds. For GC-MS, the DB1-MS column can be used (see Subheading 2.8). After GC separation, lipids are ionized and fragmented in the electron impact (EI) ion source of the mass spectrometer. Resulting fragment ions are accelerated in an electrical field toward the quadrupole mass selector and detected according to their mass-to-charge ratio (m/z). The recorded mass spectrum is characteristic for each substance. 1. Place the sample vials in the autosampler using a random organization (see Note 9). 2. Use hydrogen as carrier gas for GC-FID and helium for GC-MS; flow rate: 2 ml/min for GC-FID or GC-MS. 3. Injection volume: 1μl, injection type: on column-injection or split/splitless-injection without split. 4. Analysis of cuticular wax with on-column injection. Oven temperature: initial 50  C hold for 2 min, 40  C/min up to 200  C hold for 2 min, 3  C/min up to 310  C hold for 30 min.

Gas Chromatographic Analysis of Wax, Cutin and Suberin

287

5. Analysis of cutin and suberin monomers with on-column injection. Oven temperature: initial 50  C hold for 2 min, 10  C/ min up to 150  C hold for 1 min, 3  C/min up to 310  C hold for 15 min. 6. Analysis of cutin and suberin monomers with split/splitless injection. Oven temperature: initial 50  C hold for 2 min, 25  C/min up to 200  C hold for 1 min, 10  C/min up to 320  C hold for 8 min. 7. Quantification of the compounds during GC-FID is based on the amount of internal standards (see Subheading 2.4) added to each sample (see Note 8). Areas of integrated peaks from GC-FID chromatograms (Figs. 2a and 3) are calculated. Amounts of the individual wax, cutin, and suberin compounds are quantified from the peak area of the compound multiplied with the ratio of the known amount of added standard to the peak area of the standard.

0,08

3,0 b - surface area per root length

a - dry weight per root length 2,5 (mm2 cm-1)

(mg cm-1)

0,06

0,04

2,0 1,5 1,0

0,02 0,5 0,00

0,0 A

B

C

A

root zones 10

C

10 d - suberin amount per area suberin amount (µg cm-2)

c - suberin amount per dry weight suberin amount (µg mg-1)

B root zones

8 6 4 2 0

8 6 4 2 0

A

B root zones

C

A

B

C

root zones

Fig. 4 Suberin monomer analysis over the length of barley roots (apical root zone A: 0–25% root length; middle root zone B: 25–50% root length; basal root zone C: 50–100% root length) exposed to osmotic stress in hydroponic culture ( 0.8 MPa). When suberin amounts (c) are related to root dry weight (a), the relative suberin amounts first increase (zone A–B) and then decrease (zone B–C). When suberin amounts (d) are related to root surface area (b), as an independent reference, suberin amounts continuously increase over the root length from zone A over B to C

288

Johanna Baales et al.

Table 1 Characteristic masses in MS spectra for the derivatized compounds in the major substance classes derived from wax, cutin, and suberin of different plant species

Samples

Substance class

Characteristic masses of fragments (m/z)

Cuticular wax

Trimethylsilylated fatty acids Trimethylsilylated primary alcohols Aldehydes Alkanes Esters

57; 73/75; 117; 129; 132; 145 73/75; 89; 102/103 82; 96 57; 71; 85; 99 57; 73/74; 87; 97; 125; 143

Cutin/suberin Primary fatty acid methyl esters Trimethylsilylated primary alcohols Trimethylsilylated ω-hydroxy fatty acid methyl esters 1,ω-dicarboxylic fatty acid methyl esters Trimethylsilylated 2-hydroxy fatty acids

74; 87 73/75; 89; 102/103 55; 75; 87; 103; 146; 159 57; 74; 84; 87; 98; 112 73; 89; 103; 129; 159

Aromatics

73; 203; 250 73; 219; 250; 280

Trimethylsilylated coumaric acid methyl ester Trimethylsilylated ferulic acid methyl ester

8. Results should be related to an independent reference parameter, preferably the surface area of the tissue (leaf epidermal area, root endodermal surface area) instead of dry weight (Fig. 4). Suberin amounts should not be related to root dry weight, since dry weight significantly changes along the root length (Fig. 4a). Because amounts of other cell wall polymers (lignin, carbohydrates) also change along the root length, relation of suberin amounts to root dry weight provides misleading patterns (Fig. 4c). Therefore, the amount of suberin should be related to the root surface area (Fig. 4b, d) giving rise to suberin patterns in agreement with the microscopical observations after suberin staining [9]. 9. Identification of the individual wax components, suberin and cutin monomers is based on the comparison of measured fragmentation patterns during GC-MS with fragmentation patterns published in literature or, if available, with in-house GC-MS libraries (Table 1). 3.11 Column Maintenance and Acid Standard Chromatography

1. For testing the quality of the column and the sensitivity of the GC detector (GC-FID and GC-MS), a TMS-derivatized acid standard mixture dissolved in chloroform can be used (see Subheading 2.6; see Note 11). 2. Before and/or after each analysis or within the analysis of a larger series of samples, the TMS-derivatized acid standard mixture can be run as control. Since acids respond very sensitively, their signal rapidly decreases if the quality of the column or the sensitivity of the detector decreases, whereas alkanes are still detected without losses in intensity.

Gas Chromatographic Analysis of Wax, Cutin and Suberin

289

3. Chromatography of the acid standard mixture for column maintenance using on-column injection (see Subheading 2.6). Oven temperature: 50  C hold for 1 min, 40  C/min up to 200  C hold for 2 min, 3  C/min up to 310  C hold for 15 min. 4. Chromatography of the acid standard mixture for column maintenance using split/splitless injection without split (see Subheading 2.6): initial 50  C hold for 1 min, 40  C/min up to 200  C hold for 2 min, 3  C/min up to 310  C hold for 15 min. 5. A correction factor is calculated by dividing the peak area of the C24 alkane with the area of the different acids. The GC column is in very good condition if the resulting value is close to 1.0. This indicates that acids can be detected almost as well as the alkane. Values lower than 1.3 indicate that the GC system needs maintenance. 6. If the peaks during GC-FID and GC-MS analysis of the acid standard mixture show tailing or shoulders, maintenance work is required. The column should be examined for black dots and irregularities on the inner coat around the injector side, which indicates that contaminations affect GC performance. In this case, the first 10–20 cm of the column containing the contaminations should to be removed by cutting. After connecting the column back to the injector, the oven is heated to 310  C at 140 kPa inlet pressure until the baseline is constant for at least 10 min. 7. Prior to further sample analysis, the success of the column shortening should be verified by running an acid standard mixture.

4

Notes 1. Sample preparation: Before analyzing extracellular plant lipids by GC-FID or GC-MS, weigh the samples individually on a micro-scale balance and determine the surface area by scanning. For dry weight determination, the samples are stored over activated silica gel in the desiccator for several days or even weeks. For suberin and cutin analyses, 1–3 mg of dry weight should be used per replicate. For wax analysis, the amount of sample depends on the chloroform volume used for extraction. The chloroform volume should be large enough to cover the sample and to ensure that all waxes are dissolved. For leaves with a low wax load, several leaves can be combined and dipped together to ensure that the amount of the extracted wax is above the detection limit of GC-FID or GC-MS.

290

Johanna Baales et al.

Table 2 Mean total amounts of cuticular wax of selected plant leaves and fruits Species

Plant organ

Mean wax coveragea (μg cm 2)

Arabidopsis thaliana

Leaf

0.5

Solanum lycopersicum

Leaf

7

Hordeum vulgare

Leaf

12

Hedera helix

Leaf

25

Prunus laurocerasus

Leaf

80

Prunus avium

Fruit

50

Malus domestica

Fruit

500

a

Data for wax coverage obtained in the author’s laboratory

2. Extraction of suberin from intact roots and seeds: intact roots and seeds need to be thoroughly extracted with chloroform– methanol, to remove membrane lipids and storage lipids. Larger seeds should be cut into two halves, smaller seeds (e.g., seeds from Arabidopsis thaliana) can best be crushed in liquid nitrogen for an efficient extraction of internal lipids. 3. Quality control of samples in autosampler vials for protecting the GC instrument: Before running samples on the GC instruments, check for unusual coloration, phase separation (two phases) or particles in the samples. Intense sample coloration might indicate a high analyte concentration. Dilute the analyte if the solvent appears milky. GC instruments have a lower detection limit in the ng range, and an upper detection limit in the μg range. For GC-FID systems, the lower detection limit is generally superior compared with GC-MS. Thus, for MS analysis, the sample concentration normally needs to be increased by two- to fourfold. Samples containing two phases should never be injected on the GC instrument. Particles (e.g., salts from the transesterification reaction) might clog the syringe of the injector or the column. All samples should first be analyzed at very low concentrations using the GC-FID. In case of poor sample quality or overloading, maintenance of the GC-FID is easier than of the GC-MS. Information on the load of wax, cutin or suberin of plant species and organs can be retrieved from the literature and from Table 2. 4. Extraction kinetics of waxes: The original protocol established for Arabidopsis thaliana leaf waxes demonstrated that dipping of leaves or isolated cuticles for 10 s in chloroform is sufficient for total wax extraction. For other plant species, a kinetic should be recorded to ensure complete wax extraction. For example, wax extraction from intact barley leaves is complete

Gas Chromatographic Analysis of Wax, Cutin and Suberin

291

Fig. 5 Scanning electron microscopic (SEM) photo of a wheat (Triticum aestivum) leaf treated with Collodion. In the lower right part, characteristic epicuticular wax structures are clearly visible, whereas in the upper left side, they have been mechanically lifted off with Collodion

after 10 s (Fig. 2b) because the extraction kinetic results in a plateau after 10 s. 5. Separating epicuticular waxes from intracuticular waxes: Collodion can be used to selectively remove epicuticular waxes without releasing intracuticular waxes as shown by scanning electron microscopy (SEM) (Fig. 5) and GC-FID analysis. Always use a soft brush when applying Collodion. Additional SEM investigations should be performed to test whether the surface appears smooth after removal. The SEM photo in Fig. 5 shows a clear border between the treated (upper left) and the untreated (lower right) side, indicating that epicuticular waxes have been fully removed. 6. Phase separation: The volumes of BF3–methanol or methanolic HCl used for transesterification of cutin and suberin need to be increased in the case of high sample concentration. Phase separation during monomer extraction should be clearly visible. To promote phase separation, add more organic solvent or more water and shake vigorously. 7. Derivatized samples and storage stability: Derivatized samples should always be analyzed without long periods of storage, since the trimethylsilyl ether bond is unstable. 8. Internal standards: Ideally one individual standard should be used for each analyte. However, when analyzing wax, cutin or suberin lipids, the total amount of individual substances can be very high (sometimes >30 peaks), and commercial standards identical to the biological compounds are not available. Therefore, often one standard compound, which is absent from the

292

Johanna Baales et al.

sample, is used. The internal standard should be spiked to the samples during the first steps of sample preparation. After addition of the internal standard, the samples can be stored at 4  C. Each further step of sample processing affects the amount of sample and the internal standard in the same way. During GC analysis, the internal standard peak should elute in the same time range, but should not overlap with the analyte peaks. Alkanes, composed of CH2 and CH3 groups, are best detected by GC-FID, since they can be fully oxidized to CO2 and H2O in the hydrogen flame. The presence of heteroatoms, for example oxygen in alcohols or acids, leads to signal reduction. Since most wax lipids, cutin and suberin monomers contain oxygen, ideally for every lipid a correction factor should be calculated in relation to the alkane standard. 9. Arrangement of the samples on the GC-FID and GC-MS autosampler: Every injected sample can contaminate the column and thus lead to a reduction in signal intensity. Therefore, the samples should be placed in random positions on the autosampler. If two sample groups (e.g., wild type and mutant) are analyzed in blocks, decreasing column quality during chromatography leads to a continuous loss of sensitivity, and results might be biased. A blank (a vial containing solvents, internal standard and derivatized and treated in the same way as the sample) which serves as negative control should be included in the middle of the sequence of samples. Normally a total of 20 wax samples can be run in one sequence. For cutin and suberin monomer analysis, only 10–15 samples per sequence are recommended, since sample preparation involves harsher chemicals, and consequently GC columns suffer more. 10. On-column vs. split/splitless injection: On-column injection of samples is needed for high molecular weight molecules (e.g., wax esters). High molecular weight molecules need high temperatures for evaporation, and they could stick to the liner of the split/splitless injector without reaching the column and would disappear from the chromatogram. For cutin and suberin monomer analyses, split/spitless injection is superior since residues from the transesterification reaction are trapped in the liner, thus preventing the contamination of the GC column. 11. Acid standard mixture before and after the runs: After every GC analysis of a sequence of samples, an acid standard mixture (C24 alkane, and C29, C30, and C31 acids, all lipids in equal amounts, and trimethylsilylated) has to be run. In the chromatogram of the acid standard mixture, calculate the ratio between the C24 alkane and the longest acid (C31 acid), and inspect the peak appearance, shape, and height. In the case of a bad acid standard value higher than 1.3 and/or bad peak

Gas Chromatographic Analysis of Wax, Cutin and Suberin

293

shapes (tailing of the peaks or split peaks), perform maintenance work. The common procedure for maintenance is to remove 10–20 cm of the column on the injector or detector side. On column-injection of a chloroform–methanol solvent control helps washing away contaminations. Also the injection of pure BSTFA onto the column helps to remove polar residues from the column. After cutting the column, the GC oven is heated to 310  C and the inlet pressure is raised to 140 kPa. The baseline should ideally decrease during heating of the column and peaks derived from contaminations and from the detector side should disappear. To test if maintenance work was successful an acid standard mixture should be run.

Acknowledgments The procedures for wax, cutin, and suberin analysis have been established over many years in many different labs and credits have to be given to Jo¨rg Scho¨nherr, Markus Riederer, Claus Markst€adter, Ju¨rgen Zeier, and Klaus-Dieter Hartmann, who all contributed to the optimization of different aspects of the protocols described here. This work was supported by grants from DFG, BASF SE, and SYNGENTA to L.S. References 1. Jetter R, Kunst L, Samuels AL (2006) Composition of plant cuticular waxes. Ann Plant Rev 23:145–181. https://doi.org/10.1002/ 9780470988718.ch4 2. Riederer M, Mu¨ller C (2006) Biology of the plant cuticle. Blackwell, Oxford 3. Kreszies T, Schreiber L, Ranathunge K (2018) Suberized transport barriers in Arabidopsis, barley and rice roots: from the model plant to crop species. J Plant Physiol 227:75–83. https://doi. org/10.1016/j.jplph.2018.02.002 4. Buschhaus C, Jetter R (2011) Composition differences between epicuticular and intracuticular wax substructures: how do plants seal their epidermal surfaces? J Exp Bot 62:841–853. https://doi.org/10.1093/jxb/erq366 5. Zeisler V, Schreiber L (2016) Epicuticular wax on cherry laurel (Prunus laurocerasus) leaves does not constitute the cuticular transpiration barrier. Planta 243:65–81. https://doi.org/10. 1007/s00425-015-2397-y 6. Riederer M, Schneider G (1989) Comparative study of the composition of waxes extracted

from isolated leaf cuticles and from whole leaves of Citrus: evidence for selective extraction. Physiol Plant 77:373–384. https://doi.org/10. 1111/j.1399-3054.1989.tb05656.x 7. Pollard M, Beisson F, Li YH, Ohlrogge JB (2008) Building lipid barriers: biosynthesis of cutin and suberin. Trends Plant Sci 13:236–246. https://doi.org/10.1016/j. tplants.2008.03.003 8. Schreiber L, Hartmann K, Skrabs M, Zeier J (1999) Apoplastic barriers in roots: chemical composition of endodermal and hypodermal cell walls. J Exp Bot 50:1267–1280. https:// doi.org/10.1016/j.tplants.2008.03.003 9. Kreszies T, Shellakkutti N, Osthoff A, Yu P, Baldauf JA, Zeisler-Diehl VV, Ranathunge K, Hochholdinger F, Schreiber L (2019) Osmotic stress enhances suberization of apoplastic barriers in barley seminal roots: analysis of chemical, transcriptomic and physiological responses. New Phytol 221:180–194. https://doi.org/10. 1111/nph.15351

Chapter 16 Isolation of Lipid Droplets for Protein and Lipid Analysis Patrick J. Horn, Kent D. Chapman, and Till Ischebeck Abstract Cytosolic lipid droplets (LDs) are organelles which emulsify a variety of hydrophobic molecules in the aqueous cytoplasm of essentially all plant cells. Most familiar are the LDs from oilseeds or oleaginous fruits that primarily store triacylglycerols and serve a storage function. However, similar hydrophobic particles are found in cells of plant tissues that package terpenoids, sterol esters, wax esters, or other types of nonpolar lipids. The various hydrophobic lipids inside LDs are coated with a phospholipid monolayer, mostly derived from membrane phospholipids during their ontogeny. Various proteins have been identified to be associated with LDs, and these may be cell-type, tissue-type, or even species specific. While major LD proteins like oleosins have been known for decades, more recently a growing list of LD proteins has been identified, primarily by proteomics analyses of isolated LDs and confirmation of their localization by confocal microscopy. LDs, unlike other organelles, have a density less than that of water, and consequently can be isolated and enriched in cellular fractions by flotation centrifugation for composition studies. However, due to its deep coverage, modern proteomics approaches are also prone to identify contaminants, making control experiments necessary. Here, procedures for the isolation of LDs, and analysis of LD components are provided as well as methods to validate the LD localization of proteins. Key words Lipid droplets, Oil bodies, Subcellular fractionation, Proteomics, Mass spectrometry, Plant organelles, Confocal microscopy, Transient expression, Triacylglycerols

1

Introduction Plant lipid droplets (LDs, also referred to as lipid/oil bodies or oleosomes) are dynamic, intracellular organelles with a unique composition [1] that are generally observed in the cytosol of all cell types [2, 3]. LDs, irrespective of cell-type, tissue source, or plant species, share a spherical structure consisting of a neutral lipid core surrounded by a phospholipid monolayer with multiple LD-associated proteins. In most tissues, such as oilseeds where LDs are abundant, the neutral lipid core consists predominantly of triacylglycerols (TAGs). In some plants, this core is enriched in other neutral lipids such as terpenoids [4, 5], sterol esters [6–8], and wax esters [9]. The composition of the LD phospholipid monolayer is similar to the cytosolic leaflet of the endoplasmic

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_16, © Springer Science+Business Media, LLC, part of Springer Nature 2021

295

296

Patrick J. Horn et al.

reticulum (ER), a result of LDs forming from ER membranes [2]. Hence, LDs contain mostly phosphatidylcholine (PC) and lower amounts of other common phospholipids [10–12]. Mature plant LDs are typically between 0.5 and 2.0 μm [3] in diameter with their physical properties making them relatively easy to stain and visualize with lipophilic dyes. However, variations in the composition of the phospholipid–protein monolayer can result in substantial size differences among cell and tissue types (e.g., avocado mesocarp LDs, 5–20 μm, [13] versus oilseeds, 0.5–2.0 μm [3]), organisms (Chlamydomonas reinhardtii, ~1.5 μm, [14] vs. Chlorella sp., >3 μm [15]), and mutants (up to
6 μm in cotyledons of oleosin mutants [16], and even larger in seipin mutants [17]). Differences in lipid composition among LD populations derived from the same tissues, such as cotton seeds, demonstrate a remarkable heterogeneity at the organelle level [18]. Given that the LD lipid composition likely drives interactions with LD-associated proteins and downstream functions, it is essential to further characterize the lipid composition of isolated LDs from plants subjected to dynamic growth conditions or in specific genetic backgrounds. The neutral lipid core of LDs, and the proportionally low protein content renders LDs more buoyant (i.e., lower density) than other cellular compartments enabling their isolation by flotation centrifugation and subsequent protein/lipid characterization not only in plants [19, 20] but also in other species [21, 22]. Multiple types of tissues (seeds, mesocarp, leaves, roots, pollen tubes, etc.) containing LDs, derived from plants grown under normal conditions [13, 18, 23, 24] or under conditions triggering LD formation [14, 25], can serve as a source of LDs. LDs can be released into a buffered sucrose solution by relatively mild tissue and cell homogenization/disruption. The homogenization solution is formulated to maintain the structural integrity of LDs and to presumably retain in vivo LD-protein interactions. Using differential centrifugation, which relies on separating particles of different densities or sizes subjected to centrifugal force, LDs float to the top of the centrifuge tube forming an enriched “fat pad.” This crude fat pad can be purified through subsequent washes/floats that vary in stringency depending on the downstream analytical methods, and are particularly important for characterizing the LD proteome. The combination of LD isolation with mass spectrometrybased proteomics has expanded the collection of known proteins associated with LDs, that is, the LD proteome. Some studies rely solely on the analysis of LD-enriched fractions in the identification of LD proteins [26]. However, modern proteomics approaches regularly identify hundreds of proteins likely to include contaminants. Therefore, many studies use controls such as the comparative analysis of other subcellular fractions [14] or cell biological methods [13] or both [25]. Especially the quantitative comparison of different fractions has led to improved identification of LD proteins that were then confirmed by monitoring their localization in

Isolation and Analysis of Plant Lipid Droplets

297

transient expression systems [23, 24]. With the help of these efforts, the number of known LD protein families increased from three in 2013 to fifteen in 2019 from the model species Arabidopsis thaliana. LDs can be associated with other membranes such as with the ER during biogenesis, or peroxisomes for their degradation, and this may lead to the presence of “contaminating” proteins in LD fractions. Such proteins might have physiological relevance, even if they are not LD proteins per se. Hence, not all “non-LD proteins” should be discarded as artifacts. Moving toward the comprehensive characterization of the plant LD lipidome/proteome has revealed new LD functions, most of which support dynamic roles of LDs beyond high-energy and carbon storage. For example, proteomic screens have identified many LD-associated proteins involved in adjusting LD size and stability such as oleosins in oilseeds [16, 27], LD surface protein (LDSP) in the alga Nannochloropsis sp. [28], and major LD protein (MLDP) in the alga Chlamydomonas reinhardtii [14]. The identification of LD-anchoring proteins has enabled the localization of novel pathways to LD surfaces [29]. The analysis of purified LDs in avocado pericarp identified new LD-associated proteins (LDAPs) [13]. A follow-up analysis in the model system A. thaliana implicated these LDAPs in regulating the LD life cycle in various physiological contexts, including stress responses and post-germinative growth [30, 31], possibly also in concert with its interaction partner, LDAP interacting protein (LDIP) [32, 33]. LDs also harbor proteins involved in oxylipin synthesis such as a lipoxygenase in cucumber [34], or α- DIOXYGENASE 1 and CALEOSIN3 in A. thaliana [35]. The latter two proteins together produce the phytoalexin 2-hydroxy-octadecatrienoic acid that is important for pathogen defense. The free fatty acids needed as substrates by these enzymes could be provided by members of the oil body lipase (OBL) family [36, 37] as indicated from work in tomato [38]. Also, other recently discovered LD proteins hint to a broader role of LDs in metabolism, based on their predicted enzymatic function [23]. The LD protein, PLANT UBX DOMAINCONTAINING PROTEIN 10 (PUX10), plays a role in protein turnover of ubiquitinated LD proteins by recruiting the unfoldase cell division control protein 48 (CDC48), indicating that LDs harbor a protein degradation pathway similar to the ER [24, 39]. We outline practical guidelines to isolate and analyze the protein/lipid composition of LDs (Fig. 1). The methods cover procedures how to isolate LDs from diverse plant tissues, analyze LD lipid classes and fatty acid composition, characterize the LD proteome by quantitative and comparative mass spectrometry-based approaches, and verify the localization of LDs and LD-protein interaction. The rationale and technical difficulties associated with key steps are discussed to provide a foundation for reproducible results. Under most circumstances, each of these methods can be carried out independently providing versatility to address most

298

Patrick J. Horn et al.

(a) LD Isolation Clarified Homogenate

Homogenate

Sucrose Step Gradient

“Isolated fat pad”

Lipid Analysis

“fat pad”

}

13000 g 30 min

Homogenization (Buffer A)

13000 g 30 min

Transfer + Layer Buffer B

}

Visualization

Purified LDs fat pad”

Centrifugation / Collection

Source Tissue

Protein Analysis

“fat pad”

Characterization

(b) Lipid Extraction & Analysis StE TAG

Extraction Solvent + Lipid Standards

TAGs

Phase Separation

Purified LDs

} }

1000 g 5 min

TLC Loading

Concentrate Lipids

Lipid Extraction

1,2-DAG

Compositional Analysis

Transmethylation Aqueous

18:3 Abundance

FFA

Std

16:0 18:2 18:1 18:0 RT(min)

Organic PE PC LDs Lipids (Lipidomics)

Extraction / Separation

Standards

Thin Layer Chromatography

Gas Chromatography

(c) Protein Processing & Analysis

Purified LDs

Trypsin Digest

Protein Precipitation Data Acquistion

(SDS-PAGE + Western Blot)

Extraction / Digestion

Abundance

Gel Loading

H

A R S

A

V

Bioinformatics

Step 1

Step 2

Step 4

Step 3

Step 5

m/z

Separation / Detection

Identification / Quantification

Fig. 1 Integrated overview of protocols for the isolation and analysis of plant LDs. (a) Protocols for the purification of LDs, from plant tissues, by differential centrifugation are described in Subheadings 3.1 and 3.2. (b) Protocols for the extraction of lipids from purified LDs, separation by TLC followed by GC analysis are described in Subheading 3.3. (c) Protocols for the precipitation of proteins from purified LDs followed by processing into peptides for LC-MS/MS analysis are described in Subheadings 3.4 and 3.5. Data generated can be processed using the MaxQuant and Perseus software packages to characterize the LD proteome. An example of visualizing in vivo LDs and associated proteins, is shown in Fig. 2 and the protocols are described in Subheadings 3.6 and 3.7. Alternative workflows are provided in parentheses. DAG diacylglycerol, FFA free fatty acids, PC, phosphatidylcholine, PE phosphatidylethanolamine, Std standard, StE sterol ester, TAG triacylglycerol

LD-related questions. Some optimization will be required for uncharacterized source tissues that vary in ease of purification and LD lipid–protein composition. This integrated approach will enable an increasing number of labs to analyze plant LDs to enhance our knowledge of these unique organelles.

