In Situ Hybridization Protocols (Methods in Molecular Biology, 2148) 1071606220, 9781071606223

This fifth edition volume expands on the previous editions with updated discussions on the many new in situ hybridizatio

136 22 14MB

English Pages 405 [391] Year 2020

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Preface
Contents
Contributors
Part I: General Applications
Chapter 1: Guidelines for the Optimization and Validation of In Situ Hybridization
1 Introduction
2 Steps to Validation
2.1 Material Preparation
2.2 Understand Target
2.3 Methodology
2.3.1 Probe Generation
2.3.2 Riboprobes
2.3.3 Oligonucleotide Probes
2.3.4 Pretreatment
2.3.5 Hybridization
2.3.6 Stringency
2.3.7 Detection
2.4 Troubleshooting
2.5 Determining Specificity
2.6 Controls
2.6.1 Positive Controls
2.6.2 Negative Controls
References
Chapter 2: Overcoming Autofluorescence (AF) and Tissue Variation in Image Analysis of In Situ Hybridization
1 Introduction
2 Materials
3 Methods
3.1 Imaging of the Slides
3.2 Analyzing Images: TMA Segmentation
3.3 Analyzing Images: Creating the Classifier
3.4 Analyzing Images: Creating the Analysis Algorithm
4 Notes
References
Part II: Methods for DNA ISH
Chapter 3: Practical Application of Fluorescent In Situ Hybridization Techniques in Clinical Diagnostic Laboratories
1 Introduction
2 Materials
2.1 Buffers and Fixatives
2.2 Ready-to-Use Commercial Reagents
2.3 Specialist Reagents, DNA Probes
2.4 Equipment
2.4.1 General Laboratory Equipment
2.4.2 Specialist Equipment
2.5 Biological Samples to Test
3 Methods
3.1 Preliminary Steps
3.2 Fixation and Pretreatment of Blood, Marrow, and Tissue Imprint Preparations
3.3 MAA Fixed Cell Suspension Preparations
3.4 FFPE 3 μm Section Preparations
3.5 Pretreatment of FFPE 3 μm Sections Using Pressure Cooker and Enzyme Digestion
3.6 Probe Preparation and Application
3.7 Post-hybridization Stringency Wash
3.8 Counterstain and Coverslipping
3.9 Turning on Microscope, Lamp Power Unit, and FISH Workstation: General Points
3.10 Examination of FISH Tests and Image Capture
3.11 Analysis of FISH Test Results
3.11.1 Numerical Analysis
3.11.2 Break-Apart Analysis
3.11.3 Dual-Fusion, Dual-Color Translocation Analysis
3.11.4 Tricolor Rearrangement Analysis
3.12 Signal Enumeration
3.13 Quality Assurance
3.14 External Quality Control
4 Notes
References
Other Useful Resources
National Guidelines
International/European Standards
Chapter 4: Fluorescent In Situ Hybridization Using Oligonucleotide-Based Probes
1 Introduction
2 Materials
2.1 MYtags Immortal Oligonucleotide Library
2.2 MYcroarray Debubbling PCR
2.3 In Vitro Transcription
2.4 Reverse Transcription
2.5 Removal of RNA
2.6 Chromosome Spread Preparation
2.7 Fluorescent In Situ Hybridization (FISH)
3 Methods
3.1 MYcroarray Debubbling PCR: First Step of Amplification
3.1.1 PCR Purification
3.2 In Vitro Transcription: Second Step of Amplification
3.2.1 RNA Purification
3.3 Reverse Transcription: Labeling Step
3.3.1 RNA:DNA Hybrid Purification
3.4 Removal of RNA
3.4.1 Single-Strand DNA Probe Purification
3.5 Chromosome Spread Preparation
3.6 FISH
3.6.1 Slide Pretreatment
3.6.2 Chromosome Denaturation
3.6.3 Probe Hybridization
3.6.4 Stringency Washes
3.6.5 Optional Higher Stringency Washes
3.6.6 Probe Detection
3.6.7 Slide Cleaning for Re-probing
4 Notes
References
Chapter 5: Visualizing Genome Reorganization Using 3D DNA FISH
1 Introduction
2 Materials
2.1 Fosmid Probe Preparation
2.2 Slide Preparation: Cells
2.3 Slide Preparation: Paraffin-Embedded Tissue Sections
2.4 FISH
3 Methods
3.1 Selection of Fosmid Probes Using Ensembl Browser
3.2 Preparation of Fosmid Probes Using Alkaline Lysis Miniprep (See Note 1)
3.3 Nick Translation
3.4 Preparation of Slides: Cells
3.5 Preparation of Slides: Paraffin-Embedded Tissue Sections
3.6 3D DNA FISH
3.6.1 Preparation
3.6.2 Probe Preparation
3.6.3 Slide Preparation
3.6.4 Slide Washing
3.7 Imaging
4 Notes
References
Part III: Methods for Cultured Cells
Chapter 6: MicroRNA In Situ Hybridization in Paraffin-Embedded Cultured Cells
1 Introduction
2 Materials
2.1 Reagents and Buffers
2.2 Equipment
3 Methods
3.1 Cell Preparation and Embedding of Cells
3.2 Tissue Sections
3.3 In Situ Hybridization
4 Notes
References
Chapter 7: Multiplexed Detection and Analysis of Low-Abundance Long Noncoding RNA Using RNAscope in Cultured Cells
1 Introduction
2 Materials
2.1 Cell Culture
2.2 RNAscope
2.3 High-Content Imaging and Analysis
3 Methods
3.1 Cell Culture
3.2 RNAscope Assay
3.3 High-Content Imaging and Analysis
4 Notes
References
Chapter 8: Multiplexed Quantitative In Situ Hybridization for Mammalian or Bacterial Cells in Suspension: qHCR Flow Cytometry ...
1 Introduction
2 Materials
2.1 Sample Preparation
2.1.1 Fixation and Permeabilization (All Sample Types)
2.1.2 Bacterial Samples
2.1.3 Mammalian Samples
2.2 HCR Reagents
2.3 Additional Equipment
3 Methods
3.1 Sample Preparation
3.1.1 Bacterial Cells
3.1.2 Mammalian Cells
3.2 Detection Stage
3.3 Amplification Stage
3.4 Flow Cytometry
3.5 Data Analysis
3.5.1 Raw Single-Cell Intensities
3.5.2 Measurement of Signal, Background, and Signal-to-Background for Transgenic Targets
3.5.3 Measurement of Signal, Background, and Signal-to-Background for Endogenous Targets
3.5.4 Normalized Single-Cell Intensities for Analog mRNA Relative Quantitation with qHCR Flow Cytometry
4 Notes
References
Chapter 9: Multiplexed Quantitative In Situ Hybridization for Mammalian Cells on a Slide: qHCR and dHCR Imaging (v3.0)
1 Introduction
2 Materials
2.1 Sample Preparation
2.2 HCR Reagents
2.3 Additional Reagents and Equipment
3 Methods
3.1 Sample Preparation
3.2 Detection Stage
3.3 Amplification Stage
3.4 Confocal Microscopy
3.5 Image Analysis
3.5.1 Raw Pixel Intensities
3.5.2 Measurement of Signal, Background, and Signal-to-Background for Transgenic Targets
3.5.3 Measurement of Signal, Background, and Signal-to-Background for Endogenous Targets
3.5.4 Normalized Voxel Intensities for qHCR Imaging: Analog mRNA Relative Quantitation with Subcellular or Single-Cell Resolut...
3.5.5 Dot Detection and Colocalization for dHCR Imaging: Digital mRNA Absolute Quantitation with Single-molecule Resolution
4 Notes
References
Part IV: Methods for Wholemounts and Plant Material
Chapter 10: Multiplexed Quantitative In Situ Hybridization with Subcellular or Single-Molecule Resolution Within Whole-Mount V...
1 Introduction
2 Materials
2.1 Sample Preparation
2.1.1 Fixation and Permeabilization (All Sample Types)
2.1.2 Whole-Mount Chicken Embryo Preparation
2.1.3 Whole-Mount Zebrafish Embryo/Larva Preparation
2.1.4 Whole-Mount Mouse Embryo Preparation
2.2 HCR Reagents
2.3 Imaging
2.4 Equipment
3 Methods
3.1 Sample Preparation
3.1.1 Whole-Mount Chicken Embryos
3.1.2 Whole-Mount Zebrafish Embryos/Larvae
3.1.3 Whole-Mount Mouse Embryos
3.2 In Situ HCR Protocol
3.2.1 Detection Stage
3.2.2 Amplification Stage
3.3 Sample Mounting for Microscopy
3.3.1 Whole-Mount Chicken Embryos
3.3.2 Whole-Mount Zebrafish Embryos/Larvae
3.3.3 Whole-Mount Mouse Embryos
3.4 Confocal Microscopy
3.5 Image Analysis
3.5.1 Raw Pixel Intensities
3.5.2 Measurement of Signal, Background, and Signal-to-Background
3.5.3 Measurement of Background Components
3.5.4 Normalized Voxel Intensities for qHCR Imaging: Analog mRNA Relative Quantitation with Subcellular Resolution in an Anato...
3.5.5 Dot Detection and Colocalization for dHCR Imaging: Digital mRNA Absolute Quantitation with Single-Molecule Resolution in...
3.5.6 Read-out/Read-In Analysis Framework for Quantitative RNA Discovery in an Anatomical Context
4 Notes
References
Chapter 11: Hybridization Chain Reaction for Quantitative and Multiplex Imaging of Gene Expression in Amphioxus Embryos and Ad...
1 Introduction
2 Materials
2.1 Reagents
2.2 Equipment
2.3 Buffers and Solutions
3 Methods
3.1 Embryo Collection and Fixation
3.2 Adult Tissue Fixation, Embedding, and Sectioning
3.3 Amphioxus HCR In Situ Hybridization (ISH) Steps
3.4 Whole-Mount Imaging
4 Notes
References
Chapter 12: RNAscope Multiplex Detection in Zebrafish
1 Introduction
2 Materials
2.1 Reagents
2.2 Buffers and Solutions
2.3 Equipment
3 Methods
3.1 Probe Design
3.2 Preparation of Embryos
3.3 Preparation of Wash Buffers
3.4 Day 1 (1 h): Probe Hybridization
3.5 Day 2 (7 h): RNA Detection and Signal Amplification
3.6 Embryo Staging for Imaging
4 Notes
References
Chapter 13: Duplex In Situ Hybridization of Virus Nucleic Acids in Plant Tissues Using RNAscope
1 Introduction
2 Materials
2.1 RNAscope Assay Reagents and Samples
2.2 Other Reagents and Materials
2.3 Equipment
3 Methods
3.1 FFPE Section Preparation
3.2 Pretreatment
3.3 Duplex RNAscope Assay
3.4 Visualization
4 Notes
References
Part V: Automated Methods for RNA ISH
Chapter 14: Automated ISH for Validated Histological Mapping of Lowly Expressed Genes
1 Introduction
2 Materials
2.1 Equipment
2.2 Reagents
2.2.1 FFPE Tissue Preparation
2.2.2 Automated RNAscope on Ventana Discovery Ultra (See Note 1)
2.3 Software
3 Methods
3.1 FFPE Tissue Preparation Using Immersion Fixation (See Note 2)
3.2 Probes
3.3 Automated ISH (See Note 1)
3.4 Quantitative Image Analysis
4 Notes
References
Chapter 15: Automation of Multiplexed RNAscope Single-Molecule Fluorescent In Situ Hybridization and Immunohistochemistry for ...
1 Introduction
2 Materials
2.1 Manual Pre-staining Treatment of Fixed Frozen Sections
2.2 In Situ Hybridization and Immunohistochemical Staining Using the Leica BOND RX
2.3 Mounting and Imaging
3 Methods
3.1 Leica BOND RX Setup
3.2 Experimental Design
3.3 Manual Pre-staining Sample Preparation
3.4 In Situ Hybridization and Immunohistochemical Staining Using the Leica BOND RX
3.5 Leica BOND RX Run Checklist
3.6 Mounting and Imaging
4 Notes
References
Chapter 16: Automated Co-in Situ Hybridization and Immunofluorescence Using Archival Tumor Tissue
1 Introduction
2 Materials
2.1 Advanced Cell Diagnostics
2.2 Roche Tissue Diagnostics
2.3 Target IHC Antibodies
2.4 Akoya Biosciences Inc.
2.5 Multiplex Assay Validation Slides
2.6 Mounting
3 Methods
3.1 Multiplex Assay Construction and Validation
3.2 Validated Protocol
4 Notes
References
Chapter 17: Automated Five-Color Multiplex Co-detection of MicroRNA and Protein Expression in Fixed Tissue Specimens
1 Introduction
2 Materials
2.1 Consumables and Bulk Leica Reagents
2.2 General Solution and Reagents
2.3 Buffers and Reagent Solutions (See Also Table 1 for Probes, Table 2 for Antibodies, Table 3 for Dyes)
2.4 Tissue Processing and Slide Preparation
2.5 Equipment for Microscopy
3 Methods
3.1 Programming 4-Plex ISH/IHC Assay
3.2 Pre-hybridization Steps
3.3 ISH Hybridization (See Table 1 for Probes)
3.4 Sequential Detection of ISH Probes (See Table 2 for Antibodies and Table 3 for Dyes)
3.5 Sequential Detection of Protein Markers (See Table 2 for Antibodies and Table 3 for Dyes)
3.6 Slide Mounting (Manually)
3.7 Image Acquisition and Computer-Assisted Image Analysis
4 Notes
References
Chapter 18: Mixed Multiplex Staining: Automated RNAscope and OPAL for Multiple Targets
1 Introduction
2 Materials
2.1 Consumables and Ancillary Leica Reagents
2.2 RNAscope Reagents
2.3 Antibodies, Blocking Solutions, and Tyramides
2.4 Universal Reagents
2.5 General Solutions and Reagents
3 Methods
4 Notes
References
Part VI: Multiplexing and Combined Methods
Chapter 19: Simultaneous Visualization of RNA and Protein Expression in Tissue Using a Combined RNAscope In Situ Hybridization...
1 Introduction
2 Materials
2.1 RNAscope Assay
2.2 Opal/TSA Dyes
2.3 Samples
2.4 Reagents
2.5 Solutions
2.6 Miscellaneous
2.7 Equipment
3 Methods
3.1 FFPE Section Preparation and Deparaffinization
3.2 Pretreatment of FFPE Sections
3.3 RNAscope ISH Multiplex Fluorescent Assay v2 (See Note 3)
3.4 Immunofluorescence
3.5 Results and Interpretation
4 Notes
References
Chapter 20: In Situ Sequencing: A High-Throughput, Multi-Targeted Gene Expression Profiling Technique for Cell Typing in Tissu...
1 Introduction
2 Materials
2.1 Padlock Probe Design
2.2 Detection Probe Design
2.3 Oligonucleotides
2.4 Reagents
2.5 Buffers and Solutions
2.6 Equipment
3 Methods
3.1 Tissue Preparation
3.2 Tissue Fixation
3.3 Reverse Transcription/cDNA Synthesis
3.4 Padlock Probe Hybridization and Ligation
3.5 Rolling Circle Amplification
3.6 RCA Product Detection and Barcode Sequencing
3.7 Image Acquisition
3.8 Anchor Probe and Barcode-Detection Probe Stripping
3.9 Optional: RCA Product Detection and ``5th´´ Base Sequencing
3.10 Analysis and Results
4 Notes
References
Chapter 21: GeoMx RNA Assay: High Multiplex, Digital, Spatial Analysis of RNA in FFPE Tissue
1 Introduction
2 Materials
2.1 Equipment
2.2 Materials for Slide Preparation
2.3 Materials for nCounter Readout
3 Methods
3.1 Preparing Reagents and Equipment
3.2 Preparing Tissues
3.3 Slide Preparation
3.4 Overnight In Situ Hybridization
3.5 Perform Stringent Washes
3.6 Apply Morphology Markers
3.7 Using the GeoMx DSP
3.8 GeoMx Hyb Code Setup
4 Notes
References
Part VII: Target Selective Methods and Single Molecule Detection
Chapter 22: In Situ Point Mutation Detection in FFPE Colorectal Cancers Using the BaseScope Assay
1 Introduction
2 Materials
2.1 Equipment
2.2 Reagents
3 Methods
3.1 Preparation of FFPE Cell Pellets
3.2 Preparation of FFPE Tumor Samples
3.3 Running the BaseScope Assay
3.4 Analysis
4 Notes
References
Chapter 23: Using In Situ Padlock Probe Technology to Detect mRNA Splice Variants in Tumor Cells
1 Introduction
2 Materials
2.1 RT Primers, PLPs, and Detection Probes
2.2 Cell Lines and Cell Culture Media
2.3 Reagents, Buffers, and Solutions
2.4 Consumables and Equipment
3 Methods
3.1 Seeding PC Cell Lines
3.2 Sample Preparation
3.3 Reverse Transcription
3.4 Hybridization and Ligation of PLPs
3.5 Rolling Circle Amplification
3.5.1 Hybridization of Detection Probes
3.5.2 Imaging
3.5.3 Image Analysis and Quantification
4 Notes
References
Chapter 24: Automated One-Double-Z Pair BaseScope for CircRNA In Situ Hybridization
1 Introduction
2 Materials
2.1 Tissue Samples
2.2 Reagents
2.3 Equipment
3 Methods
3.1 Preparation of Tissue Sections
3.2 Preparation of the Ventana DISCOVERY ULTRA Instrument
3.3 Programming the Ventana DISCOVERY ULTRA Instrument
3.4 Running BaseScope
4 Notes
References
Correction to: In Situ Sequencing: A High-Throughput, Multi-Targeted Gene Expression Profiling Technique for Cell Typing in Ti
Index
Recommend Papers

In Situ Hybridization Protocols (Methods in Molecular Biology, 2148)
 1071606220, 9781071606223

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Methods in Molecular Biology 2148

Boye Schnack Nielsen · Julia Jones Editors

In Situ Hybridization Protocols Fifth Edition

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-bystep 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.

In Situ Hybridization Protocols FIFTH EDITION

Edited by

Boye Schnack Nielsen Molecular Histology, Bioneer A/S, Horsholm, Denmark

Julia Jones Cancer Research UK, University of Cambridge, Cambridge, UK

Editors Boye Schnack Nielsen Molecular Histology Bioneer A/S Horsholm, Denmark

Julia Jones Cancer Research UK University of Cambridge Cambridge, UK

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-0622-3 ISBN 978-1-0716-0623-0 (eBook) https://doi.org/10.1007/978-1-0716-0623-0 © Springer Science+Business Media, LLC, part of Springer Nature 2020, Corrected Publication 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. 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 Welcome to the 5th edition of In Situ Hybridization Protocols. Since the publication of the 4th edition back in 2014, many new in situ hybridization (ISH) techniques and applications have been developed. Innovation is thus found among new probe designs, detection systems, and signal amplification, improving specificity, sensitivity, as well as multiplexing combinations and automation. At the same time, the variety of RNA species has enlarged, and in addition to the well-known mRNA, lncRNA, and microRNA, we also embrace the distinctive group of circular RNAs. In this edition, we have included three methods for DNA in situ hybridization. O’Connor et al. (Chapter 3) provide an overview of DNA FISH with approaches on both liquid samples and paraffin-embedded samples, Braz et al. (Chapter 4) present a DNA FISH method based on multiple oligonucleotide probes, and Jubb and Boyle (Chapter 5) present a protocol to use 3D imaging to demonstrate and quantify FISH signal output. All the remaining chapters are dealing with in situ hybridization for RNA detection. In Part 1, the first chapter discusses issues in signal validation (Howat & Jones, Chapter 1) and provides guidelines for implementation of ISH, and the second chapter discusses how to deal with autofluorescence signal when performing image analysis on fluorescence-stained slides (Brodie, Chapter 2). In the 4th edition, the branched-DNA (bDNA) signal amplification technology was taking over the traditional mRNA in situ hybridization procedures based on radioactively labeled antisense probes. Thanks to the development of the in situ detection methods based on bDNA technology; several new derived applications of the technique have been reported and are presented in this edition. The 22 protocols are coarsely divided into 6 parts that we found most appropriate: (Part 2) Methods for DNA ISH, (Part 3) methods for cultured cells, (Part 4) methods for wholemounts and plant material, (Part 5) automated methods, (Part 6) multiplexing and combined methods, and (Part 7) targetselective methods. We not only believe that the seven parts will make sense to the readers, but we also encourage readers to take a look around the chapters because of the evidently overlapping procedures and applications. Novel in situ hybridization techniques are introduced in this edition: hybridization chain reaction (HCR) in applications on cultured cells, wholemounts, bacteria and tissue sections (Schwarzkopf et al, Chapters 8 and 9, Choi et al., Chapter 10), digital spatial profiler (DSP) using high-end multiplexing (Zollinger et al., Chapter 21), and padlock probe technology (Hilscher et al., Chapter 20 and Hofmann et al., Chapter 23). In addition, a method for in situ sequencing for expression profiling is presented (Hilscher et al., Chapter 20). ISH procedures on cultured cells are also presented in this edition and include the use of LNA probes for detection of microRNA (James et al., Chapter 6), multiplexing using bDNA methods (Jones et al., Chapter 7), HCR methods (Schwarzkopf et al., Chapters 8 and 9), detection of point mutations using bDNA methods (Baker & Graham, Chapter 22), and padlock probe technology (Hofmann et al., Chapter 23). Although cultured cells cannot replace tissue sections as biologic material, the cells can be engineered and processed to become a basis for optimizing probe specificity and signal amplification. Applications on wholemounts are presented using HCR technology in vertebrate embryos

v

vi

Preface

(Choi et al., Chapter 10), and bDNA technology in amphioxus (Andrews et al., Chapter 11) and zebrafish (Gross-Thebing, Chapter 12). In addition, Sheat et al. (Chapter 13) present a bDNA-based protocol for detection of virus transcript in plant material. Automation of RNA ISH has been a major relief by replacing the time-consuming and cumbersome procedures and has provided the basis for more reproducible results. Several chapters in this issue are presenting automated procedures of which four include advanced techniques (Pyke, Chapter 14, Roberts & Bayraktar, Chapter 15, Officer et al., Chapter 16, Sempere et al., Chapter 17). The advanced ISH techniques with improved sensitivity and specificity allows duplexing, multiplexing (e.g., Millar, Chapter 18), and combination with immunohistochemistry (e.g., Dikshit et al., Chapter 19). More than half of the chapters in this edition present combination methods on duplex or multiplexing ISH. The use of combination techniques or multiplexing ISH allow the use of reference transcript and cell markers to follow the dynamics in transcript expression in various cellular compartments. Finally, the development of highly sensitive assays using enhanced sequence selectivity in probe design and signal amplification has enabled detection of unique sequences such as single point mutations (Baker & Graham, Chapter 22), circRNAs (Nielsen et al., Chapter 24), and other unique splice sites (Hofmann et al., Chapter 23). We are confident that the current edition presents both well-known up-to-date methods and a series of new methods and applications based on more recently developed technologies. The continued development of the new ISH techniques indicates that the field is advancing and provides exciting promise to the future. Sincere thanks to all the contributing authors for taking part in the realization of this 5th edition of In Situ Hybridization Protocols. Horsholm, Denmark Cambridge, UK

Boye Schnack Nielsen Julia Jones

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

PART I

GENERAL APPLICATIONS

1 Guidelines for the Optimization and Validation of In Situ Hybridization . . . . . . Julia Jones and William J. Howat 2 Overcoming Autofluorescence (AF) and Tissue Variation in Image Analysis of In Situ Hybridization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cara Brodie

PART II

3

19

METHODS FOR DNA ISH

3 Practical Application of Fluorescent In Situ Hybridization Techniques in Clinical Diagnostic Laboratories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sheila J. M. O’Connor, Kathryn R. Turner, and Sharon L. Barrans 4 Fluorescent In Situ Hybridization Using Oligonucleotide-Based Probes . . . . . . . Guilherme T. Braz, Fan Yu, Lı´via do Vale Martins, and Jiming Jiang 5 Visualizing Genome Reorganization Using 3D DNA FISH . . . . . . . . . . . . . . . . . . Alasdair Jubb and Shelagh Boyle

PART III

v xi

35 71 85

METHODS FOR CULTURED CELLS

6 MicroRNA In Situ Hybridization in Paraffin-Embedded Cultured Cells . . . . . . . 99 Jaslin P. James, Laura Johnsen, Trine Møller, and Boye Schnack Nielsen 7 Multiplexed Detection and Analysis of Low-Abundance Long Noncoding RNA Using RNAscope™ in Cultured Cells . . . . . . . . . . . . . . . . . . . . . 111 Julia Jones, Heather Zecchini, and Sankari Nagarajan 8 Multiplexed Quantitative In Situ Hybridization for Mammalian or Bacterial Cells in Suspension: qHCR Flow Cytometry (v3.0) . . . . . . . . . . . . . . 127 Maayan Schwarzkopf, Harry M. T. Choi, and Niles A. Pierce 9 Multiplexed Quantitative In Situ Hybridization for Mammalian Cells on a Slide: qHCR and dHCR Imaging (v3.0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Maayan Schwarzkopf, Harry M. T. Choi, and Niles A. Pierce

PART IV 10

METHODS FOR WHOLEMOUNTS AND PLANT MATERIAL

Multiplexed Quantitative In Situ Hybridization with Subcellular or Single-Molecule Resolution Within Whole-Mount Vertebrate Embryos: qHCR and dHCR Imaging (v3.0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Harry M. T. Choi, Maayan Schwarzkopf, and Niles A. Pierce

vii

viii

11

12 13

Contents

Hybridization Chain Reaction for Quantitative and Multiplex Imaging of Gene Expression in Amphioxus Embryos and Adult Tissues . . . . . . . 179 Toby G. R. Andrews, Giacomo Gattoni, Lara Busby, ` lia Benito-Gutie´rrez Michael A. Schwimmer, and E RNAscope™ Multiplex Detection in Zebrafish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Theresa Gross-Thebing Duplex In Situ Hybridization of Virus Nucleic Acids in Plant Tissues Using RNAscope®. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Samar Sheat, Stephan Winter, and Paolo Margaria

PART V AUTOMATED METHODS FOR RNA ISH 14

15

16

17

18

Automated ISH for Validated Histological Mapping of Lowly Expressed Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Charles Pyke Automation of Multiplexed RNAscope Single-Molecule Fluorescent In Situ Hybridization and Immunohistochemistry for Spatial Tissue Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kenny Roberts and Omer Ali Bayraktar Automated Co-in Situ Hybridization and Immunofluorescence Using Archival Tumor Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leah K. Officer, Kleopatra E. Andreou, Ana V. Teodo sio, Zhangyi He, and John P. Le Quesne Automated Five-Color Multiplex Co-detection of MicroRNA and Protein Expression in Fixed Tissue Specimens . . . . . . . . . . . . . . . . . . . . . . . . . . Lorenzo F. Sempere, Erin Zaluzec, Elizabeth Kenyon, Matti Kiupel, and Anna Moore Mixed Multiplex Staining: Automated RNAscope™ and OPAL™ for Multiple Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael Millar

PART VI 19

20

21

219

229

245

257

277

MULTIPLEXING AND COMBINED METHODS

Simultaneous Visualization of RNA and Protein Expression in Tissue Using a Combined RNAscope™ In Situ Hybridization and Immunofluorescence Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 Anushka Dikshit, Hailing Zong, Courtney Anderson, Bingqing Zhang, and Xiao-Jun Ma In Situ Sequencing: A High-Throughput, Multi-Targeted Gene Expression Profiling Technique for Cell Typing in Tissue Sections . . . . . . . . . . . . 313 Markus M. Hilscher, Daniel Gyllborg, Chika Yokota, and Mats Nilsson GeoMx™ RNA Assay: High Multiplex, Digital, Spatial Analysis of RNA in FFPE Tissue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Daniel R. Zollinger, Stan E. Lingle, Kristina Sorg, Joseph M. Beechem, and Christopher R. Merritt

Contents

PART VII

ix

TARGET SELECTIVE METHODS AND SINGLE MOLECULE DETECTION

22

In Situ Point Mutation Detection in FFPE Colorectal Cancers Using the BaseScope Assay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349 Ann-Marie Baker and Trevor A. Graham 23 Using In Situ Padlock Probe Technology to Detect mRNA Splice Variants in Tumor Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361 Lilli Hofmann, Thomas Kroneis, and Amin El-Heliebi 24 Automated One-Double-Z Pair BaseScope™ for CircRNA In Situ Hybridization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 Boye Schnack Nielsen, Trine Møller, and Jørgen Kjems Correction to: In Situ Sequencing: A High-Throughput, Multi-Targeted Gene Expression Profiling Technique for Cell Typing in Tissue Sections . . . . . . . . . . . C1 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