Isolation and Analysis of Plant Lipid Droplets

2

299

Materials

2.1 General Lab Supplies

1. 18 MΩ water (ddH2O). 2. Hot plate with magnetic stirring. 3. Borosilicate glass Pasteur pipettes, 5–3/400 (14.5 cm) length. 4. Plastic transfer pipettes. 5. Standard pipettors and tips. 6. Aluminum foil. 7. Single-edge razor blades. 8. Water bath or heat block.

2.2

LD Isolation

1. Homogenization apparatus. Small seeds, siliques, seedlings, roots (e.g., Arabidopsis thaliana): mortar and pestle and sea sand or Dounce homogenizer. Large seeds (e.g., cotton, castor, soybean), fruits: singleedge razor blades, flat container such as Tupperware or weigh boat 12.5 cm  12.5 cm. Leaves: single-edge razor blades or mechanical tissue homogenizer (e.g., Waring blender or handheld homogenizer) or mortar and pestle and sea sand. 2. Buffer A (minimal components, suitable for general LD isolation for lipid analysis): 600 mM sucrose, 100 mM KCl in 100 mM potassium phosphate buffer (pH 7.2) (see Note 1). Filter-sterilize the buffer. Store at 4  C for up to 3 months. 3. Buffer A* (suitable for proteomics): same as above but with the protease inhibitor 200 μM phenylmethylsulfonyl fluoride (PMSF) and the cross-linker 0.5 M dithiobis (succinimidyl propionate). Caution: PMSF and chemical cross-linkers are highly toxic and should be handled according to manufacturer’s instructions. PMSF is unstable in aqueous solutions and should be added fresh. 4. Buffer B: same as Buffer A but with 400 mM sucrose (see Note 2). 5. Buffer B* (proteomics): same as Buffer A* but with 400 mM sucrose and without cross-linker (see Note 2). 6. Miracloth. 7. 5- or 10-ml borosilicate glass disposable serological pipettes. 8. Stainless steel spatula. 9. High-speed, refrigerated centrifuge (e.g., Sorvall Lynx 4000 or similar) with swinging bucket rotor (e.g., Sorvall HB-6 or similar). Fixed angle rotors can also be used but produce less than ideal separations.

300

Patrick J. Horn et al.

10. Kimble high-strength 15 or 30 ml glass centrifuge tubes (preferred). Similar Teflon (PTFE) tubes can be also used (e.g., Nalgene™ Oak Ridge High-Speed Centrifuge Tubes). 11. (Optional) Microcentrifuge with standard microcentrifuge tubes for small-volume separations. 12. (Optional) Ultracentrifuge with fixed-angle (e.g., TLA-100 rotor with appropriate centrifuge tubes) capable of 100,000–150,000  g. 2.3 LD Lipid Extraction and Analysis (See Note 3)

1. Silica gel G-60, 20  20 cm, TLC plates (e.g., EMD Millipore 1.05721.0001). 2. Glass TLC chamber. 3. Large Whatman qualitative filter paper (VWR 10034-772). 4. Test tube racks. 5. Pencil. 6. Chloroform, ACS grade. Caution: All organic solvents should be used according to Material Safety Data Sheets (MSDS) due to acute toxicity inhalation. 7. Isopropanol, ACS grade. 8. Hexane, ACS grade. 9. 0.9% (w/v) KCl. 10. Butylated hydroxytoluene (BHT). 11. 1 N methanolic HCl (100 ml can be prepared from ~12 ml of 12 N HCl and 88 ml methanol). 12. Chloroform–methanol–ddH2O (5:5:1, v/v/v). 13. Neutral lipid separation solvent (~100 ml): hexane/diethyl ether/acetic acid (80:20:1, v/v/v). All solvents ACS grade. 14. Polar lipid separation solvent (~100 ml): chloroform/methanol/acetic acid–ddH2O (85:15:10:3.5, v/v/v/v). All solvents ACS grade. 15. Table-top centrifuge with swinging bucket rotor (e.g., Thermo Scientific Legend X1R, TX-400 rotor). 16. Disposable culture tubes with screw cap, 13  100 mm. Caps should be PTFE lined (e.g., ThermoFisher Scientific 055694). 17. Disposable culture tubes, 13  100 mm. 18. High-quality nitrogen gas source (i.e., gas cylinder or generator). 19. GC with flame ionization detection (e.g., Agilent 7890B), appropriate glass autosampler vials. 20. Appropriate lipid standards: Odd chain fatty acids (15:0, 17:0) for GC analysis. TLC lipid class mixtures (e.g., Nu-Chek Prep

Isolation and Analysis of Plant Lipid Droplets

301

TLC 16-1 A). Other representative standards for each lipid class may be purchased from multiple suppliers, although we recommend Nu-Chek Prep and Avanti Polar Lipids. 2.4 Protein Processing

1. Cold acetone, ACS Grade. 2. 6 M urea + 5% (w/v) sodium dodecyl sulfate (SDS). 3. 2 Laemmli sample buffer (4% (w/v) SDS, 20% (v/v) glycerol, 120 mM Tris–HCl, pH 6.8, 0.02% (w/v) bromophenol blue). 4. SDS–polyacrylamide gel electrophoresis (SDS-PAGE) precast gels (e.g., Bio-Rad Mini-PROTEAN TGX) or prepare homemade gels [40]. 5. 2-Mercaptoethanol. Caution: acute toxicity inhalation. 6. Protein electrophoresis equipment (e.g., Bio-Rad Mini-PROTEAN® System). 7. Protein quantification kit (e.g., bicinchoninic acid, BCA kit). 8. Kimble high-strength 15 or 30 ml glass centrifuge tubes (preferred). 9. Trypsin, proteomics grade. See Shevchenko et al. [41] and Rappsilber et al. [42] for helpful notes on materials for in-gel digest and peptide purification. 10. Acetonitrile, ACS grade. 11. Ammonium bicarbonate. 12. Iodoacetamide. 13. Dithiothreitol (DTT). 14. Formic acid, ACS grade. 15. Methanol, ACS grade. 16. Empore solid phase extraction discs C18 (3M, Saint Paul, USA). 17. Eppendorf Protein LoBind Tubes.

2.5 Proteomic Analysis

1. MaxQuant software package [43, 44] can be downloaded free of charge at www.maxquant.org/download_asset/maxquant/ latest. Accessed 5 Mar 2020. The workflow below was developed with version 1.6.2.10 but should also work with newer versions. 2. Perseus software [45] can be downloaded free of charge under www.maxquant.org/download_asset/perseus/latest. Accessed 5 Mar 2020. The workflow below was developed with version 1.6.2.2 but should also work with newer versions. 3. A proteome database of the species analyzed in FASTA format. A nonredundant protein sequence database from Arabidopsis [46] can be found for example under https://bar.utoronto.ca/ thalemine/dataCategories.do. Accessed 5 Mar 2020.

302

Patrick J. Horn et al.

2.6 Verification of Protein-LD Localization by Expression in N. benthamiana Leaves

1. Nicotiana tabacum or Nicotiana benthamiana plants, 4-weekold grown under standard conditions (~14 h light, ~23  C). 2. Agrobacterium tumefaciens (strain LBA4404 or GV3101) transformed with appropriate binary vectors (e.g., pMDC32, pMDC43, pMDC84 [47], or similar, with coding sequence of interest fused to fluorescence tag). 3. pOREO4-P19 vector (viral-encoded suppressor of gene silencing, the p19 protein of tomato bushy stunt virus) [48]. 4. pORE04-LEC2 vector (enhances production of leaf LDs). 5. Shaking incubator, 28  C. 6. Luria–Bertani (LB) broth supplemented with appropriate antibiotics. 7. Infiltration medium: 5 mM 2-(N-morpholino)ethanesulfonic acid (MES)-NaOH, 5 mM MgSO4, pH 5.7, with 100 μM acetosyringone freshly added. 8. 1 ml disposable plastic syringes (e.g., BD 309659). 9. Standard bacterial culturing supplies: 250 ml sterile Erlenmeyer flasks, 1.5 ml microcentrifuge tubes, 50 ml standard centrifuge tubes or similar. 10. Single hole punch. 11. 1 M piperazine-N,N0 -bis(2-ethanesulfonic acid) (PIPES)NaOH, pH 7.0, stock solution. Store at 4  C for up to 3 months. 12. BODIPY™ 493/503 stock solution: 4 mg/ml BODIPY 493/503 (ThermoFisher Scientific D3922) in dimethyl sulfoxide (DMSO). Store 20  C up to 1 year in aliquots, protected from light with aluminum foil. 13. BODIPY 493/503 working solution: 2 μg/ml in 50 mM PIPES-NaOH buffer. Prepare fresh and protect from light. 14. Rotational shaker. 15. Standard microscope slides and coverslips. 16. Kimwipes® tissues. 17. Standard fingernail polish, clear. 18. Confocal microscope (e.g., Zeiss LSM710 confocal laser scanning microscope).

2.7 Verification of LD Localization by Expression in N. tabacum Pollen Tubes

1. Flowering Nicotiana tabacum (Samsun NN) plants grown under 14 h of light at 23  C. Freshly opened anther buds from six flowers are needed for one construct. Planting six plants every 3 weeks gives sufficient flowers for biolistic transformation of three constructs per day. 2. Plasmid suitable for pollen tube transformation containing the coding sequence of interest. Expression is reliably driven by a

Isolation and Analysis of Plant Lipid Droplets

303

Lat52 promoter. Monomeric fluorophores are preferable to prevent aggregation. Smaller nonbinary vectors give much better results. Suitable vectors are pUCLAT52-mVenusC, pUC-LAT52-mVenusN and pUC-LAT52-mCherryC [49, 50] or the Gateway vectors pLatMVC-GW, pLatMVNGW, and pLatMCC-GW [51]. 3. Midi-prep kit (e.g., Plasmid Midikit from Qiagen (Venlo, Netherlands)). 4. Gold particle stock (60 mg/ml). For preparation, wash 60 mg of 1 μm gold particles (Bio-Rad) once with 1 ml 70% ethanol and three times with 1 ml ddH2O by vortexing and 5 s spins at ~20,000  g. Resuspend in 1 ml of 50% glycerol (v/v). 5. 2.5 M CaCl2 in ddH2O. Store at room temperature (RT). 6. 0.1 M spermidine in ddH2O (do not use spermidine trihydrochloride). Store at 20  C up to 3 months. 7. Pollen tube (PT) medium slightly modified from [52]. 5% (w/v) sucrose, 12.5% (w/v) polyethylene glycol (PEG)-6000, 15 mM MES-KOH, pH 5.9, 1 mM CaCl2, 1 mM KCl, 0.8 mM MgSO4, 1.6 mM H3BO3, 30 μM CuSO4. Sterile filtered, it can be used for at least a year when stored at 4  C. 8. Filter paper. 9. Nitrocellulose filter. 10. An apparatus for sterile filtration. 11. Gene gun (Bio-Rad). 12. Macro carrier, rupture disks (1350 psi), stopping screens (all Bio-Rad). 13. Large petri dishes. 14. Microscope slides. 15. Formaldehyde solution 36–38%. 16. LD stain (suitable are monodanyslpentane, BODIPY 505/515, Nile Red). Stock solutions 1 mg/ml in DMSO. 17. Epifluorescence or confocal microscope.

3

Methods

3.1 Sample Collection and LD Isolation

1. Tissue harvesting Desiccated Seeds—Seeds can be harvested for LD isolation and analyses with no preservation steps, as they are already metabolically quiescent. It is helpful to excise the seed coat from oilseeds (e.g., cotton, canola, camelina), before homogenization. This can be performed by imbibing seeds for up to 15 min to soften the seed coat. Imbibition can result in the

304

Patrick J. Horn et al.

induction of metabolic events, so if possible excise the seed coat (without imbibition treatment) with a single-edge razor blade to isolate seed embryos. Metabolically Active Tissues—After harvesting, fresh tissues (roots, leaves, developing seeds, pollen tubes, etc.) should be placed on ice in appropriate homogenization solution (see Note 1). Tissues should not be frozen prior to isolation due to potential fractures of LD. 2. Tissue Homogenization: see Note 1. All steps should be carried out on ice, wherever feasible. We recommend a ratio of 5:1 of ice-cold Buffer A (or A*) volume (ml) to gram tissue as a starting point for isolation. Homogenization should be carried out in a timely manner. We provide three homogenization options depending on the tissue type: Small seeds, siliques, seedlings, roots, pollen tubes—Grind tissues in Buffer A (A*) in an ice-old mortar and pestle or Dounce homogenizer until a fine homogenate is formed (approximately 1–2 min). (Optional) Sea sand can be included to improve homogenization. Large seed embryos (seed coat excised) can be resuspended in ice-cold Buffer A (or A*) and subsequently chopped with single-edge razor blades until 90%, regardless of the TLC-developing solvent used, and the equivalent of as little as 8 pmol lipid has been detected using GC with flame-ionization detector. Besides the analysis of phosphoinositides, this method can easily be adapted to analyze lipids from other fractions enriched by solid-phase adsorption chromatography.

382

2

Larissa Launhardt et al.

Materials All solvents should be of analytical grade. All solutions should be prepared using filtered and deionized water, here specified as H2O. Regulations for proper waste disposal of organic solvents should be strictly followed.

2.1 Acidic Extraction of Phosphoinositides (See Note 1)

1. Liquid nitrogen (N2). 2. Mortar and pestle (porcelain). 3. Sample concentrator (to remove organic solvent with N2 gas stream). 4. HCl solution: 2.4 M HCl in H2O. 5. EDTA solution: 0.45 M ethylenediaminetetraacetic acid (EDTA), pH 8 with NaOH in H2O. 6. Backwashing solution: 0.5 N HCl in 50% (v/v) methanol. 7. Extraction solvent: chloroform–methanol (1:2, v/v). 8. Redissolving solution: chloroform.

2.2 Separation and Enrichment of Lipid Classes by Solid-Phase Adsorption Chromatography

1. Silica solid-phase extraction (SPE) columns, 50 or 100 mg silica material (e.g., Agilent BondElutSI, Agilent Waldbronn, Germany). 2. Solvent A for the elution of neutral lipids: chloroform. 3. Solvent B for the elution of glycolipids: acetone–methanol (9:1, v/v). 4. Solvent C for the elution of phospholipids: methanol–acetic acid (100:1, v/v).

2.3 Chromatographic Separation of Phosphoinositides

1. Silica Si60 plates for thin-layer chromatography (TLC), 20 cm
20 cm or 20 cm
10 cm. 2. Oven at 180  C, best with a metal rack for stacking TLC plates. 3. Glass TLC developing chamber with tight sealing lid. 4. Potassium oxalate solution: 5% (w/v) potassium oxalate in H2O. 5. CDTA solution: stir 4.55 g of disodium trans-1,2-diaminocyclohexane-N,N,N0 ,N0 -tetraacetic acid (CDTA) in a mixture of 165 mL H2O, 330 mL ethanol, and 3 mL of 10 N NaOH until CDTA is fully dissolved [6]. 6. Developing solvent I: dissolve 12 g boric acid in 180 mL of chloroform–methanol–pyridine (75:60:45, v/v/v), add 7.5 mL H2O, 3 mL of 88% (v/v) formic acid, and 75 μL of ethoxyquin (antioxidant, technical grade) [6].

Quantification of Unlabeled Phosphoinositides from Plants

383

7. Developing solvent II: chloroform–methanol–saturated aqueous ammonium hydroxide–H2O (45:45:4:11, v/v/v/v) [7]. 8. Developing solvent III: methylacetate–2-propanol–chloroform–methanol–0.25% (w/v) aqueous KCl solution (25:25:25:10:9, v/v/v/v/v). 2.4 Visualization of Lipid Standards on TLC Plates

1. Glass cutter. 2. CuSO4 solution: 10% (w/w) CuSO4, 8% (w/w) H3PO4 in H2O. 3. Heating plate with a flat surface of sufficient size to uniformly heat a 20 cm
20 cm TLC plate at 180  C.

2.5 Derivatization of PhosphoinositideAssociated Fatty Acids

1. Internal lipid standard: 5 μg of tripentadecanoin or another lipid, for example phosphatidylcholine with nonnatural fatty acids (15:0 or 17:0). 2. Toluene–methanol (1:2, v/v). 3. 0.5 M sodium methoxide (NaOCH3) in methanol. 4. 5 M NaCl. 5. n-Hexane.

2.6 GC Analysis of Fatty Acid Methyl Esters (FAMEs)

1. Gas chromatograph. 2. Column for separation of nonpolar analytes (e.g., DB23 or Innovax, 30 m, Agilent). 3. Fatty acid standard mixture (e.g., Merck Polyunsaturated Fatty Acid Mix No. 3 from Menhaden Oil or any mixture of authentic fatty acid standards).

3

Methods Many steps in the described protocol involve organic solvents, which should be handled in a fume hood. Organic solutions used during lipid analysis workflows should always be handled using glass vials and glass pipettes to avoid contamination of the samples and damage to the plastic equipment. TLC and GC should also be set up so that exposure of researchers to toxic solvent fumes is minimized. Required safety precautions should always be adhered to, including wearing safety glasses for all steps involving acids or liquid N2.

3.1 Acidic Extraction of Phosphoinositides [8]

1. Grind 1–2 g (fresh weight) of plant tissue in liquid N2 to a fine powder using a mortar and a pestle. Cultured plant cells or yeast (up to 2 g cell pellet) can be harvested by centrifugation and can be directly used for lipid extraction or alternatively first homogenized with a Dounce homogenizer (see Note 2).

384

Larissa Launhardt et al.

2. To the ground plant tissue or cell pellet successively add 5–10 mL chloroform–methanol (1:2, v/v), 800–1600 μL EDTA solution and 1.5–3 mL HCl solution with intermittent mixing (see Note 1). 3. Incubate biological material in extraction solution for at least 2 h at 4  C while slowly shaking (see Note 2). 4. Add 1.5–3 mL chloroform and mix vigorously. 5. Spin mixture at 600
g for 2 min to separate phases. 6. Transfer lower (organic) phase to fresh open glass tube. 7. Reextract aqueous phase twice as described above and combine organic phases. 8. Add 3.5–7 mL of backwashing solvent to combined organic phases, and mix. 9. Spin mixture at 600
g for 2 min to separate phases. 10. Remove upper (aqueous) phase and discard. 11. Backwash residual organic phase two more times (steps 7–9). 12. Transfer lower (organic) phase into fresh glass tube. 13. Evaporate solvent under N2 gas flow (see Note 3). 14. Dissolve lipids in 1 mL chloroform and proceed with Subheading 3.2. 3.2 Separation and Enrichment of Lipid Classes by Solid-Phase Adsorption Chromatography

1. Place SPE column (see Note 4) in glass tube (see Note 5). 2. Equilibrate column with 1 mL of Solvent A and change glass tube afterward. 3. Load the SPE columns with lipid extract (see Note 6) and let the solvent pass through the column completely. 4. Elute neutral lipids from the column by successively adding 1 mL of Solvent A and let the solvent pass through the column completely. Repeat this step until the neutral lipids have been eluted in 10 mL of Solvent A (see Note 7). 5. Evaporate solvent of neutral lipid fraction under N2 flow and store dried neutral lipids at 20  C until further use (if desired). 6. Use a fresh glass tube for the elution of glycolipids from the column by adding 1 mL of Solvent B. Repeat this nine more times and let the solvent pass through the column completely (see Note 8). 7. Evaporate the solvent from the glycolipid fraction under N2 flow and store dried glycolipids at 20  C until further use (if desired). 8. Use a fresh 50 mL glass tube for the elution of phospholipids from the column by adding 1 mL of Solvent C. Repeat this nine more times and let the solvent pass through the column completely.

Quantification of Unlabeled Phosphoinositides from Plants

385

9. Add 9 mL chloroform and 7 mL H2O to the phospholipid fraction and mix. Centrifuge at 600
g for 2 min. 10. Transfer lower (organic) phase into fresh glass tube (see Note 9). 11. Reextract the aqueous phase one more time by adding 5 mL chloroform and combine the organic phases. 12. Evaporate the solvent from the phospholipid fraction under N2 gas flow and store dried phospholipids at 20  C until further use. 13. Redissolve dried lipids in 100 μL chloroform before proceeding with Subheading 3.3. 3.3 Chromatographic Separation of Phosphoinositides

Thin-layer chromatography is performed using standard silica Si60 plates in 20
20
10 cm vertical glass chambers filled with 100 mL of developing solvent and containing one sheet (20 cm
20 cm) of filter paper to maintain a constant solvent atmosphere. Alternatively, horizontal high-performance (HP)TLC setups of smaller scale can also be used as specified by the respective manufacturers. 1. Activate TLC plate. For separation of phosphatidylinositolmonophosphates, submerge plate for 10 s in CDTA solution. Alternatively, for separation of phosphoinositides and other phospholipids, submerge plate for 10 s in potassium oxalate solution. After either treatment, dry the plate in an oven at 180  C for at least 1 h. Let the plate cool at room temperature. 2. Apply lipids eluted from SPE column onto activated TLCplate. 3. Also spot lipid standards corresponding to the lipid classes of interest. 4. Develop TLC plate in appropriate developing solvent. separation of phosphatidylinositol-monophosphates, developing solvent I [6], for other phosphoinositides developing solvent II [7], and for other phospholipids developing solvent III (see Notes 10–12).

3.4 Visualization of Lipid Standards

For use use use

1. After development, dry TLC plate in the fume hood. 2. Separate plate section containing lipid standards using a glass cutter (see Note 13). 3. Submerge dried plate briefly in CuSO4-solution (see Note 14). 4. Let plate dry upright on paper towels. 5. Heat plate to 180  C on a heating plate (see Note 15) until charred lipid bands appear. 6. Allow plate to cool in the fume hood.

386

Larissa Launhardt et al.

3.5 Isolation of Lipids from Silica Plates

1. Place separated part of TLC plate with stained standard lipids next to unstained part containing the separated plant lipids. Mark positions of separated plant lipids according to migration of standards using a pencil. 2. Scrape areas with plant lipids from the unstained portion of the plate and collect silica powder for each spot into fresh glass tubes (see Notes 16–18). 3. Extract lipids from collected silica material twice with 2 mL of the respective developing solvent. Perform the second extraction in closed glass tubes with screw caps and Teflon septa overnight at 20  C. 4. Combine organic phases from extraction and reextraction, and evaporate solvent under N2 gas stream and proceed to Subheading 3.6. 5. For quantitative analysis, add internal lipid standard to samples before transesterification (Subheading 3.6).

3.6 Transesterification of Lipid-Bound Fatty Acids for GC Analysis [9]

1. Add 0.5 mL toluene–methanol (1:2, v/v) to dried lipids in a closed glass tube with screw cap and Teflon septum. 2. Add 0.25 mL 0.5 M NaOCH3 in methanol and mix. 3. Incubate at room temperature for 20 min while slowly shaking. 4. Add 100 μL 5 M NaCl in H2O. 5. Add 2 mL n-hexane and mix. 6. Briefly spin mixture at 600
g to separate phases. 7. Transfer n-hexane (upper) phase into fresh closed glass tube with a screw cap and Teflon septum. 8. Reextract lower phase twice with n-hexane. 9. Evaporate solvent under N2 gas flow. Fatty acid methyl esters (FAMEs) will form a film at the bottom of the tube.

3.7 Analysis of PhosphoinositideAssociated Fatty Acids

FAME analysis can be carried out using most GC equipment currently available. The chromatography column should be chosen to enable the separation of hydrophobic compounds. Examples for suitable columns include types DB-23 or Innowax columns (both from Agilent). Columns with other matrices may also be used, but the chromatography conditions may have to be adapted. 1. Redissolve dried lipids in 5–100 μL of acetonitrile and transfer to autosampler vial for GC analysis. 2. Inject 1 μL of the lipid sample (splitless) in the gas chromatograph at 250  C injector temperature. 3. Perform chromatography for example on a 30 m DB-23 column (Agilent) using helium as a carrier gas as follows: Hold oven temperature at 150  C for 1 min, then raise temperature to 200  C at a rate of 25  C/min, then to 250  C at 4  C/min, and then keep temperature at 250  C for 6 min (see Note 19).

Quantification of Unlabeled Phosphoinositides from Plants

3.8

Data Analysis

387

1. Identify FAME peaks according to retention times of FAME standards prepared from authentic fatty acid mixtures, such as menhaden oil, or by mass spectrometry (see Note 20). 2. Integrate FAME peaks for each lipid sample and calculate molar amounts in relation to the internal standard (see Note 21). 3. The molar amount of lipid is equivalent to the molar sum of all FAMEs present in the sample for lysolipids. The sum of all fatty acids divided by 2 represents the molar amount of the phospholipids. The sum of all fatty acids divided by 3 is equivalent to the molar amount of triacylglycerols.

4

Notes 1. Phosphoinositides are extracted under acidic conditions. Protonation of phosphoinositides is required for successful extraction because no quantitative extraction is achieved with neutral or alkaline protocols. 2. Extraction of cultured cells is aided by first grinding the extraction mixture in a 50 mL glass Dounce homogenizer on ice with 20 slow movements. 3. Dried lipids can be stored over night at 20  C in glass tubes under an inert atmosphere of N2 or argon gas. 4. It is important not to overload the SPE column. Choose the column size according to the amount of plant material. For lipids extracted from up to 10 mg (fresh weight) of material use 50 mg SPE columns. For lipids extracted from larger amounts of plant material use 100 mg SPE columns. It may be necessary to distribute the plant extract to several SPE columns. 5. Use another rack placed on top of the glass tube rack to position your SPE columns to avoid direct contact between the SPE column and the solvent within the glass tube. 6. If you want to quantify the lipid amounts, add another 500 μL of chloroform to the empty glass tube of your extract and load it on the same column to make sure that no residual lipids are left in the glass tube. 7. The elution of chlorophyll from the column is a useful indicator for a successful elution of neutral lipids. 8. The successful glycolipid fractionation will be indicated by the coelution of yellow pigments from the column. 9. Keep the lower phase collected at first on ice while separating the other one.

388

Larissa Launhardt et al.

10. Activation of TLC plates is not strictly necessary, but improves resolution. 11. For improved resolution, use a spotting robot if available. 12. Upon development, dried TLC plates can be redeveloped in the same or a different developing solvent to further improve resolution if needed. 13. Standard lipids should be run on the very same—not on a separate—plate to ensure comigration with lipids from biological material. 14. Submerge the plate steadily in the CuSO4 solution to avoid tide marks. An automated immersion device can aid in uniform dipping. 15. Lipid standards as low as 1 μg can be visualized using the CuSO4 solution. If no heating plate is available, plates can also be heated more slowly in an oven. 16. To verify even migration in all lipid lanes, the scraped remainder of the plate can be stained by submerging in CuSO4 solution and subsequent heating. 17. The silica material must not be inhaled. Therefore, make sure to wear protective gear covering your mouth and nostrils. By using sheets of folded cardboard, the TLC powder can easily be collected and transferred to glass tubes for extraction. 18. The use of glass ware is highly recommended when working with lipids and organic solvents. While glass ware is not so critical for the steps prior to TLC separation of lipids, all steps after TLC should be strictly carried out in glass ware. For lipid extraction ensure that screw caps contain absolutely no rubber septa but Teflon septa. Otherwise, GC chromatograms will be encumbered by the presence of artefact peaks contaminating the biological samples. 19. The chromatography column and temperature gradient provided here are only an example for a method that has worked reliably over the past years. There are numerous published variations for other columns and temperature gradients that can be applied. 20. When using a mass detector for FAME analysis, reference spectra corresponding to many fatty acids of biological relevance can be found at the website of the American Oil Chemists’ Society (AOCS; http://lipidlibrary.aocs.org/). 21. When calculating lipid concentrations according to the internal standard, be aware that different standards may contain different stoichiometric amounts of fatty acids, for example 3 in triacylglycerols vs. 2 in glycerophospholipids.

Quantification of Unlabeled Phosphoinositides from Plants

389

Acknowledgments We gratefully acknowledge financial support for L. Launhardt through the International Graduate Schools in Agricultural and Polymer Sciences AgriPoly (“Plant Oil Deposition” to M.H. and I.H.) funded by the European Social Fund (ESF), and for M. Matzner by the German Research Foundation (DFG, grant 400681449/GRK2498 TP01 to I.H.). References 1. Palmer S, Hawkins PT, Michell RH, Kirk CJ (1986) The labelling of polyphosphoinositides with [32P]Pi and the accumulation of inositol phosphates in vasopressin-stimulated hepatocytes. Biochem J 238:491–499. https://doi. org/10.1042/bj2380491 2. Gerth K, Lin F, Menzel W, Krishnamoorthy P, Stenzel I, Helmann M, Helmann I (2017) Guilt by association: a phenotype-based view of the plant phosphoinositide network. Annu Rev Plant Biol 68:349–374. https://doi.org/10. 1146/annurev-arplant-042916-041022 3. Heilmann I (2016) Phosphoinositide signaling in plant development. Development 143:2044–2055. https://doi.org/10.1242/ dev.136432 4. Heilmann M, Heilmann I (2013) Mass measurement of polyphosphoinositides by thin-layer and gas chromatography. Methods Mol Biol 1009:25–32. https://doi.org/10.1007/978-162703-401-2_3 5. Ko¨nig S, Hoffmann M, Mosblech A, Heilmann I (2008) Determination of content and fatty acid composition of unlabeled phosphoinositide species by thin layer chromatography and gas chromatography. Anal Biochem 378:197–201. https://doi.org/10.1016/j.ab.2008.03.052

6. Walsh JP, Caldwell KK, Majerus PW (1991) Formation of phosphatidylinositol 3-phosphate by isomerization from phosphatidylinositol 4-phosphate. Proc Natl Acad Sci U S A 88:9184–9187. https://doi.org/10.1073/ pnas.88.20.9184 7. Perera IY, Davis AJ, Galanopoulou D, Im YJ, Boss WF (2005) Characterization and comparative analysis of Arabidopsis phosphatidylinositol phosphate 5-kinase 10 reveals differences in Arabidopsis and human phosphatidylinositol phosphate kinases. FEBS Lett 579:3427–3432. https://doi.org/10.1016/j.febslet.2005.05. 018 8. Cho MH, Chen Q, Okpodu CM, Boss WF (1992) Separation and quantification of [3H] inositol phospholipids using thin-layer-chromatography and a computerized 3H imaging scanner. LC-GC 10:464–468 9. Hornung E, Korfei M, Pernstich C, Struss A, Kindl H, Fulda M, Feussner I (2005) Specific formation of arachidonic acid and eicosapentaenoic acid by a front-end Δ5-desaturase from Phytophthora megasperma. Biochim Biophys Acta 1686(3):181–189. https://doi.org/10. 1016/j.bbalip.2004.11.001

Chapter 22 Studying Lipid–Protein Interactions Using Protein–Lipid Overlay and Protein–Liposome Association Assays Guido Ufer, Peter Do¨rmann, and Dorothea Bartels Abstract The study of lipid–protein interactions is crucial for understanding reactions of proteins involved in lipid metabolism, lipid transport, and lipid signaling. Different detection methods can be employed for the identification of lipid-binding interactions. Isothermal titration calorimetry (ITC) and surface plasmon resonance (SPR) spectroscopy enable real-time monitoring of lipid protein interactions and provide thermodynamic parameters of the interacting partners. However, these technologies depend on the availability of the large equipment, limiting the practicability in many laboratories. Protein–lipid overlay assays are a simple first approach to screen for protein interactions with different lipids or lipid intermediates and are independent of large equipment. Subsequently, specific interactions can be analyzed in detail using protein–liposome association assays. Key words Lipid–protein interaction, Membrane lipid strips, Liposomes

1

Introduction Lipids establish the bilayer membranes of the cell and serve as storage compounds in cytosolic lipid droplets and chloroplastic plastoglobules. Lipids are able to recruit proteins transiently and thus affect conformation and activity of intracellular proteins and metabolites. Because lipids are not soluble in aqueous environments, they are often bound in a noncovalent manner to proteins like acyl-CoA binding proteins (ACBP), lipid transfer proteins (LTP, Sec14 proteins) and others. Different enzymes involved in lipid metabolism bind lipid intermediates during catalysis. Membrane localized transporters (e.g., ATP binding cassette (ABC) transporters) and receptors are involved in lipid binding. Lipids can act as important signaling molecules, and lipid-based protein interactions are among the primary events in signaling cascades [1]. Detection of lipid binding using in vitro techniques is based on the availability of purified fractions of the protein of interest and of pure lipid ligands. The protein is commonly produced by

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_22, © Springer Science+Business Media, LLC, part of Springer Nature 2021

391

392

Guido Ufer et al.

recombinant expression in a heterologous host such as Escherichia coli. For large proteins or proteins with extensive hydrophobic sequences, it is often of advantage to limit the expression to shorter peptides. Using this strategy, it is possible to map the lipid-binding domain to the protein sequence. Protein purification is often performed by taking advantage of fusion tags, for example during the purification of His6-tagged proteins via nickel-NTA affinity chromatography. Different methods for the analysis of lipid–protein interactions have been developed. The binding of a protein to ligands is accompanied with the exchange of energy in the form of heat. Isothermal titration calorimetry (ITC) takes advantage of this effect, as the energy exchange is recorded giving rise to data in the form of a binding curve. Thus, ITC monitors real time lipid protein interaction and provides thermodynamic results, which are used to calculate kinetic binding parameters (Chapter 23). Surface plasmon resonance (SPR) spectroscopy is based on changes in light reflection during the interaction of a protein with its ligand, which takes place on the metal surface of a chip. In principle, either the protein or the ligand can be immobilized on the chip, with the other binding partner being dissolved in buffer, which floats over the surface with a constant flow rate [2, 3]. Similar to ITC, SPR spectroscopy provides quantitative data on binding constants and thermodynamic parameters. While ITC analysis and SPR spectroscopy provide quantitative data on binding parameters, these technologies require the availability of expensive equipment and consumables. Alternative methods were developed that are based on the physical separation of protein–lipid complexes from nonbound proteins or lipids. For example, after performing the binding assay in solution, free fatty acids can be separated from protein-bound fatty acids using Lipidex 1000, a polymer of hydroxypropyl dextran substituted with ~10% by long-chain (~C15) alkyl ethers. Lipidex 1000 selectively binds hydrophobic molecules like fatty acids and other lipids [4]. Two protein–lipid binding assays are presented here, the protein–lipid overlay assay which can be used to screen for the capacity of a protein to bind to a variety of lipids immobilized on a nitrocellulose membrane (Fig. 1), and the protein–liposome association assay employed to confirm binding of the protein to a specific lipid (Fig. 2) [5–7]. For a protein–lipid overlay assay, a range of lipid classes and/or molecular species are first immobilized on a nitrocellulose membrane. Different lipids, including membrane lipids (phospholipids, galactolipids, sphingolipids, sterol lipids), storage lipids (triacylglycerol), signaling lipids (phosphatidic acid, phosphoinositides), or lipid intermediates (fatty acids, diacylglycerol), can be spotted onto the nitrocellulose membrane. Then free binding sites of the membrane are blocked with BSA (bovine serum albumin), and the membrane is “overlaid” with a

Lipid-Protein Interactions

393

a)

b)

c) 16:0 (DAG16) 18:3 (DAG18)

CL

PE

PG

DGDG

PA

PC

MGDG

Fig. 1 Protein–lipid overlay assay. Lipids were spotted on a nitrocellulose membrane which was overlaid with the recombinant PLDrp1 protein [9]. (a) The full-size PLDrp1 protein was used for the overlay, and the protein was detected with anti-PLDrp1 antiserum. (b) The membrane was incubated with the PLDrp1 protein carrying a Nus-tag and a His-tag and detected with antiHis antiserum. Protein binding was visualized using the horseradish peroxidase (HRP) assay and chemiluminescence. (c) Scheme of lipids loaded on the nitrocellulose membrane. 16:0 (DAG), dipalmitoylglycerol; 18:3 (DAG18), dilinolenoylglycerol; PA, phosphatidic acid; PC, phosphatidylcholine; CL, cardiolipin; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; DGDG, digalactosyldiacylglycerol; MGDG, monogalactosyldiacylglycerol

buffer containing the protein of interest. The protein binding to lipids on the membrane is detected via immunochemistry, using antibodies directed against the protein of interest or a protein fusion tag. Results obtained from the protein–lipid overlay assay are often confirmed by protein binding to lipids, which have been integrated into liposomes. Liposomes are produced from an inert

394

Guido Ufer et al.

PLDrp1 1

2

Liposomes (PA+PC)

PLDrp1 + PA

3

4

5

S

P

6

Full-size-PLDrp1

Nt-PLDrp1

Ct-PLDrp1 S

P

S

P

Fig. 2 Protein–lipid association assay. Liposomes were prepared from PC (phosphatidylcholine) and PA (phosphatidic acid) and binding of PA to the full-length recombinant protein PLDrp1, to the N-terminal peptide and to the C-terminal peptide [9]. Purified PLDrp1 proteins (without liposomes; lanes 1, 2) and pure liposomes containing PA and PC (without protein, lanes 3, 4) were used as binding controls. Lanes 5 and 6 show the binding of PLDrp1 protein to liposomes containing PC and PA. Liposome-associated protein and free protein were separated into pellet (P) and supernatant (S) by centrifugation and subsequently separated on a polyacrylamide gel. The figure shows the protein gel stained with Coomassie R250.The figure was adapted from [9]

bilayer-forming lipid (e.g., phosphatidylcholine, PC) with different concentrations of the lipid of interest. After incubation of the protein with liposomes, liposome-bound and free proteins are separated by centrifugation. Then, the proteins in the two fractions are separated by SDS polyacrylamide gel electrophoresis (SDS-PAGE), and the proteins in the bound and nonbound fractions are detected by Coomassie staining of the gel (in case that protein abundance is high), or by Western blotting (with low amounts of protein). The combination of protein–lipid overlay and protein–liposome association assays represents a robust technology to screen proteins for their lipid-binding capacities. This strategy is well affordable and can be performed in any laboratory setting.

2

Materials

2.1 Membrane Lipid Strips and Protein– Lipid Overlay Assay

1. Lipids for binding assays can be selected from phospholipids (phosphatidylcholine, PC, phosphatidylethanolamine, PE; phosphatidylglycerol, PG; phosphatidylserine, PS; phosphatidic acid, PA), galactolipids (monogalactosyldiacylglycerol, MGDG; digalactosyldiacylglycerol, DGDG), sulfolipid (sulfoquinovosyldiacylglycerol, SQDG), storage lipid

Lipid-Protein Interactions

395

(triacylglycerol, TAG), lipid intermediates (fatty acids; diacylglycerol, DAG; acyl-CoAs), phytosterols (stigmasterol, sitosterol, campesterol). Natural lipids or lipids with defined molecular species composition can be obtained from different commercial suppliers. 2. Glass vials with screw caps (lined with Teflon septa) and glass test tubes. 3. Sample concentrator for drying lipids with N2 gas. 4. Nitrocellulose membrane (e.g., Amersham Protran 0.2 NC, GE Healthcare). 5. Spotting buffer for membrane lipid strips: 250 μL chloroform, 500 μL methanol, 200 μL 50 mM HCl, 2 μL aqueous 1% (w/v) Ponceau S in 3% (w/v) trichloroacetic acid (TCA) solution. 6. Alternative option: Ready-to-use membrane lipid strips (Echelon Bioscience, Salt Lake City, Utah, USA) (see Note 1). 7. Protein of interest (e.g., purified recombinant protein or purified native protein). 8. TBST buffer for immunodetection: 20 mM Tris–HCl, pH 8, 0.05% (v/v) Tween 20, 150 mM NaCl. 9. Blocking solution: 2% (w/v) milk powder or 1% (w/v) bovine serum albumin (BSA). 10. Primary antibody raised against protein of interest, or antiHis6-tag antibody. 11. Secondary antibody coupled to alkaline phosphatase (AP) or horseradish peroxidase (HRP) and the corresponding detection system. 2.2 Preparation of Liposomes and Protein–Liposome Association Assay

1. Lipids for binding assay: phosphatidylcholine (PC) as inert matrix lipid, and specific lipid (e.g., phosphatidic acid, PA). 2. Protein of interest (e.g., purified recombinant protein). 3. TBS buffer: 50 mM Tris–HCl, pH 7, 100 mM NaCl. 4. Laemmli sample buffer: 100 mM Tris, 2% (w/v) SDS, 10% (v/v) glycerol, 0.005% (w/v) bromophenol blue, 1 mM EDTA [8]. 5. Equipment for SDS protein gel electrophoresis, followed by Coomassie R250 staining or Western blotting.

3

Methods Home-made custom strips can easily be produced by spotting commercially available lipids on nitrocellulose membranes (see

396

Guido Ufer et al.

3.1 Preparation of Membrane Lipid Strips for Protein– Lipid Overlay Assays

Note 1). It is economical to produce a considerable number of membrane lipid strips, which can be stored for later use. 1. Dissolve the lipid of interest in chloroform–methanol (2:1, v/v) at a concentration of ~10 μM. 2. Take an aliquot of the lipid and quantify the exact concentration via gas chromatography with flame ionization detector (GC-FID) after transmethylation in the presence of an internal standard (e.g., pentadecanoic acid, 15:0; see Chapters 1 and 4) (see Note 2). Adjust concentration of lipids to exactly 10 μM by adding more solvent, or by drying the lipid with N2 gas and dissolving it in a smaller volume. Lipids can be stored at 20  C for several months. 3. Transfer 50 μL (500 nmol) of the lipid into a new glass vial, remove solvent with N2 gas in a sample concentrator, and dissolve it in 50 μL spotting buffer. 4. Spot 1 μL (10 nmol) of the lipid on the nitrocellulose membrane (see Note 3). The same procedure is performed with the other lipids. Make sure that you keep the same spotting pattern on the membrane lipid strips and make a note in your lab book to record the identity of the different spots. Also spot 1 μL of lipid-free spotting buffer as negative control. 5. Wait until all solvent has evaporated. Membrane lipid strips can directly be used for protein–lipid overlay assays (Subheading 3.2) or can be wrapped in aluminum foil and stored at a dry place for several months.

3.2 Protein–Lipid Overlay Assay

The protein–lipid overlay assay follows a protocol published in [6, 7]. In this procedure, the protein of interest dissolved in the aqueous buffer binds to a lipid film immobilized on a solid support, that is, a nitrocellulose membrane (Fig. 1) (see Note 4). 1. Place the membrane lipid strips in a container with blocking buffer (TBST buffer with 1% (w/v) BSA or 2% (w/v) milk powder). Shake on a rotary shaker for 2 h. Alternatively, the strips can be incubated in blocking buffer overnight at 4  C. 2. Add the purified protein (final concentration 1μg/mL) and incubate overnight at 4  C with shaking (see Note 5). 3. Wash three times in TBST buffer (10 min each). 4. The protein bound to lipids on the strip is visualized by immunodetection. For this, the strip is incubated in blocking buffer containing the primary antibody (protein-specific antibody or anti-His6-tag antibody at an appropriate concentration) for 1–2 h at room temperature. 5. Wash the strip three times in TBST buffer (10 min each).

Lipid-Protein Interactions

397

6. Optional: Incubate the strip again in blocking solution (see step 1) to saturate nonspecific binding sites. 7. Incubate the strip in TBST buffer containing the secondary antibody linked to alkaline phosphatase (AP) or horseradish peroxidase (HRP) for 1–2 h at room temperature. Use an antibody concentration according to manufacturer’s recommendation (see Note 6). 8. Wash the strip three times in TBST buffer (10 min each). 9. Detect protein either via the color reaction with 5-bromo-4chloro-3-indolyl phosphate (BCIP) and nitro blue tetrazolium (NBT) for AP, or by reaction of luminol and H2O2 giving rise to chemiluminescence which can be recorded using a CCD camera or on X-ray films (HRP). 3.3 Preparation of Liposomes

Large multilamellar liposomes are produced by vortexing [5–7]. In most experiments, liposomes are produced containing an inert lipid (PC) and the lipid of interest (e.g., phosphatidic acid, PA) with different concentrations, for example 0%, 25%, 50%, and 75% PA. Therefore, it is important to determine the concentrations of the lipids accurately prior to liposome production, for example, via GC-FID of fatty acid methyl esters (see Subheading 3.1, step 2). 1. Add lipids (total 250μg of lipids per assay; e.g., 150μg of PC plus 100μg of PA for a ratio of 3:2) in a glass tube. Evaporate the solvent under nitrogen gas. 2. Add 500 μL TBS buffer to the lipid film to obtain a final concentration of 0.5μg lipids/μL TBS. 3. Incubate lipids in a 30–37  C water bath for 1 h for hydration and emulsification. 4. Vortex at maximal speed for 5 min. The solution should turn cloudy. Do not use a sonicator. 5. Transfer the lipid emulsion to a microcentrifuge tube. Spin at 13,000  g for 10 min at 4  C. 6. Remove supernatant. Add ice-cold TBS buffer to a final concentration of ~0.5μg lipids/μL to the liposome pellet. Resuspend the pellet with a pipette. Centrifuge again. This washing step serves to remove lipid structures that cannot be pelleted at 13,000  g. 7. Remove supernatant, then resuspend liposomes to 4μg/μL in ice-cold TBS buffer. Use 50 μL of liposome emulsion (200μg lipids) per assay.

3.4 Protein– Liposome Association Assay

The protein–liposome association assay is carried out according to the protocol of [5, 6] (Fig. 2).

398

Guido Ufer et al.

1. Adjust purified protein to a concentration of 1μg in 10 μL of ice-cold TBS buffer. 2. Mix 50 μL liposomes (4μg/μl of lipid) with 50 μL protein solution (5μg of protein) to a final volume of 100 μL in a microfuge tube. 3. Incubate the mixture in a water bath at 30  C (or the temperature of choice for the binding experiment) for 30 min with gentle shaking (~100 rpm). 4. Centrifuge down liposomes at 13,000  g and 4  C for 10 min. 5. Remove supernatant, which can be used to detect nonbound protein. 6. Resuspend liposome pellet in 400 μL of ice-cold TBS buffer and vortex briefly. 7. Centrifuge again. Repeat steps 5 and 6, and spin again. 8. Remove and discard supernatant. Resuspend liposome pellet in 10–20 μL of 1 Laemmli sample buffer [8]. 9. Protein samples can be stored at 20  C or can be directly analyzed by SDS-PAGE. The protein is detected by staining the gel with Coomassie R250, silver nitrate or by Western blotting.

4

Notes 1. Membrane lipid strips for protein–lipid overlay assays are commercially available from Echelon Biosciences (Salt Lake City, Utah, USA). However, these strips are expensive, and they are targeted at studying binding characteristics of animal and human proteins. Therefore, no plant-specific lipids (i.e., galactolipids, sulfolipid, phytosterols, etc.) are found on these strips. 2. Acyl lipids (phospholipids, galactolipids, triacylglycerol, acylCoA, etc.; see Chapters 1 and 4) can be measured by GC-FID after transmethylation. Sterol lipids can be quantified by gas chromatography–mass spectrometry (GC-MS) after hydrolyzation and trimethylsilylation (Chapter 11). For sphingolipid quantification, see Chapter 10. 3. Instead of using the same concentration, a series of different amounts of one lipid can be spotted to study concentrationdependent binding of the protein. 4. Protein binding to lipids immobilized on solid nitrocellulose membranes does not reflect the natural mode of protein–lipid binding, because in the cellular environment proteins associate either with soluble lipids or lipids integrated in membranes or lipid droplets. Therefore, the protein–lipid association study (Subheading 3.4) where proteins bind to lipids in a bilayer

Lipid-Protein Interactions

399

membrane, might better reflect the natural binding characteristics of the proteins. 5. The protein of interest should be pure and should be free from contaminations like imidazole. Therefore, protein solutions obtained after nickel-NTA affinity chromatography might need to be dialyzed. The TBST buffer contains 0.05% Tween20 that should help to keep the protein of interest in solution. 6. In some commercial detection systems, the primary antibody is already coupled to the AP or HRP enzyme. In addition, an antibody independent His6-tag detection system based on a nickel–NTA conjugate is available (Kirkegaard & Perry Laboratories, Gaithersburg, MD, USA). In these cases, incubation with the secondary antibody is not necessary.

Acknowledgments The authors thank Rebecca L. Roston (University of Nebraska, Lincoln, NE, USA) and Christoph Benning (Michigan State University, East Lansing, MI, USA) for their help with establishing lipid–protein interaction assays. References 1. Hou Q, Ufer G, Bartels D (2016) Lipid signalling in plant responses to abiotic stress. Plant Cell Environ 39:1029–1048. https://doi.org/ 10.1111/pce.12666 2. Gopinath SCB (2010) Biosensing applications of surface plasmon resonance-based Biacore technology. Sens Actuat B 150:722–733. https://doi.org/10.1016/j.snb.2010.08.014 3. Patching SG (2014) Surface plasmon resonance spectroscopy for characterisation of membrane protein–ligand interactions and its potential for drug discovery. Biochim Biophys Acta 1838:43–55. https://doi.org/10.1016/j. bbamem.2013.04.028 4. Vork MM, Glatz JFC, Surtel DAM, van der Vusse GJ (1990) Assay of the binding of fatty acids by proteins: evaluation of the Lipidex 1000 procedure. Mol Cell Biochem 98:111–117. https://doi.org/10.1007/BF00231374 5. Sano H, Kuroki Y, Honma T, Ogasawara Y, Sohma H, Voelker DR, Akino T (1998) Analysis of chimeric proteins identifies the regions in the carbohydrate recognition domains of rat lung collectins that are essential for interactions with phospholipids, glycolipids, and alveolar type II

cells. J Biol Chem 273:4883–4789. https://doi. org/10.1074/jbc.273.8.4783 6. Awai K, Xu C, Tamot B, Benning C (2006) A phosphatidic acid-binding protein of the chloroplast inner envelope membrane involved in lipid trafficking. Proc Natl Acad Sci U S A 103:10817–10822. https://doi.org/10.1073/ pnas.0602754103 7. Wang Z, Xu C, Benning C (2012) TGD4 involved in endoplasmic reticulum-to-chloroplast lipid trafficking is a phosphatidic acid binding protein. Plant J 70:614–623. https://doi. org/10.1111/j.1365-313X.2012.04900.x 8. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685. https://doi. org/10.1038/227680a0 9. Ufer G, Gertzmann A, Gasulla VF, Ro¨hrig H, Bartels D (2017) Identification and characterization of the phosphatidic acid binding A. thaliana phosphoprotein PLDrp1 which is regulated by PLD alpha1 in a stress dependent manner. Plant J 92:276–290. https://doi.org/ 10.1111/tpj.13651

Chapter 23 Investigations of Lipid Binding to Acyl-CoA-Binding Proteins (ACBP) Using Isothermal Titration Calorimetry (ITC) Ze-Hua Guo and Mee-Len Chye Abstract Isothermal titration calorimetry (ITC) is a quantitative, biophysical method to investigate intermolecular binding between biomolecules by directly measuring the heat exchange in the binding reaction. The assay is carried out in solution when the molecules interact in vitro. This allows to determine values for binding affinity (Kd), binding stoichiometry (n), as well as changes in Gibbs free energy (ΔG), entropy (ΔS), and enthalpy (ΔH ). This method also addresses the kinetics of enzymatic reactions for a substrate during conversion to a product. ITC has been used to study the interactions between proteins and ligands such as those of acyl-CoA-binding proteins (ACBPs) and acyl-CoA thioesters or ACBPs with protein partners. ITC has also been used in investigating interactions between antiserum and antigen, as well as those involving RNA and DNA and other macromolecules. We describe the methods used to isolate and purify a recombinant rice ACBP (OsACBP) for ITC. To study OsACBP binding to long-chain acyl-CoA thioesters, a microcalorimeter was used at 30
C, and the ligand (acyl-CoA thioesters or a protein partner in the first cell), was mixed with the ACBP protein solution in a second cell, for more than 40 min comprising 20 injections. Subsequently, the binding parameters including the heat-release data were analyzed and various thermodynamic parameters were calculated. Key words Acyl-CoA thioesters, His-tagged proteins, Liquid chromatography, Oryza sativa, Protein–ligand interaction, Protein–protein interaction, Protein purification

1

Introduction Isothermal titration calorimetry (ITC) represents a method to study the interaction parameters between two molecules (protein, ligand, etc.) by measuring heat changes arising from binding [1]. ITC records the energetics associated with reactions or processes occurring at a constant temperature. Therefore, it is the method of choice for investigating biochemical reactions and molecular interactions [2, 3]. ITC provides information on the thermodynamics of enzyme-catalyzed reactions, ligand binding between macromolecules, and ligand- or pH-induced macromolecular conformational changes [2]. Starting from the 1970s, ITC has been used to study nucleotide binding to proteins, antibody to

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_23, © Springer Science+Business Media, LLC, part of Springer Nature 2021

401

402

Ze-Hua Guo and Mee-Len Chye

antigen, ligand binding to human hemoglobin, peptide–lipid association, and toxin binding to proteins [2]. Recently, ITC experiments have been used to study the affinity between acyl-CoA thioesters and acyl-CoA-binding proteins (ACBPs) from rice [4] as well as Arabidopsis thaliana [5–7]. This method has been extended to study the protein-protein interactions between an ACBP and its protein partner [8, 9]. Although other binding assays such as Lipidex assays [10] and filter-binding assays [11] have been used to study binding involving ACBPs [12– 19], ITC provides more information and it is more reliable because the thermodynamics of binding parameters are documented [5–9]. Using curve-fitting parameters and an appropriate binding site mode, the thermodynamic parameters, such as changes in Gibbs free energy (ΔG), enthalpy (ΔH) and entropy (ΔS), can be calculated from the ITC results [20]. The ΔG value represents a measure for the affinity of binding between the ligand and the protein [20]. The ΔH value provides the heat energy changes while a complex is formed at a constant temperature [20]. The equilibrium binding constant K can be determined from the amount of free or bound ligand at any point during the titration [20]. Given that ITC requires very pure recombinant protein, the steps prior to ITC analysis such as recombinant protein isolation and purification are important and have been included herein.