389

Contributors COURTNEY ANDERSON • Advanced Cell Diagnostics, A Bio-Techne Brand, Newark, CA, USA KLEOPATRA E. ANDREOU • MRC Toxicology Unit, University of Cambridge, Cambridge, UK TOBY G. R. ANDREWS • Department of Zoology, University of Cambridge, Cambridge, UK ANN-MARIE BAKER • Barts Cancer Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, UK SHARON L. BARRANS • Haematological Malignancy Diagnostic Service (HMDS), Level 3 Bexley Wing, Leeds Cancer Centre, St James’s University Teaching Hospital, Leeds, UK OMER ALI BAYRAKTAR • Wellcome Sanger Institute, Hinxton, UK JOSEPH M. BEECHEM • NanoString Technologies Inc., Seattle, WA, USA E`LIA BENITO-GUTIE´RREZ • Department of Zoology, University of Cambridge, Cambridge, UK SHELAGH BOYLE • MRC Human Genetics Unit, MRC Institute of Genetics & Molecular Medicine, University of Edinburgh, Edinburgh, UK GUILHERME T. BRAZ • Department of Plant Biology, Michigan State University, East Lansing, MI, USA CARA BRODIE • Histopathology and ISH Core Facility, Cancer Research UK/Cambridge Institute, University of Cambridge, Li Ka Shing Centre, Cambridge, UK LARA BUSBY • Department of Zoology, University of Cambridge, Cambridge, UK HARRY M. T. CHOI • Division of Biology & Biological Engineering, California Institute of Technology, Pasadena, CA, USA ANUSHKA DIKSHIT • Advanced Cell Diagnostics, A Bio-Techne Brand, Newark, CA, USA LI´VIA DO VALE MARTINS • Department of Plant Biology, Michigan State University, East Lansing, MI, USA; Departamento de Gene´tica, Universidade Federal de Pernambuco, Recife, PE, Brazil AMIN EL-HELIEBI • Division of Cell Biology, Histology & Embryology, Gottfried Schatz Research Center, Medical University Graz, Graz, Austria; Center for Biomarker Research, CBmed, Graz, Austria GIACOMO GATTONI • Department of Zoology, University of Cambridge, Cambridge, UK TREVOR A. GRAHAM • Barts Cancer Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, UK THERESA GROSS-THEBING • Institute of Anatomy and Vascular Biology, University of Muenster, Muenster, Germany DANIEL GYLLBORG • Molecular Diagnostics, Science for Life Laboratory, Department of Biochemistry and Biophysics, Stockholm University, Solna, Sweden ZHANGYI HE • MRC Toxicology Unit, University of Cambridge, Cambridge, UK MARKUS M. HILSCHER • Molecular Diagnostics, Science for Life Laboratory, Department of Biochemistry and Biophysics, Stockholm University, Solna, Sweden LILLI HOFMANN • Division of Cell Biology, Histology & Embryology, Gottfried Schatz Research Center, Medical University Graz, Graz, Austria; Center for Biomarker Research, CBmed, Graz, Austria WILLIAM J. HOWAT • Abcam plc, Discovery Drive, Cambridge Biomedical Campus, Cambridge, UK JASLIN P. JAMES • Bioneer A/S, Molecular Histology, Hørsholm, Denmark

xi

xii

Contributors

JIMING JIANG • Department of Plant Biology, Michigan State University, East Lansing, MI, USA; Department of Horticulture, Michigan State University, East Lansing, MI, USA; Michigan State University AgBioResearch, East Lansing, MI, USA LAURA JOHNSEN • Bioneer A/S, Molecular Histology, Hørsholm, Denmark JULIA JONES • Histopathology and ISH Core Facility, Cancer Research UK/Cambridge Institute, University of Cambridge, Li Ka Shing Centre, Cambridge, UK ALASDAIR JUBB • Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, UK; CRUK Cambridge Institute, University of Cambridge, Cambridge, UK ELIZABETH KENYON • Precision Health Program, Michigan State University, East Lansing, MI, USA MATTI KIUPEL • Veterinary Diagnostic Laboratory, Michigan State University, Lansing, MI, USA JØRGEN KJEMS • Department of Molecular Biology and Genetics, Aarhus University, Aarhus, Denmark THOMAS KRONEIS • Division of Cell Biology, Histology & Embryology, Gottfried Schatz Research Center, Medical University Graz, Graz, Austria JOHN P. LE QUESNE • MRC Toxicology Unit, University of Cambridge, Cambridge, UK; Leicester Cancer Research Centre, University of Leicester, Leicester, UK; Leicester University Hospitals NHS Trust, University of Leicester, Leicester, UK STAN E. LINGLE • NanoString Technologies Inc., Seattle, WA, USA XIAO-JUN MA • Advanced Cell Diagnostics, A Bio-Techne Brand, Newark, CA, USA PAOLO MARGARIA • Leibniz Institute DSMZ—German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany CHRISTOPHER R. MERRITT • NanoString Technologies Inc., Seattle, WA, USA MICHAEL MILLAR • MRC Centre for Reproductive Health, University of Edinburgh, Queens Medical Research Institute, Edinburgh, UK TRINE MØLLER • Bioneer A/S, Molecular Histology, Hørsholm, Denmark ANNA MOORE • Precision Health Program, Michigan State University, East Lansing, MI, USA SANKARI NAGARAJAN • Nuclear Hormone Receptor Laboratory, Cancer Research UK/Cambridge Institute, University of Cambridge, Li Ka Shing Centre, Cambridge, UK BOYE SCHNACK NIELSEN • Bioneer A/S, Molecular Histology, Hørsholm, Denmark MATS NILSSON • Molecular Diagnostics, Science for Life Laboratory, Department of Biochemistry and Biophysics, Stockholm University, Solna, Sweden SHEILA J. M. O’CONNOR • Haematological Malignancy Diagnostic Service (HMDS), Level 3 Bexley Wing, Leeds Cancer Centre, St James’s University Teaching Hospital, Leeds, UK LEAH K. OFFICER • MRC Toxicology Unit, University of Cambridge, Cambridge, UK NILES A. PIERCE • Division of Biology & Biological Engineering, California Institute of Technology, Pasadena, CA, USA; Division of Engineering & Applied Science, California Institute of Technology, Pasadena, CA, USA; Weatherall Institute of Molecular Medicine, University of Oxford, Oxford, UK CHARLES PYKE • Pathology & Imaging, Global Discovery and Development Sciences, Novo Nordisk A/S, Copenhagen, Denmark KENNY ROBERTS • Wellcome Sanger Institute, Hinxton, UK MAAYAN SCHWARZKOPF • Division of Biology & Biological Engineering, California Institute of Technology, Pasadena, CA, USA MICHAEL A. SCHWIMMER • Department of Zoology, University of Cambridge, Cambridge, UK

Contributors

xiii

LORENZO F. SEMPERE • Precision Health Program, Michigan State University, East Lansing, MI, USA SAMAR SHEAT • Leibniz Institute DSMZ—German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany KRISTINA SORG • NanoString Technologies Inc., Seattle, WA, USA ANA V. TEODO´SIO • MRC Toxicology Unit, University of Cambridge, Cambridge, UK KATHRYN R. TURNER • Haematological Malignancy Diagnostic Service (HMDS), Level 3 Bexley Wing, Leeds Cancer Centre, St James’s University Teaching Hospital, Leeds, UK STEPHAN WINTER • Leibniz Institute DSMZ—German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany CHIKA YOKOTA • Molecular Diagnostics, Science for Life Laboratory, Department of Biochemistry and Biophysics, Stockholm University, Solna, Sweden; In situ sequencing facility, Science for Life Laboratory, Department of Biochemistry and Biophysics, Stockholm University, Solna, Sweden FAN YU • Department of Plant Biology, Michigan State University, East Lansing, MI, USA; National Engineering Research Center for Sugarcane, Fujian Agriculture and Forestry University, Fuzhou, Fujian, China ERIN ZALUZEC • Precision Health Program, Michigan State University, East Lansing, MI, USA HEATHER ZECCHINI • Light Microscopy Core Facility, Cancer Research UK/Cambridge Institute, University of Cambridge, Li Ka Shing Centre, Cambridge, UK BINGQING ZHANG • Advanced Cell Diagnostics, A Bio-Techne Brand, Newark, CA, USA DANIEL R. ZOLLINGER • NanoString Technologies Inc., Seattle, WA, USA HAILING ZONG • Advanced Cell Diagnostics, A Bio-Techne Brand, Newark, CA, USA

Part I General Applications

Chapter 1 Guidelines for the Optimization and Validation of In Situ Hybridization Julia Jones and William J. Howat Abstract As RNA in situ hybridization (ISH) moves into the mainstream lab and increasingly into clinical adoption and additional multiplexing techniques are developed to enable further RNA ISH identification, a set of guidelines on the validation of ISH is required. These guidelines include choice of methods, appropriate controls, and protocol optimization as well as a central core message of understanding the target, understanding the ISH technique, and using the most appropriate controlling mechanisms to enable reproducible and trustworthy data to be obtained. Key words Controls, In situ hybridization, Probes, Reproducibility, Validation

1

Introduction In situ hybridization (ISH), the detection of RNA or DNA in tissue sections and whole-mount preparations, has been in use since 1969 [1] where it was being used for the identification of DNA in Xenopus. It is a method in common employment across the scientific field and is in routine use in the clinic for the detection of DNA amplifications or fusions, such as ERBB2 or TMPRSS2/ERG, respectively, where ISH can be the final predeterminer of a patient outcome and treatment decision. In contrast, RNA ISH has been largely confined to a research tool for many years. It has a base popularity in the detection of messenger RNA (mRNA) species in whole-mount and developmental biology and continues to be used worldwide in the detection of mRNA and variants in plant and mammalian research [2] in both whole-mount and tissue sections [3]. Indeed, a number of searchable expression databases have been created, such as the murine developmental expression Atlas—Eurexpress (www. eurexpress.org/ee/), EMAGE [4], and the Allen Brain Atlas, where mapping of all available genes in the mouse and human brain has been performed [5, 6].

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_1, © Springer Science+Business Media, LLC, part of Springer Nature 2020

3

4

Julia Jones and William J. Howat

However, RNA ISH has often been seen as a difficult technique to establish, and, for many years, was the provision of a few experienced labs. Initially, a preference for the use of riboprobes for RNA detection required molecular biology skills to necessitate the production of complementary RNA, cloned and expressed in plasmids, in combination with radioactive 35S, or other radioactive isotopes; tagging of the riboprobe; and emulsion dipping of the resultant slide preparations [7]. The end result being silver grain deposition to provide the level of sensitivity required to see individual RNA species expression. The size of these probes allowed incorporation of many labeled nucleotides along the probe, making this type of probe more sensitive than a corresponding shorter oligo probe. However, they were sensitive to RNA degradation. As methodologies moved on, it became clear that the use of small 20–40 bp oligonucleotides created via PCR and with hapten tagging, such as digoxigenin or biotin, and use of enzyme-linked detection, such as alkaline phosphatase (AP) with nitroblue tetrazolium/bromo chloro indole phosphate (NBT/BCIP), could provide results of similar sensitivity and were subsequently used predominately in the whole-mount environment [8]. The creation of locked nucleic acids (LNA) and peptide nucleic acids (PNA) as detection probes moved the technology forward, particularly for the detection of small RNA species, such as microRNA. LNA probes are mixmers of standard oligonucleotides and oligonucleotide analogs, in which the ribose ring is constrained by a methylene bridge between the 20 -oxygen and the 40 -carbon reducing flexibility of the ribose ring. The probes have increased binding affinity for the complementary RNA or DNA and also allow lower hybridization temperatures and higher stringency washing conditions [9]. As such, a single small probe could be used to detect a 22 bp miRNA with high affinity and when aligned with amplification chemistry, such as tyramide signal amplification (TSA), gave the sensitivity required for detection [10, 11]. In early 2000, a further advancement in the form of branched DNA for detection of RNA and, more recently, RNA single-point mutations was introduced. This technology utilized “double-Z probes”: multiple pairs of 18–25 nucleotide probes that hybridize adjacent to one another. In addition to a target-specific sequence and a linker sequence, each probe also had an additional tail sequence that, when adjacent to its partner, forms a binding site for a preamplifier and creates an amplification structure with an enzymatic endpoint of horseradish peroxidase (HRP) or AP [12] in combination with chromogens or fluorescence to provide detection. Commercial detection kits were made available, such as

Guidelines for ISH

5

ViewRNA1 or RNAscope®2, coupled with predesigned sets of probes and aligned control probes. These could be run manually or on automated platforms which enabled labs that would have previously not had the technical expertise to undertake ISH staining to be able to do so with investment in materials. Lastly, a raft of ISH methods, such as hybridized chain reaction [13], Stellaris FISH [14], SABER [15], Codex [16], MERFISH [17], STARmap [18], and RollFISH [19], have entered into the academic and commercial environment focused on the requirement for multiplexing many probes within a single section/whole mount and enabling the understanding of the diversity of biological networks. Increasingly, RNA ISH is transitioning from research labs into clinical use with the demonstration that mRNA can be used as a surrogate biomarker and can outperform traditional IHC methods in certain circumstances [20–23]. Allied with this is the growing ability to use ISH on automated staining platforms with minimal development required. While this should lead to increased rates of reproducibility, ISH will in the future be more often used outside of the specialized ISH labs where it has traditionally existed. As such, there is a need for a set of guidelines for the inexperienced ISH researcher to follow to ensure that the best possible quality of research can be produced and interpreted. A summary of the key features of the guidelines can be found in Table 1.

2

Steps to Validation

2.1 Material Preparation

1 2

RNA is a labile molecule and has been shown to be lost over time, even when stored in paraffin blocks [24, 25]. While pre-analytical variables, surprisingly, have not been shown to be a significant factor in this, with Chung et al. [26] demonstrating that RNA yield remains equivalent when compared at RT to 4  C over a 12-h time period, Evers et al. [27] have shown that the paraffin embedding process does cause harm to RNA both in RNA yield and RNA quality. Formalin fixation alone causes RNA strand breaks into small 100–200 bp fragments [28] as well as cross-linking of RNA to protein and extended fixation in formaldehyde reduces quality of RNA significantly [29]. However, short fixation in formaldehyde can be equally damaging to the RNA molecule [30], with the optimal fixation time of 24–48 h preferred for preservation of RNA as well as protein. It is a known fact that RNases present in the environment and bodily fluids such as saliva, mucus, and perspiration can degrade

ViewRNA is a product from Thermo Fisher Scientific. RNAscope® is a registered trademark of Advanced Cell Diagnostics.

6

Julia Jones and William J. Howat

Table 1 Guide to preparing and validating an in situ hybridization assay Step

Considerations

Prepare material

Fix for 24–48 h in neutral buffered formaldehyde. Minimize exposure to endogenous RNase

Understand target

Full literature review Understand subcellular localization, cell expression, and biologic relevance to choose appropriate tissue and cell controls

Choose appropriate ISH technique

Understand the best choice or probe, pretreatment, hybridization, and stringency conditions

Determine specificity

Use cell lines to determine specificity before testing on known positive and negative tissue

Use appropriate controls

Determine positive control probe for comparison to the target and tissue positive controls Use negative controls for each sample Use run controls to determine variability of the assay over time

RNA and that, in order to perform RNA ISH successfully, the user has to be extremely careful in their preparation of the materials. This generally includes using RNase-removing wipes on the microtome blade, forceps, and water bath; use diethyl pyrocarbonate (DEPC)-treated water to ensure that the water bath is free from RNases and that gloves are worn at all times. It is possible that some of this level of paranoia is overstated [31], however, it has been difficult to verify. It is also true that RNA degrades over time, potentially degrading within the block itself and certainly with prolonged exposure to formaldehyde fixation [32, 33], and one of the reasons why small oligo formats or branched DNA has been successful is the redundancy factor of applying multiple probes to a fragmented RNA, allowing for a loss of some probe binding. 2.2 Understand Target

In all cases of validation, it is essential to understand the target and the expression of that target in your tissue of interest. Firstly, there is the requirement to understand the subcellular localization of the RNA species being targeted. While RNA is most likely present in the cytoplasm this may not always be the case, as certain RNA species, such as long noncoding RNAs, enhancer RNAs, and small nuclear RNAs, would be expected to be nuclear [34, 35]. Secondly, there is the understanding of the cell type expression. In some cases, such as targets expressed in the endo- or epithelium, the cell type is morphologically distinct and can be easily identified. However, where a target has a limited expression knowledge because of its novel nature or where the expression is in a subset of a generic cell, for example activated T cells, there should be no reliance on morphology alone in determining specificity. Central

Guidelines for ISH

7

databases, such as GeneCards (www.genecards.org) or miRbase (www.miRBase.org) [36], are helpful for gaining information on this. Lastly, the researcher should expend considerable effort in understanding the biological relevance of the target and the target expression as this will help to drive the methodology employed in its detection, as not all ISH staining methods have the same level of sensitivity. It must be noted that protein is not a surrogate for mRNA. While protein expression can provide a useful guide to the expression, it cannot be relied on as a predeterminer of mRNA expression or vice versa. 2.3

Methodology

A key driver in the validation of an ISH protocol is an understanding of the technology itself. All ISH protocols follow the same basic principles: 1. A nucleic acid probe to the target must be generated. 2. The sample must undergo pretreatment with heat and/or enzymes such as proteinase K or pepsin in order to combat tissue fixation processes and allow access of the probe to the target. 3. Hybridization of the probe to the target. 4. Removal of unbound probe through stringency washing. 5. Detection of the labeled probe to allow visualization of the target location within the cells/tissues. The probe type and thus the method to be used need to be carefully considered and will be based upon factors including sample type, target DNA or RNA species, number of targets that will be detected simultaneously, as well as cost and availability of specialist equipment required. Table 2 details the sensitivity, ease of use, and target suitability for each commonly used ISH method.

2.3.1 Probe Generation

A critical step in obtaining useful RNA localization data is the design of a probe that is specific to the target of interest only. It is imperative that regions of close homology to other RNAs are avoided in order to avoid cross-hybridization with other targets. Some vendors offer probes designed in-house and available for purchase direct from the catalog while others will design probes for you. Those skilled in molecular biology can design their own probes.

2.3.2 Riboprobes

Single-stranded antisense riboprobes can be created from doublestranded DNA templates by in vitro transcription using one of the three DNA-directed RNA polymerases (T7, T3, or SP6) while incorporating hapten-labeled uridine triphosphate bases. The simplest way of obtaining probe templates is to purchase Integrated

8

Julia Jones and William J. Howat

Table 2 Comparison of the sensitivity, ease of use, and target suitability of different probe types/ISH assays Target suitability

Method

Sensitivity

Ease of use

Riboprobe ISH

Medium. Long probes allow incorporation of many labels. Labels can be fluorescent, haptens, or enzymes that can be detected using labeled antibodies or substrates, respectively

Low. Generation of probes using mRNA, lncRNA plasmids. Requires knowledge of molecular biology in the probe design and probe generation. Can be time consuming

Branched DNA

High. Highly sensitive method that utilizes multiple pairs of probes to the target and subsequent rounds of amplification to produce high signal-to-noise ratio

High. Multiple kits, pre-optimized, available for chromogenic single-plex and duplex assays and fluorescent multiplex assays, for manual and automated workflows. Probe design and synthesis performed by vendor

Medium. Can be designed and Multiple oligos Medium. Useful where the ordered by the user or are target is long allowing for large available from commercial sets of fluorescently labeled vendors oligos. Best for use in cells and frozen tissues. Less successful for FFPE tissues Padlock probe Medium. A low-efficiency assay, due to the need for reverse and rolling transcription of mRNA to circle cDNA in the tissue prior to amplification hybridization of a padlock probe. Detection of mRNAs containing SNVs is possible

mRNA, lncRNA, eRNA

mRNA, lncRNA

Low. Involved protocol contains Highly similar sequences many steps and requires such as singlemolecular biology skills within point the lab for troubleshooting mutations purposes

LNA probes

Medium. 1-day protocol. Probes miRNA, Medium. Can be suitable for can be designed by the user or VtRNA, low-abundance targets if used purchased predesigned mRNA, in tandem with amplification lncRNA techniques such as TSA. Directly labeled probe suitable for high-abundance targets

Hybridization chain reaction

mRNA, High. Probe sets consist of probe Medium. Has been lncRNA, demonstrated in many sample pairs that must bind in tandem. eRNA preparations and is suitable for Two fluorescently labeled, multiplexing up to five colors metastable hairpins drive a currently cascade of polymerization when exposed to the initiator sequence contained within each probe pair resulting in signal amplification. High signal:noise due to conditional polymerization in the presence of probe pairs and both hairpins

Guidelines for ISH

9

Molecular Analysis of Genomes and their Expression (I.M.A.G.E.) Consortium cDNA clones [37] from a commercial vendor where it saves significant time and effort if you purchase sequence-verified clones. The probe template must be linearized using restriction enzymes to avoid transcription of plasmid sequences. Restriction enzymes that give a 50 overhang or blunt ends are best as 30 overhangs can lead to the generation of probes containing multiple copies of the target sequence. Alternatively, PCR amplification using cDNA from an appropriate tissue is another method for probe generation. The resulting PCR product may be ligated into a suitable vector, containing two of T3, T7, or SP6. Alternatively, T7, T3, or SP6 sequence may be incorporated into the primer sequence to enable in vitro transcription directly from a PCR product. The probe design template from Whitehead Institute for Biomedical Research can be of assistance (http://biotools.umassmed.edu/bioapps/primer3_www.cgi). 2.3.3 Oligonucleotide Probes

For the generation of oligonucleotide probe set designs, ProbeMaker [38] http://probemaker.sourceforge.net/ is a very useful open-source software that supports generation of target-specific sequences as well as additional sequence elements, or “tags” that allow multiplex detection of different probe sets.

2.3.4 Pretreatment

Pretreatment can follow two principal forms, protease retrieval or heat-mediated retrieval, or a combination of both. Pretreatment with proteases is the most common single form of retrieval used where it increases the accessibility of the target RNA by degrading proteins. Optimal concentration of protease and length of incubation must be determined for each tissue type and for different methodologies. For example, long riboprobes are likely to require more extensive protease digestion than shorter oligo probes. Heat-mediated retrieval is less common in ISH protocols but is required for branched DNA protocols in certain sample preparations, such as FFPE tissue sections. Heat-mediated retrieval is the same as is used in immunohistochemistry and relies on the use of a sodium citrate pH 6 or Tris-EDTA pH 9 buffers or derivatives of the same. These are generally used in a pressure cooker, microwave, or rice steamer to boil the tissue with retrieval buffer and break the disulfide bonds created during the formaldehyde fixation process. Different tissue preparations will require alternative pretreatments. For example, frozen tissue may not need any pretreatment, while whole-mount preparation, such as embryos, may require a combination of heat, protease, or additional acid and detergentbased pretreatments. It is recommended that when optimizing pretreatment conditions the concentration of protease or the incubation time is changed individually but not together. Similarly, heat

10

Julia Jones and William J. Howat

retrieval conditions should be optimized by altering either the temperature or the incubation time and in all cases assessment of the sensitivity of the probe staining must be aligned with the morphology of the tissue as these are closely related, in that a good morphology can be easily lost when trying to strengthen the signal from the probe. 2.3.5 Hybridization

A number of factors can be manipulated for probe hybridization when not using a predefined kit and probe concentration. Where probe staining is suboptimal and the morphology is intact, the probe concentration can be increased and may require to be increased when transitioning from cells to tissues in order to ensure that there is an excess of probe. However, caution should be exercised as too much probe may give rise to an increase in the level of background signal. Hybridization is generally performed at a temperature around 25  C below the melting temperature of the probe; however, this can be deleterious to the tissue morphology. To mitigate this, the addition of formamide to the hybridization buffer allows stringent hybridization at a lower temperature and can additionally allow lengthened hybridization incubation times to boost the signal without affecting the morphology.

2.3.6 Stringency

Once hybridized, careful manipulation of the washing conditions can remove nonspecific probe binding that persists despite manipulation of the hybridization conditions. Stringency can be modulated during post-hybridization washing steps by reducing salt concentration in the wash buffer, as well as increasing the temperature by 2–3  C above the hybridization temperature. The addition of formamide to the wash buffer allows higher stringency to be achieved without damage to the tissue.

2.3.7 Detection

Probe detection follows four main methods. Firstly, radioactively labeled probes with emulsion dipping of the slides enable silvergrain deposition on the slide and subsequent visualization (see Fig. 1). However, as use of radioactivity for ISH has decreased due to storage, disposal, and availability issues, most current methods use enzymatic or fluorescent detection. For enzymatic methods, primarily AP is used combined with NBT/BCIP providing a bluepurple deposition of signal around the probe due to the ability to leave AP substrates on for prolonged periods of time for sensitivity. However, HRP with diaminobenzidine (DAB) can be used as an alternative with additional signal amplification provided by TSA. Finally, fluorescence is generally employed when attempting to multiplex targets.

Guidelines for ISH

11

Fig. 1 Radioactive ISH. Bright-field (a) and dark-field (b) images of Olfm4 mRNA expression in mouse small intestine showing deposited silver grains in black and white, respectively 2.4

Troubleshooting

2.5 Determining Specificity

Table 3 details advice on the causes and solutions to poor ISH staining. The use of cell lines in pathology for validating antibody expression has been steadily increasing and a large number of cell lines are available, either through academic collaborations or commercially where the expression characteristics have been determined, e.g., NCI60 cell line panel (https://dtp.cancer.gov/discovery_develop ment/nci-60/cell_list.htm). Once positive and negative cells have been chosen, such cells can be grown, fixed, and pelleted before processing into paraffin wax. At this point they can be treated in the same manner as would tissue, mimicking the behavior of the tissue with regard to the probe and the technique. Additionally, siRNA knockdown (KD) or CRISPR/CAS9 knockout (KO) procedures can be applied to positive cell lines to further enhance the validation and definitively demonstrate that the probe is specific for the RNA it is targeting. Where KD is not feasible, overexpression techniques can also be used (see Fig. 2). However, it must be recognized that the expression of the RNA in the cells is not the same as the expression of the RNA in tissue and following cell line specificity screening, the application of the probe to known positive and negative tissue is essential for determining specificity and ensuring a lack of nonspecific probe binding. The Human Protein Atlas (HPA) can be a useful tool for choosing positive and negative tissue as RNA data from HPA (https://www.proteinatlas.org/), Genotype-Tissue Expression project (https://www.gtexportal.org/home/), and FANTOM5

12

Julia Jones and William J. Howat

Table 3 Troubleshooting guide for common problems in ISH Problem

Possible causes

Solutions

No signal

1 2 3 4

1 Use a tissue known to express the target 2 Verify that the RNA quality is good by using a housekeeping control probe 3 Increase probe concentration, where possible 4 Modify hybridization conditions such as longer incubation time, less stringent hybridization buffer, lower hybridization temperature 5 Increase salt concentration of wash buffer or lower incubation temperature 6 Check that the probe sequence is correct 7 Change detection method to increase sensitivity

High background

1 Decrease probe concentration 1 Probe too concentrated 2 Decrease salt concentration and/or addition 2 Washing conditions not of formamide to wash buffer, try increasing stringent enough the incubation temperature. For riboprobes, 3 Probe sequence—riboprobes a post-hybridization RNase A treatment will may contain plasmid sequence degrade any remaining single-stranded RNA 4 Probe sequence—may be 3 Ensure that probe template is linearized suboptimal before start of plasmid sequence 4 Design the probe to a different target region

High signal

1 For hapten-labeled probes, enzymatic detection may not be optimal

1 Optimize detection conditions, such as antibody/substrate concentration and incubation times

Background in wrong cell/ tissue types

1 Some tissues are more adherent than others

1 Try a different probe, decrease probe concentration, adjust the hybridization and washing conditions

Nuclear signal

1 Detection of DNA sequence 2 RNA may also be nuclear or may be detecting immature RNAs

1 Incubation with DNase and RNase on separate slides prior to probe hybridization will aid determination of signal origin. Ensure that sections are washed adequately, in order that probes are not destroyed

Absence of target Poor RNA quality in sample Probe concentration too low Suboptimal hybridization conditions 5 Post-hybridization washes too stringent 6 Poor probe sequence 7 Low sensitivity of detection method

project (http://fantom.gsc.riken.jp/5/) that have been incorporated into the protein expression profiles. Some additional key differences remain between using cell lines for specificity and tissue. Firstly, the fixation of the cell lines, while designed to mimic tissue, is often not as long and likely better controlled than that experienced by tissue, particularly in the clinical context. Additionally, it is probable that the cell lines will be freshly prepared, in contrast to tissue blocks that may be decades old. Overall, the use of cell lines with known or engineered expression of a specific RNA is a highly valuable tool in the validation of the ISH technique.

Guidelines for ISH

13

Fig. 2 Use of cell lines in determining probe specificity. miR-15a expression detected in cell lines. (a) Overexpression, (b) wild type, (c) knockout 2.6

Controls

Critical to determining the sensitivity of ISH are the controls employed. Using the correct control is vital to the overall experimental design and the following controls are recommended: 1. Positive controls in the form of tissue positive controls and probe positive controls should be employed in all circumstances. 2. Negative controls in the form of negative probe controls should be employed for every piece of tissue. 3. Run controls.

2.6.1 Positive Controls

A housekeeping gene with uniform expression in the tissue of interest is strongly recommended. Without such a control, the researcher is unable to determine whether the “test” tissue is negative due to a lack of expression of the target or due to the loss of the RNA during the preparation technique, staining technique, or fixation. This investigation should be performed at the earliest stage when investigating the target and the positive tissue. The positive control probe should be carefully selected in order that it is of a similar expression level to the target. This is particularly relevant when the target RNA is expressed at a low level as highly expressed control RNAs, such as ubiquitin C, may be detected even using suboptimal pretreatment and hybridization conditions. GAPDH is a good general control and has been recommended in astrocytomas [39], but has been shown to have a variable expression in other circumstances [40] while the DNA-directed RNA polymerase gene, POLR2A, has been shown to be conserved in tissues and can function as a useful control [41]. An alternative, used frequently for branched DNA, is PPIB, a cyclophilin which is expressed reproducibly but at relatively high level in most samples (see Fig. 3). It is therefore important to choose the most appropriate control for your specific tissue type from the available literature. Furthermore, when running a control probe in

14

Julia Jones and William J. Howat

Fig. 3 High-, medium-, and low-abundance mRNA control probes. Ubiquitin C (UBC) expression in mouse kidney (a) and mouse small intestine (d), peptidyl-prolyl cis-trans isomerase B (PPIB) expression in mouse kidney (b) and mouse small intestine (e), and DNA-directed RNA polymerase II subunit RPB1 (POLR2A) expression in mouse kidney (c) and mouse small intestine (f)

a multiplexing format, the inclusion of this control within the multiplex itself is recommended, eliminating any doubt over issues of penetration of fixation in large tissue sections or heterogeneity of the control probe itself, as this will be evident. Alternatively, a cocktail of control probes to multiple housekeeping targets, all stained at the same time, could additionally eliminate any bias from any individual control. In addition to the positive control probe, it is recommended that a positive control tissue is utilized. This control can be placed onto the same slide as the “test” tissue or alongside within the same experiment as a run control. Using such controls helps to determine whether the overall experiment has performed successfully, tests the kit variability, and ensures that the result from each slide is fully verifiable. For example, Fig. 4 shows the staining obtained using PPIB over time within a single lab, demonstrating no drift in ISH sensitivity, despite lot of differences between kits and probes. 2.6.2 Negative Controls

Negative controls in the form of a probe that is not expected to bind to a target in the tissue of interest should be used to demonstrate the level of background staining present in the system. While the overall background should not vary greatly, the negative control

Guidelines for ISH

15

Fig. 4 Inter-run variability controls. PPIB control probe was hybridized to mRNA within sections from the same block of mouse small intestine in every run. Images shown are from March 2018 (a, e), April 2018 (b, f), July 2018 (c, g), and October 2018 (d, h)

helps to determine whether the background in all cases is zero especially where there may be variability in prefixation and fixation conditions and is particularly relevant when using image analysis as a quantifiable measurement of success. Dihydrodipicolinate reductase (DapB) mRNA, a soil bacterium transcript unlikely to be found in normal human tissue, is frequently used in branched DNA methods. However, other probes such as scrambled or nonsense probes can be designed. In each case, scrambled probes must be BLAST searched against the genome to ensure that there are no similarities resulting in genuine positive signal for the negative control.

Acknowledgments We thank members of the Histopathology/ISH Core Facility, CRUK-Cambridge Institute, and Nathan Benaich, Doug Winton, and Nikki March for kind permission to use images. References 1. Gall JG, Pardue ML (1969) Formation and detection of RNA-DNA hybrid molecules in cytological preparations. Proc Natl Acad Sci U S A 63:378–383. https://doi.org/10.1073/ pnas.63.2.378 2. Wojcik AM, Mosiolek M, Karcz J et al (2018) Whole mount in situ localization of miRNAs and mRNAs during somatic embryogenesis in

Arabidopsis. Front Plant Sci 9:1277. https:// doi.org/10.3389/fpls.2018.01277 3. Koshiba-Takeuchi K (2018) Whole-mount and section in situ hybridization in mouse embryos for detecting mRNA expression and localization. Methods Mol Biol 1752:123–131. https://doi.org/10.1007/978-1-4939-77147_12

16

Julia Jones and William J. Howat

4. Richardson L, Stevenson P, Venkataraman S et al (2014) EMAGE: electronic mouse atlas of gene expression. Methods Mol Biol 1092:61–79. https://doi.org/10.1007/9781-60327-292-6_5 5. Hochheiser H, Yanowitz J (2007) If I only had a brain: exploring mouse brain images in the Allen Brain Atlas. Biol Cell 99:403–409. https://doi.org/10.1042/BC20070031 6. Shen EH, Overly CC, Jones AR (2012) The Allen Human Brain Atlas: comprehensive gene expression mapping of the human brain. Trends Neurosci 35:711–714. https://doi. org/10.1016/j.tins.2012.09.005 7. Mahmood R, Mason I (2008) In-situ hybridization of radioactive riboprobes to RNA in tissue sections. Methods Mol Biol 461:675–686. https://doi.org/10.1007/ 978-1-60327-483-8_45 8. Pringle JH, Primrose L, Kind CN et al (1989) In situ hybridization demonstration of polyadenylated RNA sequences in formalin-fixed paraffin sections using a biotinylated oligonucleotide poly d(T) probe. J Pathol 158:279–286. https://doi.org/10.1002/ path.1711580403 9. Silahtaroglu AN, Tommerup N, Vissing H (2003) FISHing with locked nucleic acids (LNA): evaluation of different LNA/DNA mixmers. Mol Cell Probes 17:165–169 10. Silahtaroglu AN, Nolting D, Dyrskjot L et al (2007) Detection of microRNAs in frozen tissue sections by fluorescence in situ hybridization using locked nucleic acid probes and tyramide signal amplification. Nat Protoc 2:2520–2528. https://doi.org/10.1038/ nprot.2007.313 11. Turnock-Jones JJ, Le Quesne JP (2014) MicroRNA in situ hybridization in tissue microarrays. Methods Mol Biol 1211:85–93. https://doi.org/10.1007/978-1-4939-14593_8 12. Player AN, Shen LP, Kenny D et al (2001) Single-copy gene detection using branched DNA (bDNA) in situ hybridization. J Histochem Cytochem 49:603–612. https://doi. org/10.1177/002215540104900507 13. Choi HMT, Schwarzkopf M, Fornace ME et al (2018) Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust. Development 145. https://doi.org/10.1242/dev.165753 14. Orjalo A Jr, Johansson HE, Ruth JL (2011) Stellaris™ fluorescence in situ hybridization (FISH) probes: a powerful tool for mRNA detection. Nat Methods 8:884. https://doi. org/10.1038/nmeth.f.349

15. Kishi JY, Lapan SW, Beliveau BJ et al (2019) SABER amplifies FISH: enhanced multiplexed imaging of RNA and DNA in cells and tissues. Nat Methods 16:533–544. https://doi.org/ 10.1038/s41592-019-0404-0 16. Goltsev Y, Samusik N, Kennedy-Darling J et al (2018) Deep profiling of mouse splenic architecture with CODEX multiplexed imaging. Cell 174(968–981):e915. https://doi.org/ 10.1016/j.cell.2018.07.010 17. Xia C, Babcock HP, Moffitt JR et al (2019) Multiplexed detection of RNA using MERFISH and branched DNA amplification. Sci Rep 9:7721. https://doi.org/10.1038/ s41598-019-43943-8 18. Wang X, Allen WE, Wright MA et al (2018) Three-dimensional intact-tissue sequencing of single-cell transcriptional states. Science 361. https://doi.org/10.1126/science.aat5691 19. Wu C, Simonetti M, Rossell C et al (2018) RollFISH achieves robust quantification of single-molecule RNA biomarkers in paraffinembedded tumor tissue samples. Commun Biol 1:209. https://doi.org/10.1038/ s42003-018-0218-0 20. Humphries MP, McQuaid S, Craig SG et al (2019) Critical appraisal of programmed death ligand 1 reflex diagnostic testing: current standards and future opportunities. J Thorac Oncol 14:45–53. https://doi.org/10.1016/j. jtho.2018.09.025 21. Bingham V, Ong CW, James J et al (2016) PTEN mRNA detection by chromogenic, RNA in situ technologies: a reliable alternative to PTEN immunohistochemistry. Hum Pathol 47:95–103. https://doi.org/10.1016/j. humpath.2015.09.009 22. Ferrone CR, Ting DT, Shahid M et al (2016) The ability to diagnose intrahepatic cholangiocarcinoma definitively using novel branched DNA-enhanced albumin RNA in situ hybridization technology. Ann Surg Oncol 23:290–296. https://doi.org/10.1245/ s10434-014-4247-8 23. Kang H, Antonarakis ES, Luo J et al (2018) Detection of AR-V7 transcript with RNA in situ hybridization in human salivary duct cancer. Oral Oncol 84:134–136. https://doi.org/ 10.1016/j.oraloncology.2018.06.026 24. Groelz D, Viertler C, Pabst D et al (2018) Impact of storage conditions on the quality of nucleic acids in paraffin embedded tissues. PLoS One 13:e0203608. https://doi.org/10. 1371/journal.pone.0203608 25. Baena-Del Valle JA, Zheng Q, Hicks JL et al (2017) Rapid loss of RNA detection by in situ hybridization in stored tissue blocks and

Guidelines for ISH preservation by cold storage of unstained slides. Am J Clin Pathol 148:398–415. https://doi.org/10.1093/ajcp/aqx094 26. Warren M, Chung YJ, Howat WJ et al (2010) Irradiated Blm-deficient mice are a highly tumor prone model for analysis of a broad spectrum of hematologic malignancies. Leuk Res 34:210–220. https://doi.org/10.1016/j. leukres.2009.06.007 27. Evers DL, He J, Kim YH et al (2011) Paraffin embedding contributes to RNA aggregation, reduced RNA yield, and low RNA quality. J Mol Diagn 13:687–694. https://doi.org/10. 1016/j.jmoldx.2011.06.007 28. Cronin M, Ghosh K, Sistare F et al (2004) Universal RNA reference materials for gene expression. Clin Chem 50:1464–1471. https://doi.org/10.1373/clinchem.2004. 035675 29. Macabeo-Ong M, Shiboski CH, Silverman S et al (2003) Quantitative analysis of cathepsin L mRNA and protein expression during oral cancer progression. Oral Oncol 39:638–647 30. Chung JY, Braunschweig T, Williams R et al (2008) Factors in tissue handling and processing that impact RNA obtained from formalinfixed, paraffin-embedded tissue. J Histochem Cytochem 56:1033–1042. https://doi.org/ 10.1369/jhc.2008.951863 31. Tongiorgi E, Righi M, Cattaneo A (1998) A non-radioactive in situ hybridization method that does not require RNase-free conditions. J Neurosci Methods 85:129–139 32. von Ahlfen S, Missel A, Bendrat K et al (2007) Determinants of RNA quality from FFPE samples. PLoS One 2:e1261. https://doi.org/10. 1371/journal.pone.0001261 33. Schmeller J, Wessolly M, Mairinger E et al (2019) Setting out the frame conditions for feasible use of FFPE derived RNA. Pathol Res Pract 215:381–386. https://doi.org/10. 1016/j.prp.2018.12.027

17

34. Zhang T, Tan P, Wang L et al (2017) RNALocate: a resource for RNA subcellular localizations. Nucleic Acids Res 45:D135–D138. https://doi.org/10.1093/nar/gkw728 35. Mas-Ponte D, Carlevaro-Fita J, Palumbo E et al (2017) LncATLAS database for subcellular localization of long noncoding RNAs. RNA 23:1080–1087. https://doi.org/10.1261/ rna.060814.117 36. Griffiths-Jones S, Grocock RJ, van Dongen S et al (2006) miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res 34:D140–D144. https://doi.org/10. 1093/nar/gkj112 37. Lennon G, Auffray C, Polymeropoulos M et al (1996) The I.M.A.G.E. Consortium: an integrated molecular analysis of genomes and their expression. Genomics 33:151–152. https://doi.org/10.1006/geno.1996.0177 38. Stenberg J, Nilsson M, Landegren U (2005) ProbeMaker: an extensible framework for design of sets of oligonucleotide probes. BMC Bioinformatics 6:229. https://doi.org/ 10.1186/1471-2105-6-229 39. Gresner SM, Golanska E, Kulczycka-Wojdala D et al (2011) Selection of reference genes for gene expression studies in astrocytomas. Anal Biochem 408:163–165. https://doi.org/10. 1016/j.ab.2010.09.010 40. Glare EM, Divjak M, Bailey MJ et al (2002) Beta-actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax 57:765–770. https://doi.org/10. 1136/thorax.57.9.765 41. Bingham V, McIlreavey L, Greene C et al (2017) RNAscope in situ hybridization confirms mRNA integrity in formalin-fixed, paraffin-embedded cancer tissue samples. Oncotarget 8:93392–93403. https://doi. org/10.18632/oncotarget.21851

Chapter 2 Overcoming Autofluorescence (AF) and Tissue Variation in Image Analysis of In Situ Hybridization Cara Brodie Abstract Fluorescent detection of nucleic acid sequences such as DNA or RNA allows for multiplexing and visualization of an increased number of targets compared with chromogenic methods. This is due to the number of chromogens available as well as the ability of image analysis software platforms to distinguish between colors. Autofluorescence (AF) can be problematic during fluorescent imaging because the AF interferes with the detection of the specific fluorescent signals especially when the target signals are weak. AF has a broad emission spectrum leading to difficulty when performing image analysis due to masking of the specific signal across multiple wavelengths. Tissue sample variation can also affect levels of AF. In this chapter we share a method for overcoming the issues caused by sample variation and AF using HALO software on RNAscope in situ hybridization images. Key words Autofluorescence, HALO, Image analysis, ISH, RNAscope

1

Introduction Fluorescent visualization of in situ hybridization (ISH) analysis has been used since the 1980s when Bauman [1] labeled the 30 ends of the probes for detection of cRNA with a fluorochrome. The use of fluorophores for the detection of ISH has a number of advantages over chromogenic visualization such as the number of probes that can be detected at once and the ability to easily visualize any co-localization. The advances in slide scanning, imaging, and downstream analysis technologies have allowed for more fluorophores to be used simultaneously, allowing for multiplexing of up to nine probes. With narrow band-pass filter sets, spectral unmixing software, and ability to synchronize sequential sections or the same section imaged a number of times with multiple rounds of hybridization and signal stripping, the opportunities for research are expanding. Technologies such as imaging mass cytometry (IMC) from Fluidigm enable visualization of up to 20 RNA targets [2].

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_2, © Springer Science+Business Media, LLC, part of Springer Nature 2020

19

20

Cara Brodie

There are a multitude of in situ hybridization (ISH) techniques that can be employed in the detection of nucleic acid sequences, including DNA, messenger RNA (mRNA), microRNA (miRNA), and long noncoding RNA (lncRNA). The method used will be dependent on the target of interest [3]. Fluorescent ISH (FISH) is a powerful tool used worldwide in both clinical diagnosis and research for the detection of DNA and investigations into chromosome structure or function [4]. The probes can be long in length or can consist of a cocktail of probes that effectively paint the entire chromosome. Within the clinic it is routinely used for the detection of Her2 to determine if a patient will be responsive to herceptin therapy [5]. Branched DNA ISH assays such as RNAscope (Bio-Techne) and ViewRNA (Thermo Fisher Scientific) provide single-cell resolution of spatial and morphological data for a variety of RNA types. They utilize highly specific Z oligonucleotide probe designs, which hybridize to the target RNA strand in pairs. The binding of the paired Z probes ensures that only the target RNA is detected as the preamplifier oligo is unable to hybridize effectively to a single bound Z probe. Target RNAs are visualized with probe labels, either a fluorescent molecule or a chromogenic enzyme in conjunction with an appropriate substrate, and appear as punctate dots, where each dot represents a single RNA molecule [6]. RNAscope kits are available as fully automated assays on the Leica Bond RX (Leica Biosystems), with detection of up to four targets in the same tissue section currently (Fig. 1). The future release of the HiPlex assay will allow visualization of 12 targets in the same section. In situ mutation detection is a technique used to identify the presence of single-point mutant transcripts or low-frequency variants [7]. The BaseScope method is based on the RNAscope technique, but one Z probe pair is amplified instead of the standard 20 pairs. This means that a single-nucleotide alteration can be detected and visualized. An alternative method has been described by Larsson et al. (2010) [8] for the detection of single-point mutations in tissues using padlock probes. These are highly selective probes that once hybridized to the target sequence are converted to circular molecules that are amplified with rolling circle amplification [8]. The variety of ISH techniques can be used in conjunction with alternative methods of molecular analysis, such as next-generation sequencing and flow cytometry data. These techniques complement each other allowing for a deeper understanding of nucleic acid sequences. Patient samples from tissue and biobanks are an invaluable source of information and have a great potential for research. A large number of these samples have follow-up details that can be

Overcoming AF with Image Analysis

21

Fig. 1 Overview of detection method with RNAscope. Probe sets are comprised of double-Z oligonucleotides; these hybridize to the target RNA strand. Serial amplification steps are performed where preamplifier strands bind only to double-Z probes bound adjacent to one another, followed by amplifier to preamplifier strands and finally the probe label to the amplifier strands. The probe labels contain horseradish peroxidase (HRP) or alkaline phosphatase (AP) which can be detected with chromogenic substrates or fluorescent-labeled substrates

useful for research studies, such as the diagnosis, type of treatment the patient received, and sometimes long-term information such as relapse and cause of death [9]. The majority of archival tissue samples are formalin fixed paraffin embedded (FFPE) as this is the gold standard (for sample longevity) used in the UK for histopathology [10]. The formaldehyde reacts with primary amines, purines, phenols, amino-methylol groups, and thiols forming cross-linking –CH2 bonds known as methylene bridges. The reaction creates a reactive iminium ion, which in turn reacts with the phenol group of tyrosine resulting in the formation of a covalent bond. These bridges preserve the tissue and affect the primary structure of proteins and the preservation of DNA and RNA [11]. There are many pre-analytical factors that can affect the downstream ISH analysis of FFPE samples, which must be considered when using them. During and post-surgery the length of time between the sample having reduced or no blood supply prior to fixation is known as the ischemia time. During this time hypoxia and metabolic stress will occur and delays of more than 3 h have been found to cause significant degradation in signal intensity when analyzing HER2 via FISH analysis [12]. This is due to the degradation of the nucleic acids [13].

22

Cara Brodie

The fixation of the sample can also have downstream effects. Optimal length of fixation is 24 h at room temperature for samples that have been prepared correctly. Variations such as an increase in the temperature will speed up fixation and could lead to overfixation of the sample. The sample size can also affect the rate at which the formalin penetrates the tissue; too large a sample that has been inadequately sliced will suffer from under-fixation in the center of the sample [13]. The use of fluorescence for ISH analysis has increased with the advances in imaging technologies. Autofluorescence (AF) remains the most common issue within experiments in fixed mammalian tissue. This can be problematic as it may mask signal and make weak signal difficult to detect; this is a particular issue for low-abundance targets. AF can also hinder the determination of co-localization, be detected falsely as positive signal, and make downstream analysis arduous [14]. AF occurs when external FFPE processing of the tissue or internal tissue components naturally fluoresce across a certain excitation and emission wavelength [15]. Endogenous factors and extracellular matrix components such as lipofuscins, red blood cells (RBC), collagen, and elastin naturally fluoresce at many wavelengths. Tissues that have high levels of RBC will autofluoresce more than others for example liver, pancreas, and any tissues within the abdominal cavity [16]. RBC cause issues when analyzing ISH images as the size and shape can be mistaken for signal by many image analysis algorithms. The hemoglobin contained within RBC absorbs light at wavelengths below 600 nm causing them to fluoresce in many channels [15]. There are a number of remedies that can limit endogenous AF such as Sudan black B, trypan blue, and pontamine sky blue; however these can shift the background staining further up the spectrum (600 nm and above). There are also products that will quench AF such as TrueView (Vector Labs). Exogenous factors come from tissue processing stages. The Schiff bases formed during formalin fixation naturally autofluoresce, the media in which the sample is embedded in can lead to AF, and some paraffin waxes are thought to alter the molecular structure within tissues leading to AF [15]. As the number of probes that can be detected within one sample increases, the need becomes greater for downstream image analysis (IA) to be performed automatically within a digital pathology setting. IA is the computer-aided, quantitative analysis of digital images for the extraction of numerical, structural, and intensity data [17]. IA has been used for a number of years; in 1969 Spriggs published a method of automatic scanning for malignant cells in cervical smears based on nuclear size [18]. It is an area that is still developing rapidly, with new techniques emerging and a variety of commercial and open-source software options available. There are a number of reviews comparing many of these software options with the pros and cons of each being described. Holzer et al. 2019

Overcoming AF with Image Analysis

23

[19] compared four platforms against the manual quantification of RNAscope PPIB probe in non-small cell lung carcinoma tissue microarray (TMA) sections; a positive correlation between the IA platforms and the manual scoring was seen [19]. Wiesmann et al. (2015) [20] compared 15 open-source platforms creating a functionality/usability graph grading each one and grading them on a number of criteria, such as the ease of use for the interface and the amount of training material available for the software [20]. Available software options vary a great deal in terms of the level of computational knowledge required to create robust methods of analyzing images, compatible image file types, and aim of the analysis. Commercial options such as HALO (Indica Labs) and Visiopharm and also the open-source software QuPath can handle a large variety of whole-slide image formats and are user friendly with detailed user guides available. Software platforms such as CellProfiler, which was developed in 2010, and FIJI/ImageJ are Java-based programs, and Icy and MATLAB are open-source software that work by creating script pipelines and plug-ins that are shared within the community to create the analysis workflow. The disadvantage to these is that they do require some knowledge of the programming language and also have image file size limitations. Whole-slide images can contain multiple fluorophores and Z-stacks with little or no compression, meaning that these files can range from 1 to 40 GB or more in size. Image size limits the available software that can perform image analysis optimally; the opensource software, such as ImageJ, are not optimized to work with large file sizes [21]. The inherent variability within human tissue samples described above can lead to increased variation in the AF within batches of samples. The human eye is able to deal with levels of variation, but when using image analysis software, considerations must be made about how to overcome the problem. A variety of methods have been described for AF identification and removal such as the “autofluorescence remover” code created by Baharlou et al. (2019) [22], which can be integrated into a multitude of protocols across different interfaces, including FIJI, MATLAB, and R [22]. To implement the code, users must have knowledge of coding within the platform. The potential for image analysis to aid both research and clinical diagnosis is invaluable; for example scoring of FISH Her2 is done manually by a pathologist, but the advances in technology mean that automation is possible once it is validated for use. Many of these issues were discussed by Pell et al. (2019) [23] for the future adoption of IA and digital pathology across the clinical trial setting. The protocol below describes options for overcoming AF and inherent sample variation using HALO software. The software has a variety of modules enabling tissue classification, cell segmentation,

24

Cara Brodie

ability to identify positive IHC or ISH signal, and counting of probe spots in both bright-field and fluorescent images. Our laboratory is a high-throughput histopathology service with different whole-slide scanners. As a result, the IA software required must be suitable for use with a variety of image formats, intuitive to use and able to manage the large throughput volume.

2

Materials 1. Slides assayed using Multiplex RNAscope reagents (Bio-Techne) detected with TSA plus-Cy3, TSA plus-Cy5 fluorophores (Akoya Biosciences) with DAPI counterstain. 2. Zeiss Axio Scan Z1 (filter set details) giving CZI image formats. 3. Indica labs HALO software. 4. Modules: FISH (dynamic module), tissue classifier, TMA segmentation, area quantification FL (dynamic module).

3

Methods

3.1 Imaging of the Slides

1. Slides were scanned at 40 magnification on the Zeiss Axio Scan Z1, with a resolution of 0.11 μm/pixel. This magnification gives the best resolution for the downstream analysis due to the size of the RNA spots. The excitation and emission filter sets that were used and the exposure times are detailed in Table 1. 2. As the scanned slides were TMA sections the scanning profile was set up to focus on every point to ensure that each core was in focus (see Note 1).

3.2 Analyzing Images: TMA Segmentation

1. Open the HALO software and import the CZI image files for analyzing.

Table 1 Details of the channels used for visualization of targets, the nuclear stain, and the background channel

Probe/stain

Excitation filter set

Emission filter set

Exposure time (ms)

LED intensity

Nuclear stain (DAPI)

330–375

430–470

15

50%

Probe 1 (Cy3)

540–560

578–640

950

100%

Probe 2 (Cy5)

630–650

670–710

350

100%

Background channel (FITC)

465–490

514–548

500

50%

Overcoming AF with Image Analysis

25

Fig. 2 TMA map overview. An image of the TMA module within HALO; the map is superimposed over the cores and is manipulated to ensure that each core is within the correct circle and box. There are a number of settings that can be adjusted to allow for accurate separation of each core

2. To analyze the TMA cores individually, the TMA module is selected. The TMA Excel map is imported into the software creating an annotation grid that is maneuvered over the image separating each individual core, shown in Fig. 2 (see Note 2). 3.3 Analyzing Images: Creating the Classifier

1. Open around 30–50% of the images within HALO; these will be used for setting up both the classifier and the downstream analysis. Using this number of images to create the algorithms will ensure that the range of AF and sample variation is captured. 2. Load the random forest tissue classifier module and create the tissue classes that will be identified, for example, tissue, background, and glass (see Note 3). 3. Within the advanced options of the module select the fluorophores that will be used to identify the different classes (see Note 4). The FITC (background channel) and DAPI (nuclear) fluorophores were selected to create this classifier; this can be seen in Fig. 3. 4. Remaining in the advanced option the probability map output was selected and the probability set to 50% (see Note 5).

26

Cara Brodie

Fig. 3 TMA core containing AF and probe signal. (a) DAPI only. (b) The DAPI and FITC, with white arrowheads highlighting the RBC and fibrosis AF. (c) All four channels with the white arrowheads highlighting the positive probe signal

Fig. 4 The HALO interface. Screenshot of the HALO interface of the classifier module, with the annotations of examples of the tissue classes. The advanced classifier options window is open showing the tabs where the output of the classifier can be selected (mask or probability map) and also the fluorophores that are used to identify the tissue classes

5. The resolution of the classifier was set to 1.04 μm/pixel, with a minimum object size of 0. 6. Using the pen tools, annotations were created giving examples for each of the tissue classes, from a variety of the images open. This ensures that a broad range of AF background and tissue examples were given. The software uses these annotations to then create decision trees based on color and texture to define which tissue class the area belongs to. Figure 4 is an image of the HALO interface with the parameters that can be adjusted while creating a classifier.

Overcoming AF with Image Analysis

27

7. Using the “real-time tuning” window the classifier is checked to ensure that only the tissue areas are being detecting and not the AF. If there are any areas that are misidentified, then further annotations are added. The classifier is saved ready to input into the analysis algorithm downstream. 3.4 Analyzing Images: Creating the Analysis Algorithm

1. The module chosen to perform the analysis is dictated by the biological question needed to be answered and also the quality of the image. The aim of this study was to identify the number of cells that were positive for both probes and determine the spatial relationship between them. The FISH v2.1.7. was used; this is a dynamic fluorescent module, which uses the nuclear stain to detect and segment the cell’s nucleus and the module allows for any number of probes to be analyzed (see Note 6). 2. The analysis magnification is set to 1, meaning the image will be analyzed at the magnification it was scanned at (see Note 7). 3. The nuclear dye is selected and the number of probes/dyes is inputted and named along with any phenotypes to identify (see Note 8). 4. The cell segmentation is carried out solely on the nuclear dye, and a number of different features can be altered to ensure that the segmentation is accurate across all samples (see Note 9). If the cell segmentation is incorrect then the analysis will be incorrect (see Note 10). 5. The cell’s cytoplasm is created by stating an arbitrary maximum radius from which the pseudo cytoplasm is formed. 6. The probe identification has similar settings to the nuclear segmentation; using the probe contrast threshold and minimal intensity settings all of the probe can be identified and segmented. 7. The probe thresholds for the scores/bins are set to the parameters stated by ACD looking at copies of probe per cell. No staining is less than 1 dot per cell, 1+ is between 1 and 3 probe spots per cell, 2+ is 4–9 spots per cell, 3+ is cells with 10–13 spots, and 4+ is any cell containing 14 spots or more. In the module the probe scoring results give a different colored cell allowing for visual representation of the scores for each probe. 8. The classifier was added to the algorithm and the class selected for analysis was the tissue, meaning that only the areas classed as tissue would be analyzed. 9. The object data is set to true; this will allow for spatial analysis to be carried out as the results for each cell will be saved rather than just the summary (see Note 11).

28

Cara Brodie

Fig. 5 Classifier markups. The markups of both the classifier and the subsequent analysis with the AF regions excluded. (a–c) 20 magnification, (d–f) 40 magnification. The classifier eliminates the AF regions and the analysis is therefore only applied to the tissue areas

10. The algorithm is checked using the real-time tuning window on a selection of images, ensuring that the sample variation is considered. This is saved and then ran on all of the TMA cores. Figure 5 shows the analysis markup. 11. Using the object data created by the algorithm, spatial analysis is performed on the results. The cells that are positive for each probe are plotted on a map; the positive probe 1 cells are assessed for their proximity to the cells positive for probe 2. A histogram is created showing the distance between the two probes; this can be seen in Fig. 6.

4

Notes 1. A tissue microarray is a composite block made up from cores taken from donor blocks; this allows for many sample types to be on one slide saving reagent and sample. 2. The TMA map is normally provided with the tissue sections whether created in-house or purchased from a commercial company. If there is no map one can be created within the

Overcoming AF with Image Analysis

29

Fig. 6 The spatial analysis results. (a) A spatial map of the cells that are positive for either probe 1 (blue and yellow) or 2 (red) and the proximity of the probe 1 positive cells to the positive probe 2 cells. (b) Histogram output with the number of probe 1 positives cells and the distance they are from the probe 2 positive cells

HALO software. When the results are exported maps that contain core information will give the analysis results along with the data contained within the TMA map. 3. The tissue class glass is added to ensure that the classifier can correctly identify everything on the image; this includes the empty space with no sample or staining. If this is not done the classifier will not be able to determine what these areas are and therefore misidentify them, creating an inaccurate classifier and affecting the downstream analysis. 4. The classifier was used to identify any AF produced by exogenous sources such as the RBC and the fibrosis that was found in high levels in a number of TMA cores. The AF could be identified as positive signal and skew the results of any of the algorithms created and therefore must be removed. The fluorophore selection within the advanced settings ensures that fluorophores which impact the identification of the tissue classes are what the classifier focuses on. A background channel was visualized when scanning these images; this fluorophore is AF only and therefore used in the classifier. The use of a background channel is dependent on the number of probes being analyzed and the number of channels that can be visualized on the scanner used. 5. When using the probability map a threshold can be set of what is defined as a particular tissue class. The probability threshold was set at 50% for these samples meaning that if the software was less than 50% certain the pixel is part of the cells tissue class then it will be ignored. This was used to ensure that the software was identifying the tissue areas as accurately as possible

30

Cara Brodie

Fig. 7 Probability and mask comparison. The difference between the probability markup and the standard mask. (a) The DAPI and FITC image (b) depicts the classifier with the markup set to mask (pink area is the tissue class and the yellow is the AF class), and (c) the classifier using the probability map set to 50%; the red and green to yellow areas will be analyzed as part of the tissue area

ensuring that no areas of AF would be incorrectly classed. When using the probability map setting the software will only identify one tissue class so this must be the class of interest and the one which will be analyzed downstream. So for this classifier it was set to the tissue, meaning only the tissue areas would be highlighted and the background AF and glass would be excluded. Figure 7 shows the differences between the mask and probability output. 6. The analysis modules within the HALO software are set up with a similar workflow, allowing you to work down through the sections to create an accurate algorithm. Each module has a user guide that is available to look at while creating the algorithm within the software. 7. This will help identify any fainter smaller spots, essential for low-expressing probes. It is worth noting that for other algorithms this can be decreased as it may improve the analysis. 8. Within this module there is an option to identify phenotypes of interest. The phenotypes are defined with the positive and negative staining of probes and once the analysis is completed the results highlight the number of cells of each phenotype created. 9. Firstly, the nuclear contrast threshold examines individual pixels to determine whether the difference in the contrast between them would class it to be included in the cell’s nucleus or not. Next, alter the minimum nuclear intensity parameter to remove areas of low-intensity staining that may have been falsely identified as a nucleus. Exclude any elongated or irregular nuclear structures by altering the minimum nuclear roundness parameters. Any falsely identified nuclei that are smaller or larger than the cells being analyzed can be eliminated using the

Overcoming AF with Image Analysis

31

cell size threshold. Finally, the nuclear segmentation aggressiveness parameter can be altered to increase or decrease the amount of segmentation. It is imperative that time is taken to obtain accurate nuclear detection and segmentation to avoid unreliable downstream analysis results. The settings should be modulated and checked frequently during the setup phase. 10. For some images, the cell segmentation may not be accurate enough to use the FISH module as the nuclear stain alone is insufficient to correctly identify the cell. There are a number of alternative modules that can be used depending on the tissue type, the probe distribution, and the biological question. If the probe is localized at or close to the cell’s nucleus, the highplex module can be used. This is a dynamic module allowing for any number of fluorophores to be analyzed. The output of this module is not probe counts, but rather probe positivity per cell. Nuclear detection is different in this module as each fluorophore can be used as a weight to determine the detection and segmentation of the nucleus. This module will give the object data required for downstream spatial analysis. If the target is a low expresser, the signal intensity of the probe may not be sufficient for detection leading to an underestimate of the number of positive cells. In such cases, it may be necessary to use the area quantification module, another dynamic module where fluorophore number is defined by the user. The output for this module is positive pixel count and provides the total probe area within the tissue without cell data. 11. The object data contains the cell-by-cell data for the analysis, for example whether the cell is positive or negative for each probe, number of copies of the probe within each cell, cell size, and coordinates. This information is used for the spatial analysis module within HALO and can be exported and used in other software. References 1. Huber D, Voith von Voithenberg L, Kaigala GV (2018) Fluorescence in situ hybridization (FISH): history, limitations and what to expect from micro-scale FISH? Micro Nanoeng 1:15–24 2. Schulz D, Zanotelli VRT, Fischer JR et al (2018) Simultaneous multiplexed imaging of mRNA and proteins with subcellular resolution in breast cancer tissue samples by mass cytometry. Cell Syst 6:25–36 3. Jenson E (2014) Technical review: in situ hybridization. Anat Rec (Hoboken) 297:1349–1353

4. Oliveira VC, Carrara RC, Simoes DL et al (2010) Sudan Black B treatment reduces autofluorescence and improves resolution of in situ hybridization specific fluorescent signals of brain sections. Histol Histopathol 25:1017–1024 5. Gaffney EF, Riegman WE, Grizzle WE et al (2018) Factors that drive the increasing use of FFPE tissue in basic and translational cancer research. Biotech Histochem 93:373–386 6. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14:22–29

32

Cara Brodie

7. Baker AM, Huang W, Wang XMM et al (2017) Robust RNA-based in situ mutation detection delineates colorectal cancer subclonal evolution. Nat Commun 8(1):1998 8. Larsson C, Grundberg I, Soderberg O et al (2010) In situ detection and genotyping of individual mRNA molecules. Nat Methods 7:395–397 9. Thompson SM, Raven RA, Nirmalan NJ et al (2013) Impact of pre-analytical factors on the proteomic analysis of formalin-fixed paraffinembedded tissue. Proteomics Clin Appl 7:241–251 10. Baena-Del Valle JA, Zheng Q, Hicks JL et al (2017) Rapid loss of RNA detection by in situ hybridization in stored tissue blocks and preservation by cold storage of unstained slides. Am J Clin Pathol 148:398–415 11. Lewis F, Maugham NJ, Smith V, Hillan K et al (2001) Unlocking the archive - gene expression in paraffin embedded tissue. J Pathol 195:66–71 12. Portier PB, Wang Z, Downs-Kelly E et al (2013) Delay to formalin fixation ‘cold ischemia time’: effect on ERBB2 detection by in-situ hybridization and immunohistochemistry. Mod Pathol 26:1–9 13. Turashvili G, Yang W, McKinny S et al (2012) Nucleic acid quantity and quality from paraffin blocks: defining optimal fixation, processing and DNA/RNA extraction techniques. Exp Mol Pathol 92:33–43 14. Baschong W, Suetterlin R, Laeng HR (2001) Control of autofluorescence of archival formaldehyde-fixed paraffin embedded tissue in confocal laser scanning microscopy (CLSM). J Histochem Cytochem 49:1565–1571

15. Whittington NC, Wray S (2017) Suppression of red blood cell autofluorescence for immunocytochemistry on fixed embryonic mouse tissue. Curr Protoc Neurosci 81:2.28.1–2.2.28 16. Clancy B, Cauller LJ (1998) Reduction of background autofluorescence in brain sections following immersion in sodium borohydride. J Neurosci Methods 83:97–102 17. Hamilton PW, Bankhead P, Wang Y et al (2014) Digital pathology and image analysis in tissue biomarker research. Methods 70:59–73 18. Spriggs AI (1969) Automatic scanning for cervical smears. J Clin Pathol Suppl Coll Pathol S2-3:1–7 19. Holzer TR, Hanson JC, Wray EM et al (2019) Cross-platform comparison of computerassisted image analysis quantification of in situ mRNA hybridization in investigative pathology. Appl Immunohistochem Mol Morphol 27:15–26 20. Wiesmann V, Frank D, Held C et al (2015) Review of free software tools for image analysis of fluorescence cell micrographs. J Microsc 257:39–53 21. Bankhead P, Loughrey MB, Fernandez JA et al (2017) QuPath: Open source software for digital pathology image analysis. Sci Rep 7 (1):16878 22. Baharlou H, Canete NP, Bertram KM et al (2019) Digital removal of autofluorescence from microscopy images. Biorxiv. https://doi. org/10.1101/566315 23. Pell R, Oien K, Robinson M et al (2019) The use of digital pathology and image analysis in clinical trials. J Pathol Clin Res 5:81–90

Part II Methods for DNA ISH

Chapter 3 Practical Application of Fluorescent In Situ Hybridization Techniques in Clinical Diagnostic Laboratories Sheila J. M. O’Connor, Kathryn R. Turner, and Sharon L. Barrans Abstract Fluorescent in situ hybridization (FISH) techniques can be used to identify a range of chromosome abnormalities that are clinically significant in many cancers. Multicolor FISH can be used to identify multiple targets, which can be simultaneously detected in individual cells using digital imaging microscopy. In an era of precision medicine there is a requirement to make a precise diagnosis and to have a molecular classification of the tumor that can guide therapy. Cancer genomics is now regarded as a sub-specialism in pathology and genomic testing needs to be robustly integrated into the routine diagnostic practice. The FISH techniques described in this chapter have been developed over many years in a busy hematopathology diagnostic laboratory. We describe robust in-house methods for both liquid samples (blood and bone marrow mainly) and formalin-fixed paraffin-embedded (FFPE) tissue biopsies that allow for large numbers of slides to be set up in batches. The techniques described are for interphase cells in tissues where metaphase chromosome techniques are generally not applicable. Some of the FISH tests need to be carried out as an “out-of-hours” emergency test to make a critical diagnosis while others provide prognostic information and are used to guide downstream patient management. Key words Cancer genomics, DNA probes, Fluorescent in situ hybridization (FISH), Gene rearrangement, Hematopathology, Molecular cytogenetics

1

Introduction It is generally accepted that cancer is an acquired genetic disease [1, 2]. The molecular basis of many types of malignancy is well documented and the demonstration of specific numerical and structural chromosomal aberrations is an essential part of the diagnostic process [3]. While there are many molecular genetic techniques (such as RQ-PCR, high-throughput sequencing, RFLP) in use for the detection of mutations and fusion genes this chapter focuses on FISH and discusses the practical application in a clinical diagnostic laboratory [4]. The FISH technique is relevant for the detection of chromosome aberrations in tissue biopsies, blood, bone marrow, and cellular fluids diagnosed or suspected of containing neoplastic cells.