2

Materials Prepare all solutions using ultrapure deionized water and analytical grade reagents. All solutions should be autoclaved and then kept at room temperature (unless indicated otherwise). For storage, 0.02% (w/v) sodium azide is added to the solutions.

2.1 Growth of Transformed Escherichia coli Cells Expressing the Recombinant Protein

1. E. coli cells harboring the expression vector with the protein of interest, for example, E. coli BL21 (DE3) Star pLysS cells expressing OsACBP2. 2. Plasmid pOS503 [4], a pRSET A (Invitrogen) [21] derivative comprising the OsACBP2 cDNA under control of the T7 promoter. 3. Luria Broth (LB broth). 4. Ampicillin (AMP) solution: 100 mg/ml for 1000 AMP stock. 5. Chloramphenicol (CAM): 34 mg/ml for 1000 CAM stock. 6. Isopropyl β-D-1-thiogalactopyranoside (IPTG): 1 M stock solution (1000). 7. Sonicator.

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC

2.2 Protein Extraction and Purification

403

1. 0.22 μm polyethersulfone (PES) membrane syringe filter (Millipore). 2. HisTrap HP Column (see Note 1) with Ni2+ loaded on a nitrilotriacetic acid (NTA) matrix from GE Healthcare. 3. Bio-Rad DuoFlow liquid chromatography (LC) purification system (see Note 2). 4. NTA Buffer A (NTA loading and washing buffer): 20 mM monosodium phosphate (NaH2PO4) pH 7.5; 300 mM sodium chloride (NaCl); 50 mM imidazole. 5. NTA Buffer B (NTA elution buffer): 20 mM NaH2PO4 pH 7.5; 300 mM NaCl; 300 mM imidazole. 6. Stripping Buffer (see Note 3): 1 M NaCl in 20 mM Tris–HCl, 50 mM EDTA pH 8.0. 7. Charging Buffer (see Note 4): 0.1 M NiCl2. 8. Amicon® Ultra-15 Centrifugal Filters Ultracel®—10 kDa. 9. Centrifuges: high speed centrifuge, and bench top centrifuge for Eppendorf tubes. 10. Biologic Duoflow software.

2.3 Isothermal Titration Calorimetry (ITC)

1. MicroCal iTC200 System from Malvern Panalytical Ltd. 2. ITC Assay Buffer: 20 mM sodium phosphate buffer (NaH2PO4), pH 7.0, 100 mM NaCl. 3. Acyl-CoA thioesters (e.g., from Sigma): 1 mM stock solution in water, for example, oleoyl-CoA (C18:1-CoA). 4. ITC200 software (Malvern Panalytical Ltd.). 5. Origin 7.0 software (OriginLab).

3

Methods All procedures are carried out at room temperature, unless otherwise specified.

3.1 Overexpression of His-Tagged Protein in E. coli and Protein Isolation

1. Transform E. coli BL21 (DE3) Star pLysS cells with plasmid pOS503 [4], derived from the pRSET A vector (see Note 5). Plasmid pOS503 harbors the OsACBP2 cDNA which encodes an N-terminally His-tagged OsACBP2 protein of 14.3 kDa [4]. Alternatively, use other E. coli cells expressing the protein of interest. 2. Inoculate 50 ml LB medium containing AMP and CAM in a flask with E. coli cells from a single colony (grown on LB agar supplemented with AMP and CAM) and grow the culture overnight at 37
C. The following day, subculture by adding 30 ml of the overnight culture to 3 l LB medium with AMP and

404

Ze-Hua Guo and Mee-Len Chye

CAM, and incubate at 37
C with shaking using an orbital shaker (see Note 6). 3. When the optical density at 600 nm (OD600) reaches 0.6, induce the expression of the recombinant protein by adding IPTG to a final concentration of 0.1 mM and culture the bacteria for 4 h (see Note 7). 4. Harvest the cells by centrifugation at 4000  g for 15 min at 4
C and resuspend the pellet in NTA Buffer A (see Note 8). 5. To homogenize the bacterial cells, transfer ~10 ml cell suspension into a 50-ml tube kept on ice, and insert the probe of a sonicator into the cell suspension. Close the sonication chamber, and lyse the cells with ultrasonic energy at 15 s on/15 s off for ten cycles on ice. Repeat twice (see Note 9). 6. Recover the supernatant after centrifugation at 20,000  g for 50 min, then pass the supernatant through a 0.22 μm polyethersulfone (PES) membrane syringe filter (Millipore). The resultant supernatant represents the total protein extract. 3.2 Purification of His-Tagged Protein

1. Purification is carried out using a Bio-Rad BioLogic DuoFlow LC purification system (Fig. 1a). Insert tubing A (connected to Pump A) to the bottle of Buffer A and tubing B to the bottle of Buffer B (connected to Pump B). Any LC system compatible with the HisTrap HP column can be used in this experiment instead. 2. Before connecting the HisTrap HP column to the LC system, wash it with ~50 ml water at 10 ml/min. Subsequently, adjust the flow rate to 3 ml/min using the Biologic Duoflow software (Fig. 1b) and connect the column using a “drop-to-drop” technique (see Note 10). 3. Equilibrate the system including the HisTrap HP column using ~20 ml Buffer A at 3 ml/min, until the A280 readout of the UV detector is stable. 4. Through tubing A, load the protein extract, followed by ~50 ml Buffer A. Using the software program on the computer connected to the DuoFlow system, the A280 readout will show a large peak (Fig. 2a), which represents background proteins without His-tags. 5. When the A280 peak for the background proteins tapers off, switch the valve to tubing B, which contains Buffer B with 300 mM imidazole, to elute the His-tagged proteins. Meanwhile, collect the eluate fractions in separate vials (see Note 11). 6. Examine the purified protein on a sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS-PAGE) system (Fig. 2b). ITC experiments require pure proteins that show a single band on the SDS gel (see Note 12).

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC

405

Fig. 1 The Bio-Rad DuoFlow system. (a) The Bio-Rad DuoFlow apparatus. Pumps A and B are labeled “A” and “B.” A tubing from each pump is connected to a

406

Ze-Hua Guo and Mee-Len Chye

7. Use an Ultracel®-10 kDa filter to exchange the Buffer B used to elute the protein from the HisTrap column with ITC Assay Buffer (see Note 13). 3.3

1. The ITC experiments are carried out using the MicroCal iTC200 system (Fig. 3a). Dilute the protein to ~20 μM and the ligand (acyl-CoA thioester) to ~200 μM (see Note 14). Degas and prewarm the protein and ligand solution to 30
C (see Note 15).

ITC

2. Launch the “ITC200” software program (Fig. 3b) on the computer. Click the “Instrument Controls” tab, and wash the ITC system using the “Cell and Syringe Wash” program in the “Accessory Station” panel of the ITC software. Then use a syringe to rinse the titration cell twice with ITC Assay Buffer, and finally load ~200 μl protein into the titration cell. a mAu

UV

Cond

pH

Pressure

Temp

b

Conc

A280

2000 1800 1600 1400 1200 1000 800 600 400 200 0 -200 -400 -600

Zoomed Curves

kD 250 150 100 75 50 37

M

1

25 20 4

3

2

1

15

14.3kD ml

100

120

140

160

180

200

220

10

Flow-through volume

Fig. 2 Purification of His-tagged OsACBP2 protein. (a) Purification of His-tagged OsACBP2 using a HisTrap HP column. The blue plot represents the real-time absorption at 280 nm (A280) of the protein solution passing through the detector, and the red plot indicates the relative salt concentration (axis not shown). Arrow, purified His-tagged OsACBP2 protein. mAu, milli-absorption unit. (b) Purified (5g) His-tagged OsACBP2 protein was separated on a 15% SDS-PAGE gel (lane 1). Precision Plus Protein™ Dual Color Standards (Bio-Rad) loaded in lane M. The size of the His-tagged OsACBP2 protein at 14.3 kDa is indicated with an arrowhead

ä Fig. 1 (continued) bottle of buffer, marked “Buffer A” and “Buffer B.” The red arrow indicates the HisTrap HP column (GE Healthcare) connected to the system. (b) The BioLogic DuoFlow software. The panels for controlling the gradient pump and the fraction collector are shown on the computer screen in red and yellow boxes, respectively, together with the real-time absorption reading at 280 nm (A280) of the protein solution passing through the detector (below). The white area displays real-time plots of protein absorption (A280) depicted with a green line and salt concentrations with a red line

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC

407

Fig. 3 The MicroCal iTC200 system (GE Healthcare). (a) The ITC200 apparatus. The ligand is loaded to the titrating syringe, while the sample cell contains the purified protein. The sample cell can hold ~200l of protein solution, and the titrating syringe loads ~50l for each run. (b) The ITC200 software. Five tabs are provided in

408

Ze-Hua Guo and Mee-Len Chye

3. Load the ligand according to the instructions of the ITC200 software program. Set the temperature to 30
C in the “Thermostat Control” panel and wait until the temperature is stable. Meanwhile, fill in the concentration for the protein and ligand in the “Advanced Experimental Design” tab (Fig. 3c), and the data file name in the “Experimental Design” tab. 4. The titration contains twenty successive 1.8-μl injections over 4 s at 150-s intervals using a stirring speed of 1000 rpm. After one experiment is completed, an “itc” file will be generated for subsequent analysis. 5. Start the titration program by clicking the “Start” button, and wait until the program finishes (see Note 16). 3.4

ITC Data Analysis

1. Open the “Origin 7.0” software, load the “itc” file by clicking “Read Data” button (Fig. 3d). 2. Enter the “RawITC” panel by going to the menu bar: “Window)1 mRawITC”. Click “Adjust integrations” and then click on the black curve to enter the base line editing interface, after which click the “Baseline” button to reveal all the movable dots (Fig. 4). Drag the dots onto the black curve, which represents the actual baseline. When all adjustments are completed, click “Integrate” and then “Quit” to return to the “RawITC” panel. 3. Enter the “DeltaH” panel at the menu bar: “Window)6 DeltaH”. Estimate if the baseline is near 0 on the y-axis, which is indicated in kcal/mol of injectant. If not, click “Math)Simple math” on the menu bar for the “Math on/between Data Set” dialogue box (Fig. 5a). Type “+” or “” in the operator input box if the baseline is lower or higher than the 0 reading. In “Y2,” type the estimated difference (see Note 17). Click “OK” to return.

ä Fig. 3 (continued) the software. “Experimental Design” and “Advanced Experimental Design” contain settings that should be completed before an experiment. “Real Time Plot” and “Setup” are used for an ongoing experiment. At the “Instrument Controls” tab, various functions of the ITC machine can be adjusted when the program is not running. The temperature control panel in the Instrument Controls tab is boxed in red, and the buttons for syringe wash and fill are boxed in blue. (c) The “Experimental Parameters” panel in the “Advanced Experiment Design” tab. The concentrations can be set in the areas highlighted by the red box, and the file name can be filled in the area highlighted in blue. Furthermore, the number of injections, cell temperature and other parameters should be determined before start of the titrations. (d) The Origin 7 software. While the ITC200 software collects raw data, this software is designed for subsequent data analysis. The Origin 7 software can process the “itc” file and eventually generate a figure with a textbox containing thermodynamic parameters. To start the analysis, click “Read Data..” (boxed in red) to load the “itc” file

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC

409

Fig. 4 Adjusting the baseline in the Origin 7 software. Drag the black dots to best fit the data curve, as shown from the red-boxed to the black-boxed area. At each gap, follow the trend to draw a straight line, which then acts as the baseline for the peak. To see other areas of the graph, press the “” buttons to move the view

410

Ze-Hua Guo and Mee-Len Chye

Fig. 5 Curve-fitting by Origin 7. (a) y-Axis translation before curve fitting. In this DeltaH view, heat generated from each injection is calculated to display as a series of dots. The Origin 7 software will calculate a curve that best fits all these dots. The bracket indicates the difference between the plateau in the curve and zero. By adding “10,000” to “DATA1_NDH”, the whole graph will be translated upward by 10 kcal/mol, so that the software will recognize the plateau accurately. (b) Following the plot translation in (a), the data points are now correctly displayed for model fitting. Click “One Set of Sites” (boxed in red) to generate the curve and calculate the parameters (yellow box). The textbox only provides the n value, K, ΔH, and ΔS, and the user needs to work out the rest parameters

4. Click “One Set of Sites..” in “Model Fitting” for the “Curve fitting” dialog box. In the “Non Linear Curve Fitting” dialogue box, click “100 Iter” (iteration) button to generate a curve that best fits, and then click “Done.” Subsequently, a text box with all the thermostats will appear on the graph (Fig. 5b). Record the parameters in a table. 5. Click “ITC)Final ITC” to generate the final figure. The parameters are worked out by the following equations [22] (Fig. 6, see Note 18): Kd ¼ K1; ΔG ¼ ΔH  TΔS.

4

Notes 1. The volume of a mini column is defined as column volume (CV). 2. Any LC system compatible with the column can be used in this experiment. 3. Ni-NTA mini column loses affinity after several experiments. To restore its affinity to the His-tagged protein, apply 2–4 CVs of Stripping Buffer, and wash the column with 6–8 CVs of

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC

411

Fig. 6 Final ITC result of OsACBP2 binding to C18:1-CoA. The final ITC figure is generated with Origin 7 and presented in (a). The top panel displays the raw ITC data, while the bottom shows a graph for ΔH for each injection. Based on the preliminary results in the box in panel (a), the thermodynamic parameters are depicted in the table in (b). n stoichiometric ratio of ligand to protein, Kd dissociation constant; ΔH, ΔS and ΔG indicate changes in enthalpy, entropy, and Gibbs free energy, respectively

NTA Buffer A. The column will turn from cyan/green to white. For storage of the stripped column, fill the column with 20% (v/v) ethanol and keep it at room temperature or 4
C.

412

Ze-Hua Guo and Mee-Len Chye

4. To supply Ni2+ ions to the stripped column (see Note 3), apply 0.5 CV of Charging Buffer followed by 4 CVs of NTA Buffer A. The column will turn cyan again due to the presence of Ni2+ ions. This technique is known as “recharging” and the column can be reused afterward. 5. The cloning vector pRSET A (Invitrogen) has two more derivatives, pRSET B and pRSET C. The user should choose the appropriate version for cloning to facilitate in-frame expression of the His-tagged protein. 6. The cell culture is transferred to conical (Erlenmeyer) flasks. To assure sufficient space and optimize aeration of the bacterial culture, use a maximal volume of medium equivalent to one-fifth of the total volume of the flask (e.g., maximum ~200 ml culture in a one-liter flask). 7. Optimal induction occurs when the OD600 is between 0.4 and 0.6, that is, when the cells are at the exponential growth phase and protein expression usually lasts for 4–6 h. It is recommended to run a pilot expression study first, with smaller volumes to identify the optimal conditions. 8. To resuspend the cell pellet from one liter of cell culture, 10 ml of NTA Buffer A is used. Insufficient buffer will reduce protein yield, because the cells are not resuspended properly. On the other hand, a too big volume will result in unnecessary dilution which is disadvantageous for subsequent purification. 9. The volume for sonication is recommended to be 5–15 ml. The probe should be placed beneath the surface and at the center of the whole cell suspension. Place the tube on ice to protect the proteins from overheating caused by the heat generated during sonication. 10. Column flow is limited to a specific flow rate as provided in the manufacturer’s protocol. The “drop-to-drop” technique will ensure that the connection of the column to the tubing with the pump running will prevent air from entering the system. 11. Usually, the His-tagged proteins elute after ~5 ml of elution with NTA Buffer B. The fraction size can be adjusted after a few runs to collect most of the purified protein in one tube. 12. In this study, a single-step purification using a HisTrap column yields a protein fraction showing a single band on SDS-PAGE gel. However, if protein purity is insufficient after purification with a HisTrap column, the protein can be further purified using an anion exchange chromatography column (e.g., Q Sepharose) or gel-filtration column. 13. The ITC Assay Buffer contains 20 mM sodium phosphate buffer, pH 7.0, 100 mM NaCl to mimic the rice cytosol. For other experiments, the ITC Assay Buffer can be adjusted to

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC

413

simulate other environments. For example, the pH can be acidified for studies on the interaction of macromolecules in peroxisomes [6]. For storage, keep aliquots of the protein frozen at 80
C. 14. The concentration of the ligand is approximately ten times higher compared with that of the protein. The sample cell contains ~200 μl protein, while the titration of the ligand encompasses ~40 μl (2 μl  20 injections). As this experiment assumes a one-site binding mode, this setting of concentration allows around ten (out of a total of 20) injections to saturate the protein (ligand to protein ¼ 1:1). In one run, data on heatrelease before saturation (before the tenth injection), during saturation (around the tenth injection), and after saturation (injections subsequent to the tenth injection) are all recorded. With numbers available for all these stages, a curve can be drawn to provide reliable results for the calculations. 15. The temperature is selected according to the temperature used for rice cultivation. For experiments with Arabidopsis thaliana or other plant species, the temperature should be adjusted [6, 8]. 16. Depending on the settings, one experiment (including washing the machine) takes ~1 h. 17. As the unit on the y-axis is represented in kcal instead of cal, adding 1000 in “Y2” will translate the whole graph upward by one unit. 18. The n value indicates the ligand to protein ratio, which is usually around one. However, when proteins form dimers during binding, this value can be ~0.5. The Kd (dissociation constant) indicates the propensity of the protein-ligand complex to dissociate into its component molecules. Thus, the smaller the Kd value the more stable the complex, indicating the affinity between the protein and ligand. As an example, OsACBP2 was shown to bind C18:1-CoA with a Kd of ~7.5 μM (Fig. 6b).

Acknowledgments This work was supported by the Wilson and Amelia Wong Endowment Fund (to M.-L.C.), Research Grants Council of the Hong Kong Special Administrative Region, China (17101818 M to M.L.C.), and a University of Hong Kong Postdoctoral Fellowship (to Z.-H.G.). Partial support from the Research Grants Council of HKSAR, China (AoE/M-403/16 and AoE/M-05/12 to M.-L. C.) and the Innovation Technology Fund of Innovation Technology Commission: Funding Support to State Key Laboratory of

414

Ze-Hua Guo and Mee-Len Chye

Agrobiotechnology (to M.-L.C.) is gratefully acknowledged. Any opinions, findings, or recommendations expressed in this study do not reflect the views of the Government of the Hong Kong Special Administrative Region or the Innovation and Technology Commission. References 1. Wiseman T, Williston S, Brandts JF, Lin LN (1989) Rapid measurement of binding constants and heats of binding using a new titration calorimeter. Anal Biochem 179 (1):131–137. https://doi.org/10.1016/ 0003-2697(89)90213-3 2. Freire E, Mayorga OL, Straume M (1990) Isothermal titration calorimetry. Anal Chem 62 (18):A950–A959. https://doi.org/10.1021/ ac00217a002 3. Jiang Z, Zhou X, Tao M, Yuan F, Liu L, Wu F et al (2019) Plant cell-surface GIPC sphingolipids sense salt to trigger Ca2+ influx. Nature 572(7769):341–346. https://doi.org/10. 1038/s41586-019-1449-z 4. Guo ZH, Chan WH, Kong GK, Hao Q, Chye ML (2017) The first plant acyl-CoA-binding protein structures: the close homologues OsACBP1 and OsACBP2 from rice. Acta Crystallogr D 73(5):438–448. https://doi.org/10. 1107/S2059798317004193 5. Xue Y, Xiao S, Kim J, Lung SC, Chen L, Tanner JA et al (2014) Arabidopsis membraneassociated acyl-CoA-binding protein ACBP1 is involved in stem cuticle formation. J Exp Bot 65(18):5473–5483. https://doi.org/10. 1093/jxb/eru304 6. Hu TH, Lung SC, Ye Z-W, Chye ML (2018) Depletion of Arabidopsis ACYL-COA-BINDING PROTEIN3 affects fatty acid composition in the phloem. Front Plant Sci 9(2). https://doi.org/10.3389/fpls.2018.00002 7. Hsiao AS, Haslam RP, Michaelson LV, Liao P, Chen QF, Sooriyaarachchi S et al (2014) Arabidopsis cytosolic acyl-CoA-binding proteins ACBP4, ACBP5 and ACBP6 have overlapping but distinct roles in seed development. Biosci Rep 34:865–877. https://doi.org/10.1042/ BSR20140139 8. Miao R, Lung SC, Li X, Li XD, Chye ML (2019) Thermodynamic insights into an interaction between ACYL-CoA–BINDING PROTEIN2 and LYSOPHOSPHOLIPASE2 in Arabidopsis. J Biol Chem 294 (16):6214–6226. https://doi.org/10.1074/ jbc.RA118.006876 9. Ye ZW, Chen QF, Chye ML (2017) Arabidopsis thaliana acyl-CoA-binding protein ACBP6

interacts with plasmodesmata-located protein PDLP8. Plant Signal Behav 12(8):e1359365. https://doi.org/10.1080/15592324.2017. 1359365 10. Rasmussen J, Bo¨rchers T, Knudsen J (1990) Comparison of the binding affinities of acylCoA-binding protein and fatty-acid-binding protein for long-chain acyl-CoA esters. Biochem J 265:849–855. https://doi.org/10. 1042/bj2650849 11. Stevenson JM, Perera IY, Boss WF (1998) A phosphatidylinositol 4-kinase pleckstrin homology domain that binds phosphatidylinositol 4-monophosphate. J Biol Chem 273 (35):22761–22767. https://doi.org/10. 1074/jbc.273.35.22761 12. Meng W, Su YCF, RMK S, Chye ML (2011) The rice acyl-CoA-binding protein gene family: phylogeny, expression and functional analysis. New Phytol 189(4):1170–1184. Corrigendum: New Phytol 190(3):807. https://doi. org/10.1111/j.1469-8137.2010.03546.x 13. Gao W, Xiao S, Li HY, Tsao SW, Chye ML (2009) Arabidopsis thaliana acyl-CoA-binding protein ACBP2 interacts with heavy-metalbinding farnesylated protein AtFP6. New Phytol 181(1):89–102 14. Leung KC, Li HY, Xiao S, Tse MH, Chye ML (2006) Arabidopsis ACBP3 is an extracellularly targeted acyl-CoA-binding protein. Planta 223 (5):871–881. https://doi.org/10.1007/ s00425-005-0139-2 15. Chye ML, Li HY, Yung MH (2000) Single amino acid substitutions at the acyl-CoA-binding domain interrupt (14) C palmitoyl-CoA binding of ACBP2, an Arabidopsis acyl-CoAbinding protein with ankyrin repeats. Plant Mol Biol 44(6):711–721. https://doi.org/ 10.1023/a:1026524108095 16. Gao W, Li HY, Xiao S, Chye ML (2010) AcylCoA-binding protein 2 binds lysophospholipase 2 and lysoPC to promote tolerance to cadmium-induced oxidative stress in transgenic Arabidopsis. Plant J 62(6):989–1003. https:// doi.org/10.1111/j.1365-313X.2010. 04209.x 17. Chye ML (1998) Arabidopsis cDNA encoding a membrane-associated protein with an

Thermodynamics of ACBP Binding to Acyl-CoA Revealed by ITC acyl-CoA binding domain. Plant Mol Biol 38 (5):827–838. https://doi.org/10.1023/ A:1006052108468 18. Xiao S, Chye ML (2011) Overexpression of Arabidopsis ACBP3 enhances NPR1dependent plant resistance to Pseudomonas syringe pv tomato DC3000. Plant Physiol 156 (4):2069–2081. https://doi.org/10.1104/ pp.111.176933 19. Li HY, Xiao S, Chye ML (2008) Ethylene- and pathogen-inducible Arabidopsis acyl-CoAbinding protein 4 interacts with an ethyleneresponsive element binding protein. J Exp Bot 59(14):3997–4006. https://doi.org/10. 1093/jxb/ern241

415

20. Ladbury JE, Klebe G, Freire E (2010) Adding calorimetric data to decision making in lead discovery: a hot tip. Nat Rev Drug Discov 9 (1):23–27. https://doi.org/10.1038/ nrd3054 21. Kroll DJ, Abdel-Hafiz HA-M, Marcell T, Simpson S, Chen C-Y, Gutierrez-Hartmann A et al (1993) A multifunctional prokaryotic protein expression system: overproduction, affinity purification, and selective detection. DNA Cell Biol 12(5):441–453. https://doi.org/10. 1089/dna.1993.12.441 22. Nelson DL, Lehninger AL, Cox MM (2008) Lehninger principles of biochemistry. Macmillan, New York, NY

Chapter 24 In Situ Localization of Plant Lipid Metabolites by MatrixAssisted Laser Desorption/Ionization Mass Spectrometry Imaging (MALDI-MSI) Drew Sturtevant, Mina Aziz, Trevor B. Romsdahl, Chase D. Corley, and Kent D. Chapman Abstract Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) has emerged as a major analytical platform for the determination and localization of lipid metabolites directly from tissue sections. Unlike analysis of lipid extracts, where lipid localizations are lost due to homogenization and/ or solvent extraction, MALDI-MSI analysis is capable of revealing spatial localization of metabolites while simultaneously collecting high chemical resolution mass spectra. Important considerations for obtaining high quality MALDI-MS images include tissue preservation, section preparation, MS data collection and data processing. Errors in any of these steps can lead to poor quality metabolite images and increases the chance for metabolite misidentification and/ or incorrect localization. Here, we present detailed methods and recommendations for specimen preparation, MALDI-MS instrument parameters, software analysis platforms for data processing, and practical considerations for each of these steps to ensure acquisition of high-quality chemical and spatial resolution data for reconstructing MALDI-MS images of plant tissues. Key words MALDI mass spectrometry imaging, Lipids, Plant tissues, Seeds, Triacylglycerol, Phosphatidylcholine, Camelina sativa, Data processing

1

Introduction Over the last two decades matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) has become an integral tool to study the spatial localization of biomolecules. Discoveries using MALDI-MSI have reemphasized an often overlooked and important theme in biology, that biomolecules can be unequally, or heterogeneously, distributed in different tissues and cell types. In a variety of plant tissues, MALDI-MSI has been used to describe distributions of metabolites including, lipids in seeds [1–4], secondary metabolites of leaves [5, 6] and stem tissues [7], and peptides in root and nodule tissues [8]. More importantly, discoveries of heterogeneous metabolite distributions using

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_24, © Springer Science+Business Media, LLC, part of Springer Nature 2021

417

418

Drew Sturtevant et al.