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_3, © Springer Science+Business Media, LLC, part of Springer Nature 2020

35

36

Sheila J. M. O’Connor et al.

The techniques described in this chapter are based on our experience with hematological malignancies; however these are transferable to other tissues types containing solid tumors/nonhematological malignancies. The key principle of FISH is the use of locus-specific (LS) DNA probes that recognize unique DNA sequences and alpha satellite or chromosome enumeration (CE) DNA probes that recognize repetitive sequences at the chromosome centromeres. These fluorescently labeled probes can be applied to interphase cells to detect structural and numerical chromosome aberrations by using different fluorochromes for specific regions. It is essential that the tissue architecture and morphology are retained—this is the “in situ” aspect and the true power of FISH is this ability to identify a molecular abnormality in a specific tumor cell and to be able to see normal patterns in adjacent normal cells. Clinical diagnostic laboratories are subject to stringent legislation and quality assurance [5]. It is essential that any DNA probes used are validated and are fit for purpose. The clinical significance of a specific chromosome aberration is normally identified and robustly tested in large clinical trials or patient series published in peer-reviewed journals. While it remains possible to manufacture DNA probes in-house (and our laboratory did this for many years prior to their commercial availability), there are many reputable companies now manufacturing a range of fluorescently labeled probes against a wide range of genetic targets. Our recommendation would be to use commercial probes where possible. Many commercial DNA probes are now CE marked (indicates compliance with EU legislation of a product) or FDA marked (indicates compliance with USA legislation of a product) and the introduction of a CE- or FDA-marked probe in the diagnostic laboratory is generally easier compared to an unmarked probe. Validation is performed by the manufacturer; individual laboratories then only need to perform verification; however, note that if there is any deviation from the manufacturer’s protocol, this change would need to be supported by additional validation. The basic principle of FISH is to utilize fluorochrome(s) to identify the presence or absence of a particular genomic aberration in specific cells related to specific diseases. There are six key stages with the process: (1) fixation, (2) pretreatment, (3) denaturation, (4) hybridization, (5) stringency wash, and finally (6) visualization. All cells and tissue require fixation; the aim is to prevent or arrest the degenerative processes that start once tissue is deprived of its blood supply. Autolysis and bacterial decomposition are processes that must be prevented. The loss of soluble cell contents (proteins, DNA, RNA, etc.) must be avoided by precipitation or coagulation of these components or by cross-linking them to other insoluble structural components. The key aim of fixation is that tissues must be preserved to retain reactivity to tinctorial stains and

FISH Techniques for Clinical Laboratories

37

retain surface, cytoplasmic, and nuclear antigens so that reactivity to antibodies can be assessed and DNA is not degraded or fragmented to enable annealing to DNA probes [6]. Liquid samples such as blood, marrow, and cytology samples are simply fixed using precipitation, e.g., methanol/acetic acid (MAA) (also known as Carnoy’s solution), whereas solid tissue usually requires crosslinking fixation to retain the structural integrity of the tissue, e.g., 10% buffered formalin. The type of fixation impacts what pretreatment is required. In general, precipitating fixatives require no or very little additional pretreatment whereas cross-linking fixatives always require pretreatment. The amount and type of pretreatment depend on many factors and this is probably the most difficult aspect of the technique to standardize—the in-house method we will describe is based on years of experience and is distilled from minor modifications and adjustments. Based on the FFPE tissue types and diseases we commonly see that the success rate of the technique is close to 95% at first attempt. The few samples that fail usually either are poor-quality samples or may not have been adequately processed. The aim of pretreatment is to reverse the formalin-induced crosslinkage and make the DNA accessible to the probe during the hybridization stage. Most pretreatment protocols use enzymes and heat to achieve this. The purpose of denaturation (often called “melt”) is to render the DNA of both the probe and the sample single stranded. There are a number of ways to make double-stranded DNA into singlestranded DNA; however, a key advantage of FISH is the in situ aspect and we want to retain the cell morphology and tissue architecture. With most molecular techniques such as polymerase chain reaction (PCR) heating is the most frequent method of denaturation. The DNA is usually heated to >95  C to achieve this, but this temperature destroys morphology and cells are unrecognizable. To overcome this we use formamide, a chemical denaturing agent that lowers the melting temperature (Tm) by competing for hydrogen bond donors and acceptors with preexisting nitrogenous base pairs. To minimize handling the formamide is normally added to the probe during manufacturing. The inclusion of formamide to the probe hybridization buffer allows the denaturation temperature to be dropped to approximately 70–75  C; this is sufficient to retain the tissue architecture and cell morphology. We carry out co-denaturation of the probe and the cells/tissue using a programmable hot plate, where the probe is suspended in the hybridization buffer which contains formamide. The probe is applied to the target area; there is sufficient formamide present to simultaneously denature the cellular DNA as well as the probe DNA once heat is applied. These precision instruments will ramp the temperature to the specified high level before returning to the lower level required for hybridization.

38

Sheila J. M. O’Connor et al.

Hybridization occurs when both a denatured probe and denatured target DNA—cells/tissue on a glass slide—are allowed to anneal and form hydrogen bonds that join the two strands resulting in a hybrid double-stranded segment consisting of probe-DNA. The probe specifically hybridizes to its complementary sequence on the chromosome. However, all DNA contains repetitive sequences which could result in the probe binding to multiple areas of the genome—so-called cross-hybridization. To prevent this and achieve a specific hybrid between the probe and the target DNA the repetitive sequence DNA must be blocked; this is done by adding short fragments of reagents such as COT-1 DNA to the hybridization buffer during probe manufacture. There are many factors that can impact hybridization efficiency; these include the degree of nuclear condensation; for example blast cells with “open” chromatin are easier to FISH compared with mature cells with “closed” chromatin (heterochromatin). The purpose of the pretreatment stage and the denaturation conditions is to ameliorate these biological variables, but some empirical adjustment may still be needed for individual samples. With modern probe manufacture we expect hybridization efficiency to be close to 100% for most samples. Stringency is the term used to indicate the strictness with which base-pairing (hybridization) is allowed under specified conditions of temperature, pH, salt concentration, and formamide. Conditions of high stringency require all bases of one nucleotide to be paired with complementarybases on the other; conditions of low stringency allow some bases to be unpaired. Therefore, the stringency wash following hybridization is a critical step. The purpose of this wash is to remove any probe partly attached by repetitive sequence binding. If the stringency is too high, then the probe is stripped off resulting in a failed FISH test and if the stringency is too low there will be multiple nonspecific fluorescent spots in each cell resulting in an uninterpretable FISH test. The stringency wash is carefully designed to be a step below the denaturation conditions for the probe used to achieve a specific result in a clean background. Hybrids or fluorescent spots are detected/visualized using a fluorescent microscope. Fluorescence microscopy requires an intense light source at the specific wavelength that will excite the fluorochromes used. The traditional method uses white light, typically from a mercury arc lamp. These broad-spectrum lamps generate strong light across the spectral range with the desired wavelengths selected by specific narrow band-pass filters and wavelengths outside of the desired range being blocked. There have been recent advances in high-performance light-emitting diode (LED) technology but in general, the intensity of the LEDs is still not as high as conventional arc lamps. When designing a FISH experiment, you need to consider whether the sensitivity and resolution of the assay lie within the technical limits of fluorescence

FISH Techniques for Clinical Laboratories

39

microscopy. Sensitivity depends on the light-gathering ability of the particular microscope, which determines the minimum size of fluorescent signal that can be seen. Light microscopy also has limits to resolution and cannot resolve objects that are separated by less than 200–250 nm, the lower limit of the visible light spectrum. The conformation of DNA within the chromosome is a factor with metaphase chromosomes thousands of times more compacted than interphase chromosomes, giving a resolution in the range of megabases for regions on metaphase chromosomes and kilobases for interphase regions in cells and tissue. Fluorescence microscope instrumentation utilizes a combination of filters to provide selective excitation of specimen fluorochromes with maximum sensitivity. Using a combination of direct microscopic visualization and an image capture system, patient samples, which have undergone FISH, are analyzed by healthcare scientists and evaluated for specific abnormalities, such as the presence of fusion genes and loss and/or gain of genetic material. In clinical diagnostic laboratories FISH tests are “blind” checked by a second healthcare scientist before integration with the main pathology report [7–9]. High-quality sample preparation is essential for high-quality FISH results. Anticoagulated, unfixed liquid biological samples will deteriorate within hours and procedures should be established in pathology laboratories for the rapid preparation and preservation of cells (see Note 1). Tissue samples, either biopsy or resection samples, require prompt fixation and processing to minimize cell degradation and to preserve the architecture (see Table 1). Pathology laboratories have well-established techniques in place. Formalin remains the fixative of choice in pathology laboratories, but there are discussions surrounding alternative more molecular friendly fixatives but to date no robust alternative to formalin has been introduced [10, 11].

2

Materials Wear powder-free nitrile gloves, protective clothing, and eye protection. Toxic substances, harmful substances, or concentrated acids should be weighed out or measured in a fume cupboard or ventilated bench. All necessary precautions to minimize the risk of spillage including restricted access to the area by other staff should be taken. Spillage kit and respirators should be readily available. Prepare all solutions using molecular grade water (MGW) (prepared by purifying deionized water to a sensitivity of 18 MΩ at 25  C or commercial MGW) and analytic grade reagents. Prepare and store all reagents at room temperature (unless otherwise indicated). All solutions should be labeled with the date, batch number, and appropriate hazard labels. Follow all waste disposal regulations when disposing waste material. We do not add sodium azide to reagents.

40

Sheila J. M. O’Connor et al.

Table 1 Common sample types in pathology laboratories Fresh sample type

Considerations

Advantage

Smears are easy to Smears prepared from Samples should have prepare, cheap, and fresh EDTA >10% neoplastic cells easy to store anticoagulated bone in the bone marrow for marrow aspirates and translocation or gene It is standard practice in hematology peripheral blood rearrangement analysis laboratories to prepare samples Samples should ideally spare blood and have >50% neoplastic marrow smears as soon cells in the bone as a sample is received marrow for deletion (see Note 19) analysis. The % of tumor cells must be considered with the cellularity of the sample—very low count samples may be unsuitable for FISH testing

Touch preparations/ imprints/dabs

Disadvantage This technique is suitable for slides that are less than 3 months old. Smears must not be exposed to formalin at any stage Very particulate bone marrow smears will need to have the particles carefully removed without damaging the trails containing the assessable cells Morphology experience is desirable Autofluorescent background is high compared to MAA fixed cell suspensions

Tissue may be damaged Fresh tissue imprint/dab Cheap and easy to during handling prepare preparations must be Process as for blood and Often not representative of the representative with marrow smears tissue as a whole—dab just normal T and B must contain lymphocytes present neoplastic cells of in the imprint interest

Fixed cell suspensions Blood or marrow samples Simple and cheap Samples can be safely identified at flow using MAA stored in MAA for cytometry as having a Preparations of cells years neoplastic population from effusions, CSF, can be rapidly fixed in or cells from other suspension. Lyse any liquid samples RBC prior to the addition of MAA

Cells may clump during the process; this can be minimized with careful handling and gentle mixing

MAA fixed immunomagnetic selected cells

Allows the selection of a Specialist equipment We have carried out needed. Needs rare population of cells immunomagnetic cell trained staff to carry in a mixed cellular selection on cases out the procedure background where the population If samples are >24 h old is around 0.5%—lower the purity can drop than this the purity tends to drop

FFPE sections

Identification of tumor cells/area needs skill

Standard histology technique Archive tissue available

Signal dropout due to sectioning Deletion analysis is difficult

FISH Techniques for Clinical Laboratories

2.1 Buffers and Fixatives

41

1. Methanol, acetic acid (3:1) MAA fixative solution: methanol, acetic acid. Take three parts methanol and one part glacial acetic acid (generally need sufficient to fill a standard-sized Coplin jar; this is 40 mL). Add acid to alcohol, mix, and cool to 4  C before use. Store in fridge for up to 1 month (see Note 2). 2. 20 Standard saline citrate (SSC)—master buffer: 3 M NaCl, 0.3 M sodium citrate. Weight out 175.32 g of 3 M NaCl and 88.2 g of 0.3 M sodium citrate; dissolve in 950 mL of H20. Adjust the pH to 7.0 (see Note 3) and make up to 1 L. This 20 SSC master buffer can be stored at RT for up to 6 months. 3. 2 SSC/0.1% NP-40 buffer: 3 M NaCl, 0.3 M sodium citrate, Tergitol NP-40. Take 100 mL of 20 SSC master buffer, add 850 mL of H20 and 1 mL of NP-40 (Tergitol NP-40) (see Note 4), adjust the pH to 7.0–7.5 (see Note 3), and make up to 1 L. 4. 0.4 SSC/0.3% NP-40 stringency wash buffer: 3 M NaCl, 0.3 M sodium citrate, Tergitol NP-40. Take 20 mL of 20 SSC master buffer, add 950 mL of H20 and 3 mL of NP-40 (Tergitol NP-40) (see Note 4), adjust the pH to 7.0–7.5 (seeNote 3), and make up to 1 L. 5. Proteinase-K enzyme solution: 2 SSC/0.1% NP-40, proteinase K (PK). Add 250 mL of 2 SSC/0.1% NP-40 to a conical flask and pre-warm at 37  C in a water bath. Cover the opening of the flask with parafilm to avoid evaporation (this would concentrate the salt which would alter the stringency of the solution). Pre-warm sufficient empty glass Coplin jars with lids on (maximum ten slides per jar) in the 37  C water bath. Immediately before use add 2 mg PK enzyme to the pre-warmed 250 mL of 2 SSC/0.1% NP-40 general buffer (see Note 5). 6. Pressure cooker pretreatment solution: Immediately before use add 1500 mL of H20 and 15 mL of high pH vector antigen unmasking solution (see Note 6).

2.2 Ready-to-Use Commercial Reagents

These are normally purchased as ready-to-use reagents: 1. Xylene. 2. Industrial methylated spirit (IMS). 3. Glue (rubber solution) (see Note 7). 4. Mounting medium for fluorescence with DAPI (see Note 8). 5. Nail varnish (see Note 9). 6. Fluorescence-specific immersion oil for microscopy (see Note 10).

42

Sheila J. M. O’Connor et al.

2.3 Specialist Reagents, DNA Probes

1. DNA probes (see Note 11) are labeled with fluorochromes and are normally validated/verified for use with specific sample types either FFPE sections (pathology use) or fresh (cytogenetic/hematology use). Most manufacturers also offer a custom design service which is a specific target gene or chromosome region probe that is required but is not currently commercially available (see Note 12).

2.4

1. Balance/weight scale (range capable of accuracy at μg level up to kg level).

Equipment

2.4.1 General Laboratory Equipment

2. Fridge/freezer. 3. Water baths (37  C and a variable bath up to 95  C). 4. Pipettes (a range accurate between 0.1 μL and 1000 μL). 5. Stop clocks. 6. Thermometer (electronic to cover a range of 20  C to 100  C). 7. Coplin jars (glass and plastic with deep screw on lids) (see Note 13). 8. Glass coverslips (12 mm diameter and 22  32 mm). 9. Microcentrifuge tubes (0.5 mL and 1.5 mL); cytocentrifuge. 10. Halogen or other hot plate, range to boiling point.

2.4.2 Specialist Equipment

1. Programmable hot plate (see Note 14). 2. Pressure cooker (see Note 15). 3. Microwave (see Note 16). 4. Fluorescent microscope fitted with appropriate “narrow bandpass” filters and/or “dual-band-pass” or “triple-band-pass” filters (see Note 17). 5. Image capture system (see Note 18).

2.5 Biological Samples to Test

1. Unfixed EDTA blood and bone marrow smears; dabs/imprints from fresh unfixed solid tissue biopsy—these will require fixation in MAA fixative solution prior to FISH; see Subheading 3 (see Note 19). 2. Cytology specimens, CSF or effusion samples, immunomagnetic cell selections, and unselected white blood cell (WBC) preparations can be fixed in MAA fixative solution early in the diagnostic process and stored at 20  C until required (see Note 20). 3. Formalin-fixed paraffin-embedded (FFPE) thin sections are prepared at 3 μm thickness on coated slides and preheated at 60  C for an hour to improve adherence of the section (see Note 21).

FISH Techniques for Clinical Laboratories

3

43

Methods There are several health and safety (H&S) factors to be considered with FISH testing. There is the potential biological hazard of the patient sample and the chemical risk from some of the reagents used. The only safe way to work in a diagnostic laboratory is to assume that all samples are a possible biohazard and handled accordingly. In general, samples cannot be rejected due to a known infectious agent but need to be processed safely [12] (see Note 22). Appropriate personal protective equipment (PPE) should be used including powder-free nitrile gloves and protective clothing. Eye protection should be worn if splashing is considered a hazard. In addition, DNA and DNA probes need protection from DNase, an enzyme secreted by human skin that digests DNA.

3.1 Preliminary Steps

1. Turn water bath(s) on to 37  C. 2. Place the bottle of 2 SSC/0.1% NP-40 into the 37  C water bath to pre-warm. 3. Presoak the humidity strips (if needed, depending on the programmable hot plate model used) in the 37  C water bath (see Note 23).

3.2 Fixation and Pretreatment of Blood, Marrow, and Tissue Imprint Preparations

1. Label the slides (patient slides already have essential patient identifiers such as name, laboratory number) with the probe name and batch number (see Note 2). 2. Place labeled slides in a clean, dry glass Coplin jar (maximum ten slides per jar, back to back) and add cold MAA fixative. 3. Incubate in the fridge for 30 min minimum (see Note 24). 4. After incubation, dispose of waste MAA fixative in a labeled waste MAA bottle stored in a suitable fume cabinet until disposal according to local regulations. 5. Wash the slides in the Coplin jar with the pre-warmed 2 SSC/0.1% NP-40 solution to remove all traces of MAA (this can be discarded in the sink). 6. Top up the Coplin jar with fresh pre-warmed 2 SSC/0.1% NP-40 solution and incubate at 37  C for at least 30 min (see Note 24). 7. After incubation, dehydrate in ethanol series 70%, 80%, and 100% for approximately 1–2 min each. 8. Once dry select a thin area of smear toward the tail, and mark the back and side of the slide with marker pen to enable location (see Note 25). 9. Slides are ready for probe application.

44

Sheila J. M. O’Connor et al.

3.3 MAA Fixed Cell Suspension Preparations

1. Remove the MAA fixed cell suspension sample from the freezer (see Note 20). 2. Spin the vials for 1 min at 10,000 rpm (9,500  g) in the microcentrifuge. 3. Tip off the MAA fixative. 4. Resuspend the pellet by gently flicking the vial; the remaining MAA fixative is usually sufficient to give optimum concentration of cells. For larger cell volumes cold MAA fixative may be added to decrease cell concentration. 1–2 drops is usually sufficient (see Note 26). 5. Mark a small circle on the back of the slide in permanent marker pen to indicate where the cells will be applied (see Note 27). 6. Pipette 1 μL of the MAA fixed cell suspension per test onto marked area of the slide and allow to air-dry. For low-cellconcentration MAA preps this step can be repeated. 7. Air-dry slides for at least 10 min keeping the slides flat as they dry. 8. Slides are ready for probe application. 9. The residual sample can be returned to storage, simply top up with approximately 1 mL of MAA fixative solution, and return to the freezer.

3.4 FFPE 3 μm Section Preparations

1. Place the labeled FFPE slides on a hot plate for approximately 60 min at 60  C. 2. Prepare for enzyme digestion while the slides are on the hot plate. 3. Add 250 mL of 2 SSC/0.1% NP-40 to a conical flask and pre-warm at 37  C in the water bath. 4. Cover the opening of the flask with parafilm to prevent evaporation. 5. Pre-warm sufficient empty glass Coplin jars with lids on (maximum ten slides per jar) in the 37  C water bath (see Note 28). 6. Take an aliquot of 2 mg pre-weighed PK enzyme from the freezer and set aside (see Note 5). 7. Following hot plating, dewax slides in three changes of xylene for approximately 5 min each. 8. Rehydrate slides in three changes of IMS for approximately 5 min each. 9. Wash in tap water and keep wet.

3.5 Pretreatment of FFPE 3 μm Sections Using Pressure Cooker and Enzyme Digestion

1. Add pressure cooker pretreatment solution (1500 mL of distilled water and 15 mL of high pH vector antigen unmasking solution) to the pressure cooker (use 12 lb. PSI weight) (see Note 29).

FISH Techniques for Clinical Laboratories

45

2. Place on the halogen hot plate and bring to the boil with the lid resting loosely. 3. Carefully place the sections in staining racks into the boiling buffer in the pressure cooker. 4. Close the lid and make sure that it is in the lid-locked position (see Note 30). 5. As the pressure increases (depending on the model used) an indicator switch will trigger; when full pressure is reached the pressure cooker will begin “hissing” and release excess steam. 6. Start the time for 90 s and fill the sink with cold water. 7. After 90 s immediately plunge the pressure cooker into cold water and run cold water over the lid until the red indicator goes down. 8. Turn the switch/dial to the steam-release position to release residual pressure and open the lid (take care as steam released from the pressure cooker will be very hot). 9. Flood the pressure cooker with cold water and allow the sections to cool. Take care not to let the sections dry out before continuing to enzyme digestion. 10. Immediately before use, add the aliquot of 2 mg PK enzyme to the 250 mL pre-warmed buffer in the flask and mix well. 11. Quickly decant the PK enzyme solution into the pre-warmed empty glass Coplin jars in the 37  C water bath. 12. Quickly add the pressure-cooked slides and incubate for 30 min at 37  C. 13. Following incubation, wash thoroughly in cold running water for at least 1 min to stop the enzyme activity. 14. Dehydrate in ethanol series 70%, 80%, and 100% for approximately 1–2 min each. 15. Air-dry for at least 10 min. 16. Once the slides are dry the area for probe application can be marked on the back of the slide using a permanent marker pen. Use the marked H + E stained section to select where to apply the probe (see Note 31). 17. Slides are ready for probe application. 3.6 Probe Preparation and Application

Fluorescent-labeled probes must be protected from photobleaching. Work under low-light conditions if possible. Probes are subject to acceptance testing in clinical laboratories (ISO 15189). 1. Remove probe(s) from the freezer or fridge approximately 5–10 min before needed. 2. Pulse microcentrifuge briefly (once thawed if frozen) to settle the contents.

46

Sheila J. M. O’Connor et al.

3. Work out the total volume of probe needed for the total number of tests being carried out and decant into an appropriately labeled 0.5 mL Eppendorf tube (see Note 32). 4. Some manufacturers’ (e.g., Cytocell) probes are incubated for at least 5 min at 37  C and held at 37  C until use. 5. Group the slides according to which probe (if more than one type of FISH test is being set up) is required (see Note 33). 6. Lay out on the bench one group of slides at a time. 7. Apply volume of probe according to the manufacturer’s instructions to the area previously identified by marker pen (see Note 34). 8. Cover with an appropriately sized coverslip. Avoid air bubbles. 9. Seal around the rim of the coverslip with rubber glue. 10. Place slides on the programmable hot plate. 11. Allow the glue to dry to the tacky stage before starting “Denaturation/Hybridization” program. 12. Select appropriate “Denaturation/Hybridization” program. See Table 2 for details of main manufacturers’ recommended temperature and times. 13. Close the lid and press start—check that the program is running before leaving the laboratory. Table 2 Denaturation and hybridization conditions Probe manufacturer

Fresh

DAKO

82C

5 mins

45C

12

Fresh

Cytocell

75C

2 mins

37C

12

Fresh

Cytocell FAST FISH

C 75

2 mins

37C

I hour

Fresh

Abbo/Vysis

73C

3 mins

37C

4

Fresh

Zytolight

75C

2 mins

37C

12

75C

5 mins

37C

12

Denaturation temperature

Denaturation time

Hybridization temperature

Hybridisation time in hours (minimum)

Sample type

All FFPE

manufacturers probes

Red colour is urgent or critical

FISH Techniques for Clinical Laboratories

3.7 Post-hybridization Stringency Wash

47

Unless it is a FAST FISH protocol the post-hybridization stringency wash is done on day 2 of FISH setup. Standard hybridization is normally at least 12 h; however, some manufacturers have developed same-day assays with a 4-h hybridization (see Note 35). Fluorescent-labeled probes must be protected from photobleaching. Work under low-light conditions. 1. Fill the appropriate number of plastic Coplin jars (deep screw on lids), (see Note 13) with post-hybridization stringency wash buffer 0.4 SSC/0.3% NP-40, enough jars to wash all slides with no more than six slides in each Coplin jar (see Note 36). 2. Place into the 72  C water bath, allow sufficient time for the stringency wash buffer to reach the required temperature, and allow at least 60 min to heat before starting the procedure (see Note 13). 3. Open the programmable hot plate (Hybrite/ThermoBrite) lid and remove the glue from the slides using forceps, taking care not to damage the cells/tissue section on the slide (see Note 37). 4. Fill sufficient glass Coplin jars (ten slides per Coplin jar back to back) with 2 SSC/0.1% NP-40 solution. 5. Place the slides into the Coplin jars containing the 2 SSC/0.1% NP-40 solution for approximately 5 min; this step is to loosen the coverslips from the slides. 6. Check Table 3 to ensure correct temperature of the posthybridization stringency wash for the sample preparation type and probe type.

Table 3 Post-hybridization stringency wash temperature and times Sample Preparation Type Fresh smears & MAA

Fresh smears & MAA

FFPE thin section

Temperature Probe Manufacturer CRITICAL

Stringency Wash time CRITICAL

Cytocell, Cytocell FAST FISH Abbo/Vysis, Zytolight

72(+/ -1)C

2 mins

Dako

65(+/ -1)C

2 mins

Dako, Cytocell, Abbo/Vysis, DAKO

72(+/ -1)C

2 mins

48

Sheila J. M. O’Connor et al.

7. Check the temperature of the pre-warmed stringency wash buffer before use (check the temperature inside one of the Coplin jars rather than the water bath as there may be some discrepancy). 8. Working quickly, transfer no more than six slides (two slides back to back) from the 2 SSC/0.1% NP-40 into one of the plastic Coplin jars containing the stringency wash buffer at the correct temperature. 9. Place the Coplin jar back into the water bath so it maintains the correct temperature and time for 2 min. 10. As soon as the 2 min is up, remove the plastic Coplin jar from the water bath and decant stringency wash buffer (see Note 38) and replace with fresh 2 SSC/0.1% NP40 wash buffer. 11. Wash for a minimum of 2 min. 12. Decant the 2 SSC/0.1% NP40 wash buffer and replace with fresh 2 SSC/0.1% NP40 wash buffer. 13. Wash for a minimum of 2 min, and keep slides in wash buffer until they are mounted with DAPI and coverslipped. 3.8 Counterstain and Coverslipping

1. Apply a drop of DAPI nuclear counterstain (see Note 8) to a 22  32 mm coverslip or larger if necessary. Slides with two FISH test areas need each area covered with mountant and a 22  40 mm coverslip can be used to cover both areas. 2. Apply slide, ensuring coverage of the area where the probe has been applied. 3. Remove air bubbles by pressing gently on the coverslip with fine-tipped forceps. 4. Press flat and blot dry with paper towel taking care not to damage the cell/tissue preparation. 5. Ensure that the coverslip is covering the probed area and seal around the edges of the coverslip with nail varnish; take care not to cover any area where probe may be with the nail varnish (see Note 9). 6. Place slides in appropriate batches in slide tray. 7. Close slide tray or cover with foil and place with the appropriate paperwork ready for visualization and reporting. 8. Allow between 10 and 30 min for even DAPI dye uptake and allow the nail varnish to dry completely prior to microscopy.