MALDI-MSI have helped inform other biological studies including the identification of a mutant allele of the fatty acid desaturase 2–1 (FAD2–1) in a wild accession of Gossypium barbadense that has both a high-oil and high-oleic seed phenotype [9]; tracking the heterogeneity of phosphatidylcholine and triacylglycerol metabolism throughout seed development [10]; and used to highlight the potential bottlenecks of metabolism in transgenic lines of Camelina sativa (C. sativa) engineered to produce eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA; [11]). In a typical MALDI-MSI experiment, an intact tissue or tissue section is coated with a chemical matrix, to assist with ionization of molecules near the surface of the tissue. The matrix-coated tissue sample is then loaded into the mass spectrometer where a laser beam is pulsed over the tissue, typically at 10–100 μm steps. A plume of ions is generated at each step and these ions are analyzed for their intensity, mass to charge ratio (m/z), and spatial position (x, y coordinates). Together these data are used to construct falsecolored images, or heat maps, representing the physical distribution of ions within a sample. While the general workflow of a MALDIMSI experiment appears streamlined, simple errors in sample preservation and preparation can thwart an experiment, leading to mis-localization and poor to no detection of targeted metabolites. The overall goal of tissue preparation for a MALDI-MSI experiment is to preserve the true spatial distribution of metabolites within a tissue without altering the chemical integrity of targeted metabolites. Preserving sample integrity for MALDI-MS imaging has been well studied in mammalian tissues [12–17] and these resources act as excellent reference points for preparation of plant tissues for MALDI-MSI experiments. In addition to these considerations learned from animal tissue preparation for MALDI-MSI, plant tissue preparation faces additional challenges that must be taken into consideration. For example, plant tissue morphology can range widely from fresh leaves, which are highly vacuolated and include many air spaces, to seed tissues, containing little in the way of vacuoles and air spaces and are desiccated to less than 10% water content. These wide-ranging morphological features of plant tissues can make preparation and cryo-sectioning of plant tissues challenging, as some frozen plant tissues can fracture, shrink, and shred during the cryo-sectioning process. Tissue deformities introduced from the cryo-sectioning process can make data interpretation more difficult, as analytes can become mis-localized if tissue structures are distorted or disrupted [18, 19]. Additionally, for MALDI-MS imaging experiments designed to image the surface of plant tissues (e.g., leaf surfaces), the plant cuticle can act as a hydrophobic barrier preventing an even matrix coating, when sprayed, resulting in decreased or unequal analyte ionization efficiency [20]. However, care in handling and preparation of plant tissues can overcome many of the technical difficulties associated with MALDI-MSI analysis.

MALDI-MS Imaging of Plant Tissues

419

Here, we outline practical guidelines for the preparation and analysis of plant tissues by MALDI-MSI. Methods include our typical work-flows for specimen preparation, tissue embedding and cryo-sectioning, matrix application by sublimation, MALDIMS imaging analysis by a MALDI-LTQ-Orbitrap-XL highresolution mass spectrometer, data analysis, and validation of MALDI-MS through other analytical methods. As an example, this methodology will be applied to MALDI-MSI imaging of C. sativa seed sections. Although some optimization will be required for the targeted tissue or specimen of choice, the methods outlined here should enable the acquisition of reproducible highspatial and chemical resolution MALDI-MS images from plant tissues and tissue sections.

2

Materials

2.1 Tissue Fixation (Optional)

1. 4.0% Paraformaldehyde solution; Paraformaldehyde powder, 50 mM PIPES–NaOH buffer, pH 7.2. 2. 50 mM PIPES–NaOH buffer, pH 7.2. 3. Sterile 18 Ω water. 4. Hot plate with magnetic stirring and stir-bar. 5. Aluminum foil. 6. Falcon tube(s); capable of holding tissue.

2.2 Tissue Embedding

1. 10.0% gelatin solution; Porcine gelatin 300 bloom, sterile 18 Ω water. 2. Embedding molds, slightly larger than size of tissue (e.g., small weigh boats, 40
40
8 mm). 3. Flask. 4. Water bath; heated to 40  C. 5. P1000 or P5000 pipettor with sterile tips. 6. Forceps; spatula or toothpick. 7. Scalpel or single-edge razor blade. 8. Sharpie® or marker. 9. Weigh boat; medium (85
85
24 mm). 10. Aluminum foil. 11. 4  C refrigerator. 12. 80  C freezer. 13. 20  C freezer.

420

Drew Sturtevant et al.

2.3 Tissue CryoSectioning

1. Tissue preparation from Subheading 2.2. 2. Cryostat; Held at temperature between 15 and 18  C. 3. Disposable stainless steel cryostat blades. 4. Precleaned charged glass slides (Orbitrap) or conductive slides (TOF). 5. Cryostat specimen discs or “chucks”. 6. Optimal Cutting Temperature (OCT) compound or other cryo-embedding/mounting medium. 7. CryoJane® adhesive tape windows (Optional, see Note 6). 8. Slides with double-sided tape on both ends (Optional if using adhesive windows). 9. Metal spatula and forceps. 10. Slide box. 11. Methanol. 12. Kimwipes®. 13. Lyophilizer and lyophilization tube. 14. Desiccator; room temperature held under vacuum.

2.4 Matrix Application by Sublimation

1. Tissue sections from Subheading 2.3. 2. Chemical matrix (see Note 7). 3. Mortar and pestle. 4. Fine mesh sifter. 5. Sublimation apparatus; Chemglass (CG-3038). 6. Sand-bath; Heated between 120 and 140  C. 7. Rough vacuum; Capable of 50–200 mTorr. 8. Two vacuum traps or one double vacuum trap. 9. Ringstand(s) and ringstand clamps and rings. 10. Vacuum tubing to fit sublimation chamber and vacuum traps. 11. Small container filled with ice and water. 12. Acetone. 13. Ice. 14. Water. 15. Tape. 16. Glass cutter. 17. Desiccator.

2.5 MALDI Mass Spectrometry

1. High-resolution mass spectrometer equipped with a matrixassisted laser desorption/ionization source, such as a MALDI-LTQ-Orbitrap XL (ThermoScientific, used for this

MALDI-MS Imaging of Plant Tissues

421

methods paper) or other commonly used MALDI instrumentation, including Bruker MALDI-FTICR-MS (solariX XR) or MALDI-TOF (rapifleX, autofleX, ultrafleXtreme), Shimadzu (Axima Confidence MALDI-TOF, Axima ID MALDI-TOF, and MALDI 7090 TOF-TOF), and Waters MALDI-SYNAPT G2-Si. 2. MALDI adaptor plate capable of holding slides from Subheading 2.3, item 4. 3. Samples from Subheading 2.4.

3

Methods

3.1 Sample Collection and Preservation

1. Desiccated seeds—Seeds can be imaged with no preservation steps as they are already metabolically quiescent. If the seed tissues have any debris or dirt on them they can be quickly rinsed with sterile water then dried in a low temperature oven (~37  C). 2. Fresh, metabolically active tissues—After harvesting, fresh tissues (roots, leaves, developing seeds, etc.) should be immediately placed in 4% paraformaldehyde solution at room temperature (see Note 1). We recommend tissues to be vacuum-infiltrated for 2–4 h depending on the tissue. After fixation, tissues should be washed 3
(15 min each) with 50 mM PIPES–NaOH buffer, pH 7.2, replacing the buffer after each wash. Fixed tissues can be embedded immediately or stored (or shipped) at 4  C overnight in the final wash buffer.

3.2 Tissue Embedding

1. Prepare 10% gelatin solution by adding gelatin and water to a flask and swirling briefly to mix (it will not dissolve). Microwave suspension for ~90 s, ensuring it does not bubble over the flask. To prevent bubbling over, once the bubbles start to rise, pause the microwave, swirl the flask, and return to the microwave. Repeat this process until the gelatin is completely dissolved, and is uniformly distributed in the solution. Immediately cover flask with foil cap and transfer to a 40  C water bath. 2. In the water bath, allow gelatin solution to equilibrate to 40  C and for bubbles on the surface of the gelatin to dissipate (~1–2 h; see Note 2). Acquire remaining materials while gelatin is equilibrating. 3. Pipet 0.5–1.0 cm of gelatin into the tissue mold(s) (Fig. 1a), ensuring there are no bubbles (see Note 3), and allow to semiset ~3.0 min. Semiset in this instance means the gelatin should just begin to solidify, but should not be liquid or completely set. To determine if the gelatin layer is semiset, use a sterile pipette tip or toothpick to test the surface of the

422

Drew Sturtevant et al.

Fig. 1 Preparation of Camelina sativa seeds for MALDI-MSI. (a) Weigh boat with 2.0 ml of semi-set gelatin, (b) C. sativa seeds placed on semiset gelatin, (c) Second layer of gelatin pipetted on C. sativa seeds, (d) Gelatin and seeds decasted from weigh boat mold, (e) Trimmed gelatin-tissue blocks, (f) Frozen and marked gelatin-

MALDI-MS Imaging of Plant Tissues

423

gelatin, which should pull away a thin strand of gelatin. Place tissue(s) on semiset gelatin with forceps in desired orientation for sectioning and transfer to a 4  C refrigerator for 5–10 min until the gelatin layer is completely set (Fig. 1b). If embedding freshly fixed tissues, carefully dry the surfaces of the tissues with a Kimwipes to remove excess moisture. 4. Remove tissue mold from refrigerator and pipet enough gelatin to completely cover the tissue by 0.5–1.0 cm and ensure to remove all bubbles. Transfer the tissue mold back to the 4  C refrigerator for an additional 5–10 min until completely set (Fig. 1c). 5. Remove tissue mold from refrigerator and remove the gelatinembedded tissue from the mold with spatula or toothpick (Fig. 1d). Trim the gelatin-embedded tissue using the scalpel/single edge blade, so there is a ~3.0–5.0 mm border of gelatin surrounding the tissue on all sides (Fig. 1e). 6. Mark one of the corners of the gelatin-embedded tissue, as seen in Fig. 1f, with the Sharpie/marker to indicate desired orientation. 7. Place gelatin-embedded tissue in a medium weigh boat, cover with aluminum foil, and place in 80  C for 3–4 h. 8. Quickly transfer gelatin-embedded tissue from 80  C to 20  C for a minimum of 48 h or up to 10 days (tissue blocks after freezing and equilibration Fig. 1f). Equilibration time will depend largely on the size, density, and storage content of tissues. For example, larger seeds (e.g., cotton) require a longer equilibration time than smaller seeds (e.g., Arabidopsis) or leaf tissues. 3.3

Cryo-Sectioning

1. Precaution: The cryostat blades are extremely sharp, use the blade guard whenever possible. 2. Chill cryostat to between 15 and 18  C. Place 20  Cequilibrated, gelatin-embedded tissue blocks, slides, slide box, forceps, and all other tools necessary in the cryostat for 20 min to equilibrate to the cryostat temperature.

ä Fig. 1 (continued) embedded tissue blocks prior to cryo-sectioning, (g) Frozen gelatin-tissue block mounted to cryostat chuck using OCT compound (h) Cryo-sectioning a tissue block (i) Thaw mounting sections onto a slide, (j) Lyophilizing slides with cyro-sections (k) Bright-field image of medial-longitudinal section of a C. sativa seed (l) Slide with tissue sections pre and post matrix sublimation (m) Thermo-MALDI-LTQ-Orbitrap XL instrument running an analysis (n) Data processing using Thermo ImageQuest software (o) Processed MALDI-MS images for an embryonic axis-enriched lipid metabolite at m/z 891.683 and cotyledon-enriched metabolite at m/z 943.715

424

Drew Sturtevant et al.

3. Carefully clean cryostat blade, cryostat glass antiroll bar, and cryostat stage using a Kimwipes and methanol and reinstall these items to their respective locations in the cryostat. 4. The following steps all occur within the cryostat unless explicitly mentioned. 5. On a cryostat chuck place a small amount of OCT compound. Using forceps, quickly and firmly press a gelatin-embedded tissue block onto the OCT compound and allow to completely freeze. Apply an additional, small amount of OCT compound around the base of the tissue block to secure the edges of the gelatin-embedded tissue to the chuck block. Allow OCT compound to completely freeze, about 2–3 min. 6. Repeat step 5 (Subheading 3.4) for remaining gelatinembedded tissue blocks (Fig. 1g; see Note 4). 7. Using the cryostat controls, move the specimen head all the way back, and set the cryostat to section at your desired thickness. We routinely use 30 μm sections for MALDI-Orbitrap analysis, however thinner sections, 20 μm are minimal or not observed. As a general guideline we recommend sectioning at the thinnest thickness possible that yields high quality and reproducible sections. 6. Tissues including leaf, root, and some starchy seed tissues can be prone to “shredding” while sectioning. High quality and intact sections of tissues that are fragile or prone to “shredding” can be collected using CryoJane adhesive windows, rather than thaw mounting. 7. There are a wide range of commercially available MALDI-MSI matrices, each with preferential ionization for class/classes of biomolecules. Prior to MALDI-MSI experiments, the matrix should be thoughtfully selected to aid in analysis of desired biomolecules. There are numerous reviews and references, for plant and mammalian tissues, covering the properties and applications of each matrix [12, 18, 38–40]. For analysis of triacylglycerols (TAG) and phosphatidylcholine (PC) in positive ionization mode we prefer the matrix 2,5-dihydroxybenzoic acid (DHB). Whereas, for the analysis of other phospholipids (e.g., phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol), galactolipids, and N-acylphosphatidylethanolamine (NAPE) in negative ionization mode we prefer the matrix 1,5-diaminonapthalene (DAN). 8. The following are recent references for instrument conditions of MALDI-MS analysis of plant tissues other than Orbitrap instruments: UltrafleXtreme MALDI-TOF/TOF, solariX MALDI-FT-ICR (Bruker Daltonics [41, 42], MALDISYNAPT G2, and MALDI-TOF micro MX (Waters) [6, 43].

436

Drew Sturtevant et al.

Acknowledgments Adaptation of MS imaging methodology was made possible by a grant from the Hoblitzelle Foundation for the purchase of cryosectioning and MALDI-MS instrumentation. Research support for tissue imaging of lipid metabolites in Camelina sativa seeds was provided by the U.S. Department of Energy (DOE) Office of Science, BES-Physical Biosciences program (DE-SC0016536). References 1. Lu S, Sturtevant D, Aziz M, Jin C, Li Q, Chapman KD, Guo L (2018) Spatial analysis of lipid metabolites and expressed genes reveals tissuespecific heterogeneity of lipid metabolism in high- and low-oil Brassica napus L. seeds. Plant J 94(6):915–932. https://doi.org/10. 1111/tpj.13959 2. Marmon S, Sturtevant D, Herrfurth C, Chapman K, Stymne S, Feussner I (2017) Two acyltransferases contribute differently to linolenic acid levels in seed oil. Plant Physiol 173(4):2081–2095. https://doi.org/10. 1104/pp.16.01865 3. Sturtevant D, Duenas ME, Lee YJ, Chapman KD (2017) Three-dimensional visualization of membrane phospholipid distributions in Arabidopsis thaliana seeds: a spatial perspective of molecular heterogeneity. Biochim Biophys Acta 1862(2):268–281. https://doi.org/10. 1016/j.bbalip.2016.11.012 4. Sturtevant D, Romsdahl TB, Yu X-H, Burks DJ, Azad RK, Shanklin J, Chapman KD (2019) Tissue-specific differences in metabolites and transcripts contribute to the heterogeneity of ricinoleic acid accumulation in Ricinus communis L. (castor) seeds. Metabolomics 15 (1):6 5. Korte AR, Yandeau-Nelson MD, Nikolau BJ, Lee YJ (2015) Subcellular-level resolution MALDI-MS imaging of maize leaf metabolites by MALDI-linear ion trap-Orbitrap mass spectrometer. Anal Bioanal Chem 407 (8):2301–2309. https://doi.org/10.1007/ s00216-015-8460-5 6. Shroff R, Schramm K, Jeschke V, Nemes P, Vertes A, Gershenzon J, Svatos A (2015) Quantification of plant surface metabolites by matrix-assisted laser desorption-ionization mass spectrometry imaging: glucosinolates on Arabidopsis thaliana leaves. Plant J 81 (6):961–972. https://doi.org/10.1111/tpj. 12760 7. Aziz M, Sturtevant D, Winston J, Collakova E, Jelesko JG, Chapman KD (2017) MALDI-MS

imaging of urushiols in poison ivy stem. Molecules 22(5). https://doi.org/10.3390/ molecules22050711 8. Gemperline E, Keller C, Jayaraman D, Maeda J, Sussman MR, Ane JM, Li LJ (2016) Examination of endogenous peptides in Medicago truncatula using mass spectrometry imaging. J Proteome Res 15(12):4403–4411. https://doi.org/10.1021/acs.jproteome. 6b00471 9. Sturtevant D, Horn P, Kennedy C, Hinze L, Percy R, Chapman K (2017) Lipid metabolites in seeds of diverse Gossypium accessions: molecular identification of a high oleic mutant allele. Planta 245(3):595–610. https://doi.org/10. 1007/s00425-016-2630-3 10. Woodfield HK, Sturtevant D, Borisjuk L, Munz E, Guschina IA, Chapman K, Harwood JL (2017) Spatial and temporal mapping of key lipid species in Brassica napus seeds. Plant Physiol 173(4):1998–2009. https://doi.org/ 10.1104/pp.16.01705 11. Usher S, Han LH, Haslam RP, Michaelson LV, Sturtevant D, Aziz M, Chapman KD, Sayanova O, Napier JA (2017) Tailoring seed oil composition in the real world: optimising omega-3 long chain polyunsaturated fatty acid accumulation in transgenic Camelina sativa. Sci Rep 7:6570. https://doi.org/10.1038/ s41598-017-06838-0 12. Thomas A, Chaurand P (2014) Advances in tissue section preparation for MALDI imaging MS. Bioanalysis 6(7):967–982. https://doi. org/10.4155/Bio.14.63 13. Puolitaival SM, Burnum KE, Cornett DS, Caprioli RM (2008) Solvent-free matrix dry-coating for MALDI imaging of phospholipids. J Am Soc Mass Spectr 19(6):882–886. https://doi.org/10.1016/j.jasms.2008.02. 013 14. Zavalin A, Yang J, Hayden K, Vestal M, Caprioli RM (2015) Tissue protein imaging at 1 μm laser spot diameter for high spatial resolution and high imaging speed using

MALDI-MS Imaging of Plant Tissues transmission geometry MALDI TOF MS. Anal Bioanal Chem 407(8):2337–2342. https:// doi.org/10.1007/s00216-015-8532-6 15. Khatib-Shahidi S, Andersson M, Herman JL, Gillespie TA, Caprioli RM (2006) Direct molecular analysis of whole-body animal tissue sections by imaging MALDI mass spectrometry. Anal Chem 78(18):6448–6456. https:// doi.org/10.1021/ac060788p 16. Casadonte R, Caprioli RM (2011) Proteomic analysis of formalin-fixed paraffin-embedded tissue by MALDI imaging mass spectrometry. Nat Protoc 6(11):1695–1709. https://doi. org/10.1038/nprot.2011.388 17. Seeley EH, Caprioli RM (2011) MALDI imaging mass spectrometry of human tissue: method challenges and clinical perspectives. Trends Biotechnol 29(3):136–143. https:// doi.org/10.1016/j.tibtech.2010.12.002 18. Dong Y, Li B, Malitsky S, Rogachev I, Aharoni A, Kaftan F, Svatos A, Franceschi P (2016) Sample preparation for mass spectrometry imaging of plant tissues: a review. Front Plant Sci 7:60. https://doi.org/10.3389/fpls. 2016.00060 19. Dong YH, Li B, Aharoni A (2016) More than pictures: when MS imaging meets histology. Trends Plant Sci 21(8):686–698. https://doi. org/10.1016/j.tplants.2016.04.007 20. Takahashi K, Kozuka T, Anegawa A, Nagatani A, Mimura T (2015) Development and application of a high-resolution imaging mass spectrometer for the study of plant tissues. Plant Cell Physiol 56(7):1329–1338. https://doi.org/10.1093/pcp/pcv083 21. Hankin JA, Barkley RM, Murphy RC (2007) Sublimation as a method of matrix application for mass spectrometric imaging. J Am Soc Mass Spectr 18(9):1646–1652. https://doi.org/10. 1016/j.jasms.2007.06.010 22. Aerni HR, Cornett DS, Caprioli RM (2006) Automated acoustic matrix deposition for MALDI sample preparation. Anal Chem 78 (3):827–834. https://doi.org/10.1021/ ac051534r 23. Nikolau B, Song ZH, Jun J, Yeung E, Lee Y (2010) High-spatial resolution metabolomics analysis: a case study of imaging plant surface metabolites by laser desorption ionization mass spectrometry using colloidal silver. In Vitro Cell Dev An 46(Suppl 1):S52–S52. https:// doi.org/10.1007/s11626-010-9338-7 24. Robichaud G, Garrard KP, Barry JA, Muddiman DC (2013) MSiReader: an open-source interface to view and analyze high resolving power MS imaging files on Matlab platform. J

437

Am Soc Mass Spectr 24(5):718–721. https:// doi.org/10.1007/s13361-013-0607-z 25. Horn PJ, Chapman KD (2014) Metabolite imager: customized spatial analysis of metabolite distributions in mass spectrometry imaging. Metabolomics 10(2):337–348. https:// doi.org/10.1007/s11306-013-0575-0 26. Race AM, Palmer AD, Dexter A, Steven RT, Styles IB, Bunch J (2016) SpectralAnalysis: software for the masses. Anal Chem 88 (19):9451–9458. https://doi.org/10.1021/ acs.analchem.6b01643 27. Fahy E, Sud M, Cotter D, Subramaniam S (2007) LIPID MAPS online tools for lipid research. Nucleic Acids Res 35(Web Server): W606–W612. https://doi.org/10.1093/ nar/gkm324 28. Du JL, Yuan ZF, Ma ZW, Song JZ, Xie XL, Chen YL (2014) KEGG-PATH: Kyoto encyclopedia of genes and genomes-based pathway analysis using a path analysis model. Mol BioSyst 10(9):2441–2447. https://doi.org/10. 1039/c4mb00287c 29. Kanehisa M, Furumichi M, Tanabe M, Sato Y, Morishima K (2017) KEGG: new perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res 45(D1):D353–D361. https://doi.org/10.1093/nar/gkw1092 30. Kanehisa M, Goto S (2000) KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28(1):27–30. https://doi.org/10. 1093/nar/28.1.27 31. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M (2016) KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res 44(D1): D457–D462. https://doi.org/10.1093/nar/ gkv1070 32. Chapman KD, Moore TS Jr (1993) N-Acylphosphatidylethanolamine synthesis in plants: occurrence, molecular composition, and phospholipid origin. Arch Biochem Biophys 301(1):21–33. https://doi.org/10. 1006/abbi.1993.1110 33. Li MY, Baughman E, Roth MR, Han XL, Welti R, Wang XM (2014) Quantitative profiling and pattern analysis of triacylglycerol species in Arabidopsis seeds by electrospray ionization mass spectrometry. Plant J 77 (1):160–172. https://doi.org/10.1111/tpj. 12365 34. Salazar C, Jones MD, Sturtevant D, Horn PJ, Crossley J, Zaman K, Chapman KD, Wrona M, Isaac G, Smith NW, Shulaev V (2017) Development and application of sub-2-μm particle CO2-based chromatography coupled to mass spectrometry for comprehensive analysis of

438

Drew Sturtevant et al.

lipids in cottonseed extracts. Rapid Commun Mass Spectrom 31(7):591–605. https://doi. org/10.1002/rcm.7825 35. Welti R, Li WQ, Li MY, Sang YM, Biesiada H, Zhou HE, Rajashekar CB, Williams TD, Wang XM (2002) Profiling membrane lipids in plant stress responses – role of phospholipase D alpha in freezing-induced lipid changes in Arabidopsis. J Biol Chem 277(35):31994–32002. https://doi.org/10.1074/jbc.M205375200 36. Korte AR, Lee YJ (2013) Multiplex mass spectrometric imaging with polarity switching for concurrent acquisition of positive and negative ion images. J Am Soc Mass Spectr 24 (6):949–955. https://doi.org/10.1007/ s13361-013-0613-1 37. Horn PJ, Silva JE, Anderson D, Fuchs J, Borisjuk L, Nazarenus TJ, Shulaev V, Cahoon EB, Chapman KD (2013) Imaging heterogeneity of membrane and storage lipids in transgenic Camelina sativa seeds with altered fatty acid profiles. Plant J 76(1):138–150. https:// doi.org/10.1111/tpj.12278 38. Feenstra AD, O’Neill KC, Yagnik GB, Lee YJ (2016) Organic-inorganic binary mixture matrix for comprehensive laser-desorption ionization mass spectrometric analysis and imaging of medium-size molecules including phospholipids, glycerolipids, and oligosaccharides. RSC Adv 6(101):99260–99268. https:// doi.org/10.1039/c6ra20469d 39. Sturtevant D, Lee YJ, Chapman KD (2016) Matrix assisted laser desorption/ionizationmass spectrometry imaging (MALDI-MSI)

for direct visualization of plant metabolites in situ. Curr Opin Biotechnol 37:53–60. https:// doi.org/10.1016/j.copbio.2015.10.004 40. Thomas A, Charbonneau JL, Fournaise E, Chaurand P (2012) Sublimation of new matrix candidates for high spatial resolution imaging mass spectrometry of lipids: enhanced information in both positive and negative polarities after 1,5-diaminonapthalene deposition. Anal Chem 84(4):2048–2054. https://doi.org/10. 1021/ac2033547 41. Gorzolka K, Bednarz H, Niehaus K (2014) Detection and localization of novel hordatinelike compounds and glycosylated derivates of hordatines by imaging mass spectrometry of barley seeds. Planta 239(6):1321–1335. https://doi.org/10.1007/s00425-014-2061y 42. Sarabia LD, Boughton BA, Rupasinghe T, van de Meene AML, Callahan DL, Hill CB, Roessner U (2018) High-mass-resolution MALDI mass spectrometry imaging reveals detailed spatial distribution of metabolites and lipids in roots of barley seedlings in response to salinity stress. Metabolomics 14(5):ARTN 63. https://doi.org/10.1007/s11306-018-13593 43. Dalisay DS, Kim KW, Lee C, Yang H, Rubel O, Bowen BP, Davin LB, Lewis NG (2015) Dirigent protein-mediated lignan and cyanogenic glucoside formation in flax seed: integrated omics and MALDI mass spectrometry imaging. J Nat Prod 78(6):1231–1242. https:// doi.org/10.1021/acs.jnatprod.5b00023

Part V Lipid Databases

Chapter 25 Plant Lipid Databases Peter Do¨rmann Abstract Along with the increase in knowledge on lipid metabolism during the last years, different lipid databases were established in a web-based system. This chapter presents an overview on plant lipid databases for simple and complex lipids focusing on nomenclature, structures as well as physical and chemical properties. Many databases provide information on methods and protocols for lipid isolation, fractionation, and analysis, including lipidomic procedures. References to the lipid literature are included in all databases. Additional data including mass spectra derived from GC-MS, LC-MS, and LC-MS/MS experiments are included in specialized lipid databases. An introduction is presented on how to use the most important lipid databases. Key words Lipidomics, Mass spectrometry, Structures, Database, Library

1

Introduction The knowledge on lipid research has seen an enormous increase in the past decades. Accordingly, the accumulation of data and information on lipids has stimulated the development of numerous databases, some of which are devoted to lipid research only, while others provide a broader scope on lipids and other small molecules. Most lipid databases are open to the public and free of charge, and they are maintained and updated by public organizations, research institutions, research consortia, or individual scientists. The development of lipid databases included the classification of lipid structures and their organization into different lipid classes or categories [1, 2]. Databases focusing on fatty acids, their structures and distribution were developed starting in the 1990s (e.g., SOFA, PlantFAdb) [3–5]. Databases presenting the classification of simple and complex lipids include Lipid Library, Cyberlipid, LipidWeb, LipidBank, and Lipid Maps (Lipid Maps Structure Database, LMSD) [6–9]. General databases on small molecules, such as PubChem or NIST Chemistry WebBook, also include a large number of lipid structures [10, 11]. These databases provide information

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_25, © Springer Science+Business Media, LLC, part of Springer Nature 2021

441

442

Peter Do¨rmann

on physical and chemical properties for many compounds, and some databases also include NMR and mass spectral data. Many databases contain data on animal as well as plant, bacterial, and yeast lipids. Furthermore, lipid biochemistry and lipid metabolic pathways have been collected, for example in the KEGG and AraLip databases [12, 13] (Table 1; Chapter 26). Another important aspect is the presentation of methods and protocols for lipid extraction, fractionation and analysis, including the recently developed protocols for lipidomic approaches [12]. It became instrumental to use mass spectra for compound identification, when GC-MS and LC-MS methods were introduced in lipid research. GC-MS mass spectra can be retrieved from LipidWeb, Lipid Maps, PubChem, and NIST Chemistry WebBook [6, 10, 11]. The latter database offers the possibility to search for matching patterns in mass spectra of an unknown compound with a reference compound in the library [10]. The Lipid Maps, PubChem or Metlin databases provide a strategy to search for masses of precursor ions obtained by LC-MS without fragmentation [6, 11, 14]. For MS/MS spectra of unknown compounds obtained on tandem mass spectrometers, it is still difficult to develop a general search strategy in a spectral library, because different instruments with high or low mass accuracy and different mass analyzers are used (Q-TOF, triple quadrupole, Q-trap, orbitrap, etc.). Furthermore, the use of different ion sources or buffers can result in formation of various adducts during ionization, and in addition, the collision energy used for fragmentation has a strong impact on the relative abundances of fragment ions in the mass spectra. The Metlin database provides a collection of mass spectra of a large number of small molecules recorded with tandem mass spectrometers at different collision energies [14]. This database can be searched with mass spectra recorded at any tandem LC-MS instrument. Many databases present overviews on historical aspects of lipid research (e.g., LipidWeb, Lipid Library, Cyberlipid), and all databases also provide an enormous collection of lipid literature of the past years. The databases included in this chapter are free of charge. Some require the registration of the user with an email address and password. While it is impossible to present all online lipid databases, it is the goal of this chapter to show a selection of highly relevant databases as a starting point. The reader is encouraged to test these databases and to start searching for additional databases to collect information on further aspects of lipid research.

2

Materials A conventional personal computer (IBM compatible or Apple) with stable connection to the internet is sufficient for this work. Some databases require personnel registration with an e-mail address prior to use.

http://cyberlipid.gerli. com/

Cyberlipid (integrated in GERLI)

General protocols for lipid isolation, analysis

Lipid analysis protocols including, lipidomics, NMR, biochemical pathways, oil processing

http://lipidbank.jp/

https://pubchem.ncbi. nlm.nih.gov/

https://www. lipidomicnet.org/

LipidBank

PubChem

LipidomicNet

William W. Christie (James Hutton Institute, Dundee)

Groupe d’Etude et de Recherche en Lipidomique (GERLI)

American Oil Chemists’ Society (AOCS)

National Institutes of Health (NIH)

Structures of lipids, protocols for lipid isolation, LipidomicNet Consortium lipidomics

Structures, masses, formulas of small molecules (including lipids) and their NMR, GC-MS, LC-MS spectra

Structures, formulas of lipids, NMR and GC-MS Japanese Conference on the spectra Biochemistry of Lipids (JCBL)

LipidWeb (also https://www.lipidhome. Structures, formulas of lipids, GC-MS mass integrated into Lipid co.uk/index.html spectra Maps) https://www.lipidmaps. org/resources/ lipidweb/

https://lipidlibrary. aocs.org/

Lipid structures and characteristics, analytical protocols

Lipid Library

Reference

Subheading 3.1 [4, 5]

See methods

Subheading 3.8 (continued)

Subheading 3.7 [11]

Subheading 3.6 [9]

Subheading 3.5

Subheading 3.4

Subheading 3.3

Based on SOFA database, fatty acids from plant John Ohlrogge and others, MSU, Subheading 3.2 [3] species with phylogenetic relationship East Lansing, USA

https://plantfadb.org/

Host

Plant Fatty Acid database (PlantFAdb)

Content

http://sofa.mri.bund.de Composition of fatty acids, sterols, tocopherols, Bertrand Mattha¨us, Max Rubner triacylglycerols in seed oils of different plants Institut (MRI), Karlsuhe, Germany

URL

Seed Oil Fatty Acids (SOFA)

Fatty acid databases

Database (name)

Table 1 Databases for plant lipid research

http://www. sphingomap.org/

index.php/Main_ Page

URL

Structures, of sphingolipids, genes of biosynthesis pathways

Content

Scripps Research Institute (La Jolla, CA)

Metlin

Minoru Kanehisa Laboratories

(Chapter 26)

Biochemical pathways, genes of different organisms

Kyoto Encyclopedia of Genes and Genomes (KEGG)

https://www.genome. jp/kegg/

Biochemical pathways, genes of lipid synthesis in John Ohlrogge and others, MSU, (Chapter 26) Arabidopsis East Lansing, USA

Arabidopsis Acyl-Lipid http://aralip. plantbiology.msu. Metabolism edu/ (AraLip)

Databases on pathways and genes

https://metlin.scripps. edu

Subheading 3.12

Subheading 3.11

Lipid Maps Consortium

Lipid Maps Lipidomics https://www.lipidmaps. Structures, masses, formulas of lipids, mass Gateway org/ spectra, genes, pathways, software tools, lipidomics Structures, formulas of lipids, tandem LC-MS/ MS mass spectra

Subheading 3.10

LipidomicNet Consortium and others

https://www. ebi.ac.uk/ metabolights/ lipidhome/

LipidHome

Database of theoretical lipid structures and masses

https://webbook.nist. gov/chemistry/

NIST Chemistry WebBook

Reference

[13]

[12]

[14]

[6–8]

[17]

Subheading 3.9 [10]

See methods

National Institute of Standards and Technology (NIST)

Alfred H. Merrill Jr., Georgia Institute of Technology

Host

Structures, masses, formulas of small molecules (including lipids) and their GC-MS mass spectra

Databases with structures and search options for mass spectra

SphinGOMAP

Database (name)

Table 1 (continued)

Plant Lipid Databases

3

445

Methods The description of the different databases follows the order as listed in Table 1. Two databases (SOFA, PlantFAdb) focus on fatty acid structures and occurrence in different plants. The next six databases (Lipid Library, Cyberlipid, LipidBank, LipidWeb, PubChem, and LipidomicNet) contain important information on lipid structures and characteristics and analytical protocols. The last four databases (NIST Chemistry WebBook, LipidHome, Lipid Maps Lipidomics Gateway, and Metlin) contain mass spectral libraries that are searchable with parental masses or fragment masses or both. The Metlin database is particularly useful for identifying lipids by searching a large collection of experimentally determined MS/MS spectra.

3.1 Seed Oil Fatty Acids (SOFA) Database

Fatty acids are the basic building blocks of numerous lipids. Seed oils of many plant species constitute a rich resource for different fatty acids, including saturated, unsaturated fatty acids, and fatty acids with triple bonds, hydroxyl, epoxy, cyclopropane rings, and so on. The SOFA database contains data on more than 300 fatty acids and their distribution in more than 7000 plant species, and includes ~1400 references [4, 5]. Kurt Aitzetmu¨ller, former director of the German Institute for Chemistry and Physics of Lipids (now: Federal Research Institute of Nutrition and Food), initiated the development of a web-based database between 1998 and 2002. The database is now supervised by Bertrand Mattha¨us at the German Max Rubner-Institute (MRI). 1. Enter the link http://sofa.mri.bund.de in your browser. Select an e-mail address and password for login. 2. To search for a specific fatty acid, select “Fatty Acids” from the menu on the left and enter the name (e.g., linoleic acid) or use the “Delta nomenclature” which is explained in the “Help” pull-down menu (see Note 1). A table will be presented with the plant species including references that reported the occurrence of this fatty acid in the seed oil. Furthermore, tables with the quantitative data derived from gas chromatography analyses are shown. 3. To search for fatty acid composition in a plant species, select “Plants” in the menu on the left. Use the Latin name (genus and/or species name). A table will open with reports and details on the quantification of the fatty acids in the seed oil of this selected species. 4. To search for sterols, tocopherols or triacylglycerols in the database, select “Sterols,” “Tocopherols,” or “Triacylglycerol.”

446

Peter Do¨rmann

3.2 Plant Fatty Acid Database (PlantFAdb)

The PlantFAdb has been developed by John Ohlrogge and others based on the data of the SOFA database to enable phylogenetic searches [3]. English names of the plant species were added, while the sterol, tocopherol, and triacylglycerol data have not yet been included. In 2017, 2000 datasets of Chinese literature were incorporated. 1. Enter the address https://plantfadb.org/ in your browser. 2. Selecting “Fatty Acids” from the pull-down menu will show a table of >300 fatty acids with structure, name, Delta notation (see Note 1), formula, mass and a link to the SOFA database. 3. Click on the structure or name of the fatty acid to present detailed information including references. 4. In the fatty acid table, click on “Tree” to show the phylogenetic groups of plants that produce this fatty acid. The content of the respective fatty acid in the seed oil is presented on the right next to the plant species name. 5. Clicking on “Plants” in the pull-down menu will provide a table with plant species names. 6. Selecting the species name in the plant species table will show detailed information on the seed oil composition of this species.

3.3

Lipid Library

The Lipid Library database of the American Oil Chemists’ Society (AOCS) provides overviews on methods of lipid analysis, lipid biosynthetic pathways and oil processing techniques. For each chapter, a comprehensive list of references is provided. 1. Enter https://lipidlibrary.aocs.org/ in your browser. 2. The pull down menu “Lipid Analysis” provides protocols for silver ion chromatography (separation of acyl lipids according to the degree of unsaturation), NMR (1H and 13C), lipidomics (shotgun, MALDI MS, 2D GC) and selective topics of the analysis of lipids (solid-phase extraction, transesterification of fatty acids, lipid extraction, separation of chiral lipids, HPLC with evaporative light scattering detector (ELSD), fatty acid analysis by GC and HPLC, IR spectroscopy for analysis of the degree of unsaturation, acyl-glycerol analysis by LC-MS, fatty acid double bond analysis after adduct formation, countercurrent chromatography, phytosterol analysis by GC, tocopherol analysis by HPLC, TAG analysis by reversed-phase HPLC or GC, TLC). 3. The pull-down menu “Chemistry/Physics,” “Plant Lipid Biochemistry Pathways” provides the pathways for de novo fatty acid synthesis, unusual fatty acid production, acyl-CoA binding proteins (ACBPs), acyl-activating enzymes, TAG synthesis in plants and algae, oil droplets, TAG mobilization,

Plant Lipid Databases

447

transcription factors of oil accumulation, polyester and rubber synthesis, carotenoid biosynthesis, oxylipin synthesis, N-acylphosphatidylethanolamines (NAPEs), phosphoinositides, lipidomics, galactolipids, and phloem lipids. 4. The pull-down menu “Resource Material,” “Scientific Milestones in Lipid Research” presents historical highlights of lipid biochemistry including the development of gas–liquid chromatography. 5. Under the same pull-down menu (“Resource Material”), the menus “Trivial names of fatty acids Part 1 and Part 2” present a detailed description of fatty acid nomenclature. A table with the trivial and systematic names and abbreviated denominations, and a table with trivial names in different languages of the most important fatty acids are included. 3.4

Cyberlipid

The cyberlipid.org database, originally created by Claude Leray (formerly French National Centre for Scientific Research, Paris), was integrated into the web pages of the Groupe d’Etude et de Recherche en Lipidomique (GERLI) in 2019. The database provides comprehensive information on classification of lipids, protocols for extraction and analysis of lipids, lipid oxidation, an overview of literature and the history of lipid research. 1. Enter http://cyberlipid.gerli.com/ in your browser. 2. Click on the pull-down menu “Description of Various Lipids,” “Classification of Lipids.” Two types of lipids are described, simple lipids (containing one or two types of different compounds) and complex lipids (containing three or more compounds). 3. If you select “Simple Lipids,” the structures and occurrence of MAG, DAG, TAG, carotenoids, ceramides, fatty alcohols, hydrocarbons, quinone lipids (e.g., tocopherol), sterols, and wax esters are described. 4. Select “Complex Lipids” to show lipids that contain three or more different compounds (e.g., phosphoglycerolipids, phosphosphingolipids, glycoglycerolipids, glycosphingolipids). 5. Select “Analysis,” “Estimation of the total amount of lipids” to show protocols for the gravitometric or calorimetric determination of lipids. 6. Click on “Analysis,” “Extraction, handling of extracts” for methods on lipid extraction from various organisms using different techniques, including the Folch and Bligh and Dyer methods [15, 16]. 7. If you select “Analysis,” “Fractionation of lipid extracts,” protocols for solvent precipitation, liquid–liquid extraction, solid-

448

Peter Do¨rmann

phase extraction (column chromatography), HPLC with ELSD detector, and TLC are shown. 8. Chose “Analysis,” “Analysis of simple lipids” to show protocols for the analysis of simple acyl lipids, ceramides, sterols, and quinone lipids. 9. The analysis of complex lipids (phospholipids, glycoglycerolipids, glycosphingolipids) is shown under “Analysis,” “Analysis of complex lipids.” 10. The pull-down menu “Lipid oxidation” describes the mechanism and products of lipid oxidation. 11. An overview of the relevant literature on lipid analysis is presented under menu “References.” 3.5

LipidWeb

The LipidWeb database was established by William W. Christie (former Professor at the James Hutton Institute, Dundee, UK), and it has recently been replicated at the Lipid Maps web pages as a first step for full integration. The database provides information on lipid classification with structures and biosynthetic pathways and the lipid composition of different organisms. Furthermore, GC-MS mass spectra of important lipids and their derivatives are presented. 1. Open the LipidWeb at https://www.lipidhome.co.uk/index. html or https://www.lipidmaps.org/resources/lipidweb/. 2. Click on “Lipid essentials” and submenus to show definition, classification of lipids, including simple and complex lipids. 3. Select “Mass spectrometry of fatty acid derivatives” to show mass spectra of fatty acid methyl esters and other derivatives obtained by GC-MS. 4. Click on “Literature surveys” to present references on lipid biochemistry, organized by year of publication.

3.6 LipidBank Database

The LipidBank database was established and is maintained by the Japanese Conference on the Biochemistry of Lipids (JCBL) [9]. It contains more than 7000 entries for lipids including names, structures, physical, and chemical properties. 1. Go to: http://lipidbank.jp/. 2. The database can be searched for lipid names, formulas, and so on (left side). 3. Alternatively, select a lipid class and a specific lipid from the central menu. 4. Click on the lipid of interest. Detailed information on the name, structure, formula, spectra (IR, UV, NMR, mass, if available), occurrence in different organisms, chemical synthesis, and metabolism are presented. Furthermore, important publications are included.

Plant Lipid Databases

3.7

PubChem

449

The PubChem database provides structures, masses, formulas of small molecules (including lipids) and their NMR, GC-MS, and LC-MS spectra [11]. The database was set up and is maintained by the National Institutes of Health (NIH). 1. Select the https://pubchem.ncbi.nlm.nih.gov/ web page in your browser. 2. In the main search field, you can enter the name, formula, CAS number, or structure in SMILE or InChI formats (see Notes 2 and 3). 3. If you click on “Draw Structure,” a window opens where you can draw the structure and subsequently search for it. 4. The search results are presented as a table with the compound hits. Click on the compound of interest. 5. A window with detailed information on this compound will open, including structure, formula, synonyms, molecular mass, SMILE and InChI structure identifiers (see Notes 2 and 3), CAS number, chemical and physical properties. NMR spectra, mass spectra (GC-MS, LC-MS, LC-MS/MS), and UV and IR spectra are shown if available. 6. The window “Substances” shows the individually submitted entries for this compound. 7. Under “Literature,” the references referring to this compound are shown.

3.8

LipidomicNet

The LipidomicNet database was developed by the LipidomicNet consortium from 2007 to 2012. The database contains information on structures, biological function, and protocols for isolation and measurements of lipids. 1. Click on https://www.lipidomicnet.org/index.php/Main_ Page. 2. Select “Main Page” on the navigation panel on the left. 3. In the “Browse LipidomicNet-Wiki” window, you can search the entries of the database by different parameters, that is, “Lipid Class,” “Lipids as constituents of biological processes,” “Lipidomics technologies,” “Nomenclature of lipid species.”

3.9 NIST Chemistry WebBook

The NIST Chemistry WebBook harbors the NIST Standard Reference Database Number 69 and is maintained by the National Institute of Standards and Technology (NIST) [10]. It contains more than 70,000 data sets (e.g., IR, UV, MS spectra). 1. Enter the Internet address https://webbook.nist.gov/chemis try/ in your browser. 2. Use the “Search” pull-down menu to search the database for the formula, name, IUPAC identifier, CAS number, structure or mass of the compound of interest.

450

Peter Do¨rmann

3. Detailed information will be provided, including GC-MS spectra. 4. A searchable library including 350,704 electron ionization (EI) spectra (GC-MS spectra) (NIST/EPA/NIH EI-MS Library, 2020 release) is available through commercial suppliers (see Note 4). 5. Similarly, a searchable LC-MS/MS spectral library (NIST Tandem Mass Spectral Library, 2020 release) can be obtained from commercial suppliers (see Note 4). 3.10

LipidHome

The LipidHome database of theoretical lipid structures and masses was developed by scientists at the EMBL Outstation/European Bioinformatics Institute and The Babraham Institute at Cambridge, UK, in association with the LipidomicNet consortium (see Subheading 3.8) [17]. The database contains lipid structures and masses calculated from the theoretical combination of different fatty acids and head groups. The lipid structures are limited to glycerolipids (MAG, DAG, TAG) and phospholipids. The database can be searched for precursor ion masses obtained in lipidomics (LC-MS) experiments. 1. Go to the web page: https://www.ebi.ac.uk/metabolights/ lipidhome. 2. Click on “Browser,” “Glycerolipids,” or “Glycerophospholipids” to see the structures and molecular species of different glycerolipids (MAG, DAG, TAG) or glycerophospholipids. 3. Select “Tools.” A new window opens. Enter masses of precursor ions and select the mass tolerance and possible adducts. Click on “Submit.” 4. The “Output” window shows the hits of molecular lipid species. You can select/deselect different glycerolipids and glycerophospholipids to change the filter. 5. For a new search, select the “Input” window. 6. Click on “Documentation” for help and background information.

3.11 Lipid Maps Lipidomics Gateway

The Lipid Maps Lipidomics Gateway represents an entry web page for different databases focusing on structures, masses, formulas of lipids, mass spectra, genes, pathways, software tools, and lipidomic protocols. The Lipid Maps database was created and is maintained by the Lipid Maps Consortium [6–8]. The Lipid Maps Lipidomics Gateway also includes a link to the LipidWeb page (see Subheading 3.5). 1. Enter the web page https://www.lipidmaps.org/ in your browser.

Plant Lipid Databases

451

2. Click on “Resources,” “Classification” to open the Lipid Classification System with the Lipid Categories. Click on any lipid category (e.g., “Fatty Acyls’) to show a table of the compound class including common and systematic names, formulas, and masses. If you select the Lipid Maps Identifier (LM_ID), detailed information including the structure of the compound will be shown. 3. Click on “Resources,” “Lipid Standard” to show lipid standards suggested for lipidomics experiments. 4. Select “Resources,” “Methods and Protocols” to show the methods organized by lipid category. 5. Select “Resources,” “Databases” to show a window with the available databases. 6. Go to “Lipid Maps Structure Database” (LMSD) to show information on more than 40,000 unique lipid structures. You can browse or search the LMSD database in various ways by clicking on “Browse/Search/Download LMSD.” 7. In the database window, click on “Lipid Maps Gene/Proteome Database” (LMPD) to show information on over 8500 lipidrelated genes and over 12,500 proteins from animals, Escherichia coli and Arabidopsis thaliana. Continue by clicking on “Browse/Search/Download LMPD.” 8. Select “Lipid Maps In-Silico Structure Database (LMISSD)” to search the database containing lipid structures obtained by computational expansion of head groups and side chains. 9. By clicking on “Resources,” “Pathways,” The Lipid Maps WikiPathways are presented. 10. Click on “Resources,” “Tools,” to see different online tools available. 11. Select “Mass Spectrometry Analysis Tools” (MS Analysis) to see tools for mass spectra analysis. 12. Click on “Multiple Lipid Classes,” “Search a computationallygenerated database. . .” for bulk structure searches of compounds for LC-MS experiments without fragmentation (precursor ions). 13. Click on any lipid class to search for specific lipid classes or predict MS/MS spectra. 3.12

Metlin Database

The Metlin database for LC-MS/MS spectra has been developed at the Scripps Research Institute (La Jolla, CA) since 2003 [14]. It now includes data on >1,000,000 small molecules. The database contains MS/MS spectra for over 500,000 compounds derived from LC-MS/MS experiments with different collision energies in positive or negative mode.

452

Peter Do¨rmann

1. Go to the web page: https://metlin.scripps.edu. 2. Register, then sign up with your email and password. 3. Go to the Metlin database (pull-down menu). 4. Select “Simple Search” from the pull-down menu to search the MS masses (without fragmentation) in the positive or negative mode. You can preselect the mass tolerance (ppm), the formation of different adducts, elemental composition etc. Remove “Peptides,” “Toxicants,” and “Drugs” if you aim to search for lipid molecules. 5. The result page will show a table with the exact masses of the adducts, names, formulas, structure, and MS/MS spectra if available. If you click on the Metlin ID number, the KEGG ID, CAS number, and PubChem numbers of the compounds are presented. 6. Click on “Advanced Search” to search using the Metlin ID (MID), the structure (in SMILE format; see Note 2), a mass range, and the sum formula of your compound. You can also preselect element composition and restrict your search to entries with MS/MS spectra. 7. The MS/MS spectra were mainly obtained by ESI Q-TOF mass spectrometry in positive or negative mode and with different collision energies. 8. Select “Batch Search” if you want to search for several compounds at a time. 9. In “Fragment Similarity Search,” you can search for fragments of a spectrum obtained in MS/MS experiments. 10. In analogy, in “Neutral Loss Search,” matches for a neutral loss of MS/MS experiments can be retrieved. 11. Select “MS/MS Spectrum Match Search” to enter the precursor m/z and the fragment masses (m/z). In addition, you need to select collision energy, the mass tolerances, positive/negative mode, and adduct formation.

4

Notes 1. The Delta (Δ) notation depicts each fatty acid by a computer readable abbreviation: The fatty acid chain is presented with X:Y, where X is the number of carbon atoms and Y the number of double bonds. Cis and trans are abbreviated with “c,” “t,” respectively, and a triple bond is shown by “a” (acetylene). Numbers after ‘delta’ indicate the position of the double bond. Functional groups are indicated by –O– (epoxy), –OH (hydroxyl), –O¼(keto), –cpe– (cyclopropene), and –cpa– (cyclopropane). See http://sofa.mri.bund.de/.