3.9 Turning on Microscope, Lamp Power Unit, and FISH Workstation: General Points

Switch-on instructions need to be established locally as these will differ depending on the make and model of the fluorescent microscope and the FISH image capture workstation (see Note 39).

FISH Techniques for Clinical Laboratories

3.10 Examination of FISH Tests and Image Capture

49

1. Allow the microscope to warm up according to the manufacturer’s instructions before use. 2. Make sure that the fluorescent microscope has the correct filter configuration for the experiment type. 3. Check name, number, and probe details on slide before placing onto the stage. 4. Put a drop of immersion oil onto the slide. Using a 63 Plan Apo oil objective, locate the area on the slide where probe has been applied. 5. Open the lamp shutter. 6. Choose a filter using the filter change buttons; single colors can be viewed using the narrow band-pass filters; both red and green can be viewed together using the dual-band-pass filter and red, green, and blue can be viewed using the triple bandpass if fitted. 7. Set the light path to the camera, using the push-pull lever on the microscope. 8. Scan the entire hybridization area and check if hybridization is even and bright. 9. Proceed to capture images through the preset filters; a prompt may be required (mouse click or any keyboard key) for each filter change. This allows for refocusing between filter changes if necessary. Consider z-stacked versus single focal plane (see Note 40). 10. Once image capture is complete and all fluorochromes are captured the image needs to be processed and saved. 11. Check that individual fluorochromes are in registration (colors are all perfectly aligned). If the error is slight it can be corrected using the appropriate command in the software. If not, then recapture. 12. Annotate the image if appropriate; for example point out the cell of interest using an arrow and a short comment.

3.11 Analysis of FISH Test Results

There are four main strategies for FISH analysis in routine clinical laboratories: 1. Numerical or single target gene probes 2. Break-apart or split-signal probes 3. Dual-fusion, dual-color translocation probes 4. Tricolor rearrangement probes

3.11.1 Numerical Analysis

Numerical analysis is the detection of loss or gain of a gene or chromosome region of interest. Select an area of the smear or cell preparation showing good nuclei distribution; count the number of

50

Sheila J. M. O’Connor et al.

Fig. 1 Schematic representation of a deletion/gain probe set with centromeric alpha satellite control probe added (a). Normal (2R2G) pattern; (b) typical deletion (1R2G) pattern; (c) monosomy (1R1G) pattern; (d) biallelic deletion (2G0R) pattern; (e) trisomy (3R3G) pattern (red ¼ critical gene or area of interest, green ¼ centromeric or other control area). Be aware that other signal patterns are possible

signals within the nuclear boundary of each evaluated nucleus. Focus up and down to find all of the signals in the individual nucleus. A normal cell will have a count of two locus-specific signals and, if present, two alpha satellite control probe signals (see Fig. 1). Very careful probe design is needed for the assessment of loss or gain in FFPE samples. In general, we avoid this strategy in FFPE but there are some very specific abnormalities that may need to be assessed in certain disease types. The problem with FFPE sections is that nuclear material is lost during tissue sectioning. This makes the interpretation of deletion/loss particularly difficult due to signal “dropout”. For numerical/deletion/gain analysis a minimum of 100 cells need to be manually counted, and results expressed as a percentage. For borderline deletion, where the abnormality is present in 5–20/ 100 cells, count an additional 100 cells if possible. A minimum of 20 cells is acceptable for myeloma CD138+ selected plasma cells as it may not be possible to reach 100 cells in some cases [13]. 3.11.2 Break-Apart Analysis

Break-apart analysis is used for the detection of rearrangement of a gene. Red and green probes are designed to hybridize upstream and downstream of the breakpoint cluster region for the gene/region of interest. Co-localization of the probes results in a red/green fusion signal (normal 2F pattern). Rearrangement in the breakpoint cluster region will split one signal resulting in separate green and red signals (rearranged 1F1R1G pattern) with retention of a normal signal on the other chromosome. A centromeric control probe may also be added to the probe set. Break-apart probes are very useful for FFPE sections where signal “dropout” may be a problem, as a high level of single red or green signals can indicate a gene rearrangement (see Fig. 2).

FISH Techniques for Clinical Laboratories

51

Fig. 2 Schematic representation of a break-apart or split-signal gene rearrangement probe set with centromeric alpha satellite control probe added. (a) Normal (2B2F) pattern; (b) typical rearrangement (2B1F1G1R) pattern (red/green ¼ critical gene or area of interest) (R/G appears as a yellowish fusion (F) signal, blue ¼ centromeric or other control area). Be aware that other signal patterns are possible 3.11.3 Dual-Fusion, Dual-Color Translocation Analysis

Dual-fusion, dual-color translocation analysis is used for the detection of a translocation between two known chromosome regions. A single-color probe is designed to hybridize to each of the genes or the breakpoint cluster regions extending both upstream and downstream. For dual color, one gene is labeled with a red probe and one with a green probe. In a normal cell this will show two red and two green signals (2R2G pattern). A control centromeric probe labeled blue may be present in some probe sets. In a normal cell this will show two red and two green signals and two blue signals (2R2G2B pattern). If a translocation is present, the most common pattern is one red signal, one green signal (representing the normal chromosomes), and two red/green (yellow) fusion signals representing the two derivative chromosomes resulting from the reciprocal translocation (1R1G2F pattern) plus two blue signals for tricolor probe sets (1R1G2F2B pattern) (see Fig. 3).

3.11.4 Tricolor Rearrangement Analysis

Tricolor rearrangement probes can be used for the detection of certain interstitial deletions or rearrangements. Co-localization of the probes results in a red/green/blue fusion signal (normal 2F). Depending on the probe design, an interstitial deletion of one of the probe targets will result in a signal pattern of one red/green/ blue fusion and one green/blue fusion, with the red probe target being deleted. There are many variations possible with tricolor probes; check manufacturers’ websites for details (see Fig. 4).

52

Sheila J. M. O’Connor et al.

Fig. 3 Schematic representation of a dual-fusion, dual-color probe set with centromeric alpha satellite control probe (blue). (a) Normal (2B2R2G) pattern; (b) typical rearrangement (2B2F1G1R) pattern (red and green ¼ represent critical genes or area of interest on the same or different chromosomes which if both genes are rearranged in a reciprocal balanced translocation will give rise to R/G (appears yellowish) fusion (F) signal, blue ¼ centromeric or other control area). Be aware that other signal patterns are possible

Fig. 4 Schematic representation of a Vysis 4q12 tricolor rearrangement probe set. This particular rearrangement is associated with myeloid neoplasms with eosinophilia. (a) Normal (2BRG fusion) pattern; (b) standard rearrangement (1BRG fusion and 1BG fusion) pattern. Blue, red, and green represent critical genes or area of interest on the same chromosome. The PDGFRA-FIP1L1 fusion rearrangement is cytogenetically cryptic as it is caused by a small interstitial deletion of CHIC2 gene—the deletion (loss of red signal) brings the blue and the green signals together. Be aware that other signal patterns are possible

FISH Techniques for Clinical Laboratories

3.12 Signal Enumeration

53

1. Avoid scoring slides or areas within slides that have excessive nonspecific autofluorescence (see Note 41); sometimes this is unavoidable. 2. Avoid scoring clumped cells where it is not possible to determine to which cell a signal belongs. It is not always possible to avoid overlapping cells due to the nature of the material, i.e., FFPE sections. 3. Signals of the same color that are touching, regardless of size, are counted as one signal. If there is a small strand of signal connecting separated signals, also count as one signal (see Note 42). 4. In a hybridization area, if the signals of the same color have no gaps greater than a signal width, count as one signal. However, if two signals of the same color are distinct, are compact, are of the same size and intensity, and are clearly separated, and no connecting signal strand is visible, count this as two signals even though the gap between signals is less than a signal width. 5. If there is any doubt as to whether or not a cell should be scored, do not score the cell. 6. If >20% of nuclei/cells within the hybridization area have no signals or patchy signals this may indicate nuclear degradation and possible poor-quality sample. We would record this as a quality concern (see Notes 43 and 44). 7. Be aware that sometimes a FISH preparation may have significantly better signals at the edge of the preparation as compared to the center of the preparation. 8. Crush artifact is a potential problem in FFPE sections, where the biopsies are physically crushed during the biopsy-taking process. The resulting FISH can appear ambiguous and be difficult to interpret. The crush artifact can pull the FISH signal and make a break-apart probe appear to be split resulting in a false-positive result. 9. Record the FISH result as the signal pattern including a count for numerical analysis (see Note 45). 10. For numerical /deletion/gain analysis 100 cells are counted. 11. If 0–4% of cells scored have an abnormal copy number the result is reported as normal as this low level is a technical limitation of the technique. 12. If 5–19% of cells with abnormal copy number are scored the result is reported as borderline. In this case, count an additional 100 cells if possible. 13. If 20–100% of cells are detected with abnormal copy number (deleted or gain depending on the probe used) the result is abnormal and reported with an appropriate comment in line with local and/or national/international guidelines.

54

Sheila J. M. O’Connor et al.

14. It is standard practice in diagnostic laboratories to have a two-person reporting system. The healthcare scientists reporting and checking results must be suitably trained and have evidence of ongoing competency. If results agree (and most of the time they do) then the FISH result is formally reported and integrated with the rest of the pathology report and released to the clinicians. Any discrepancies between the reporters must be referred for a third check (see Note 46). 15. While there is an International System for Human Cytogenetic Nomenclature (ISCN) [14] that uses standard nomenclature to describe any genomic rearrangement identified by techniques ranging from karyotyping to FISH, microarray, various region-specific assays, and DNA sequencing it is generally not recommended for use in clinical diagnostic reports. A normal FISH result presented in ISCN can be easily misunderstood by non-cytogeneticists due to the style; for example a chronic lymphocytic leukaemia (CLL) FISH panel consisting of SEC63, MYB, ATM, TP53, alpha 12 centromere probe, and 13q14 with normal results for all probes would be displayed as “nuc ish (D6Z1,SEC63,MYB) x2[200], (D11Z1,ATM)x2 [200], (D12Z3,MDM2)x2[200], (DLEU1, LAMP1)x2[200], (IGHx2)[200], (TP53,NF1)x2[200]”. Plain English (or local language) is clear and much safer for interpretation and the professional guidelines state that ISCN does not have to be used for FISH reports. 3.13 Quality Assurance

Clinical diagnostic laboratories are usually assessed against ISO 15189 standards; this covers most operational aspects of the laboratory including staff training and competency. Additional quality checks are carried out during each lot of FISH testing. Each hybridized slide is evaluated for quality of both the test material and the FISH hybridization (see Fig. 5). 1. Signals must be visible down the microscope. A slide with no or very weak signals should be repeated if possible. If the probe cannot be seen down the microscope, but is seen after image capture, this is regarded as poor and the test should be repeated (see Note 47). 2. Signals should be distinct; diffuse or fuzzy signals can be misinterpreted and should be repeated if possible, or evaluated with care if repeat testing is not feasible. 3. There should be sufficient cells to evaluate at least 100 nuclei in MAA cell suspension preparations. However, CD138+ plasma cell selected samples may fall below this level—if cells are sparse then count as many cells as possible and state on the worksheet that only x number of cells were scored. In some cases, the yield

Fig. 5 Pitfalls and problems: (a) Lymph node biopsy: Poor fixation and processing are obvious on the H&E but may be less clear by fluorescent microscopy. The darker zone around the edge is well fixed but the central pale area is not fully fixed and will deteriorate further in processing. FISH may fail in the central area. A FISH result is still possible in many cases as the disease may be represented in the outer rim (confirmed by immunohistochemistry. (b) Lymph node biopsy: Good fixation and processing—clear nuclear detail. (c) Bone marrow aspirate smear: Eosinophils can be increased in many conditions. The cells degranulate readily and the highly fluorescent granules may lie over other cells and mimic FISH signals. (d) Bone marrow aspirate smear: Formalin fixation during transit to the laboratory can happen if the aspirate sample is stored near a formalin histology sample (the trephine); vapor leaking is sufficient to partly fix smears. We always suggest separate specimen bags to prevent this from happening. FISH will fail on the smears. (e) Myeloma: First “pull” bone marrow aspirate on left; this is cellular with high levels of plasma cells. The second “pull” aspirate sent for genetic testing is hypocellular. FISH is possible but the cells need to be purified by a cell selection process. (f) Detail of a binucleate plasma cell post-immunomagnetic cell selection; the Perls stain shows the presence of iron on the cell surface. If the iron is not fully removed FISH will be inhibited and false deletion results are possible

56

Sheila J. M. O’Connor et al.

Fig. 6 B-cell chronic lymphocytic leukemia MAA cell suspension: patient has a biallelic deletion of DLEU (loss of green signals). There is a single normal cell in the upper left showing a normal pattern of 2B2R2G. Cytocell deletion/gain probe for alpha 12 (blue), RB1 (red), and DLEU (green). RB1 and DLEU are adjacent on 13q14 so the red and green signals appear fused

of cells post-selection may be very low but as the population is pure then >20 but 80 μg. At least 42 μg of RNA is necessary for the reverse transcription procedure. If the reverse transcription will be performed within 1–2 h, leave the RNA on ice. 4. The final volume of the reverse transcription product is 82 μL. The quantification is not required in this step. 5. The expected final concentration of the ssDNA probe is >150 ng/μL. 6. The treatment of the root tips with nitrous oxide is essential to provide a desired chromosome condensation. The time of this treatment varies according to the species. For example, in potato it takes about 20 min and for maize it takes about 2 h. 7. The time for meristem digestion depends on the diameter and age of the tissue and it is species related. But generally, for example, we digest the potato and maize root tips for 1 h and 2–3 h, respectively. 8. The final volume of the probe master mix is 20 μL. 9. At this time, the slide is placed on the top of iron sheet and do not directly touch the hot plate surface.

Acknowledgments This work was supported by National Science Foundation (NSF) grants MCB-1412948 and IOS-1444514 to J.J. References 1. Jiang JM, Gill BS (2006) Current status and the future of fluorescence in situ hybridization (FISH) in plant genome research. Genome 49:1057–1068 2. McClintock B (1929) Chromosome morphology in Zea mays. Science 69:629

3. Jiang JM, Gill BS, Wang GL et al (1995) Metaphase and interphase fluorescence in situ hybridization mapping of the rice genome with bacterial artificial chromosomes. Proc Natl Acad Sci U S A 92(10):4487–4491

Oligo-FISH: Amplification, Labeling, and Hybridization 4. Lysak MA, Fransz PF, Ali HBM et al (2001) Chromosome painting in Arabidopsis thaliana. Plant J 28:689–697 5. Mukai Y, Yumiko N, Maki Y (1993) Simultaneous discrimination of the three genomes in hexaploid wheat by multicolor fluorescence in situ hybridization using total genomic and highly repeated DNA probes. Genome 36:489–494 6. Kato A, Lamb JC, Birchler JA (2004) Chromosome painting using repetitive DNA sequences as probes for somatic chromosome identification in maize. Proc Natl Acad Sci U S A 101:13554–13559 7. Lamb JC, Danilova T, Bauer MJ et al (2007) Single-gene detection and karyotyping using small-target fluorescence in situ hybridization on maize somatic chromosomes. Genetics 175:1047–1058 8. Danilova TV, Friebe B, Gill BS (2012) Singlecopy gene fluorescence in situ hybridization and genome analysis: Acc-2 loci mark evolutionary chromosomal rearrangements in wheat. Chromosoma 121(6):597–611 9. Beliveau BJ, Joyce EF, Apostolopoulos N et al (2012) Versatile design and synthesis platform for visualizing genomes with oligopaint FISH probes. Proc Natl Acad Sci U S A 109:21301–21306 10. Han YH, Zhang T, Thammapichai P et al (2015) Chromosome-specific painting in Cucumis species using bulked oligonucleotides. Genetics 200:771–779 11. Jiang JM (2019) Fluorescence in situ hybridization in plants: recent developments and future applications. Chromosome Res 27. https://doi.org/10.1007/s10577-01909607-z 12. Braz GT, He L, Zhao H et al (2018) Comparative Oligo-FISH mapping: an efficient and powerful methodology to reveal karyotypic

83

and chromosomal evolution. Genetics 208:513–523 13. He L, Braz GT, Torres GA et al (2018) Chromosome painting in meiosis reveals pairing of specific chromosomes in polyploid Solanum species. Chromosoma 127:505–513 14. Qu MM, Li K, Han Y et al (2017) Integrated karyotyping of woodland strawberry (Fragaria vesca) with oligopaint FISH probes. Cytogenet Genome Res 153:158–164 15. Hou LL, Xu M, Zhang T et al (2018) Chromosome painting and its applications in cultivated and wild rice. BMC Plant Biol 18:110 16. Meng Z, Zhang Z, Yan T et al (2018) Comprehensively characterizing the cytological features of Saccharum spontaneum by the development of a complete set of chromosome-specific oligo probes. Front Plant Sci 9:1624 17. Albert PS, Zhang T, Semrau K et al (2019) Whole-chromosome paints in maize reveal rearrangements, nuclear domains, and chromosomal relationships. Proc Natl Acad Sci U S A 116:1679–1685 18. Murgha YE, Rouillard J-M, Gulari E (2014) Methods for the preparation of large quantities of complex single-stranded oligonucleotide libraries. PLoS One 9:e94752 19. De Carvalho CR, Saraiva LS (1993) An Air Drying Technique for Maize Chromosomes without Enzymatic Maceration. Biotech Histochem 68:142–145 20. Schubert I, Fransz PF, Fuchs J, de Jong JH (2001) Chromosome painting in plants. Methods Cell Sci 23:57–69 21. Ross KJ, Fransz P, Jones GH (1996) A light microscopic atlas of meiosis in Arabidopsis thaliana. Chromosom Res 4:507–516

Chapter 5 Visualizing Genome Reorganization Using 3D DNA FISH Alasdair Jubb and Shelagh Boyle Abstract Understanding how the genome is organized within the cell nucleus is increasingly recognized to be important to understand gene regulation. In 3D DNA fluorescence in situ hybridization (3D DNA FISH) labeled probes complementary to specific loci of interest are hybridized to the genome. The samples are then imaged using fluorescence microscopy, collecting z-stacks through the nuclei, and the relative positions of the hybridized probes are analyzed in the reconstructed 3D images. In this way 3D DNA FISH provides a powerful tool to interrogate how the organization of specific genomic loci changes in response to stimuli. This chapter describes protocols which have allowed us to produce consistent data in cultured cells and paraffin-embedded tissue sections. Key words DNA, Chromatin, FISH, Genetics, Gene regulation, Nuclear organization

1

Introduction The two meters of linear DNA in our cells is extensively folded to fit into each cell’s nucleus—nuclear size being measured in microns— and the way this folding occurs is not random. Organization in higher order genome structure was recognized long before anything more than fragments of the linear sequence of the genome was known, initially through identification of differential Giemsa staining of partially denatured and enzymatically digested DNA. Chromosome bands defined by these Giemsa-based methods, combined with recognition that differential digestion related to density of cytosine bases adjacent to guanine bases (CpG, a surrogate for gene density) which also correlated with patterns of replication banding, demonstrated that the genome is functionally compartmentalized such that gene-rich, early-replicating regions are generally clustered together [1]. In the postgenomic era, the discovery that gene regulation involves sites many thousands of base pairs along the linear genome from the associated coding genes leads to a further question relating to higher order structure: How do distant sites enact control of

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_5, © Springer Science+Business Media, LLC, part of Springer Nature 2020

85

86

Alasdair Jubb and Shelagh Boyle

their target genes? DNA fluorescence in situ hybridization (DNA FISH) played a major role in understanding how the genome is arranged at a large scale in terms of both linear sequence and organization in the nuclear space [2–6]. Precisely mapping the relative positions of genes and their regulatory sites by 3D DNA FISH is now a powerful tool to help understand how stimuli cause acute and persistent changes at specific genomic loci [7–9]. DNA FISH techniques have evolved to address these increasingly detailed biological questions [7–13]. Key steps are sample preparation for 3D imaging, probe preparation and labeling, and image capture and analysis. Many methods have been reported for each element and the overall procedure can become extremely complex [11– 14]. This chapter includes protocols to prepare samples for and undertake 3D DNA FISH using probes made from directly labeled fosmids; see Fig. 1. We describe probe design and preparation, slide preparation from cultured cells, cells in suspension as well as paraffin-embedded tissue sections, and subsequent hybridization.

Fig. 1 Decompaction of chromatin visualized by 3D DNA FISH in mouse macrophages. (a) Genome browser image showing positions of fosmid probes (red and green blocks) relative to genes Klhl6, B3gnt5, and Klhl24 that respond to glucocorticoid treatment. GR ChIP (tags/base pair) shows the position of a glucocorticoid receptor-bound enhancer, identified using chromatin immunoprecipitation followed by sequencing, associated with gene induction [9]. (b) Extended focus example images of 3D DNA FISH before (left) and 24 h after (right) treatment with dexamethasone 100 nM. Scale bars ¼ 1 μm

Visualizing Genome Reorganization with 3D DNA FISH

87

Our streamlined sample preparation protocols, simplified through experience, are combined with descriptions of crucial variables and suggestions for assay optimization. The principles for imaging and quantifying the prepared material are also briefly discussed. Probes for DNA FISH are formed from stretches of nucleic acid complementary to the site of interest. Several methods to generate targeted probes have been reported; here we describe a well-established method using individual fosmid clones labeled using nick translation. Fosmids are large single-copy plasmids that are extremely stable over cell division and are used to make comprehensive genomic libraries in bacteria, each piece 32–48 kb in size [15]. In nick translation DNase I is used to generate random cuts and the 50 –30 exonuclease activity of E. coli DNA polymerase I transcribes new DNA, incorporating labeled nucleotides, converting the fosmid DNA into a large pool of small (~500 bp) labeled fragments complementary to the site of interest. Direct incorporation of fluorophore-conjugated nucleotides during nick translation is an effective approach and can be utilized alongside superresolution imaging [16, 17]. It is also possible by this method to incorporate other haptens such as digoxigenin for detection with sequential layers of antibodies to produce much stronger fluorescent signals [11]. Pools of labeled oligonucleotides have also been shown to be effective as probes to allow precise mapping of chromatin. These oligo pools can be designed to avoid repeat regions and can be either commercially synthesized [18], generated by PCR from genomic DNA [13], or amplified selectively from a commercially prepared oligo library [19]. Labeling of these pools can be either direct or using additional hybridization steps of complementary labeled oligonucleotides. In our hands oligo probes of this type have also performed well alongside the methods described below.

2

Materials

2.1 Fosmid Probe Preparation

1. Fosmid probes complementary to region of genome [20]. 2. L-broth. 3. Chloramphenicol. 4. GTE buffer: 50 mM Glucose, 25 mM Tris pH 8, 10 mM EDTA. 5. Lysozyme. 6. Lysis buffer: 0.2 M NaOH, 1% SDS. 7. Acetate buffer (fresh): 3 M Potassium acetate, 11.5% acetic acid, 100 mL ¼ 60 mL 5 M potassium acetate, 11.5 mL glacial acetic acid, 28.5 mL dH2O. 8. Phenol/chloroform.

88

Alasdair Jubb and Shelagh Boyle

9. Chloroform. 10. Isopropanol. 11. Tris-EDTA. 12. RNaseA at 20 mg/mL. 13. 50 mL Falcon. 14. 1.5 mL Eppendorf tubes. 15. Benchtop centrifuge with 50 mL Falcon inserts. 16. Microcentrifuge. 17. Fume hood. 18. Nick translation salts 10: 0.5 M Tris pH 7.5, 0.1 M MgSO4, 1 mM DTT, 0.5 mg/mL BSA fraction V. 19. DNase I (Roche). 20. DNA polymerase I (Invitrogen). 21. Green496-dUTP (ENZO Life Sciences). 22. Chroma Tide Alexa Fluor 594-5-dUTP. 23. Sterile nuclease-free water. 24. dATP, dCTP, dGTP at 0.5 mM. 25. 0.5 M EDTA pH 8. 26. 20% SDS. 27. Quick Spin column, Sephadex G50 (Roche). 28. Qubit 3.0 and reagents. 29. Agarose gel, ladder, and imaging equipment. 2.2 Slide Preparation: Cells

1. Superfrost microscope slides (Thermo). 2. Quadriperm slide dishes for cell culture. 3. Cytospin (Shandon). 4. 1 Phosphate-buffered saline (PBS). 5. 4% Paraformaldehyde/1 PBS. 6. 0.5% Triton X/1 PBS. 7. Coplin jars (50 mL) and/or glass dishes (200 mL).

2.3 Slide Preparation: Paraffin-Embedded Tissue Sections

1. Oven. 2. Microwave. 3. Xylene. 4. Citrate buffer: 0.1 M Citric acid monohydrate pH 6. 5. 2% Pepsin. 6. HCl 36% w/v. 7. 1 PBS/MgCl2 50 mM.

Visualizing Genome Reorganization with 3D DNA FISH

2.4

FISH

89

1. 20 Trisodium citrate + sodium chloride (SSC). 2. RNase A (20 mg/mL). 3. Denaturant: 2 SSC/70% formamide, pH 7.5. 4. Ethanol. 5. Hybridization mix: 50% Deionized formamide, 2 SSC, 10% dextran sulfate, 1% Tween 20, for 100 μL ¼ 50 μL deionized formamide, 10 μL 20 SSC, 19 μL dH2O, 20 μL 50% dextran sulfate, 1 μL Tween 20. 6. Cot I DNA (Invitrogen). 7. Sonicated salmon sperm (Invitrogen). 8. Rubber solution (e.g., REMA TipTop). 9. 40 ,6-Diamidino-2-phenylindole 50 mg/mL.

(DAPI),

stock

solution

10. Mounting medium (e.g., Vectashield, Prolong Gold). 11. Nail varnish. 12. Coverslips. 13. Water baths. 14. Hybridization chamber. 15. Heat block. 16. pH meter. 17. 100 Wide-field fluorescence microscope with piezoelectric motorized stage. 18. Deconvolution software (e.g., Huygens, Volocity (PerkinElmer), Imaris, Fiji).

3

Methods

3.1 Selection of Fosmid Probes Using Ensembl Browser

1. Navigate to the region of interest using the relevant organism browser at www.ensembl.org. 2. Select region in detail > Configure this page > Sequence and assembly > Clones & misc regions. 3. Select WIBR-2 library to show fosmids in browser and choose appropriate fosmid clones to use as probes. Sites closer than 50 kb may not resolve well with fosmid probes.

3.2 Preparation of Fosmid Probes Using Alkaline Lysis Miniprep (See Note 1)

1. Prepare fosmids as recommended by the supplier. Set up overnight culture in 50 mL Falcon in 5 mL L-broth plus chloramphenicol 25 μg/mL. 2. Spin culture down for 30 s at 4000  g 4  C, and discard supernatant.

90

Alasdair Jubb and Shelagh Boyle

3. Resuspend in 200 μL GTE + 50 mg lysozyme + 1 μL RNase A 20 mg/mL by vigorous mixing—it is very important that the pellet is well resuspended. Leave at room temperature for 5 min. 4. Add 400 μL of lysis buffer, mix by inversion, and leave on ice for 5 min. The mixing must be gentle, or the bacterial DNA will begin to shear and contaminate your fosmid preparation. 5. Centrifuge for 10 min at 4  C maximum speed. 6. Transfer the supernatant to a fresh Eppendorf and in a fume hood add an equal volume of phenol:chloroform. Mix well by inversion and centrifuge for 5 min at 4  C max speed (see Note 2). 7. Transfer the top layer to a fresh Eppendorf, add an equal volume of chloroform, mix, and centrifuge for 3 min at 4  C. 8. Transfer the top layer to a fresh Eppendorf, add an equal volume of isopropanol, leave at 20  C for >1 h, and then centrifuge for 15 min at 4  C. 9. Remove and discard supernatant, add 500 μL 70% ethanol, and re-pellet by centrifuging for 5 min at 4  C. 10. Remove supernatant and air-dry; do not overdry as it will become very difficult to resuspend the pellet. 11. Add 30 μL TE and allow to resuspend. Fosmid DNA can be stored at this point at 20  C. 12. Quantify DNA using Qubit as per the manufacturer’s instructions; take on 1 μg into a nick translation reaction. 3.3

Nick Translation

1. For each probe combine in a clean Eppendorf: l

1 μg Fosmid DNA.

l

2 μL 10 Nick translation salts.

l

2.5 μL of each of 0.5 mM dATP, dCTP, and dGTP.

l

2.5 μL of labeled dUTP, using a different fluorophore for each probe to be used within an experiment to allow identification by fluorescence microscopy; for example for two probes use ChromaTide™ Alexa Fluor™ 594-5-dUTP for one and Green496-dUTP (ENZO life sciences) for the other.

l

1 μL 1:5 Dilution of DNase I diluted in ice-cold water.

l

1 μL DNA polymerase I.

l

Make up to 20 μL with nuclease-free water.

2. Incubate at 16  C for 90 min; this step may benefit from optimization (see Note 3).

Visualizing Genome Reorganization with 3D DNA FISH

91

3. Place on ice, take an aliquot of 3 μL, denature at 95  C for 5 min, place on back on ice for 3 min, and run on an agarose gel. There should be a smear with a peak around 500 kb. 4. Once the correct size is obtained stop the reaction with 3 μL 0.5 M EDTA and 2 μL 20% SDS and then add 65 μL TE for a total volume of 90 μL. 5. Put through a Quick Spin Sephadex G50 column following the manufacturer’s instructions to remove unincorporated nucleotides. 6. Quantify labeled DNA using Qubit as per the manufacturer’s instructions. Probes can be stored at 20  C protected from light. 3.4 Preparation of Slides: Cells

1. For cells growing in a monolayer: wash once in 1 PBS and then proceed to step 3. 2. For cells growing in suspension: collect cells and spin onto slides using a Cytospin (Shandon) keeping the forces below those used when handling that cell type in culture. For example, for human monocytes 1000 rpm (¼112  g) for 3 min at 0.5  106/mL using a Superfrost slide in a EZ Megafunnel is effective (see Note 4). 3. Fix with 4% paraformaldehyde/1 PBS for 10 min. 4. Wash 3 2 min 1 PBS. 5. Permeabilize cells by washing for 10 min in 0.5% Triton X/1 PBS (see Note 5). 6. Wash 3 2 min 1 PBS. 7. Slides can be stored in cold PBS for 2–3 weeks. If longer storage is required, then air-dry and keep at 80  C.

3.5 Preparation of Slides: Paraffin-Embedded Tissue Sections

1. Heat sections in oven at 60  C for 20 min to melt wax. 2. Move to xylene for 10 min 3 (see Note 6). 3. 100% Ethanol for 10 min 3. 4. 95% Ethanol for 5 min 2. 5. 70% Ethanol for 5 min 2. 6. Put into water. 7. Microwave at full power in generous volume (600–800 mL) of citrate buffer for 20–60 min. This time is a critical variable (see Note 7) and dependent on the age and thickness of the sections and each set of sections may require a different time. Where the time is at the high end of this range adding pepsin treatment can be effective—see Subheading 3.5, step 10. 8. Allow to cool for 20 min in the buffer.

92

Alasdair Jubb and Shelagh Boyle

9. Wash and store in water before proceeding to denaturing steps in Subheading 3.6. If using pepsin then: 10. Incubate for 10 min to 1 h at 37  C in 50 mL water with 125 μL 2% pepsin stock and 43 μL concentrated HCl. This time will vary with the age, thickness, and tissue type of the samples. Tissues with dense stroma need more heat and pepsin, e.g., liver sections. 11. Wash slides with 1 PBS/MgCl2 50 mM. 12. Proceed to denaturing steps in Subheading 3.6. 3.6

3D DNA FISH

3.6.1 Preparation

3.6.2 Probe Preparation

Put 50 mL 2 SSC/RNaseA 100 μg/mL in a Coplin jar at 37  C, or a 200 mL slide dish if more than five slides. Make up 200 mL denaturant and place at 80  C (see Note 8) (if the water bath is already at temperature then pre-warm the jar/dish to 45  C to avoid breaking). Prepare ethanol series 70%, 90%, and 100% in 200 mL glass slide dishes for slide dehydration at room temperature, and ensure fresh 70% ethanol for every FISH. Place another aliquot of 70% ethanol on ice. 1. Dispense 100 ng of each directly labeled fosmid probe per slide. 2. Add 6 μg Cot I DNA per fosmid (to bind repeats in the probe) and 5 μg sonicated salmon sperm (nonspecific block). 3. Add 2 volumes of ethanol and spin vac to pellet and dry—once the probe is fragmented, they will dissolve after drying completely. 4. Add 40 μL hybridization mix to the probe per slide (for a 40  22 coverslip). Mix well and spin down. Leave for >45 min to dissolve at room temperature, protecting from light.

3.6.3 Slide Preparation

1. Rinse slides briefly in 2 SSC. 2. Place in pre-warmed 2 SSC/RNaseA 37  C for 1 h. 3. Wash for 2 min in 2 SSC; then put through alcohol series 70%, 90%, and 100% for 2 min each; and then air-dry. 4. Heat slides at 70  C for 5 min in oven and then denature for 40 min in the preheated denaturant (2 SSC/70% formamide pH 7.5) (see Note 8). 5. Place into ice-cold ethanol for 2 min and then 90% and 100% ethanol at room temperature for 2 min each and air-dry. 6. Vortex the prepared probes well, spin down, and denature the probes and competitor DNA for 5 min at >70  C. Reanneal at 37  C for 15 min for all probes where not wanting repeats— i.e., using Cot I.

Visualizing Genome Reorganization with 3D DNA FISH

93

7. Warm slide and coverslip to 37  C on a hot plate. Dispense reannealed probe to coverslip and pick up with slide quickly returning this to 37  C—it is important to keep the temperature as consistent as you can. Seal the edges with a removable rubber solution (e.g., REMA TipTop) and hybridize overnight. This can be done either in a hybridization oven or in a metal dish in a water bath with a foil cover. 3.6.4 Slide Washing

1. Prepare 1000 mL 2 SSC at 45  C and 1000 mL 0.1 SSC at 60  C. 2. Wash 4 3 min 2 SSC at 45  C. 3. Wash 4 3 min 0.1 SSC at 60  C. 4. Final wash in 1 PBS + 1:1000 dilution of 50 mg/mL DAPI stock (see Note 9). 5. Drain excess fluid from slide. Dispense 25 μL Vectashield onto 40  22 mm coverslip and pick up with slide (can also use curing mounting medium such as Prolong Gold but then need to allow it to cure before imaging). Seal edges with nail varnish.

3.7

Imaging

This depends entirely on the setup of local equipment and the software packages but in general: 1. Image using a 100  wide-field fluorescence microscope equipped with a cMOS or CCD camera with a piezoelectrically driven stage. Where the assay has worked well most nuclei should have both signals visible through the eyepiece (see Note 10). 2. We routinely capture z-stacks with 0.2 μm frequency, capturing a complete stack for each channel. 3. Deconvolve the wide-field data using a calculated point spread function for each channel using appropriate software, e.g., Volocity (PerkinElmer) and Huygens (SVI). 4. Measure the distance in 3D either manually or (preferably) using automated image analysis software to identify spots with manual quality control.

4

Notes 1. Column-based mini, midi, and maxi prep kits such as those supplied by major manufacturers are not in our hands effective for fosmid preparation for this assay. There are large plasmid/ fosmid/BAC-specific kits which may be appropriate, but we have not assessed this.

94

Alasdair Jubb and Shelagh Boyle

2. Steps involving phenol and/or chloroform should be carried out using appropriate personal protective equipment and in a fume extraction hood. 3. The required duration may vary and need titration; the usual duration is 90–120 min. Varying the dilution of the DNase I may be necessary if adequate labeling is not obtained within this timescale. 4. Cytospin conditions will need to be optimized per cell type; the minimal centrifugal force that results in adhesion is desirable as higher forces risk distortion of the nuclear architecture. 5. Permeabilization may not always be essential for 3D DNA FISH. 6. Steps including xylene should be performed using appropriate personal protective equipment and in a fume extraction hood. 7. This time is a critical variable and is dependent on the age and thickness of the sections. Each set of sections studied is likely to require a different time. The volume of buffer needs to be sufficient to ensure that liquid does not evaporate to below the level of the slides in the rack during heating. 8. The temperature of denaturant may need to be varied. If probes are not getting into most nuclei when viewed then increase the duration of denaturation and if still not getting consistent probe entry the temperature can be increased too; for example 85  C for 45 min was required for mouse bone marrow-derived macrophages [9]. Formamide is a respiratory sensitizer and carcinogen; steps should be carried out using appropriate personal protective equipment and in a fume extraction hood. 9. If wishing very strong signals then use probes labeled with haptens such as biotin or digoxigenin; at this point layers of fluorescent antibodies can be hybridized for 30–60 min interspersed with 3 2 min washes in 4 SSC + 0.1% Tween, as described in for example [5]. 10. DNA FISH signals for directly labeled probes should ideally require exposure times less than 800 ms when collecting on bin 1. Where the background is low longer exposure times may be tolerated but with an appropriate degree of skepticism—for example if clear pairs of spots are localizing in proximity then more confidence might be inferred, whereas if unexpectedly widely dispersed spots are discovered with very long exposure times the result should be doubted.

Visualizing Genome Reorganization with 3D DNA FISH

95

References 1. Craig JM, Bickmore WA (1993) Chromosome bands – flavours to savour. BioEssays 15:349–354. https://doi.org/10.1002/bies. 950150510 2. Pinkel D, Landegent J, Collins C et al (1988) Fluorescence in situ hybridization with human chromosome-specific libraries: detection of trisomy 21 and translocations of chromosome 4. Proc Natl Acad Sci U S A 85:9138–9142. https://doi.org/10.1073/pnas.85.23.9138 3. Cremer T, Lichter P, Borden J et al (1988) Detection of chromosome aberrations in metaphase and interphase tumor cells by in situ hybridization using chromosome-specific library probes. Hum Genet 80:235–246. https://doi.org/10.1007/BF01790091 4. Cross SH, Lee M, Clark VH et al (1997) The chromosomal distribution of CpG Islands in the mouse: evidence for genome scrambling in the rodent lineage. Genomics 40:454–461. https://doi.org/10.1006/geno.1996.4598 5. Gilbert N, Boyle S, Fiegler H et al (2004) Chromatin architecture of the human genome. Cell 118:555–566. https://doi.org/10.1016/ j.cell.2004.08.011 6. Fraser P, Bickmore W (2007) Nuclear organization of the genome and the potential for gene regulation. Nature 447:413–417. https://doi.org/10.1038/nature05916 7. Williamson I, Eskeland R, Lettice L et al (2012) Anterior-posterior differences in HoxD chromatin topology in limb development. Development 139:3157–3167. https:// doi.org/10.1242/dev.081174 8. Rafique S, Thomas JS, Sproul D et al (2015) Estrogen-induced chromatin decondensation and nuclear re-organization linked to regional epigenetic regulation in breast cancer. Genome Biol 16:145. https://doi.org/10.1186/ s13059-015-0719-9 9. Jubb AW, Boyle S, Hume DA et al (2017) Glucocorticoid receptor binding induces rapid and prolonged large-scale chromatin decompaction at multiple target loci. Cell Rep 21:3022–3031. https://doi.org/10.1016/j. celrep.2017.11.053 10. Chambeyron S, Bickmore WA (2004) Chromatin decondensation and nuclear reorganization of the HoxB locus upon induction of transcription. Genes Dev 18:1119–1130. https://doi.org/10.1101/gad.292104

11. Morey C, Da Silva NR, Perry P et al (2007) Nuclear reorganisation and chromatin decondensation are conserved, but distinct, mechanisms linked to Hox gene activation. Development 134:909–919. https://doi.org/ 10.1242/dev.02779 ˜ o MS et al 12. Beliveau BJ, Boettiger AN, Avendan (2015) Single-molecule super-resolution imaging of chromosomes and in situ haplotype visualization using Oligopaint FISH probes. Nat Commun 6:7147. https://doi.org/10.1038/ ncomms8147 13. Bienko M, Crosetto N, Teytelman L et al (2013) A versatile genome-scale PCR-based pipeline for high-definition DNA FISH. Nat Methods 10:122–124. https://doi.org/10. 1038/nmeth.2306 14. Giorgetti L, Galupa R, Nora EP et al (2014) Predictive polymer modeling reveals coupled fluctuations in chromosome conformation and transcription. Cell 157:950–963. https:// doi.org/10.1016/j.cell.2014.03.025 15. International Human Genome Sequencing Consortium (2004) Finishing the euchromatic sequence of the human genome. Nature 431:931–945. https://doi.org/10.1038/ nature03001 16. Benabdallah NS, Williamson I, Illingworth RS, et al (2017) PARP mediated chromatin unfolding is coupled to long-range enhancer activation. bioRxiv 155325. https://doi.org/10. 1101/155325 17. Williamson I, Lettice LA, Hill RE et al (2016) Shh and ZRS enhancer colocalisation is specific to the zone of polarising activity. Development 143:2994–3001. https://doi.org/10.1242/ dev.139188 18. Boyle S, Rodesch MJ, Halvensleben HA et al (2011) Fluorescence in situ hybridization with high-complexity repeat-free oligonucleotide probes generated by massively parallel synthesis. Chromosom Res 19:901–909. https://doi. org/10.1007/s10577-011-9245-0 19. Beliveau BJ, Joyce EF, Apostolopoulos N et al (2012) Versatile design and synthesis platform for visualizing genomes with Oligopaint FISH probes. Proc Natl Acad Sci U S A 109:21301–21306. https://doi.org/10. 1073/pnas.1213818110 20. BACPAC Children’s Hospital Oakland Research Institute. https://bacpacresources. org/

Part III Methods for Cultured Cells

Chapter 6 MicroRNA In Situ Hybridization in Paraffin-Embedded Cultured Cells Jaslin P. James, Laura Johnsen, Trine Møller, and Boye Schnack Nielsen Abstract MicroRNA-21 (miR-21) is one of the most abundant microRNAs in cancer tissues and is considered a strong prognostic biomarker. In situ hybridization (ISH) analyses using locked nucleic acid (LNA) probes have shown that miR-21 is expressed in stromal fibroblastic cells and in subsets of cancer cells. Image analysis of the miR-21 ISH signal has shown that increased expression estimate is associated with poor prognosis in colon cancer. However, assessment of the ISH signal by image analysis to obtain quantitative estimates has been done in retrospective studies without normalization of the expression estimates to reference parameters. The ISH signal output is sensitive to several experimental parameters, including hybridization temperature, probe concentration, and pretreatment, and therefore improved standardized procedures are warranted. We considered the use of paraffin-embedded cultured cells (PECCs) as reference standards that potentially can accompany staining of clinical cancer samples. We found that the cancer cell lines HT-29, CACO-2, and HeLa cells express miR-21 when measured by ISH, and used those cell lines to obtain PECCs. In this methods chapter we present a fixation and embedding procedure to obtain PECCs suitable for microRNA ISH and a double-fluorescence protocol to stain microRNAs together with protein markers in the PECCs. Key words Double immunofluorescence, HT-29 cells, In situ hybridization, LNA, microRNA, miR-21

1

Introduction MicroRNAs (miRNAs) are a group of small, highly preserved, single-stranded noncoding RNAs that regulate gene expression at the posttranscriptional level [1–3]. MiRNAs play important roles during development and disease including cancer, neurological, viral, and metabolic diseases [4–6]. They are expressed as precursor transcripts containing regular short hairpin structures that are processed into short strands of RNA typically consisting of 19–22 nucleotides. These mature miRNAs are loaded into the RNA-induced silencing complex (RISC), which enables the miRNAs to bind to the complementary regions at the 30 UTR of mRNAs for the regulation of protein expression [7]. The dynamic

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_6, © Springer Science+Business Media, LLC, part of Springer Nature 2020

99

100

Jaslin P. James et al.

expression of miRNAs can be used to detect and monitor disease development, thus assisting in taking decisions in patient care. MiRNAs can be measured in total RNA obtained from tissue samples by a variety of methods like RT-qPCR, microarrays, and sequencing methods [8–11]. Loaded into RISC, the miRNAs show high stability compared to mRNAs [12], which allows miRNAs also to be measured in liquid biopsies such as blood and urine [13]. MicroRNA-21 (miR-21) is one of the most commonly upregulated miRNAs in cancer and is considered an oncogenic miRNA that regulates expression of tumor suppressors such as PTEN and PDCD4 [14]. In colon cancer, miR-21 is upregulated in early stages of disease progression, and high expression levels are associated with increased risk of recurrence [15, 16]. Schetter et al. measured the relative upregulation of miR-21 in total RNA from cancer and normal tissue samples using quantitative RT-PCR [15], whereas Nielsen et al. used in situ hybridization followed by image analysis [16]. The prognostic importance of miR-21 in stage II colon cancer has been further validated by similar analyses in a population cohort both by ISH and qPCR [17, 18]. A preferred method to measure miRNA expression levels is qPCR that is a robust and reproducible quantitative method; however, the cellular context and the cellular origin of expression are missing. The ISH analyses of colon cancer samples have revealed that miR-21 is primarily expressed in tumor-associated fibroblastic cells and a subset of cancer cells [16]. This histological information is important for interpretation of the role of miR-21 in tumor biology. In a recent study, Thorlacius-Ussing et al. identified miR-21 as a potential biomarker, measured both by qPCR and ISH, in inflammatory bowel disease (IBD), and showed that quantitative ISH may be similar or superior in discriminating ulcerative colitis from Crohn’s disease [19]. Therefore, in situ hybridization and associated image analysis is an attractive alternative method to obtain quantitative expression data in tissue samples that could be useful in the differential diagnosis. Detection of miRNAs by in situ hybridization was first described in whole-mount specimens [20], and took advantage of DNA oligonucleotides with locked nucleic acids (LNA) incorporated, which was found to increase the binding affinity and specificity between complementary nucleotide sequences [21]. The chimeric DNA:LNA oligonucleotides used for ISH analyses are typically conjugated with haptens like digoxigenin (DIG) or 6-carboxyfluorescein (FAM), enabling detection with enzymeconjugated antibodies in a typical immunohistochemical workflow [22]. The LNA-based ISH method has been developed for both paraffin-embedded [22, 23] and OCT-embedded (frozen) [24] tissue samples in manual [22] and automated [16, 25] procedures, and in combination with immunohistochemistry [26, 27] and RNAscope [28].

MicroRNAs in PECCs

101

In diagnostic immunohistochemistry, where antibodies are used for detection of protein biomarkers, the potential use of cellbased reference standards is considered a future tool to improve scoring precision [29]. For example in the determination of PD-L1 positivity, strategies involving reference standards based on formalin-fixed and paraffin-embedded engineered cell lines [30] enable controlled scoring and quantitation of staining intensities in tissue sections. Aiming at a clinically applicable diagnostic miR21 ISH assay, we are exploring the possibility to establish a cellbased reference standard for miRNAs, and here present a manual 1-day fluorescence-based protocol for staining of miR-21 in paraffin-embedded cultured cells (PECCs).

2

Materials

2.1 Reagents and Buffers

1. Cell culture: HT-29 (human colorectal adenocarcinoma cell line), HeLa (human cervical cancer cell line), CACO2 (human colorectal adenocarcinoma cell line). Cells were grown to 70% confluence on a T175 adherent cell culture flask in DMEM (Dulbecco’s modified Eagle’s medium). 2. Neutral-buffered formalin (NBF). 3. Histogel (Thermo Fisher Scientific). 4. Double-labeled LNA™ probe (Exiqon/QIAGEN): In this study we use miRCURY™ double-DIG-labeled LNA™ probes specific for miR-17, miR-21, miR-122, miR-124, and miR-126 as well as the negative control probe, scramble (see Note 1). 5. In situ hybridization buffer (microRNA ISH buffer set, Exiqon/QIAGEN). 6. 1 PBS, RNase-free quality. 7. Tween-20 and a 10% solution in RNase-depleted water. 8. RNase-depleted water, for example, RNase-free Milli-Q water (Millipore). 9. 20 SSC buffer, RNase-free quality. 10. Sheep anti-DIG-POD: Diluted 1:400 times in blocking solution. 11. Primary antibodies: In this study we have used a mouse monoclonal antibody against cytokeratin (Dako). 12. FITC-conjugated Immunoresearch).

goat-anti-mouse

(Jackson

13. Tyramine signal amplification (TSA) reagent, for example, TSA–Cy3 (PerkinElmer). 14. Anti-fade Prolong Gold with DAPI (Invitrogen).

102

Jaslin P. James et al.

15. SuperFrost® Plus (Thermo Fisher Scientific). 16. For RNase-depleting solution: RNase ZAP, RNase Away, or similar. 17. Hydrogen peroxide, 30% (VWR). 18. Proteinase-K buffer: To 900 mL Milli-Q water, add 5 mL of 1 M Tris–HCl (pH 7.4), 2 mL 0.5 M EDTA, and 0.2 mL 5 M NaCl. Adjust volume to 1000 mL. 19. 15 μg/mL Proteinase-K reagent: To 10 mL of 1 proteinaseK buffer, add 7.5 μL proteinase-K stock of 20 mg/mL. 20. 0.1 SSC buffer: To 995 mL Milli-Q water, add 5 mL 20 SSC. The SSC buffer should be autoclaved. 21. 0.1% PBST: To 1 L of PBS (pH 7.4), add 1 mL Tween-20. 22. Blocking solution: To 37.5 mL of Milli-Q water, add 5 mL 1 M Tris–HCl (pH 7.5), 1.5 mL 5 M NaCL, 5 mL fetal bovine serum (FBS), and 100 mL diluted Tween-20 (10% solution). 23. 3% Hydrogen peroxide: Add 3 mL of H2O2 to 80 mL of Milli Q water and make up to 100 mL with Milli-Q water. 2.2

Equipment

1. Cell culture hood and 37  C incubator. 2. Biopsy cassettes. 3. Plastic cryomolds or similar. 4. Hybridizer (Dako). 5. Shandon Sequenza Slide Racks (Thermo Scientific). 6. Horizontal humidifying chamber. 7. Glassware: All glassware, including Coplin jars, glass-staining racks, and stacks of cover glass and bottles for buffers, were heat treated in an oven at 180  C for 8 h. The items were covered by aluminum foil before being placed in the oven in order to prevent contamination when removing the items afterward.

3

Methods

3.1 Cell Preparation and Embedding of Cells

1. Harvest HT-29 cells and then spin down at 300  g for 10 min in PBS. 2. Wash the cells in PBS to remove the medium residue. 3. Add 10 mL of 10% NBF to the cell pellet and gently pipette up and down to resuspend the cell pellet. Ensure that there are no cell clumps. 4. Add an additional 40 mL of 10% NBF to the 50 mL tube. 5. Incubate the cells in NBF for 3 days (see Note 2).

MicroRNAs in PECCs

103

Table 1 Procedure to dehydrate and paraffin embed the solid cell pellets Step

Cycles

Time (min)

Reagents

Temperature

1

2

30

70% Ethanol

RT

2

2

30

96% Ethanol

RT

3

2

30

99% Ethanol

RT

4

1

30

Xylene

RT

5

1

60

Xylene

RT

6

1

60

Paraffin

60  C

7

1

Until next day

Paraffin

60  C

8

Block embedding

6. Spin down the cells at 300  g for 10 min, and decant the NBF into an appropriate waste container without disturbing the pellet. 7. Resuspend the cells in 10 mL PBS, and then transfer the cell suspension to a 15 mL tube. Centrifuge at 300  g for 10 min and aspirate off the PBS without disturbing the pellet. 8. Suspend the cells in 70% ethanol, centrifuge at 300  g for 10 min, and pour off the supernatant. 9. Warm up Histogel approximately for 1 min until it becomes liquefied. 10. Gently resuspend the cell pellet in 5–10 drops of liquid Histogel. 11. Divide the Histogel with the mass of cells using a disposable Pasteur pipette into plastic cryomolds. 12. Cool the Histogel to RT and transfer it to a pre-labeled biopsy cassette. 13. Dehydrate the solid cell pellet into a paraffin block by following the procedure listed in Table 1. 14. Remove the paraffin-embedded pellets from the cassette and use traditional paraffin-embedding molds to obtain the paraffin blocks. 3.2

Tissue Sections

1. Clean the workstation, including bench top, microtome, blade holder, brushes, tweezers, cooling plate, water bath, etc. using RNase ZAP. 2. Set the cooling plate to 15  C and then place the FFPE block on the plate.

104

Jaslin P. James et al.

3. Fill the warm water bath with RNase-free Milli-Q water and set the temperature to 40–50  C depending on the type of paraffin used for embedding. 4. Prepare another water bath in an RNase-free Coplin jar containing RNase-free Milli-Q water at room temperature. 5. Insert a new disposable blade in the knife carrier, and place the block in the cassette clamp. The first couple of sections, which might be contaminated during previous handling such as from tissue preparation or embedding, should be avoided by trimming the block. 6. Cut 5 or 6 μm sections and place them into a jar with RNasefree Milli-Q water at room temperature (RT), in order to reverse the folding. 7. Transfer the slide to a heated water bath of about 40–50  C, in order to avoid the folds and overlaps in the structure by stretching the tissue, and then mount sections immediately on SuperFrost® Plus glass slides. Allow water to slide away from in between the paraffin section and the glass slide to avoid sections falling off during deparaffinization. 8. Let the slides dry for 2 h at RT and store at 4  C in a dry box containing silica gel. 9. Melt paraffin at 60  C on the day prior to the in situ hybridization experiment and store dry at RT. 3.3 In Situ Hybridization

1. Deparaffinize slides in xylene and ethanol solutions in Coplin jars ending up in PBS. Place slides in xylene for 15 min (via 2–3 Coplin jars) and then hydrate through ethanol solutions 99% (3 Coplin jars), 96% (2 Coplin jars), and 70% (2 Coplin jars) to PBS (2 Coplin jars), 5 min in each jar. In parallel, prepare a water bath and SSC buffer to be heated to 55  C (or the hybridization temperature). 2. Pre-digestion of tissue sections is done by applying 300 μL/ slide of the proteinase-K reagent at 5–15 μg/mL (see Note 3) directly on the tissue and incubate for 10 min at 37  C in a horizontal humidifying chamber or in a hybridizer. 3. Discard the proteinase-K reagent and wash twice with PBS. 4. Remove excess PBS, immediately apply 50 μL DIG probe solution, and gently shield with cover glass. The probe solution is prepared as follows: denature the microRNA, scramble LNA™ probe, and dilute the probe in hybridization buffer. LNA™ probes for other miRNAs may require optimization of the concentration (see Note 4). 5. Place the slides in the hybridizer and start a preset hybridization program for 1 h at 55–57  C (see Note 5).

MicroRNAs in PECCs

105

6. Place slides into 55–57  C pre-warmed 0.1 SSC in a Coplin jar. The cover slides will easily detach. Then transfer slides to another casket with 55  C pre-warmed 0.1 SSC. Wash slides thrice using 55  C pre-warmed 0.1 SSC. 7. Discard the SSC washing buffer and wash twice with PBS in Coplin jar. 8. Remove PBS, add freshly prepared 3% hydrogen peroxide, and incubate at room temperature for 15 min. 9. Transfer slides to PBS-T and mount the slides into Shandon Sequenza® Slide Racks. Avoid air bubbles during mounting. 10. Incubate with 300 μL blocking solution for 15 min at RT. 11. Detect DIG probes by applying sheep anti-DIG-POD diluted 1:400 in blocking solution and incubate for 30 min at RT. 12. Wash each slide with 300 μL PBS twice for 2 min. 13. Incubate 150 μL freshly prepared TSA-Cy3 reagent for 4–8 min at RT (see Note 6). Protect from light during development. 14. Wash each slide with 300 μL PBS once for 2 min. 15. Detection of protein markers is done by incubating 200 μL primary antibody diluted in PBS containing 1% BSA for 30 min at room temperature. 16. Wash each slide with 300 μL of PBS. 17. The protein marker antibodies are detected by incubating FITC-conjugated detecting antibody (e.g., anti-mouse) diluted 1:400 in PBS for 30 min at room temperature. 18. Wash each slide twice with 300 μL of PBS. 19. Mount slides with anti-fade ProLong Gold containing DAPI. Store slides in the dark and at 4  C. Evaluate staining result after 24 h. 20. Evaluate slides using a fluorescence microscope with filters allowing detection of FITC, Cy3, and DAPI emission. Standardize the exposure time according to the optimized intensity for the target probe signal (here miR-21), and use the same for comparison with staining intensities in the other control sections including scramble. The exposure time used for snRNAU6 was adjusted separately. 21. Considerations on the specificity of the signal are required for the evaluation of miRNA ISH results (see Note 1 and Note 7).

106

4

Jaslin P. James et al.

Notes 1. Standard LNA-based miRNA ISH analyses include negative and sometimes positive control probes. The negative control probe, scramble, is designed from a generic target sequence not found among human transcripts [23]. In the case of the colon cancer cell line HT-29 it would be expected to be negative also for cell-specific microRNA such as the endothelial cell-specific miR-126. On the other hand, miR-17 is a potential positive control probe for the HT-29 cells, which is of epithelial origin. Alternative reference probes in the HT-29 cells could be the neuronal miR-124 and the epithelial miR-200b/c as negative and positive reference probes, respectively. The LNA probe against snRNA-U6 represents an overall positive control often used during implementation of the ISH assay and the ISH signal is seen in the nuclei (see Fig. 1).

Fig. 1 ISH for various microRNAs. LNA probes to miR-21, miR-17, and miR-126 along with scramble were tested at 10 nM on HT-29 PECC samples. ISH was performed at 57  C after pretreatment with 5 μg/mL proteinase-K. Exposure time for the LNA probes was standardized to the miR-21 signal. Epifluorescence microscopy. Scale bar, 50 μm

MicroRNAs in PECCs

107

Fig. 2 Effect of fixation on miRNA ISH signal intensity. LNA probes to miR-21 and scramble were tested on HT-29 PECCs that had been fixed for 3 days (F3), 4 days (F4), and 6 days (F6). After the miRNA ISH, the sections were stained for cytokeratin (CK) using immunofluorescence. ISH was performed after pretreatment with 5 μg/mL proteinase-K at 57  C. LNA probes to miR-21 and scramble were used at 10 nM. Exposure time for the miR-21 and scramble probes was that of the miR-21 signal in the F3 sample. Epifluorescence microscopy. Scale bar, 50 μm

2. Routinely processed tissue specimens in clinical pathology departments are typically fixed in NBF for 1–3 days and then paraffin embedded. Such FFPE samples are usually suitable for microRNA ISH analyses. Here, we have tested fixation of the cultured cells in NBF for 3, 4, 6, and 7 days followed by paraffin embedding. The effect of fixation time had no major effect on the performance of the miR-21 ISH assay (see Fig. 2). 3. Proteinase-K treatment is one of the key parameters to optimize in any miRNA ISH assay on FFPE tissue sections. The amount of proteinase-K treatment will depend on the degree of fixation and the type of tissue. If the fixation is hard, extended proteinase-K treatment may be required. For the analysis of PECCs, we tested protease-K at 5 and 15 μg/mL, and found that the best miR-21 ISH signal intensity was obtained using 5 μg/mL (see Fig. 3). However, the probe to snRNA-U6 consistently performed better at 15 μg/mL. This suggests that different RNA target types may require different protease treatments for optimal performance. 4. A second key parameter to optimize is the probe concentration, and should normally be tested together with variation of the hybridization temperature. In the fluorescence assay presented here, we tested the miR-21 probe at 0, 2, 5, 10, and 20 nM on the PECCS at 57  C, and found that 10 nM gave the best result in terms of dynamics in the staining between cells.

108

Jaslin P. James et al.

Fig. 3 Optimization of protease-K concentration. Different protease-K concentrations, 5 μg/mL (PK5) and 15 μg/mL (PK15), were tested prior to hybridization with LNA probes to miR-21 (10 nM), scramble (10 nM), and snRNA-U6 (1nM) at 57  C. Protease-K at 5 μg/mL is giving good signal-to-noise ratio with respect to miR21, whereas snRNA-U6 performed better at 15 μg/mL. Exposure time for the LNA probes was that of the miR21 signal at PK5. Epifluorescence microscopy. Scale bar, 50 μm

5. The third key parameter to optimize is the hybridization temperature. In order to avoid potential cross-hybridization, the highest possible hybridization temperature should be used. Using the Qiagen ISH buffer, the typical hybridization temperature range is 52–57  C, which depends on the RNA Tm of the probe. The optimal hybridization temperature is usually 25–30  C below the theoretical RNA Tm of the LNA probe. Here we tested hybridization at 55  C and 57  C, and found for both the miR-21 and miR-17 probes that 57  C gave the best result in terms of dynamics in the staining between cells. 6. The standard incubation time of the TSA substrate in the presented protocol is 8 min. The incubation time, typically varying from 2 to 15 min, can be used to regulate the intensity of ISH signal relative to the autofluorescence and background staining intensity obtained with the negative control probe. We tested 4- and 8-min incubation at RT and obtained the most reproducible ISH signal after 8 min. 7. In this method study we also tested the colon cancer cell line, CACO-2, and the cervix cancer cell line, HeLa, with the miR17 and miR-21 probes, and found that both cell lines are positive for miR-17 and miR-21 and thus suitable further studies of PECCs (see Fig. 4).

Acknowledgments This study was supported by The Danish Agency for Science and Higher Education.