Plant Lipid Databases

453

2. The Simplified Molecular Input Line Entry System (SMILE) represents a format for the abbreviation of a molecular structure in a computer readable way. Atoms are represented by C, O, N, P, S, while H is omitted. Single bonds are omitted, and double and triple bonds are represented by ¼ and #, respectively. Branches are indicated in parenthesis. See https:// archive.epa.gov/med/med_archive_03/web/html/smiles. html. 3. Similar to SMILE, the International Chemical Identifier (InChI) denotes chemical structures of compounds in an abbreviated format. Each standard InChI starts with “InChI ¼ 1S/” followed by the “layers” and “sub-layers”separated by “/”. The layers start with the formula, followed by the connection of the atoms (numbered, with prefix c), then the hydrogen atoms (prefix “h”) with their position on the different carbon atoms. See Heller et al. [18]. 4. These mass spectral libraries are not available online but can be obtained in combination with the GC-MS or LC-MS analysis software from commercial suppliers. See: https://www.nist. gov/programs-projects/nist20-updates-nist-tandem-and-elec tron-ionization-spectral-libraries, https://chemdata.nist.gov/ dokuwiki/doku.php?id¼chemdata:distributors. References 1. Liebisch G, Vizcaı´no JA, Ko¨feler H et al (2013) Shorthand notation for lipid structures derived from mass spectrometry. J Lipid Res 54 (6):1523–1530. https://doi.org/10.1194/jlr. M033506 2. Fahy E, Subramaniam S, Brown HA et al (2005) A comprehensive classification system for lipids. J Lipid Res 46(5):839–862. https:// doi.org/10.1194/jlr.E400004-JLR200 3. Ohlrogge J, Thrower N, Mhaske V et al (2018) PlantFAdb: a resource for exploring hundreds of plant fatty acid structures synthesized by thousands of plants and their phylogenetic relationships. Plant J 96(6):1299–1308. https:// doi.org/10.1111/tpj.14102 4. Mattha¨us B (2012) The new database Seed Oil Fatty Acids (SOFA). Lipid Technol 24 (10):230–234. https://doi.org/10.1002/lite. 201200227 5. Aitzetmu¨ller K, Mattha¨us B, Friedrich H (2003) A new database for seed oil fatty acids—the database SOFA. Eur J Lipid Sci Technol 105(2):92–103. https://doi.org/10. 1002/ejlt.200390022 6. Fahy E, Cotter D, Sud M et al (2011) Lipid classification, structures and tools. Biochim

Biophys Acta 1811(11):637–647. https:// doi.org/10.1016/j.bbalip.2011.06.009 7. Sud M, Fahy E, Cotter D et al (2012) LIPID MAPS-Nature Lipidomics Gateway: an online resource for students and educators interested in lpids. J Chem Educ 89(2):291–292. https://doi.org/10.1021/ed200088u 8. Sud M, Fahy E, Cotter D et al (2007) LMSD: LIPID MAPS structure database. Nucleic Acids Res 35(Database):D527–D532. https://doi.org/10.1093/nar/gkl838 9. Watanabe K, Yasugi E, Oshima M (2000) How to search the glycolipid data in “LIPIDBANK for Web” the newly developed lipid database in Japan. Trends Glycosci Glycotechnol 12 (65):175–184. https://doi.org/10.4052/ tigg.12.175 10. Linstrom PJ, Mallard WG (2001) The NIST Chemistry WebBook: a chemical data resource on the internet. J Chem Eng Data 46 (5):1059–1063. https://doi.org/10.1021/ je000236i 11. Kim S, Thiessen PA, Bolton EE et al (2016) PubChem substance and compound databases. Nucleic Acids Res 44(D1):D1202–D1213. https://doi.org/10.1093/nar/gkv951

454

Peter Do¨rmann

12. Li-Beisson Y, Shorrosh B, Beisson F et al (2013) Acyl-lipid metabolism. Arabidopsis Book 22:e0161. https://doi.org/10.1199/ tab.0161 13. Aoki KF, Kanehisa M (2005) Using the KEGG database resource. Curr Protoc Bioinformatics Chapter 1:Unit 1.12. https://doi.org/10. 1002/0471250953.bi0112s11 14. Benton HP, Wong DM, Trauger SA et al (2008) XCMS2: processing tandem mass spectrometry data for metabolite identification and structural characterization. Anal Chem 80 (16):6382–6389. https://doi.org/10.1021/ ac800795f 15. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37(8):911–917

16. Folch J, Lees M, Sloane Stanley GH (1957) A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 226(1):497–509 17. Foster JM, Moreno P, Fabregat A et al (2013) LipidHome: a database of theoretical lipids optimized for high throughput mass spectrometry lipidomics. PLoS One 8(5):e61951. https://doi.org/10.1371/journal.pone. 0061951 18. Heller SR, McNaught A, Pletnev I et al (2015) InChI, the IUPAC international chemical identifier. J Cheminform 7:23. https://doi. org/10.1186/s13321-015-0068-4

Chapter 26 Lipid Pathway Databases with a Focus on Algae Naoki Sato and Takeshi Obayashi Abstract Pathways of lipid biosynthesis are highly complex and have been established in model organisms such as Arabidopsis thaliana and Chlamydomonas reinhardtii, whereas in other organisms, we need bioinformatic tools to map individual enzymes onto reference pathways. In this chapter, we explain representative tools that are useful in identifying algal orthologs of lipid biosynthetic enzymes and finding new enzymes that are possibly involved in the pathway of interest. All descriptions in this chapter refer to in silico (i.e., computerbased) methods rather than laboratory experiments. Key words Lipid pathway, Algae, Bioinformatics

1

Introduction Algal lipids are the focus of numerous biotechnological studies, because algae are potential resources of useful compounds and biofuels. The term algae is used here to denote both cyanobacteria and eukaryotic algae. Algae are phylogenetically diverse, and the life styles and living environments are highly variable. Full exploitation of the productive capability of algae requires complete knowledge of the metabolic pathways of “useful” compounds including complex lipids, fatty acids, and pigments, among others. The engineering of algal production consists of “push” and “pull” strategies. The “push” strategy enhances the basic productive power by increasing the activity of the pathway leading to the accumulation of the product of interest, and the “pull” method powers up the effective production of the compound of interest. The “push” strategy is based on biosynthetic pathways common in various plants and algae, whereas the “pull” strategy requires a search for useful enzymes in specific organisms. Bioinformatic tools support both types of strategies by providing information on orthologs and potential homologs with a new function, respectively. In this chapter, we will explain the methodology of identifying useful orthologs

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7_26, © Springer Science+Business Media, LLC, part of Springer Nature 2021

455

456

Naoki Sato and Takeshi Obayashi

of relevant enzymatic steps, potential homologs, as well as co-expressed enzymes in related pathways. Here, the term homologs is used to denote genes or proteins in different organisms, which are “similar” in their nucleotide or amino acid sequences, respectively. The similarities of two genes or proteins are quantitatively assessed by alignment and scoring, and represented by a similarity score and an E-value (index of expectation value assuming random match) after homology search, conveniently performed by BLAST (Basic Local Alignment Search Tool) search in different bioinformatic websites. Two sequences are more similar if the similarity score is higher or the E-value is lower. The term homolog was originally defined for genes, but is currently also applied to proteins. There are two types of homologs: ortholog and paralog. Two homologs are orthologs (in two different species) if they are descendants of a common ancestral gene. Paralogs typically result from gene duplication within a species. In practice (at least in bioinformatics), orthologs are defined by bidirectional best hit (BBH): namely, in an ortholog pair, the ortholog A in an organism X is the most similar among all genes in X with the ortholog B in another organism Y as a query, and vice versa. The definition of ortholog is not simple, but we usually understand that orthologs have an identical function in different organisms, even though orthologs defined by BBH could have somewhat different functions. Algae represent diverse organisms, more diverse than land plants. The genome of the red alga Cyanidioschyzon merolae was sequenced in 2004 [1]. Subsequently, genome sequences of two diatoms [2, 3] and two green algae [4, 5] have been reported. Currently, many other algae of diverse phyla are sequenced, and the genomic data are available from the Joint Genome Institute (JGI) and the Ensembl database of European Bioinformatic Institute (EBI), among others (see Table 1). In addition, sequencing of laboratory strains and useful isolates employing commercial sequencing services has become easy, due to developments in next generation sequencing (NGS). The purpose of this chapter is to present tools on how to infer pathways of lipid biosynthesis in strains for which genome information has been obtained that are useful in, for example, oil production. In this chapter, we provide a point-by-point manual for the bioinformatic identification of orthologs and related genes starting from known pathway information. In general, a strategy including the following steps can be applied: 1. A general starting point to infer pathways in algal strains for which genome sequence information has recently been obtained is the knowledge of enzymes in model algae, such as Chlamydomonas reinhardtii and Cyanidioschyzon merolae. For lipid studies, Arabidopsis thaliana could be another starting point.

Algal Lipid Pathway Databases

457

Table 1 Useful databases in algal lipid pathway analysis Subheadin Content

Name

URL

Reference

3.1, 3.2

General genome information

NCBI

https://www.ncbi.nlm.nih.gov/

–

3.1

General genome information

EMBL-EBI

https://www.ebi.ac.uk/

–

3.1

General genome information

JGI genome portal

https://genome.jgi.doe.gov/ portal/

–

3.1

General genome information

DDBJ

https://www.ddbj.nig.ac.jp/

–

3.1, 3.3

Genomes of Eukaryotic Algae

EnsemblePlants

http://plants.ensembl.org/

–

3.1, 3.3

Genomes of Prokaryotic Algae

EnsembleBacteria https://bacteria.ensembl.org/

–

3.1

General algal information

AlgaeBase

https://www.algaebase.org/

–

3.1

Plant and algal genome information

Phytozome

https://phytozome.jgi.doe.gov/ pz/portal.html

[6]

3.1, 3.5

Arabidopsis AcylLipid Metabolism

ARALIP

http://aralip.plantbiology.msu. edu/pathways/pathways

[7]

3.3

Cyanobacterial database

CyanoBase

http://genome.microbedb.jp/ cyanobase/

[8]

3.3

Microbial Genome Database

MBGD

http://mbgd.genome.ad.jp/

[9]

3.3

Protein families

Pfam

http://pfam.xfam.org/

–

3.4

Orthologs/homologs Gclust

http://gclust.c.u-tokyo.ac.jp/

[10]

3.5

Carbohydrate-active enzymes

CAZy

http://www.cazy.org/

–

3.6

Metabolic pathway

KEGG

http://www.genome.jp/kegg/

[11]

3.7

Coexpression

ALCOdb

http://alcodb.jp/

[12]

4

Protein–protein interaction

STRING

https://string-db.org/

[13]

4

Genome annotation

DFAST

https://dfast.nig.ac.jp/

[14]

“–” in reference indicates “see the web site,” because these are very common tools

2. Similarity search with BLAST tools. 3. Ortholog search with comparative genomic tools. 4. Mapping of orthologs in pathways. 5. Identification of coexpressed genes.

458

2

Naoki Sato and Takeshi Obayashi

Materials All manipulations in this chapter are performed with a computer connected to the internet. We ask the readers to follow the protocols by working with their own computers to better understand the description provided here. All methods in this chapter can be performed using any type of browser software, and they are not dependent on the different operating systems, such as macOS (Apple) or Windows OS (PC compatible). Useful databases are listed in Table 1. It includes general genome information, as well as databases of metabolic pathways, orthologs/homologs, protein families, and coexpression. References and web page addresses (uniform resource locators, URLs) are also shown in Table 1.

3

Methods

3.1 General Comments

A common starting point for pathway search and pathway construction includes the identification of enzymes of interest, such as lipid biosynthesis enzymes. We are usually not able to obtain sequence information directly from textbooks, but instead from databases. We can readily obtain sequence information from the NCBI (National Center for Biotechnology Information) or EBI websites. General genome databases are listed in the first part of Table 1. However, the data retrieved from general genome databases could be sometimes ambiguous, for example, because of the presence of many homologs in a single organism. Therefore, to obtain more reliable information on enzymes involved in plant/ algal lipids, use specialized databases such as Phytozome [6], EnsemblPlants, EnsemblBacteria, and ARALIP [7]. We advise the readers to obtain descriptions from different websites and compare them. Many annotations in sequence databases are based on inferences relying on similarity information, but do not represent experimentally confirmed results. Identification of orthologs or functionally identical enzymes depends on sophisticated analysis using different tools, but the final confirmation requires experimentation. Bioinformatics is just an entrance point to functional studies. The database AlgaeBase is sometimes useful in obtaining background information on an algal strain of interest, including strain history and major characteristics.

3.2 General Homology Search for Identification of Orthologs: NCBI BLAST

A basic search for orthologs is simply performed by web-based BLAST services, such as the one available in the NCBI website. This is commonly performed in any laboratory, but some precautions need to be taken. In this section, we explain methods of homology search using NCBI BLAST. Similar tools are available at the EBI and DDBJ databases.

Algal Lipid Pathway Databases

459

1. Use a browser to access the NCBI home page. 2. Open the BLAST search panel by clicking on the “BLAST” link in the “Popular Resources” on the right. 3. Click “Protein BLAST” panel, and then a standard protein BLAST query suite will open. 4. Paste your amino acid query sequence into the top text box under “Enter Query Sequence”. Multiple queries can be used. In this case, you will have to fill the text box with query sequences in FASTA format. Alternatively, upload a text file of query sequences in FASTA format. 5. “Database” may be “non-redundant protein sequences (nr)” for most searches. To limit the range of organisms to search, use “Organism” selection. A species name of organism or a phylum name may be put into the text box, and then, candidate items will appear. Just choose one of them. You may “exclude” an organism (or a phylum) by checking the small box. You may increase the number of your selection by clicking “+” (plus) button on the right. 6. In most cases, just clicking the large “BLAST” button located below will be enough to start a search. 7. In very exceptional cases, you have to open the “Algorithm parameters” panel by clicking “+” button below the large “BLAST” button. “Expect threshold” (default ¼ 10) may be as small as 1e20 for searching orthologs. In an ortholog search for common enzymes, you may leave “Filters and Masking” unselected. Touching other parameters is not recommended. After setting appropriate parameters, just click either of the two large “BLAST” buttons. 8. A temporary window will be displayed until the results window will open. Four different types of results may be selected. “Descriptions” is a list of homologs. “Graphic summary” shows a graphical representation of homologous regions. “Alignments” shows a sequence alignment of the query and the target. “Taxonomy” presents taxonomical distribution of detected homologs. In most cases, you have to exclude identical sequences by carefully examining the %identity values. 9. NCBI BLAST usually detects too many homologs. We recommend limiting the range of organisms or setting a lower E-value threshold in the second run of the search, once you evaluated the initial search results. 3.3 Detection of Orthologs in CyanoBase, MBGD, Ensembl, and Pfam

Ortholog information of cyanobacteria has been assembled in CyanoBase (cyanobacterial database), originally developed by the Kazusa DNA Research Institute, but now hosted by the National Institute of Genetics, Mishima, Japan [8]. Information on

460

Naoki Sato and Takeshi Obayashi

microorganisms, mostly prokaryotes, is available at MBGD (Microbial Genome Database), in the National Institute of Basic Biology, Okazaki, Japan [9]. These databases will be useful in finding prokaryotic homologs of algal or plant enzymes. Information specific to plants or bacteria is obtained at the Ensembl database implemented in EBI. Pfam is a protein family database, and provides domain structures of proteins, which is helpful in detecting orthologs in different organisms. 3.4 Ortholog Finding by Genome Clustering: Gclust

Gclust is a comparative genomics database created by one of the authors (NS), and presents homolog clusters within preselected major genomes [10]. It was constructed to find orthologs in photosynthetic organisms, and will be useful in reconstructing a metabolic pathway in an organism based on a pathway in another organism. 1. Access the website with a browser. 2. Click “Start from here! Select Dataset” button on the left. 3. On the next page, Gclust and Cyanoclust datasets are presented. Check one of the radio buttons to select a dataset. Gclust2012_42 is recommended as an initial trial. 4. “Search Menu” provides various ways of data search in “Basic Search.” For a search for orthologs, you might or might not have a seed (or query) sequence. If you have a seed sequence to find orthologs, use the BLAST function available at the top of the page. Follow step 5. If you do not have a seed sequence, you might want to obtain a homolog cluster by a single search. Follow step 7. 5. Paste your query sequence into the text box. For a protein query, use the program “blastp” (default) above the box. For a nucleotide query, select “blastx.” The default value of “Expect” is “1e2”. This could be lower, such as “1e20”. Leave other options, and click “Run BLAST” button above the text box. After a while, results will be displayed. 6. A list of many homologs is shown in the results. Many of the top-ranking homologs could belong to an identical homolog cluster, but this is not clear at this point. Check the cluster number by clicking the name of the protein on the left. If all of the top-ranking proteins belong to an identical cluster, this is the cluster of your choice. If several different clusters appear, then you will have to check all of them to evaluate the members if they are orthologs that you are trying to find. In both cases, proceed to step 8. 7. To find a protein cluster of your interest, use “annotations” or “sequence ID” in “Search Menu.” To use one of these search modes, click a radio button on the left representing a function

Algal Lipid Pathway Databases

461

name, then, fill in the text box on the right of the respective function name. The text box of “annotations” does not accept quotation marks. Only a very simple text search is possible. You will be able to select appropriate clusters in the next panel. “Sequence ID” is sometimes useful, if you know the gene name. To search homologs of atpA gene product, write “*atpA” in the text box. An asterisk “*” is used as a wild card corresponding to any characters. In the Gclust data, a gene name is attached to the locus tag for easy identification of protein function. To perform a search, click “submit” button. 8. “Search Results” presents a list of clusters. Click “Cluster Number” to see cluster members. 9. A square table will appear if the number of members is not very large. For a very large cluster, only a list of homologs will appear, and the following explanation of the table is not applicable. Protein names are shown both on the left and above the table. The names above the table are color-coded according to organism groups defined in the “Basic Search” menu. The red shading in the table with “0” and “1” indicates the similarity of proteins. 10. On the right of the list or table, you will find annotations in the original database which the sequence data were obtained from. Most of the annotations might not be informative, but some of them will indicate a protein function. If an identical annotation appears for several items, then this annotation is likely to be reliable. 11. Below the list or table, you will find two or three buttons: “Related Groups” (if present), “ClustalW,” and “Get All Sequences.” Since ortholog clustering always leaves some homologies that are not used for clustering, “Related Groups” provides such minor homologies. Some of the related groups could be orthologs that have subtle structural differences. “ClustalW” button will show a multiple alignment, which is precalculated (for large clusters) or calculated upon request. A “Start Jalview” button is available on top for only old browsers. A Java applet will show a colored alignment and a phylogenetic tree. This is not available in new macOS or Windows 10, in which Java is no longer integrated into the browser. Instead, you will be able to perform phylogenetic analysis after retrieving all member sequences by the “Get All Sequences” button. 3.5 Metabolite-Oriented Databases: CAZy and ARALIP

CAZy (carbohydrate active enzymes) is an encyclopedia of sugarrelated enzymes, including glycosylhydrolases and glycosyltransferases. Enzymes are classified into many families, such as GT28 for A. thaliana monogalactosyldiacylglycerol synthase, and GT4 for digalactosyldiacylglycerol synthase. The website provides

462

Naoki Sato and Takeshi Obayashi

comprehensive information on each enzyme family. All enzymes of a certain glycosyltransferase family share the same anomeric reaction mechanisms; for example, all GT28 family members show an inverting mechanism (producing β anomeric bonds), and all GT4 members are retaining (producing α anomeric bonds). A large list of family members in each organism group is provided, with links to GenBank (www.ncbi.nlm.nih.gov/genbank), UniProt (universal protein resource; www.uniprot.org), and PDB (protein data bank; www.rcsb.org) entries. The website is easy to use, and will be useful for glycolipid researchers. ARALIP (Arabidopsis acyl-lipid metabolism pathway) provides a way of transferring information on lipid pathways from plants to algae [7]. The web interface is self-explanatory, and easy to use, but the information is limited to A. thaliana. 3.6 Pathway Analysis with KEGG

3.6.1 Inferring Pathways in Organisms with Sequenced Genomes

KEGG (Kyoto Encyclopedia of Genes and Genomes) provides information on genes, genomes, and chemicals used in medicine, but the unique function is pathway analysis [11]. Almost all important biochemical pathways are implemented. Enzymes and metabolites can be found that are probably present in model organisms whose genomes have been sequenced. This tool is useful to find metabolic pathways in nonmodel organisms, if the genome sequence has been determined. 1. Access the KEGG website with a browser. 2. Click “KEGG PATHWAY”. 3. Select one of the pathways (metabolism, genetic information processing, environmental information processing, cellular processes). Many of the other menu items are specific to human or animal genomes. For lipid biochemists, “1.3 Lipid metabolism” may be a section of interest. 4. We choose “Glycerolipid metabolism,” as an example. 5. A large metabolic map is presented. Each small circle represents a metabolite. Each arrow corresponds to an enzymatic reaction. The EC number of each enzyme is shown in a rectangle. Neighboring pathways are also shown in rectangles with rounded corners. Any of these (metabolites, enzymes, pathways) are clickable. But note that this is a reference pathway; namely, this is an artificially assembled pathway from all known pathways involving the metabolites of interest. 6. You may find a flattened button (“Reference pathway”) on top of the window. You can select any model organism from this menu button. In the default settings, the names of the organisms are arranged in an unknown order, and it is very difficult to select the favorite organism. The organism names may be sorted by selecting “Sort below by alphabet” (you have to

Algal Lipid Pathway Databases

463

click “Go”). But, if you type the first two or three characters of the organism name, then a candidate name will appear, and you will be able to select it. Here, let us choose “Chlamydomonas reinhardtii” as an example. Then you click the “Go” button. Alternatively, you may choose an organism from the “Organism menu” on the topmost list of menus. 7. Then, the enzymes that are inferred to be present in the organism are highlighted. You may realize that only a part of the pathway is functional in this alga. Frequently, a single enzyme is missing in a pathway. This could be the result of incomplete gene annotation due to less conserved sequences. Alternatively, the reaction catalyzed by the missing enzyme is catalyzed by an unknown or nonhomologous enzyme. To address this question, you might need to search for homologs using known enzymes in the nucleotide sequence database by TBLASTN or other tools. If you are sure that the missing enzyme is really absent, then you have a chance to discover a new enzyme. 8. If you click the EC number, then you will find detailed information (including amino acid and nucleotide sequences) on the enzyme of the organism that you selected. If you click the EC number in the Reference pathway, then you will find a list of homologs in various model organisms in “Genes” box. This will help to identify orthologs in different organisms. 9. Another way of search starts from the first page of the KEGG website. You just type the name of the enzyme or metabolite in the box on top of the page and click the “Search” button. Next, you will find a long list of items provided by the KEGG system. 3.6.2 Mapping a Sequence of Interest onto a Pathway

1. In the top menu of KEGG, click the “BLAST” link within the “BLAST/FASTA” field near the bottom of the page (Note that BLAST and FASTA are different links). 2. Paste your sequence of interest (either protein or nucleotide) in the box “Sequence data.” Then select a radio button, “BLASTP” or “BLASTX,” respectively. Select “KEGG GENES,” and then click the “Compute” button located on top. 3. A list of homologous proteins will be shown. You may choose the top ten or more sequences, and select “Map Genes to KEGG Ortholog Cluster” from the “Select operation” menu button. Then click the “Exec” button. 4. A page entitled “OC Viewer” (Ortholog Cluster Viewer) will appear. If you choose an OC number displayed as a BLAST Search Result, a list of orthologs will be shown. This may be a large list, and useful in constructing a phylogenetic tree. 5. You will find the “KOs in cluster” link in the middle of the page (KO stands for KEGG orthology). The “K number” identifies

464

Naoki Sato and Takeshi Obayashi

the enzyme function. If you click the “Pathway search” link, then a list of pathways involving the enzyme of interest will appear. Clicking “ko number” (“ko” is in lowercase here) in the list, then the pathway map will be shown with the enzyme of interest highlighted. For sequences that have not yet been mapped to an enzymatic step, no “KOs in cluster” can be presented. 3.6.3 Detection of a Gene of Interest in Newly Sequenced Genomes

KEGG provides a batch search service to identify putative functions of proteins in newly sequenced genomes. They are available under the heading “Genome analysis” of the main home page of the KEGG. 1. “KofamKOALA” is an HMM (hidden Markov model) search for the KO number in the KEGG Orthology database. To use this function, open the KofamKOALA page, and paste your query sequence in the text box or upload a sequence file. Do not forget to enter “>comment” in the headline (sequence in FASTA format). Entering your E-mail address is mandatory. Then, click the “Compute” button. 2. An E-mail requesting your confirmation will be sent. Click the link to submit your job. 3. Another E-mail will be sent after a while when the job is completed. Click the link to open the search result page. You will find the KO number with enzyme definition. By following the link of “KEGG Mapper,” you will be able to map your protein of interest in the KEGG pathway map. 4. “KAAS” (KEGG automatic annotation server; www.genome. jp/kegg/kaas) is a much more powerful service that provides an automatic annotation of many proteins encoded by a new genome. In the home page of KAAS, several different options are available: namely, complete genome, partial genome, or metagenome. For all of them, you will need protein sequences in FASTA format. Fill in the sequence box and E-mail address field. You will have to select the appropriate data set in the GENES database. Then, click the “Compute” button. 5. Upon the arrival of an E-mail, click the link to submit your job. A web page will open, showing the URL of your results. You will also receive an E-mail describing the same URL. 6. When another E-mail confirms the completion of the job, click the link to open the result page. By clicking the “html” or “text” link under the “result” header, you will find a list of KEGG maps. If you click the pathway ID, then you will find a metabolic map with your query protein(s) highlighted. This is an easy way of constructing metabolic maps encoded by the newly sequenced genome.