MicroRNAs in PECCs

109

Fig. 4 MicroRNA ISH in different cell lines. LNA probes to miR-21, miR-17, and scramble were tested at 10 nM on sections from PECCs with CACO-2, HeLa, and HT-29 cells. Different ISH signal intensities for miR-17 and miR-21 are observed on the three cell lines. All LNA probes were tested at 10 nM probe concentration and hybridized at 57  C. Exposure time for the LNA probes was that of the miR-21 in HT-29 cells. Confocal slide scanning microscopy, see ref. 28. Scale bar, 50 μm References 1. Ambros V (2001) microRNAs: tiny regulators with great potential. Cell 107:823–826. https://doi.org/10.1007/978-3-642-001505_33 2. Lee RC, Feinbaum RL, Ambros V (1993) The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to &II-14 rosalind. Cell 75:843–854. https://doi. org/10.1016/0092-8674(93)90529-Y 3. Pillai RS, Bhattacharyya SN, Artus CG et al (2005) Inhibition of translational initiation by Let-7 MicroRNA in human cells. Science (80-) 309:1573–1576 4. Sempere LF, Kauppinen S (2010) Translational implications of MicroRNAs in clinical diagnostics and therapeutics. In: Bradshaw RA, Dennis EA (eds) Handbook of cell signaling, 2nd edn. Academic Press, Cambridge, pp 2965–2981

5. Xuan Y, Yang H, Zhao L et al (2015) MicroRNAs in colorectal cancer: small molecules with big functions. Cancer Lett 360:89–105 6. Thakral S, Ghoshal K (2015) miR-122 is a unique molecule with great potential in diagnosis, prognosis of liver disease, and therapy both as miRNA mimic and antimir. Curr Gene Ther 15:142–150 7. Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function review. Cell 116:281–297 8. Ferdin J, Kunej T, Calin GA (2010) Non-coding RNAs: identification of cancerassociated microRNAs by gene profiling. Technol Cancer Res Treat 9:123–138. https://doi. org/10.1177/153303461000900202 9. Sørensen KD, Ørntoft TF (2010) Discovery of prostate cancer biomarkers by microarray gene expression profiling. Expert Rev Mol Diagn

110

Jaslin P. James et al.

10:49–64. https://doi.org/10.1586/erm. 09.74 10. Jensen SG, Lamy P, Rasmussen MH et al (2011) Evaluation of two commercial global miRNA expression profiling platforms for detection of less abundant miRNAs. BMC Genomics 12:435. https://doi.org/10.1186/ 1471-2164-12-435 11. Anderson AL, Stanger SJ, Mihalas BP et al (2015) Assessment of microRNA expression in mouse epididymal epithelial cells and spermatozoa by next generation sequencing. Genomics Data 6:208–211. https://doi.org/ 10.1016/j.gdata.2015.09.012 12. Filipowicz W, Bhattacharyya SN, Sonenberg N (2008) Mechanisms of post-transcriptional regulation by microRNAs: are the answers in sight? Nat Rev Genet 9:102–114 13. To¨lle A, Jung M, Rabenhorst S et al (2013) Identification of microRNAs in blood and urine as tumour markers for the detection of urinary bladder cancer. Oncol Rep 30:1949–1956. https://doi.org/10.3892/or. 2013.2621 14. Feng Y, Tsao C (2016) Emerging role of microRNA-21 in cancer (review). Biomed Rep 5:395–402. https://doi.org/10.3892/br.2016. 747 15. Schetter AJ, Leung SY, Sohn JJ et al (2008) MicroRNA expression profiles associated with prognosis and therapeutic outcome in colon adenocarcinoma. JAMA 299:425–436. https://doi.org/10.1001/jama.299.4.425 16. Nielsen BS, Jørgensen S, Fog JU et al (2011) High levels of microRNA-21 in the stroma of colorectal cancers predict short disease-free survival in stage II colon cancer patients. Clin Exp Metastasis 28:27–38. https://doi.org/10. 1007/s10585-010-9355-7 17. Kjaer-Frifeldt S, Hansen TF, Nielsen BS et al (2012) The prognostic importance of miR-21 in stage II colon cancer: a population-based study. Br J Cancer 107:1169–1174. https:// doi.org/10.1038/bjc.2012.365 18. Hansen TF, Nielsen BS, Joergensen S et al (2012) The prognostic importance of miR-21 in stage II colon cancer: a population-based study. Br J Cancer 107:1169–1174. https:// doi.org/10.1038/bjc.2012.365 19. Thorlacius-Ussing G, Schnack Nielsen B, Andersen V et al (2017) Expression and localization of miR-21 and miR-126 in mucosal tissue from patients with inflammatory bowel disease. Inflamm Bowel Dis 23:739–752. https://doi.org/10.1097/MIB. 0000000000001086

20. Kloosterman WP, Wienholds E, de Bruijn E et al (2005) In situ detection of miRNAs in animal embryos using LNA-modified oligonucleotide probes. Nat Methods 3:27 21. Vester B, Wengel J (2004) LNA (locked nucleic acid): high-affinity targeting of complementary RNA and DNA. Biochemistry 43:13233–13241. https://doi.org/10.1021/ bi0485732 22. Gould BR, Damgaard T, Nielsen BS (2017) Chromogenic in situ hybridization methods for microRNA biomarker monitoring of drug safety and efficacy, Methods in molecular biology. Springer, New York, pp 399–412 23. Jørgensen S, Baker A, Møller S, Nielsen BS (2010) Robust one-day in situ hybridization protocol for detection of microRNAs in paraffin samples using LNA probes. Methods 52:375–381. https://doi.org/10.1016/j. ymeth.2010.07.002 24. Nielsen BS, Møller T, Holmstrøm K (2014) In: Nielsen BS (ed) Chromogen detection of microRNA in frozen clinical tissue samples using LNA™ probe technology BT - in situ hybridization protocols. Springer, New York, pp 77–84 25. Sempere LF, Korc M (2013) In: Su GH (ed) A method for conducting highly sensitive MicroRNA In situ hybridization and immunohistochemical analysis in pancreatic cancer BT pancreatic cancer: Methods and protocols. Humana Press, Totowa, NJ, pp 43–59 26. Nielsen BS, Holmstrøm K (2019) Combined MicroRNA in situ hybridization and immunohistochemical detection of protein markers, Methods in molecular biology. Springer, New York, pp 271–286 27. Sempere LF, Preis M, Yezefski T et al (2010) Fluorescence-based codetection with protein markers reveals distinct cellular compartments for altered microRNA expression in solid tumors. Clin Cancer Res 16:4246–4255. https://doi. org/10.1158/1078-0432.CCR-10-1152 28. Møller T, James JP, Holmstrøm K et al (2019) Co-detection of miR-21 and TNF-α mRNA in budding cancer cells in colorectal cancer. Int J Mol Sci 20:1907. https://doi.org/10.3390/ ijms20081907 29. Torlakovic EE, Nielsen S, Vyberg M, Taylor CR (2015) Getting controls under control: the time is now for immunohistochemistry. J Clin Pathol 68:879–882. https://doi.org/10. 1136/jclinpath-2014-202705 30. Solutions HI. Cell PD-L1 Reference Standards. https://www.horizondiscovery.com/ref erence-standards/ihc/pd-l1-referencestandards. Accessed 01 July 2019

Chapter 7 Multiplexed Detection and Analysis of Low-Abundance Long Noncoding RNA Using RNAscope™ in Cultured Cells Julia Jones, Heather Zecchini, and Sankari Nagarajan Abstract Detecting low-abundance long noncoding RNAs (lncRNAs) is extremely difficult due to their expression levels. Deeper sequencing with extensive protocols is required to detect these RNAs and high-throughput screens to examine the regulation of these RNAs are challenging. This protocol provides a multiplexed and robust method of detecting low-abundance RNAs, with improved signal-to-noise ratio using RNAscopebased RNA-FISH which utilizes a series of amplification steps. We have validated this protocol for investigating the regulation of low-abundance lncRNAs, which would be ideal for in vitro screening in 96-well plates. Key words lncRNAs, Multiplex RNA-FISH, Operetta, RNA-FISH, RNAscope

1

Introduction lncRNAs are becoming increasingly studied due to their versatile function in many cellular processes including regulation of gene expression, DNA repair, development, differentiation, etc. [1, 2]. The majority of these RNAs are enriched in nuclei comparative to mRNAs, which are mostly transported to cytoplasm after RNA processing [3]. Detecting highly abundant lncRNAs is straightforward using qPCR and next-generation sequencing. However, the majority of noncoding RNAs are expressed at low levels and are highly unstable, but possess crucial functions in development and disease including cancers and neurodegenerative diseases [4]. Nascent RNA sequencing provides more robust information about these RNAs, but the protocols are labor intensive, have low success rates, and require deeper sequencing to detect low copy numbers [5, 6]. Such protocols are difficult to be scaled for high-throughput screening experiments. RNA-FISH, using fluorescently labeled oligos can be utilized to overcome these problems, but detection of low-abundance RNAs may still be problematic due to high levels of fluorescent background signals.

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_7, © Springer Science+Business Media, LLC, part of Springer Nature 2020

111

112

Julia Jones et al.

RNAscope technology from Advanced Cell Diagnostics provides high specificity and sensitivity due to the use of up to 20 Z-probe pairs per RNA target in a single probe set. Briefly, each probe contains an 18–25 base sequence complementary to the target RNA, a linker sequence, and a 14-base tail sequence that becomes a 28-base binding site for a preamplifier when the pair of probes are contiguously hybridized to the target. The preamplifiers contain 20 binding sites for the amplifiers and amplifiers can bind up to 20 label probes. Therefore, each probe may have up to 400 label probes. Hybridization of only 3 Z-probe pairs is sufficient to detect the RNAs in tissue sections, and so a probe set containing 20 Z-probe pairs gives robustness and redundancy that allows visualization of even partially degraded RNAs [7]. Using this technology, we were able to detect low-copy-number nuclear RNAs at the single molecule level and identify the locations of the actively transcribing regions in the nuclei and perinuclear bodies. We then used this technique to detect multiple mRNAs regulated by the lncRNAs in the same cell in 96-well plate format. Overall, this protocol provides a cutting-edge solution for the detection of up to four low-abundance RNA targets simultaneously at the single-cell level, and is suitable for use in high-throughput screens.

2

Materials All solutions during RNAscope protocol should be made using DEPC-treated or RNase-free water.

2.1

Cell Culture

1. Biological safety cabinet—CL1. 2. MCF7 cells—80% confluency, grown under passage below 30. 3. DMEM: high glucose with phenol red and pyruvate, with 10% fetal bovine serum, 2 mM glutamine, and penicillin and streptomycin. 4. 0.05% Trypsin-EDTA: 0.5% Trypsin-EDTA (Thermo Fisher Scientific) diluted with PBS. 5. CellCarrier-96 Ultra Microplates, tissue culture treated, black, 96-well with lid (Perkin-Elmer, Cat# 6055302) (see Note 1). 6. Cell-Tak (Corning™ Cat# 354240): 1.83 mg/mL in 5% acetic acid (see Note 2). Cell-Tak buffer solution made in 0.1 M sodium bicarbonate buffer pH 8.0 as 1.12 μg Cell-Tak/well in a 96-well plate. 1 N NaOH should be added to the buffer solution as ½ volume of Cell-Tak stock to maintain the pH. 7. Cell culture incubator, 37  C, 5% CO2.

Multiplex lncRNA Detection in Cultured Cells

113

8. Hemocytometer (Neubauer’s chamber). 9. T175 flasks. 10. 10% Neutral-buffered formalin. 2.2

RNAscope

1. HybEZ hybridization system (Bio-Techne). 2. 20 Sodium saline citrate (SSC): Dissolve 175.3 g sodium chloride and 88.2 g of sodium citrate in 800 mL distilled water. Adjust the pH to 7.0 with 1 M hydrochloric acid. Adjust the volume to 1 L with distilled water. Sterilize by autoclaving. 3. RNAscope Multiplex Fluorescent V2 kit containing RNAscopeHydrogen Peroxide, Protease Plus, Protease III, Protease IV, 10 Target retrieval, AMP1, AMP2, AMP3, HRP-C1, HRP-C2, HRP-C3, HRP Blocker, DAPI, 50 wash buffer, and Multiplex TSA Buffer (see Note 3) (Bio-Techne). 4. Pre-warm RNAscope 50 wash buffer to 40  C for 10–20 min. Prepare 1 wash buffer by adding 60 mL pre-warmed RNAscope 50 wash buffer to 2.94 L distilled water. 5. RNAscope target probes in channel C1, channel C2, and channel C3 (Bio-Techne). 6. Species-specific RNAscope 3-plex positive control probes and RNAscope® 3-plex negative control probe (Bio-Techne) (see Note 4). 7. TSA plus Fluorescein kit, TSA plus cyanine 3 kit, and TSA plus cyanine 5 kit from Akoya Biosciences (see Note 5). 8. Multichannel pipette.

2.3 High-Content Imaging and Analysis

3 3.1

1. Operetta CLS imaging system (Perkin Elmer, UK) with Harmony 4.9 PhenoLOGIC software.

Methods Cell Culture

1. Dispense 10 μL/well of Cell-Tak buffer solution directly into the wells of the CellCarrier 96-well plate and add 40 μL/well of bicarbonate buffer. Incubate at room temperature for 20–40 min. Rinse twice with distilled water and dry the plate completely inside the safety cabinet before seeding the cells. 2. Grow cells up to 80% confluency in a T175 flask. 3. Wash the cells once with 7 mL of sterile PBS. Add 3 mL of 0.05% trypsin-EDTA to cells, and incubate at 37  C incubator for a few minutes, until the cell dislodges. 4. Neutralize the trypsin by adding 7 mL of media with FBS. Suspend them carefully, without bubbles, and count the cells in a hemocytometer.

114

Julia Jones et al.

5. Seed 8000 cells per well in a CellCarrier-96-well plate. Cell number depends on the cell line. 8000 cells will become confluent in 2–3 days for the cell lines with doubling time of 24 h. 6. Incubate the cells in the incubator for 48 h. Any desired treatments can be performed the next day. 7. After 48 h, fix the cells with 150 μL/well 10% neutral-buffered formalin for 30 min. 8. Wash the cells with 200 μL/well of PBS for 3 min, and repeat. 3.2

RNAscope Assay

1. Pre-warm RNAscope target probes, and control probes to 40  C for 10–20 min. 2. Switch on HybEZ oven, insert the humidity control tray with a piece of moistened blotting paper, and warm to 40  C. 3. Remove PBS from wells and perform protease digestion of the cells by adding 200 μL/well Protease III reagent diluted 1:15 in PBS. Incubate at room temperature for 10 min (see Note 6). 4. Remove protease reagents and rinse cells with PBS for 2 min; repeat twice. 5. Dilute the pre-warmed RNAscope target probes C1:C2:C3 in a ratio of 50:1:1 (see Note 7). 6. Remove PBS from each well and carefully apply prepared probes to corresponding wells according to the experimental plan. Put on the plate lid and place the plate inside the HybEZ humidity control tray. Return the humidity control tray to the HybEZ oven and incubate at 40  C for 3 h. 7. Remove the plate from the humidity control tray in the HybEZ oven and then carefully remove the probes (see Note 8). 8. Wash the cells with 200 μL/well of 1 wash buffer for 2 min; repeat twice. 9. At this point the protocol may be stopped overnight by storing the cells in 200 μL/well of 5 SSC (prepared from 20 SSC) at room temperature. 10. Remove the AMP1, AMP2, AMP3, HRP-C1, HRP-C2, HRP-C3, and HRP blocker components of the RNAscopeMultiplex Fluorescent v2 kit from the refrigerator and warm to room temperature. Ensure that HybEZ oven and humidity control tray are warmed to 40  C. 11. Remove SSC from wells and wash cells with 200 μL/well 1 wash buffer for 2 min; repeat. 12. Remove excess 1 wash buffer from wells and apply three drops of 90 μL AMP1 reagent to each well (see Note 9). Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min.

Multiplex lncRNA Detection in Cultured Cells

115

13. Take the plate from the humidity control tray and remove AMP1 reagent (see Note 8) from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 14. Remove the 1 wash buffer from each well and apply three drops of 90 μL AMP2 reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 15. Take the plate from the humidity control tray and remove AMP2 reagent (see Note 8) from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 16. Remove the 1 wash buffer from each well and apply three drops of 90 μL AMP3 reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 15 min. 17. Take the plate from the humidity control tray and remove AMP3 reagent (see Note 8) from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 18. Dilute TSA plus fluorophores to 1:1500 using the TSA dilution buffer (see Note 10). 19. Remove the 1 wash buffer from each well and apply three drops of 90 μL HRP-C1 reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 15 min. 20. Take the plate from the humidity control tray and remove HRP-C1 reagent (see Note 8) from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 21. Remove the 1 wash buffer from each well and apply 100 μL TSA plus fluorescein (or user’s choice of TSA plus/Opal fluorophore) reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 22. Take the plate from the humidity control tray and remove TSA reagent from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 23. Remove the 1 wash buffer from each well and apply three drops of 90 μL HRP blocker reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 24. Take the plate from the humidity control tray and remove HRP blocker reagent from each well. Wash the cells with 200 μL/ well 1 wash buffer for 2 min; repeat twice. 25. Remove the 1 wash buffer from each well and apply three drops of 90 μL HRP-C2 reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 15 min.

116

Julia Jones et al.

26. Take the plate from the humidity control tray and remove HRP-C2 reagent (see Note 8) from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 27. Remove the 1 wash buffer from each well and apply 100 μL TSA plus cyanine 3 (or user’s choice of TSA plus/Opal fluorophore) reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 28. Take the plate from the humidity control tray and remove TSA reagent from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 29. Remove the 1 wash buffer from each well and apply three drops of 90 μL HRP blocker reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 30. Take the plate from the humidity control tray and remove HRP blocker reagent from each well. Wash the cells with 200 μL/ well 1 wash buffer for 2 min; repeat twice. 31. Remove the 1 wash buffer from each well and apply three drops of 90 μL HRP-C3 reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 15 min. 32. Take the plate from the humidity control tray and remove HRP-C3 reagent (see Note 8) from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 33. Remove the 1 wash buffer from each well and apply 100 μL TSA plus cyanine 5 (or user’s choice of TSA plus/Opal fluorophore) reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 34. Take the plate from the humidity control tray and remove TSA reagent from each well. Wash the cells with 200 μL/well 1 wash buffer for 2 min; repeat twice. 35. Remove the 1 wash buffer from each well and apply three drops of 90 μL HRP blocker reagent to each well. Place the plate into the humidity control tray within the HybEZ oven and incubate at 40  C for 30 min. 36. Take the plate from the humidity control tray and remove HRP blocker reagent from each well. Wash the cells with 200 μL/ well 1 wash buffer for 2 min; repeat twice (see Note 11). 37. Remove the 1 wash buffer from each well, apply three drops of 90 μL of DAPI to each well, and incubate for 30 s. 38. Replace DAPI with two changes of PBS. 39. Place a black sticker lid over the plate and store in the dark at 4  C for up to 48 h before imaging.

Multiplex lncRNA Detection in Cultured Cells

3.3 High-Content Imaging and Analysis

117

1. Image cells with the Operetta CLS imaging system with Harmony 4.9 PhenoLOGIC software (Perkin Elmer, UK). If cell numbers are low, the plate should be scanned using the PreciScan mode (see Note 12). 2. The PreciScan mode requires three steps; prescan—scanning step at low resolution, prescan analysis—the analysis step to define the parameters for the selection of the regions for the subsequent higher resolution scan, and rescan—the final higher resolution scan. 3. Define the prescan settings: (a) Use the “Set Up” tab to insert the desired parameters for the prescan (see Note 13). Use the test option to run a test scan to check the settings and adjust if necessary: l

Plate type: 96 Perkin Elmer CellCarrier Ultra

l

Objective: 5 Air, NA 0.16

l

Optical mode: Non-confocal

l

Binning: 2

(b) Apply channel selection as shown in Table 1: l

Define layout: Plate: Select all wells – Well: Select all nine fields so that the entire well is covered. – Stack: no stack.

4. Define prescan analysis settings: Once the test scan has been acquired, select the “image analysis” tab to set up the prescan analysis for the determination of the areas to rescan with the 40 water objective: Online jobs: Apply prescan analysis settings as shown in Table 2. 5. Once the prescan analysis settings have been optimized, save the analysis protocol and apply this in the “Online Jobs” block in the prescan protocol. Save the whole prescan protocol. 6. Define rescan settings: Use the “set up” tab again to insert the desired parameters for the rescan. Use the test option to run a test scan to check the settings and adjust if necessary.

Table 1 Channel selection settings for the prescan Channel

Excitation wavelength

Emission wavelength

LED power

Exposure time

DAPI

355–385

430–500

30%

20 ms

118

Julia Jones et al.

Table 2 Settings for the prescan analysis protocol Input image

Input

Method

Output

Find image region

Channel: DAPI ROI: None

Method: Absolute threshold Lowest intensity: >310 Highest intensity: 226,000 μm2 Fill holes

Output population: Image region Output region: Image region

Find nuclei

Output population: Method: B Channel: DAPI Nuclei ROI: Image region Common threshold: 0.4 ROI Region: Area: >30 μm2 Image region Splitting coefficient: 7 Individual threshold: 0.3 Contrast: >0.07

Calculate intensity properties

Channel: DAPI Population: Nuclei Region: Nucleus

Method: Standard Mean

Property prefix: Intensity nucleus DAPI

Calculate morphology properties

Population: Nuclei Region: Nucleus

Method: Standard Area

Property prefix: Morphology properties Nucleus

Select population

Population: Nuclei

Method: Filter by property Nucleus area [μm2]: 60 μm Splitting coefficient: 4.5 Individual threshold: 0.4 Contrast: >0.1

Output population: Nuclei

Select Population: Nuclei population

Method: Remove border objects Region: Nucleus

Output population: Nuclei whole

See Note 18

Find cytoplasm

Channel: Cy5 Nuclei: Nuclei whole

Method: A Individual threshold: 0.02

Find spots

Channel: FITC ROI: Nuclei whole ROI region: Nucleus

Method: A Relative spot intensity: >0.09 Splitting sensitivity: 0.96 Calculate spot properties

Output population: FITC spot nucleus

See Note 19

Find spots [2]

Channel: FITC ROI: Nuclei whole ROI region: Cytoplasm

Method: A Relative spot intensity: >0.09 Splitting sensitivity: 0.96 Calculate spot properties

Output population: FITC spot cytoplasm

Find spots [3]

Channel: Cy3 ROI: Nuclei whole ROI region: Nucleus

Output population: Method: A Cy3 spot nucleus Relative spot intensity: >0.1 Splitting sensitivity: 0.9 Calculate spot properties (continued)

Multiplex lncRNA Detection in Cultured Cells

121

Table 4 (continued) Analysis block

Input

Method

Output

Find spots [4]

Channel: Cy3 ROI: Nuclei whole ROI region: Cytoplasm

Output population: Method: A Cy3 spot cytoplasm Relative spot intensity: >0.1 Splitting sensitivity: 0.9 Calculate spot properties

Find spots [5]

Channel: Cy5 ROI: Nuclei whole ROI region: Nucleus

Output population: Method: A Cy5 spot nucleus Relative spot intensity: >0.12 Splitting sensitivity: 0.9 Calculate spot properties

Find spots [6]

Channel: Cy5 ROI: Nuclei whole ROI region: Cytoplasm

Output population: Method: A Cy5 spot cytoplasm Relative spot intensity: >0.12 Splitting sensitivity: 0.9 Calculate spot properties

Define results

Method: List of outputs Population: Nuclei whole Number of objects Apply to all: Mean + StdDev Method: Object results Population: Nuclei whole: ALL

Note

See Note 20

cells grow nicely on these plates and do not detach from the plastic bottom compared to glass. The ultralow skirted (0.2 mm) plates provide better access to the objective. 2. Cell-Tak coating is preferred due to the substantial number of washes, but this will depend on the cell line used. We recommend testing the concentration of Cell-Tak coating for each cell line before your experiment. 3. The RNAscope®Multiplex fluorescent v2 kit is recommended for formalin-fixed paraffin-embedded tissues, low-abundance targets in fixed frozen or fresh frozen tissues, and cultured cells. Each kit provides sufficient reagent for 50 wells.

122

Julia Jones et al.

4. The inclusion of the positive controls is important for the assessment of optimal fixation and protease steps. Negative controls are essential for determining the level of nonspecific signal obtained. 5. While we used TSA plus reagents from Akoya Biosciences, alternative Opal dyes (also from Akoya Biosciences) are now recommended by Bio-Techne. These Opal dyes have narrower emission spectra to reduce spectral overlap and increase the number of color channels that can be detected. 6. We recommend optimizing the concentration of the protease III reagent for each cell type/preparation. In a pilot study, we tested 1:5, 1:10, 1:15, and 1:30 dilutions of the protease III reagent with an incubation time of 10 min at room temperature and found 1:15 to aid the penetration of reagents into the cells while maintaining the cellular morphology. 7. Briefly spin C2 and C3 probes and take care not to mix probes of the same channel. 120 μL probe/well is sufficient volume. Unused, diluted probes can be stored at 4  C for 6 months. 8. The probes, AMP1, 2, and 3 and HRP-C1, -C2, and -C3 reagents all contain formamide and should therefore be preserved for appropriate disposal. 9. The reagents within the RNAscopeMultiplex Fluorescent v2 kit are provided in dropper bottles to remove the necessity for pipetting. However, in the case of multi-well plates, it is more efficient to dispense the required volume into reagent reservoirs and use a multichannel pipette to apply to each well. 10. The TSA plus reagents from Akoya Biosciences are reconstituted with DMSO, according to the manufacturer’s instructions and are then diluted using TSA dilution buffer. The dilution should be optimized, with a recommended starting range of 1:750–1:3000. 11. While we have used this kit for detection of three targets simultaneously, it is possible to purchase RNAscopeChannel C4 target probes and an RNAscope4-Plex ancillary kit from Bio-Techne which allows detection of a fourth target in the same sample. This is simply a kit containing HRP-C4, HRP blocker, and TSA dilution buffer. This also requires a fourth TSA plus or Opal dye and the associated imaging capability. 12. The PreciScan is an automated dual-scan mode employing a low-resolution prescan, a preset analysis step to locate the objects of interest, followed by a rescan with a higher magnification objective. The prescan is a non-confocal, 5 magnification, single-channel, single-plane scan of the whole well area (Fig. 1). This is followed by an analysis step to locate, in this case, the nuclei. The layout for the rescan is then automatically

Multiplex lncRNA Detection in Cultured Cells

123

Fig. 1 Prescan images showing: (a) Typical cell distribution across a whole well scanned using the DAPI channel at 5 magnification, (b) one field scanned at 5 showing the selection of 5 fields for rescanning at 40 magnification and (c) graphic of the whole well showing the distribution of all 45 fields for the rescan (5 fields per 5 prescan field)

determined by the software to scan only fields that contain nuclei and therefore dramatically reduces scan time and file size. A maximum number of fields per well can be set. 13. Only one channel, DAPI, is required for the prescan as it is for nuclei detection only. This reduces unnecessary light exposure for the cells. 14. The maximum number of fields here can be varied depending on the requirements for the final cell count in the analysis. This field number is the number per each 5 field and therefore will give us a total of 45 fields per well. It is possible to scan many more fields; however this will result in a much longer scan time and much larger file size. For example, a scan of 18 wells at 40 magnification, 4 channels, and 45 fields per well results in a file size of 94 GB. 15. The results here are not required for the rescan, but are useful as they give the total cell count per well. This is not available from the final analysis as that gives us only the total cell number scanned in the selected fields at the higher resolution, which is not the entire well. 16. Exposure times here should be varied according to the signal intensity with positive controls. 17. It is not essential to generate a global image at this stage as it is not used in the analysis. However, it can be useful to have an overview of all the fields scanned. 18. It is important that only whole cells are included in the analysis; therefore cells touching the edge of the field image (border objects) are removed beforehand to avoid erroneous results.

124

Julia Jones et al.

Fig. 2 Images and analysis examples from the 40 rescan showing: (a) Four-color image, (b) nuclear lncRNA segmentation, (c) nuclear mRNA-1 segmentation, (d) cytoplasmic mRNA-2 segmentation, (e) percentage of RNA colocalized in either nucleus or cytoplasm shown as mean values from two technical replicates for a lncRNA and two different mRNAs

19. Sequential spot detection steps are added to allow spots in the three channels and the different cell regions to be counted and expressed as a number per cell (Fig. 2). 20. Results output is defined to give the results for each selected measured parameter. Object results for a population, in this case for whole nuclei, gives the single-cell results for each parameter measured per each whole nucleus.

Acknowledgments We would like to thank Jason S. Carroll, Susan G Komen leadership grant, Light Microscopy Core Facility, Histopathology and ISH Core Facility, and Morgane Rouault (ACD/Bio-Techne). References 1. Quinn JJ, Chang HY (2016) Unique features of long non-coding RNA biogenesis and function. Nat Rev Genet 17:47–62

2. Fatica A, Bozzoni I (2014) Long non-coding RNAs: new players in cell differentiation and development. Nat Rev Genet 15:7–21

Multiplex lncRNA Detection in Cultured Cells 3. Sun Q, Hao Q, Prasanth KV (2018) Nuclear Long noncoding RNAs: key regulators of gene expression. Trends Genet 34:142–157 4. Sa´nchez Y, Huarte M (2013) Long non-coding RNAs: challenges for diagnosis and therapies. Nucleic Acid Ther 23:15–20 5. Lam MTY, Li W, Rosenfeld MG et al (2014) Enhancer RNAs and regulated transcriptional programs. Trends Biochem Sci 39:170–182

125

6. Kashi K, Henderson L, Bonetti A et al (2016) Discovery and functional analysis of lncRNAs: methodologies to investigate an uncharacterized transcriptome. Biochim Biophys Acta 1859:3–15 7. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14:22–29

Chapter 8 Multiplexed Quantitative In Situ Hybridization for Mammalian or Bacterial Cells in Suspension: qHCR Flow Cytometry (v3.0) Maayan Schwarzkopf, Harry M. T. Choi, and Niles A. Pierce Abstract In situ hybridization based on the mechanism of hybridization chain reaction (HCR) enables highthroughput expression profiling of mammalian or bacterial cells via flow cytometry. Third-generation in situ HCR (v3.0) provides automatic background suppression throughout the protocol, dramatically enhancing performance and ease of use. In situ HCR v3.0 supports analog mRNA relative quantitation via qHCR flow cytometry. Here, we provide protocols for multiplexed qHCR flow cytometry for mammalian or bacterial cells in suspension. Key words Accuracy, Autofluorescence, Automatic background suppression, Bacterial cells, Fluorescence in situ hybridization (FISH), High-throughput expression profiling, Hybridization chain reaction (HCR), In situ HCR v3.0, Mammalian cells, mRNA flow cytometry, Multiplexed, Precision, qHCR flow cytometry, Quantitative, Signal amplification

1

Introduction In situ HCR v3.0 offers a unique combination of multiplexing, quantitation, sensitivity, versatility, and robustness for mRNA analysis in diverse sample types (Table 1) [1]. The same two-stage enzyme-free protocol is used independent of the number of target mRNAs (Fig. 1). In situ HCR v3.0 uses probes and amplifiers that combine to provide automatic background suppression throughout the protocol, ensuring that reagents will not generate amplified background even if they bind nonspecifically within the sample. Automatic background suppression dramatically enhances performance (signal-to-background and quantitative precision) and ease of use (no probe set optimization for new targets and organisms). Figures 2 and 3 illustrate mRNA flow cytometry for transgenic and endogenous targets in mammalian and bacterial cells, achieving high signal-to-background without probe set optimization.

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_8, © Springer Science+Business Media, LLC, part of Springer Nature 2020

127

128

Maayan Schwarzkopf et al.