Algal Lipid Pathway Databases

3.7 Coexpression Analysis with ALCOdb

3.7.1 Retrieval of Coexpressed Genes with a Single Guide Gene

465

Another approach to find gene-to-gene functional associations is using gene coexpression information, which is based on similar gene expression profiles across a large number of cellular conditions. ALCOdb (Algae Gene Coexpression database) was developed by one of the authors (TO), and provides gene coexpression information for Chlamydomonas reinhardtii and Cyanidioschyzon merolae [12]. 1. Access the ALCOdb website with a browser. 2. Click “Data” in the menu bar, and a search page will open. Three ways of searches are available to select a guide gene. “Keyword Search” accepts a gene ID, a gene name, or a keyword. A nucleotide or amino acid sequence is used in “Sequence Similarity Search.” “Select From Available Gene List” provides a list of gene IDs for each coexpression platforms called, Cre, Cme1, and Cme2. As an example, here, we focus on MGDG synthase in Chlamydomonas reinhardtii. Let us fill “monogalactosyl” in the box of Keyword Search. 3. In the matched gene list, one gene (Cre13g585301) is found, which is one Chlamydomonas homolog of the A. thaliana MGDG synthase gene. To retrieve a coexpressed gene list for this gene, click the icon of this gene in the “Information” column (http://alcodb.jp/coexpression/Cre/ Cre13g585301/list). You may locate the “Coexpressed Gene List” icon by passing a mouse pointer over the icons. 4. In ALCOdb, the degree of coexpression is represented as MR (mutual rank) index, which is the geometric average of the bidirectional correlation ranks between a gene pair. Therefore, MR ¼ 10 roughly indicates a gene-to-gene relationship in a gene module composed of ten genes. A larger value of MR indicates a lower similarity of coexpression. In the coexpressed gene list for Cre13g585301, the top three coexpressed genes are Cre12g489050 (phosphatidate cytidylyltransferase), Cre11g467723 and Cre04g216950 (3-ketoacyl-ACP synthases I and III) with MR values of 1.0, 5.9, and 6.7, respectively. All three genes are involved in lipid biosynthesis. 5. In this list, coexpression information in other species is also provided in the right columns to infer evolutionary conservation. In the current example, the coexpression in Chlamydomonas reinhardtii is not conserved in A. thaliana (MR values >1000), suggesting a species-specific association between MGDG synthesis and enzymes of its upstream lipid metabolism. 6. To change the guide gene in the gene list, click the Gene ID. For example, clicking the second coexpressed gene, Cre11g467723 (3-ketoacyl-ACP synthase I), will present a

466

Naoki Sato and Takeshi Obayashi

coexpressed gene list with this gene. Unlike the case of the MGDG synthase gene, the coexpression profile with the 3-ketoacyl-ACP synthase I gene is conserved in Chlamydomonas reinhardtii and A. thaliana, but not in Cyanidioschyzon merolae. 7. To draw a coexpressed gene network around the guide gene, click the “Coexpressed Gene Network” button in the top area of this page. 8. To draw a gene network, the coexpression threshold (MR) should be specified. The bar on the right may be used to set a threshold for identifying an edge to find an appropriate network structure. The threshold value may be varied by dragging the slider, but fine tuning may be performed by arrow buttons in the keyboard after clicking the slide bar. Genes with a particular function may be highlighted by selecting a pathway below the bar. 9. The position of each gene may be displaced by manually dragging its icon. You may thus draw a more comprehensive network map. The network map can be downloaded by first clicking the “Convert to PNG” button at the bottom and then clicking the new button “Save PNG File.” 3.7.2 Drawing a Coexpressed Gene Network for a Set of Genes

ALCOdb also provides a network-drawing function for a set of selected genes. This is useful for the functional classification of genes in a gene family or of differentially expressed genes. 1. Access the ALCOdb website with a browser. 2. Select “Network Analyzer” in the “Tools” pull-down menu in the top menu. 3. Fill the “Query Form” with the genes of interest, or click the “Example” button. Then, click the “Submit” button. 4. Check the list of available and nonavailable genes in the tables. Only “Available Genes” are used to construct a network. Click the “Draw Network” button. 5. An interactive gene network will appear for the set of query genes. You may change the threshold and arrangement of genes, and highlight a pathway, as explained in Subheading 3.7.1.

Algal Lipid Pathway Databases

4

467

Notes The tools that we explain in this chapter are only a small part of all available bioinformatics tools. Algal information is not commonly implemented in many tools. The tools that we present here are especially useful in studying algal lipid research. Other databases that may be useful in constructing pathways are included in Table 1, for example, the STRING and DFAST databases on protein–protein interaction and genome annotation, respectively [13, 14].

References 1. Matsuzaki M, Misumi O, Shin-i T, Maruyama S, Takahara M, Miyagishima S, Mori T, Nishida K, Yagisawa F, Nishida K, Yoshida Y, Nishimura Y, Nakao S, Kobayashi T, Momoyama Y, Higashiyama T, Minoda A, Sano M, Nomoto H, Oishi K, Hayashi H, Ohta F, Nishizaka S, Haga S, Miura S, Morishita T, Kabeya Y, Terasawa K, Suzuki Y, Ishii Y, Asakawa S, Takano H, Ohta N, Kuroiwa H, Tanaka K, Shimizu N, Sugano S, Sato N, Nozaki H, Ogasawara N, Kohara Y, Kuroiwa T (2004) Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428:653–657. https://doi.org/10.1038/nature02398 2. Armbrust EV, Berges JA, Bowler C, Green BR, Martinez D, Putnam NH, Zhou S, Allen AE, Apt KE, Bechner M, Brzezinski MA, Chaal BK, Chiovitti A, Davis AK, Demarest MS, Detter JC, Glavina T, Goodstein D, Hadi MZ, Hellsten U, Hildebrand M, Jenkins BD, Jurka J, Kapitonov VV, Kro¨ger N, Lau WW, Lane TW, Larimer FW, Lippmeier JC, Lucas S, Medina M, Montsant A, Obornik M, Parker MS, Palenik B, Pazour GJ, Richardson PM, Rynearson TA, Saito MA, Schwartz DC, Thamatrakoln K, Valentin K, Vardi A, Wilkerson FP, Rokhsar DS (2004) The genome of the diatom Thalassiosira pseudonana: ecology, evolution, and metabolism. Science 306:79–86. https://doi.org/10.1126/science.1101156 3. Bowler C, Allen AE, Badger JH, Grimwood J, Jabbari K, Kuo A, Maheswari U, Martens C, Maumus F, Otillar RP, Rayko E, Salamov A, Vandepoele K, Beszteri B, Gruber A, Heijde M, Katinka M, Mock T, Valentin K, Verret F, Berges JA, Brownlee C, Cadoret JP, Chiovitti A, Choi CJ, Coesel S, De Martino A, Detter JC, Durkin C, Falciatore A, Fournet J, Haruta M, Huysman MJ, Jenkins BD, Jiroutova K, Jorgensen RE, Joubert Y, Kaplan A, Kro¨ger N, Kroth PG, La Roche J, Lindquist E, Lommer M, Martin-Je´ze´quel V,

Lopez PJ, Lucas S, Mangogna M, McGinnis K, Medlin LK, Montsant A, Oudot-Le Secq MP, Napoli C, Obornik M, Parker MS, Petit JL, Porcel BM, Poulsen N, Robison M, Rychlewski L, Rynearson TA, Schmutz J, Shapiro H, Siaut M, Stanley M, Sussman MR, Taylor AR, Vardi A, von Dassow P, Vyverman W, Willis A, Wyrwicz LS, Rokhsar DS, Weissenbach J, Armbrust EV, Green BR, Van de Peer Y, Grigoriev IV (2008) The Phaeodactylum genome reveals the evolutionary history of diatom genomes. Nature 456:239–244. https://doi.org/10.1038/nature07410 4. Derelle E, Ferraz C, Rombauts S, Rouze´ P, Worden AZ, Robbens S, Partensky F, Degroeve S, Echeynie´ S, Cooke R, Saeys Y, Wuyts J, Jabbari K, Bowler C, Panaud O, Pie´gu B, Ball SG, Ral JP, Bouget FY, Piganeau G, De Baets B, Picard A, Delseny M, Demaille J, Van de Peer Y, Moreau H (2006) Genome analysis of the smallest free-living eukaryote Ostreococcus tauri unveils many unique features. Proc Natl Acad Sci U S A 103:11647–11652. https://doi.org/ 10.1073/pnas.0604795103 5. Merchant SS, Prochnik SE, Vallon O, Harris EH, Karpowicz SJ, Witman GB, Terry A, Salamov A, Fritz-Laylin LK, Mare´chalDrouard L, Marshall WF, Qu LH, Nelson DR, Sanderfoot AA, Spalding MH, Kapitonov VV, Ren Q, Ferris P, Lindquist E, Shapiro H, Lucas SM, Grimwood J, Schmutz J, Cardol P, Cerutti H, Chanfreau G, Chen CL, Cognat V, Croft MT, Dent R, Dutcher S, Fernández E, Fukuzawa H, González-Ballester D, GonzálezHalphen D, Hallmann A, Hanikenne M, Hippler M, Inwood W, Jabbari K, Kalanon M, Kuras R, Lefebvre PA, Lemaire SD, Lobanov AV, Lohr M, Manuell A, Meier I, Mets L, Mittag M, Mittelmeier T, Moroney JV, Moseley J, Napoli C, Nedelcu AM, Niyogi K, Novoselov SV, Paulsen IT, Pazour G, Purton S, ˜o-Pacho´n DM, Riekhof W, Ral JP, Rian Rymarquis L, Schroda M, Stern D, Umen J,

468

Naoki Sato and Takeshi Obayashi

Willows R, Wilson N, Zimmer SL, Allmer J, Balk J, Bisova K, Chen CJ, Elias M, Gendler K, Hauser C, Lamb MR, Ledford H, Long JC, Minagawa J, Page MD, Pan J, Pootakham W, Roje S, Rose A, Stahlberg E, Terauchi AM, Yang P, Ball S, Bowler C, Dieckmann CL, Gladyshev VN, Green P, Jorgensen R, Mayfield S, Mueller-Roeber B, Rajamani S, Sayre RT, Brokstein P, Dubchak I, Goodstein D, Hornick L, Huang YW, Jhaveri J, Luo Y, Martı´nez D, Ngau WC, Otillar B, Poliakov A, Porter A, Szajkowski L, Werner G, Zhou K, Grigoriev IV, Rokhsar DS, Grossman AR (2007) The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science 318:245–250. https://doi.org/ 10.1126/science.1143609 6. Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, Mitros T, Dirks W, Hellsten U, Putnam N, Rokhsar DS (2012) Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res 40(D1):D1178–D1186. https://doi. org/10.1093/nar/gkr944 7. Li-Beisson Y, Shorrosh B, Beisson F, Andersson M, Arondel V, Bates P, Baud S, Bird D, DeBono A, Durrett T, Franke R, Graham I, Katayama K, Kelly A, Larson T, Markham J, Miquel M, Molina I, Nishida I, Rowland O, Samuels L, Schmid K, Wada H, Welti R, Xu C, Zallot R, Ohlrogge J (2013) Acyl-lipid metabolism in The Arabidopsis Book. American Society of Plant Biologists, Rockville, MD. https://doi.org/10.1199/tab.0133 8. Fujisawa T, Narikawa R, Maeda SI et al (2017) CyanoBase: a large-scale update on its 20th anniversary. Nucleic Acids Res 45(D1):D551–

D554. https://doi.org/10.1093/nar/ gkw1131 9. Uchiyama I, Mihara M, Nishide H, Chiba H, Kato M (2019) MBGD update 2018: microbial genome database based on hierarchical orthology relations covering closely related and distantly related comparisons. Nucleic Acids Res 47:D382–D389. https://doi.org/ 10.1093/nar/gky1054 10. Sato N (2009) Gclust: trans-kingdom classification of proteins using automatic individual threshold setting. Bioinformatics 25:599–605. https://doi.org/10.1093/bioinformatics/ btp047 11. Kanehisa M, Sato Y (2020) KEGG Mapper for inferring cellular functions from protein sequences. Protoc Sci 29:28–35. https://doi. org/10.1002/pro.3711 12. Aoki Y, Okamura Y, Ohta H, Kinoshita K, Obayashi T (2016) ALCOdb: gene coexpression database for microalgae. Plant Cell Physiol 57:e3. https://doi.org/10.1093/pcp/pcv190 13. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, Simonovic M, Doncheva NT, Morris JH, Bork P, Jensen LJ, von Mering C (2019) STRING v11: proteinprotein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47:D607–D613. https://doi.org/ 10.1093/nar/gky1131 14. Tanizawa Y, Fujisawa T, Nakamura Y (2018) DFAST: a flexible prokaryotic genome annotation pipeline for faster genome publication. Bioinformatics 34:1037–1039. https://doi. org/10.1093/bioinformatics/btx713

INDEX A

D

Acetylation..................................136, 137, 142, 160, 250 Acyl-carrier protein (acyl-ACP)...................219–246, 466 Acyl-CoA ............................................158, 203–217, 221, 225, 227, 231, 234, 244, 391, 395, 398, 401–413, 446 Alcohol.................................... 43, 44, 46, 47, 49, 50, 56, 85, 107, 251, 280, 285, 288, 292, 447 Algae ................................ 30, 81, 82, 137, 446, 455–467 Anomeric configuration...............................249–251, 259 Arabidopsis thaliana .................4, 17, 18, 21, 23, 25, 34, 47, 48, 54, 55, 59, 60, 62, 65–67, 70–72, 76, 77, 101, 113, 118, 126, 147, 153, 158–160, 175, 249, 290, 297, 299, 312, 323, 325, 327, 328, 338, 339, 347, 353, 355, 356, 359, 366, 368, 372, 374, 402, 413, 451, 456, 461, 465 Aspartyl protease ........................................................... 221

Databases ..............................................56, 124, 128, 130, 131, 270, 276, 301, 309, 358, 427, 434, 441–453, 455–467 Data processing ................................. 127, 138, 152, 191, 213, 216, 423, 427 Deoxysphinganine....................................... 159, 160, 171 Double bond position ............................................ 44, 47, 52–54, 217, 452

B

Fatty acid ...............................3, 5, 10, 11, 17–21, 46, 48, 49, 54, 60–62, 67, 69, 70, 72, 83, 85, 86, 92, 103, 107, 112, 113, 118, 119, 121, 126, 127, 158, 159, 167–169, 171, 174, 200, 204, 206, 220, 254, 255, 260, 262, 264, 267, 269, 270, 286, 288, 304, 322, 324, 325, 331, 332, 383, 387, 397, 418, 443, 445–448, 452 analysis ..................................................................... 446 composition.............................................60, 117–131, 297, 307, 381, 445 synthesis .....................................60, 61, 219–246, 446 Fatty acid methyl esters (FAMEs) ........................ 4, 5, 10, 11, 38, 44, 48, 67, 69, 85, 92, 103, 107, 113, 118, 152, 288, 307, 325, 332, 381, 383, 386, 397, 448 Fatty acid phytyl esters................................ 107, 322, 331 Flotation ..............................................296, 323, 325–328 Free fatty acids (FFA) ............................3, 16, 17, 24, 30, 35, 46, 54, 55, 71, 72, 107, 112, 117, 118, 123, 127, 203, 204, 279, 297, 298, 322, 324, 331, 392 Free sterols (FS) ..............................................93, 95, 112, 142, 146, 179, 181, 183, 185, 187, 192

Betaine lipid......................................................30, 82, 144 Bioinformatics ............................450, 455, 456, 458, 467

C Camelina sativa ........................215, 221, 418, 422, 424, 428, 430–434 Carbohydrates ...........................250, 254, 255, 257, 259, 260, 270, 288, 457, 461 Carbon-14 (14C) .............................................. 59–78, 232 Cell cultures................................................ 137, 338, 339, 342, 347, 368, 412 Cell signaling ......................................................................v Ceramide ..................................... 29, 142, 146, 158–160, 162, 164, 165, 168, 173–175, 447 Chloroform/methanol ..........................4, 6, 7, 9, 19, 24, 32–34, 36, 38, 45, 54, 69, 72, 74–76, 84, 85, 87, 88, 103, 105–107, 121, 123, 124, 162, 183, 185, 276, 277, 290, 293, 300, 306, 307, 322, 324, 326, 329, 330, 354, 355, 382–384, 396 Collision-induced dissociation (CID).........................102, 113, 117–131, 196, 353, 355, 358, 359 Confocal fluorescence microscopy ............. 302, 303, 313 Cuticle.........................................275, 276, 285, 290, 418 Cutin ..................................................................... 275–293

E Electrospray ionization-tandem mass spectrometry (ESI-MS/MS) ..................................................136, 179–200, 353, 354 Extracellular lipids .................................95, 279, 280, 284

F

G Galactolipids .....................................31, 34, 62, 103–108, 110, 112–114, 117, 126, 219, 249, 250, 329, 358, 359, 392, 394, 398, 435, 447

Dorothea Bartels and Peter Do¨rmann (eds.), Plant Lipids: Methods and Protocols, Methods in Molecular Biology, vol. 2295, https://doi.org/10.1007/978-1-0716-1362-7, © Springer Science+Business Media, LLC, part of Springer Nature 2021

469

PLANT LIPIDS: METHODS

470 Index

AND

PROTOCOLS

Gas chromatography (GC) ..........................................4, 5, 17, 24, 43–56, 64, 67, 69, 85, 86, 92, 94, 95, 118, 126, 147, 179–200, 275, 315, 322, 325, 331, 332, 379–388, 396, 398, 445 Gas chromatography to mass spectrometry (GC-MS)................................... 11, 44–47, 49–54, 56, 118, 182, 184, 275–278, 281–286, 288–290, 292, 398, 442, 443, 448–450, 453 Gas chromatography with a flame ionization detector (GC-FID) ................................... 5, 13, 44, 46–48, 103, 113, 118, 152, 153, 180–182, 184, 191, 275–278, 281–292, 322, 396–398 Glucosylceramide ..................................... 29, 33, 34, 158, 159, 162, 164, 168, 173 Glycerolipids................................ 4, 5, 15–26, 29, 60–62, 82, 101–103, 108, 112, 137, 144, 152, 158, 163, 322, 337, 450, 462 Glycolipid .........................................29–31, 37, 110, 145, 250–252, 270, 380, 382, 384, 387, 462 Glycosylglycerolipids............................................ 249–270 Glycosylinositolphosphoceramides (GIPCs)............................................. 29, 157, 159, 160, 162, 171, 175, 176

H High‐performance liquid chromatography (HPLC)................................................5, 6, 19, 45, 63, 64, 68, 70, 72, 73, 75, 110, 112, 127, 138, 147, 161, 180–183, 186, 187, 199, 204–207, 210, 217, 226, 239, 277, 283, 345, 354, 365–375, 446, 448 His-tagged protein.....................403, 404, 406, 409, 412

I Inositol polyphosphates (InsPs) ......................... 365–368, 370–372, 374 Iodine staining .............................................................. 306 Isotope dilution................................................... 221, 226, 231, 240, 243, 245

L Lipases.......................................... 3, 5, 12, 15–26, 33, 34, 65, 72, 74, 75, 106, 297 Lipid analysis ....................................3, 13, 30, 43–56, 64, 84–87, 101–114, 143, 148, 151, 152, 295–316, 321–332, 337–348, 353, 355, 383, 443, 446, 448 Lipid droplets (LDs) .................................. 101, 295–316, 321, 322, 325, 391, 398 Lipid extraction ............................................ 3, 4, 6, 9, 11, 15–17, 19, 21–24, 26, 33, 34, 38, 40, 45–47, 54, 63, 64, 67–69, 84, 102, 103, 105, 106, 112, 136, 137, 140, 153, 161, 171, 173, 175, 277, 300, 304, 306, 307, 324, 333, 383, 388, 434, 442, 446, 447

Lipid metabolism ............................................59–78, 102, 135, 221, 391, 444, 457, 462, 465 Lipidomics ......................................... 122, 124, 130, 135, 136, 153, 338, 339, 381, 442–447, 449–451 Lipid pathway ........................................81, 455, 457, 462 Lipid-protein interaction ..................................... 391–399 Liposomes............................................393–395, 397, 398 Liquid chromatography-mass spectrometry (LC-MS) ........................................... 31, 102, 103, 107, 112, 137, 138, 167, 180, 199, 210, 213, 214, 217, 219–246, 250, 307–311, 354, 355, 442, 443, 446, 449–451, 453 Liquid scintillation counting ...................... 61, 69, 71, 72 Long-chain base (LCB) ............................. 142, 145–147, 158–160, 162, 164, 167, 169, 171, 173, 174

M MALDI-Mass Spectrometry Imaging (MALDI-MSI) .................................102, 417–435 Mass spectrometry (MS)...............................31, 102–105, 107–110, 112, 113, 118, 121–124, 135–153, 175, 181–183, 186, 187, 191, 197, 205, 207, 210, 211, 213, 217, 219–246, 250, 288, 290, 298, 308, 309, 339, 347, 434, 435, 442, 444–446, 449, 451, 452 Matrix assisted laser desorption/ionization (MALDI) .................................102, 204, 417–435 Membrane enrichment ........................................ 139, 140 Membrane lipids ........................................ 15, 30, 81–85, 87–90, 219, 276, 290, 337, 392 Membrane lipid strips ..................................394–396, 398 Metabolic flux............................................................60–62 Methylation ....................... 46, 47, 54, 55, 136, 137, 142 Microalgae ............................................... 81–96, 203–218 Mitochondria isolation ................................................. 344 Monoacylglycerol (MAG) ................................30, 33, 35, 44, 46, 49–51, 56, 72, 107, 109, 146, 324, 331, 447, 450 Monophasic extraction ........................................ 140, 141 Multiple reaction monitoring (MRM) .......................120, 148, 158, 163–165, 168, 170, 171, 174–176, 180, 182, 187, 188, 191, 200, 204, 205, 207, 210, 211, 213–215, 222, 239 Multiplexed collision-induced dissociation (Multiplexed CID) ................................... 353, 355

N Nanoelectrospray ionization (nanoESI) .....................136, 138, 146, 147 Nanospray-direct infusion ............................................ 102 N-Methyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA) ...................................45, 47, 49, 50, 55 NMR spectroscopy.........................................38, 249–270 Non-polar lipid.................................................32, 35, 103

PLANT LIPIDS: METHODS

AND

PROTOCOLS Index 471

O

S

Oil bodies ............................................................. 295, 297

Saponification .............................181, 182, 184, 191, 199 Seeds ..........................................4, 25, 29, 46, 47, 54, 55, 59–62, 65–70, 72, 76, 77, 102, 105, 113, 137, 179–200, 215, 221, 275–293, 296, 299, 303, 304, 325, 332, 368, 370, 417–419, 421–424, 428, 430–436, 443, 445, 446, 460 Serine palmitoyltransferase (SPT).................................................158–161, 171 Silylation .............................................................. 137, 147, 180–182, 184, 185, 191 Solid-phase adsorption chromatography ....................................... 379–389 Solid-phase-extraction (SPE) ................................. 30, 34, 44–47, 50, 102, 103, 105–107, 112, 186, 204, 301, 380, 382, 384, 385, 387, 446, 448 Sorghum bicolor......................................16, 17, 20, 21, 23 Spectrometry .............................................. 16, 30, 31, 38, 43–56, 101–114, 117–131, 135, 138, 142–144, 147, 148, 151, 157–176, 207, 211, 213, 217, 221, 245, 275, 296, 297, 315, 338, 345, 355, 387, 398, 417–436, 448, 451, 452 Sphingolipids .............................4, 29, 82, 117, 136, 142, 143, 145, 152, 157–176, 337, 392, 398, 444 Sterol lipids ............................................29, 146, 392, 398 Steryl esters (SEs)............................................91–95, 146, 180, 181, 183, 186–188, 191–193, 197–199, 322, 324, 330, 331, 341 Storage lipids ......................................................... 29, 101, 219, 290, 392, 394 Strong anion exchange HPLC (SAX-HPLC) ............................................ 365–375 Subcellular fractionation............................................... 296 Suberin.................................................................. 275–293 Sucrose gradients ................................322, 325, 327, 333 Surface lipid ........................................... 82, 84–88, 91, 92

P Phloem..........................................................351–360, 447 Phosphatidylcholine (PC).................................17, 19, 21, 24, 29, 33, 34, 61, 62, 71, 72, 75, 81–83, 101, 104, 105, 107, 108, 110, 112, 113, 121, 145, 146, 296, 298, 324, 354, 355, 381, 383, 393–395, 397, 418, 428–433, 435, 458 Phosphatidylethanolamine (PE).................................... 19, 21, 29, 33, 34, 71, 72, 81–83, 104, 105, 107, 110, 121, 145, 146, 298, 324, 331, 381, 393, 394, 435 Phosphatidylglycerol (PG) ..................................... 29, 33, 34, 39, 60, 71, 72, 81–83, 104, 105, 110, 145, 146, 331, 358, 393, 394, 435 Phosphatidylinositol-4,5-bisophosphate (PtdIns(4,5)P2) ............................... 366, 373, 381 Phosphoinositides ............................................... 365–375, 379–388, 392, 447 Phospholipid ........................................16, 29, 30, 33, 81, 103–108, 110, 113, 117, 121, 136, 144, 146, 295, 296, 329, 337, 338, 351, 359, 366, 380–382, 384, 385, 387, 392, 394, 398, 426, 427, 435, 448, 450 Phytic acid ..................................................................... 366 Plant organelle............................................................... 295 Plant tissues ....................................... 3, 4, 15, 16, 21, 23, 25, 29, 30, 33, 34, 39, 46, 47, 49, 55, 60–62, 105, 106, 112, 126, 136, 137, 204, 221, 225, 231, 232, 238, 239, 297, 298, 380, 383, 384, 417–419, 427, 435 Plastid membranes ...................................... 322, 326, 339 Plastoglobules...............................................321–334, 391 Polar lipids ................................... 4, 5, 24, 30, 32–36, 64, 70–72, 103, 105, 162, 171, 174, 182, 183, 205, 251, 300, 301, 321, 322, 324, 326, 329–331, 354 Protein-protein interaction......................... 402, 457, 467 Protein purification ..................................... 227, 244, 392 Proteomics......................................... 236, 296, 297, 299, 301, 304, 307–309, 315, 316, 325 Pulse-chase labeling ........................................................ 62

Q Quadrupole time-of-flight (QTOF) ...........................102, 110, 118, 122–124, 126, 127, 200 Qualitative analysis ...................................... 104, 108, 113 Quantitative analysis ........................................46, 47, 386

R Radiolabel ..................................................................61, 62

T Thin-layer chromatography (TLC)........................ 84, 85, 251, 252, 298, 300, 304, 306, 307, 322, 324, 326, 329, 330, 345, 355, 358, 381–383, 385, 386, 388 Total acyl lipid collision induced dissociation time-offlight (TAL-CID-TOF) mass spectrometry ............................................. 117–131 Transient expression............................................. 296, 316 Triacylglycerols (TAG)......................................5, 6, 8, 10, 12, 16, 30, 33, 35, 54, 60–62, 66, 67, 70–72, 74, 75, 78, 85, 91, 101, 103–107, 109, 111, 113, 117, 136, 144, 146, 186, 199, 219, 295, 324, 331, 387, 388, 392, 395, 398, 418, 426, 428, 429, 431, 433–435, 443, 445–447, 450

PLANT LIPIDS: METHODS

472 Index

AND

PROTOCOLS

Trimethylsilyl (TMS) ...........................45, 47, 49, 52, 55, 56, 86, 93, 95, 137, 152, 182, 277, 283, 285, 291

W Wax .......................................................30, 33, 35, 82, 85, 86, 91, 92, 95, 107, 275–293, 295, 447

Wax ester (WE) .......................................... 33, 35, 82, 85, 86, 91, 92, 94, 292, 295, 447