Table 1 mRNA analysis using in situ HCR v3.0 Property

Details

Simple

Two-stage enzyme-free protocol independent of number of targets

Amplified

Boosts signal above autofluorescence

Multiplexed

Simultaneous one-stage signal amplification for up to five target mRNAs

Quantitative

Signal scales linearly with target abundance

Penetrating

Whole-mount vertebrate embryos

Resolved

Subcellular or single-molecule resolution as desired

Sensitive

Single molecules detected in thick autofluorescent samples

Portable

Suitable for use with diverse targets in diverse organisms

Versatile

Multiple assay formats: imaging, flow cytometry, blotting

Compatible

Combine with tissue clearing as desired

Robust

Automatic background suppression throughout the protocol

Adapted with permission from Development [1]

With qHCR flow cytometry, the signal is analog in the form of single-cell fluorescence intensities that scale approximately linearly with the number of target molecules per mammalian or bacterial cell [1]. The instrument integrates both signal and background over the volume of the cell. To overcome autofluorescent background in the cell, we recommend maximizing the number of probe pairs per target mRNA (e.g., 20+ or 30+ probe pairs) as permitted by the length of the target mRNA (signal-to-background and precision increase with probe set size) [2]. Figure 4 illustrates high accuracy and precision using qHCR flow cytometry for analog mRNA relative quantitation for mammalian and bacterial cells [1]. These in situ HCR v3.0 protocols for qHCR flow cytometry of mammalian and bacterial cells in suspension are adapted with permission from Development [1].

2

Materials Make sure that your environment and reagents are DNase and RNase free. Lab surfaces and pipettes should be cleaned using RNaseZAP. Prepare all solutions using ultrapure (resistivity of 18 MΩ-cm at 25  C) or molecular grade water.

qHCR Flow Cytometry

a

129

In situ HCR v3.0 Protocol Summary Detection stage

•••

mRNA target

H1

DNA probe set

Hybridize probe set and wash

•••

HCR initiator I1 split between pair of probes HCR initiator I1

H2

Tethered fluorescent amplification polymers

In situ HCR and wash •••

HCR initiator I1

mRNA target

•••

b

•••

Amplification stage Metastable DNA HCR hairpins

Multiplexing Timeline Amplification stage

Detection stage

•••

•••

•••

Add all HCR amplifiers

Add all HCR probe sets

Wash

Incubation 0

16 18 Experimental time (h)

Incubation

Wash 34 36

Fig. 1 qHCR flow cytometry (v3.0): multiplexed quantitative mRNA flow cytometry with automatic background suppression throughout the protocol. (a) Two-stage enzyme-free protocol. Detection stage: probe sets hybridize to mRNA targets and unused probes are washed from the sample. Amplification stage: specifically bound probe pairs trigger self-assembly of a tethered fluorescent amplification polymer and unused hairpins are washed from the sample. Probes and amplifiers combine to provide automatic background suppression throughout the protocol, ensuring that reagents will not generate amplified background even if they bind nonspecifically in the sample. During the detection stage, individual probes that bind nonspecifically do not trigger HCR since each probe carries only a fraction of initiator I1. During the amplification stage, individual hairpins that bind nonspecifically do not trigger HCR since they are kinetically trapped. (b) Multiplexing timeline. The same two-stage protocol is used independent of the number of target mRNAs. Adapted with permission from Development [1] 2.1 Sample Preparation 2.1.1 Fixation and Permeabilization (All Sample Types) 2.1.2 Bacterial Samples

1. RNaseZAP. 2. Phosphate-buffered saline (PBS; 10): 1.37 M NaCl, 27 mM KCl, 80 mM Na2HPO4, and 20 mM KH2PO4.

1. LB media: Weigh 5 g of Novagen Broth Miller powder and place in a PYREX wide-mouth storage bottle (see Note 1). Add 200 mL of ultrapure water. Autoclave at 121  C for 20 min. 2. 100% Methanol (MeOH).

Maayan Schwarzkopf et al.

Cell count

a 2000

WT cells

1000

0 1 10

b

HEK cells d2eGFP mRNA

Cell count

130

GFP+ cells

10

2

10

3

10

4

10

800

E. coli cells eGFP mRNA WT cells

GFP+ cells

400

5

0 0 10

HCR fluorescence intensity

10

1

10

2

10

3

10

4

10

5

HCR fluorescence intensity

Fig. 2 qHCR flow cytometry with high signal-to-background for transgenic target mRNAs in mammalian and bacterial cells. Raw single-cell HCR fluorescence intensity distributions. (a) HEK cells. Target mRNA d2eGFP, probe set with 12 probe pairs (no probe set optimization), amplifier B3-Alexa594, signal-to-background 101.8  0.5 (mean  standard error, N ¼ 55,000 HEK cells). (b) E. coli cells. Target mRNA eGFP, probe set with 12 probe pairs (no probe set optimization), amplifier B3-Alexa594, signal-to-background 26  5 (mean  standard error, N ¼ 18,000 E. coli cells). Adapted with permission from Development [1]

3. Fixation solution: 4% Formaldehyde in PBS. In a 15 mL conical tube combine 6.5 mL ultrapure water, 2.5 mL of 16% formaldehyde (w/v), methanol free (see Note 2), and 1 mL 10 PBS pH 7.4 (see Note 3). Prepare fresh before use. 2.1.3 Mammalian Samples

1. 5% CO2 incubator at 37  C. 2. Tissue culture plate. 3. Dulbecco’s PBS (DPBS) without calcium chloride or magnesium chloride (see Note 3). 4. Trypsin-EDTA (0.25%). 5. Cell growth media. 6. Fixation solution: 4% Formaldehyde in PBST. In a 15 mL conical tube combine 6.4 mL ultrapure water, 2.5 mL of 16% formaldehyde (w/v), methanol free (see Note 2), 1 mL 10 PBS pH 7.4 (see Note 3), and 100 μL 10% Tween20. Prepare fresh before use. 7. PBST: 1 PBS, 0.1% Tween 20. 8. Trypan blue stain 0.4% (see Note 4). 9. Hemocytometer slide (see Note 4). 10. Ice-cold 70% ethanol.

2.2

HCR Reagents

1. HCR probes, amplifiers, and buffers for cells in suspension (probe hybridization buffer, probe wash buffer, amplification buffer) are available from Molecular Instruments (www. molecularinstruments.com) (see Note 5). 2. PBST: 1 PBS, 0.1% Tween 20. 3. 20 Sodium chloride sodium citrate (SSC): 3 M Sodium chloride and 0.3 M sodium citrate. 4. 5 SSCT: 5 SSC, 0.1% Tween 20.

qHCR Flow Cytometry

a

131

HEK cells

Ch2 fluorescence intensity

106

GAPDH + PGK1 104

2

1000

10

PGK1

GAPDH

Autofluoresence

Counts

Counts 10

0

102

104

106

1000 Ch1 fluorescence intensity

b

E. coli cells

6

10 icd + fusA

Ch2 fluorescence intensity

icd 4

10

2

10

0

fusA

1200

10 Autofluoresence -2

Counts

Counts 10

0

10

2

10

4

10

6

10

1200 Ch1 fluorescence intensity

Fig. 3 Multiplexed qHCR flow cytometry for endogenous target mRNAs in mammalian and bacterial cells. Raw single-cell fluorescence intensity scatterplots and distributions. (a) HEK cells: GAPDH vs. PGK1 expression. Ch1: target mRNA GAPDH, probe set with 10 probe pairs, amplifier B4-Alexa488, signal-to-background 29.6  0.1 (mean  standard error, N ¼ 18,000 cells). Ch2: target mRNA PGK1, probe set with 18 probe pairs, amplifier B2-Alexa594, signal-to-background 6.77  0.05. (b) E. coli cells: fusA vs. icd expression. Ch1: target mRNA fusA, probe set 18 probe pairs, amplifier B3-Alexa488, 20.1  0.2 (mean  standard error, N ¼ 35,000 cells). Ch2: target mRNA icd, probe set with 20 probe pairs, amplifier B1-Alexa594, signal-to-background 38.9  0.5.

132

Maayan Schwarzkopf et al.

1. Thermocycler.

2.3 Additional Equipment

2. Centrifuge. 3. Hybridization oven, incubator, or heat block at 37  C. 4. 35 or 40 μm mesh or cell strainer. 5. Flow cytometer.

3

Methods The following protocol has been optimized for E. coli and for human cell lines. Additional modifications may be needed for other cells. Throughout the protocol, bacterial cells are centrifuged at 4000  g and solutions are removed by gentle pipetting (do not aspirate). For mammalian cells, centrifugation is done at 180  g and solutions may be removed by aspiration or gentle pipetting.

3.1 Sample Preparation

1. Grow E. coli cells (from frozen glycerol stock or from a plate) in 2–3 mL of LB overnight in a 37  C shaker at 250 rpm.

3.1.1 Bacterial Cells

2. Dilute to make a 5 mL liquid culture with OD600 ¼ 0.05. 3. Incubate in a 37  C shaker until OD600  0.5 (exponential phase). 4. Aliquot 1 mL of cells and centrifuge for 10 min. 5. Remove supernatant and resuspend cells in 750 μL of 1 PBS. 6. Add 250 μL of fixation solution and incubate overnight at 4  C. Caution: Use fixation solution with extreme care as it contains formaldehyde, a hazardous material. 7. Centrifuge for 10 min and remove supernatant. 8. Resuspend cells in 150 μL of 1 PBS. 9. Add 850 μL of 100% methanol and store cells at 20  C before use (see Notes 6 and 7). 1. Aspirate growth media from culture plate and wash cells with DPBS.

3.1.2 Mammalian Cells

2. Add 3 mL of trypsin per 10 cm plate and incubate in a 5% CO2 incubator at 37  C for 5 min. 3. Quench trypsin by adding 3 mL of growth media.  Fig. 3 (continued) For signal-to-background calculations, the signal estimate SIG was calculated using the background approximation BACK  NSA + AF (experiment of type 2 in Table 3) for both mammalian and bacterial cells. Adapted with permission from Development [1]

qHCR Flow Cytometry •• •







••

••

••

a

133

•••

2-channel redundant detection

b

Ch2 fluoresence intensity

c

14 ×10 HEK cells

Ch2 normalized signal

1 HEK cells GAPDH mRNA r = 0.99

GAPDH mRNA

7

4 0 3 6×10 0 Ch1 fluorescence intensity

3 ×10

0

1 Ch1 normalized signal

5

1

E. coli cells fusA mRNA

1.5

00

0

Ch2 normalized signal

Ch2 fluorescence intensity

4

3×10 1.5 Ch1 fluoresence intensity

5

E. coli cells fusA mRNA r = 0.98

0 0

1 Ch1 normalized signal

Fig. 4 mRNA relative quantitation with high accuracy and precision for mammalian and bacterial cells via qHCR flow cytometry. (a) Two-channel redundant detection of a target mRNA using two probe sets, each initiating an orthogonal spectrally distinct HCR amplifier. (b, c) Left: Raw single-cell fluorescence intensity scatterplots (dashed lines represent BOT and TOP values used for normalization; outliers excluded from normalized scatterplots marked with ellipses). Right: Normalized single-cell fluorescence intensity scatterplots representing estimated normalized signal (Pearson correlation coefficient, r). (b) HEK cells: Target mRNA GAPDH. Ch1: probe set with 10 probe pairs, amplifier B5-Alexa488, signal-to-background 12.9  0.1 (mean  standard error, N ¼ 20,000 cells). Ch2: probe set with 10 probe pairs, amplifier B4-Alexa594, signal-to-background 14.4  0.1. (c) E. coli cells: fusA target mRNA. Ch1: probe set with 18 probe pairs, amplifier B3-Alexa488, signal-to-background 99  3 (mean  standard error, N ¼ 3400 cells). Ch2: probe set with 18 probe pairs, amplifier B2-Alexa594, signal-to-background 135  7. For signal-to-background calculations, the signal estimate SIG was calculated using the background approximation BACK  NSA + AF (experiment of type 2 in Table 3) for both mammalian and bacterial cells. Adapted with permission from Development [1]

134

Maayan Schwarzkopf et al.

4. Transfer cells to a 15 mL conical tube and centrifuge for 5 min. 5. If counting cells follow steps 6 and 7; otherwise skip to step 8. 6. Aspirate supernatant and resuspend cells in growth media or DPBS (see Note 8). 7. Mix cells with trypan blue in a 1:1 ratio and count cells (see Note 9). 8. Aspirate supernatant and resuspend cells in fixation solution to reach approximately 106 cells/mL. Caution: Use fixation solution with extreme care as it contains formaldehyde, a hazardous material. 9. Fix cells for 1 h at room temperature. 10. Centrifuge for 5 min and remove supernatant. 11. Wash cells four times with PBST (use the same volume as fixative solution). Pellet your cells by centrifugation between washes. 12. Resuspend cells in ice-cold 70% ethanol (use the same volume as fixative and wash solutions). 13. Store cells at 4  C overnight or 20  C until use (see Note 6). 3.2

Detection Stage

1. Transfer desired volume of cells into a 1.5 mL Eppendorf tube (see Notes 10 and 11). 2. Centrifuge for 5 min and remove supernatant. 3. Wash cells with 500 μL of 1 PBST and remove the solution by centrifugation (see Note 12). 4. Resuspend the pellet with 400 μL of probe hybridization buffer and pre-hybridize for 1 h at 37  C (see Notes 13 and 14). Caution: Probe hybridization buffer contains formamide, a hazardous material. 5. Prepare probe solution by adding 2 pmol of each probe mixture (e.g., 2 μL of 1 μM stock) to 100 μL probe hybridization buffer at 37  C. 6. Add the probe mixture directly to the sample to reach a final probe concentration of 4 nM. 7. Incubate the sample overnight at 37  C. 8. Centrifuge for 5 min and remove the probe solution (see Note 15). 9. Resuspend the cell pellet with 500 μL of probe wash buffer. Caution: Probe wash buffer contains formamide, a hazardous material. 10. Incubate for 5 min at 37  C and remove the wash solution by centrifugation for 5 min (see Note 16).

qHCR Flow Cytometry

135

11. Repeat steps 9 and 10 for two additional times but with 10-min incubation (see Note 17). 12. Mammalian cells only: Resuspend the cell pellet with 500 μL of 5 SSCT. Incubate for 5 min at room temperature. Centrifuge for 5 min and remove solution. 13. Proceed to amplification stage. 3.3 Amplification Stage

1. Resuspend the cell pellet with 150 μL of amplification buffer and pre-amplify for 30 min at room temperature (see Note 18). 2. Separately prepare 15 pmol of hairpin h1 and 15 pmol of hairpin h2 by snap cooling 5 μL of 3 μM stock. Heat individual amplifier hairpins in a thermocycler to 95  C for 90 s, remove from thermocycler, and cool to room temperature in a dark drawer for 30 min (see Note 19). 3. Prepare hairpin mixture by adding all snap-cooled hairpins to 100 μL of amplification buffer at room temperature. 4. Add the hairpin mixture directly to the sample to reach a final hairpin concentration of 60 nM. 5. Incubate the sample overnight (>12 h) in the dark at room temperature (see Note 20). 6. Centrifuge for 5 min and remove the amplification solution (see Note 21). 7. Resuspend the cell pellet with 500 μL of 5 SSCT. 8. Incubate for 5 min at room temperature and remove the wash solution by centrifugation for 5 min (see Note 22). 9. Repeat steps 7 and 8 for two additional times but with 10-min incubation (see Notes 22 and 23). 10. Resuspend the cells in desired volume and buffer. 11. Store the sample at 4  C protected from light before flow cytometry.

3.4

Flow Cytometry

Prior to flow cytometry, cells should be filtered through a 35 μm or a 40 μm mesh (see Note 24). For HEK cells, apply two gates to the data: a first gate of forward scatter area (FSC-A) versus side scatter area (SSC-A) to select cells, and a second gate of FSC-A versus forward scatter height (FSC-H) to select single cells. Only cells satisfying both gates should be used for the analysis. For E. coli cells, apply one gate of FSC-A versus SSC-A to select cells (see Note 25).

3.5

Data Analysis

The quantitative analysis framework presented here was developed over a series of publications [1–5]. For convenience, here we provide a self-contained description of the details relevant to the present protocols.

136

Maayan Schwarzkopf et al.

The total fluorescence within a cell is a combination of signal and background. Fluorescent background (BACK) arises from three sources in each channel:

3.5.1 Raw Single-Cell Intensities

l

Autofluorescence (AF): fluorescence inherent to the sample

l

Nonspecific detection (NSD): probes that bind nonspecifically in the sample and subsequently trigger HCR amplification

l

Nonspecific amplification (NSA): HCR hairpins that bind nonspecifically in the sample

Fluorescent signal (SIG) in each channel corresponds to: l

Signal (SIG): probes that hybridize specifically to the target mRNA and subsequently trigger HCR amplification

For cell n, we denote the background: ¼ X NSD þ X NSA þ X AF X BACK n n n n

ð1Þ

X SIG n

ð2Þ

the signal:

and the total fluorescence (SIG + BACK): BACK X SIGþBACK ¼ X SIG n n þ Xn

ð3Þ

For a transgenic target mRNA, signal (SIG) and background (BACK) are characterized based on experiments of types 1a and 1b (Table 2) with SIG + BACK measured in transgenic cells containing the target and BACK measured in wild-type (WT) cells lacking the target. Performance across a population of cells is BACK characterized by calculating the sample means (X and

3.5.2 Measurement of Signal, Background, and Signal-to-Background for Transgenic Targets

SIGþBACK

X ) and standard errors (S X BACK and S X SIGþBACK ). The mean signal is then estimated as X

SIG

¼X

SIGþBACK

X

BACK

ð4Þ

with the standard error estimated via uncertainty propagation as

Table 2 Experiment types for qHCR flow cytometry of a transgenic target mRNA Experiment type

Quantity

Probes

Hairpins

Cell type

1a

SIG + NSD + NSA + AF¼SIG + BACK

x

x

Transgenic

1b

NSD + NSA + AF ¼ BACK

x

x

WT

2

NSA + AF

x

Transgenic

3

AF

Adapted with permission from Development [1]

Transgenic

qHCR Flow Cytometry

s X SIG

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2ffi  s X SIGþBACK þ s X BACK

137

ð5Þ

The signal-to-background ratio is estimated as X

SIG=BACK

¼X

SIG

=X

BACK

ð6Þ

with standard error estimated via uncertainty propagation as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi s X BACK 2 s X SIG 2 SIG=BACK SIG=BACK s X þ SIG BACK X X

ð7Þ

These upper bounds on estimated standard errors hold under the assumption that the correlation between SIG and BACK is nonnegative. If desired, additional control experiments that omit certain reagents can be used to characterize the individual components of background (AF, NSA, NSD). A type 2 experiment (no probes, NSAþAF hairpins only) yields X and a type 3 experiment (no probes, AF no hairpins) yields X (see Note 26). The background components can then be estimated via calculations analogous to (Eq. 4) and (Eq. 5). The estimated means are X

NSD

X

¼X

NSA

BACK

¼X

X

NSAþAF

NSAþAF

X

ð8Þ

AF

ð9Þ

with estimated standard errors: rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi S X NSD 

S X NSA 

S X BACK

2

2

þ S X NSAþAF

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi S X NSAþAF

2

þ S X AF

2

ð10Þ ð11Þ

These upper bounds on estimated standard errors hold under the assumption that the correlations are nonnegative for the different components of background. If a type 1 experiment demonstrates SIG BACK, as is typically the case using in situ HCR v3.0, then there is little motivation to perform type 2 and type 3 experiments to characterize the individual background components (AF, NSA, NSD) as these are all bounded above by BACK. 3.5.3 Measurement of Signal, Background, and Signal-to-Background for Endogenous Targets

For an endogenous target mRNA, signal and background are characterized using experiments of types 1a and 1b (Table 3) with SIG + BACK measured using a probe set that addresses the target in WT cells and BACK measured using a probe set that addresses a different transgenic target absent from WT cells. Use of a previously validated transgenic probe set to measure background in WT cells ensures that a low measured fluorescence value does not simply

138

Maayan Schwarzkopf et al.

Table 3 Experiment types for qHCR flow cytometry of an endogenous target mRNA Experiment type

Quantity

Probes

Hairpins

Cell type

1a

SIG + NSD + NSA + AF¼SIG + BACK

x

x

WT

1b

NSD + NSA + AF ¼ BACK

x(transgenic)

x

WT

2

NSA + AF

x

WT

3

AF

WT

Adapted with permission from Development [1]

indicate a dysfunctional probe set, but indeed represents low background generated by a probe set that is known to be functional if the target is present in the sample. If desired, additional control experiments of types 2 and 3 (Table 3) can be performed to characterize the components of background (AF, NSA, NSD) using the calculations of Subheading 3.5.2. In lieu of an experiment of type 1b, an experiment of type 2 can be used to obtain the partial background estimate BACK  AF + NSA. 3.5.4 Normalized Single-Cell Intensities for Analog mRNA Relative Quantitation with qHCR Flow Cytometry

To facilitate relative quantitation between cells, we estimate the normalized HCR signal of cell n as xn

X SIGþBACK  X BOT n X TOP  X BOT

ð12Þ

which translates and rescales the data so that the voxel intensities in each channel fall in the interval [0,1]. Here, X BOT X

BACK

ð13Þ

is the mean background across cells and X TOP max X SIGþBACK n n

ð14Þ

is the maximum total fluorescence for a cell. Figure 4 demonstrates two-channel redundant detection of target mRNAs in mammalian and bacterial cells. For the normalized signal in a redundant detection experiment (panels b and c), accuracy corresponds to linearity with zero intercept, and precision corresponds to scatter around the line [2].

4

Notes 1. Alternative bottles may be used. Make sure that the bottle is autoclavable and contains a cap.

qHCR Flow Cytometry

139

2. Formaldehyde is a hazardous material. Use appropriate protective equipment and dispose as hazardous waste. 3. Avoid using calcium chloride and magnesium chloride in PBS as this leads to increased autofluorescence in the samples. 4. Trypan blue and a hemocytometer slide are needed only if counting cells using a hemocytometer. Cell count may also be obtained by alternative methods such as using a flow cytometer. 5. Select a different HCR amplifier (e.g., B1, B2, ...) for each target RNA that will be imaged in the same sample (for example, amplifier B1 for target 1, amplifier B2 for target 2, ...). Choose a different fluorophore label (e.g., Alexa647, Alexa594, ...) for each HCR amplifier that will be imaged in the same sample (e.g., B1-Alexa647, B2-Alexa594, ...). For each target mRNA, maximize the number of probe pairs as permitted by target length (e.g., 30+ probe pairs). Select your optimal dye for your lowest expression target; this will depend on the light source and filters available on the flow cytometer but in general higher wavelength dyes have less sample autofluorescence (e.g., use Alexa488 for a higher expression target, use Alexa647 for a lower expression target). 6. Samples can be stored at 20  C for up to a week. Longer time points should be validated experimentally. 7. Additional permeabilization (e.g., lysozyme) may be needed for gram-positive bacteria. 8. If a fraction of the cells will be reseeded, resuspend the cells in growth media. 9. We use an automated cell counter to determine cell counts. If manually counting cells, follow recommendations for cell counting of your hemocytometer slide supplier. 10. For bacteria transfer 150 μL. For mammalian cells use a volume corresponding to 0.5–1  106 cells. 11. See Tables 2 and 3 for experiment types to characterize signal, background, signal-to-background, and/or background components for transgenic and/or endogenous target mRNAs. 12. For mammalian cells repeat this step an additional time for a total of two washes. 13. Preheat probe hybridization and probe wash buffers to 37  C prior to use. Mix contents by swirling. 14. For mammalian cells a 30-min prehybridization step is sufficient. 15. For bacterial cells, add 1 mL of probe wash buffer (preheated to 37  C) to the sample prior to centrifugation to dilute the sample. Caution: Probe wash buffer contains formamide, a hazardous material.

140

Maayan Schwarzkopf et al.

16. For mammalian cells incubate at 37  C for 10 min. 17. For mammalian cells repeat three times for a total of four washes. 18. Equilibrate amplification buffer to room temperature before use. Mix contents by swirling. 19. HCR amplifier hairpins h1 and h2 are provided in hairpin storage buffer ready for snap cooling. h1 and h2 should be snap cooled in separate tubes. 20. We amplify the sample in a dark drawer. 21. For bacterial cells, add 1 mL of 5 SSCT to the sample to dilute the solution prior to centrifugation. 22. Mammalian cells do not require an incubation time and may be centrifuged immediately. 23. For mammalian cells, repeat washes five additional times for a total of six washes. 24. Samples should be handled gently; avoid fast runs and harsh sample mixing. Samples may also be mounted and imaged with microscopy. 25. See Figs. S1 and S2 of Ref. 1 for examples gating HEK and E. coli cells. 26. If a flow cytometer generates non-negligible fluorescence intensities in the absence of sample, this so-called instrument noise (NOISE) should be taken into consideration when calculating background and signal contributions, leading to four experiment types ((1a) SIG + BACK + NOISE, (1b) BACK + NOISE, (2) NSA + AF + NOISE, (3) AF + NOISE, (4) NOISE; cf. Tables 2 and 3).

Acknowledgments We thank A. Acharya and G. Artavanis for performing preliminary studies. Within the Beckman Institute at Caltech, we thank the following for assistance: C.R. Calvert and G.J. Shin (Molecular Technologies), R. Diamond, and D. Perez (Flow Cytometry Facility). This work was funded by the National Institutes of Health (National Institute of Biomedical Imaging and Bioengineering R01EB006192), by the Defense Advanced Research Projects Agency (HR0011-17-2-0008; the findings are those of the authors and should not be interpreted as representing the official views or policies of the US Government), by the Beckman Institute at Caltech (Programmable Molecular Technology Center, PMTC), by the Gordon and Betty Moore Foundation (GBMF2809), by the National Science Foundation Molecular Programming Project (NSF-CCF-1317694), by a Professorial Fellowship at Balliol

qHCR Flow Cytometry

141

College, University of Oxford, and by the Eastman Visiting Professorship at the University of Oxford. Competing Interests: The authors declare competing financial interests in the form of patents, pending patent applications, and a startup company (Molecular Instruments). References 1. Choi HMT, Schwarzkopf M, Fornace ME et al (2018) Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust. Development 145: dev165753. https://doi.org/10.1242/dev. 165753 2. Trivedi V, Choi HMT, Fraser SE et al (2018) Multidimensional quantitative analysis of mRNA expression within intact vertebrate embryos. Development 145:dev156869. https://doi.org/10.1242/dev.156869 3. Choi HMT, Chang JY, Trinh LA et al (2010) Programmable in situ amplification for

multiplexed imaging of mRNA expression. Nat Biotechnol 28:1208–1212. https://doi.org/10. 1038/nbt.1692 4. Choi HMT, Beck VA, Pierce NA (2014) Nextgeneration in situ hybridization chain reaction: higher gain, lower cost, greater durability. ACS Nano 8:4284–4294. https://doi.org/10. 1021/nn405717p 5. Choi HMT, Calvert CR, Husain N et al (2016) Mapping a multiplexed zoo of mRNA expression. Development 143:3632–3637. https:// doi.org/10.1242/dev.140137

Chapter 9 Multiplexed Quantitative In Situ Hybridization for Mammalian Cells on a Slide: qHCR and dHCR Imaging (v3.0) Maayan Schwarzkopf, Harry M. T. Choi, and Niles A. Pierce Abstract In situ hybridization based on the mechanism of hybridization chain reaction (HCR) enables multiplexed quantitative mRNA imaging in diverse sample types. Third-generation in situ HCR (v3.0) provides automatic background suppression throughout the protocol, dramatically enhancing performance and ease of use. In situ HCR v3.0 supports two quantitative imaging modes: (1) qHCR imaging for analog mRNA relative quantitation with subcellular resolution and (2) dHCR imaging for digital mRNA absolute quantitation with single-molecule resolution. Here, we provide protocols for qHCR and dHCR imaging in mammalian cells on a slide. Key words Accuracy, Autofluorescence, Automatic background suppression, dHCR imaging, Fluorescence in situ hybridization (FISH), Hybridization chain reaction (HCR), In situ HCR v3.0, Mammalian cells, mRNA imaging, Multiplexed, qHCR imaging, Precision, Quantitative, Signal amplification, Single-molecule imaging

1

Introduction In situ HCR v3.0 offers a unique combination of multiplexing, quantitation, sensitivity, versatility, and robustness for mRNA analysis in diverse sample types (Table 1) [1]. The same two-stage enzyme-free protocol is used independent of the number of target mRNAs (Fig. 1). In situ HCR v3.0 uses probes and amplifiers that combine to provide automatic background suppression throughout the protocol, ensuring that reagents will not generate amplified background even if they bind nonspecifically within the sample. Automatic background suppression dramatically enhances performance (signal-to-background and quantitative precision) and ease of use (no probe set optimization for new targets and organisms). With qHCR imaging, the signal is analog in the form of fluorescence voxel intensities that scale approximately linearly with the number of target molecules per voxel [1, 2]. Precision increases

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_9, © Springer Science+Business Media, LLC, part of Springer Nature 2020

143

144

Maayan Schwarzkopf et al.

Table 1 mRNA analysis using in situ HCR v3.0 Property

Details

Simple

Two-stage enzyme-free protocol independent of number of targets

Amplified

Boosts signal above autofluorescence

Multiplexed

Simultaneous one-stage signal amplification for up to five target mRNAs

Quantitative

Signal scales linearly with target abundance

Penetrating

Whole-mount vertebrate embryos

Resolved

Subcellular or single-molecule resolution as desired

Sensitive

Single molecules detected in thick autofluorescent samples

Portable

Suitable for use with diverse targets in diverse organisms

Versatile

Multiple assay formats: imaging, flow cytometry, blotting

Compatible

Combine with tissue clearing as desired

Robust

Automatic background suppression throughout the protocol

Adapted with permission from Development [1]

with voxel size (e.g., averaging pixels to obtain 2  2  2 μm voxels enhances precision while retaining subcellular resolution) [2]. Figure 2 illustrates multiplexed qHCR imaging of four target mRNAs in mammalian cells, achieving high signal-to-background without probe set optimization. With dHCR imaging, the signal is digital in the form of diffraction-limited single-molecule dots representing individual target molecules [1, 3]. Figure 3 illustrates use of dHCR imaging for digital mRNA absolute quantitation with single-molecule resolution in mammalian cells [1]. qHCR and dHCR quantitative imaging modes are complementary (Table 2), with qHCR suitable for medium- and highcopy targets (where the quantitative signal dominates autofluorescent background) and dHCR suitable for medium- and low-copy targets (where the signal from individual target molecules can be spatially separated) [1]. Because the qHCR signal per imaging voxel is analog, it will naturally decrease to zero as the number of targets per voxel decreases to zero. On the other hand, because the dHCR signal per target molecule is digital, it remains the same as the number of target molecules decreases to zero. We recommend 20+ probe pairs for qHCR imaging (signal-to-background and precision increase with probe set size) and 30+ probe pairs for dHCR imaging (fidelity increases with probe set size), as permitted by the length of the target mRNA. The same reagents can be used for either imaging mode, so imaging can be performed in qHCR mode (longer amplification time, lower magnification) or dHCR

qHCR and dHCR Imaging: Mammalian Cells on a Slide

a

145

In situ HCR v3.0 Protocol Summary Detection stage

Amplification stage

H1 DNA probe set

Hybridize probe set and wash

•••

HCR initiator I1 split between pair of probes

H2 Tethered fluorescent amplification polymers

In situ HCR and wash •••

HCR initiator HCR initiator I1 I1

mRNA target

•••

b

•••

mRNA target

•••

Metastable DNA HCR hairpins

Multiplexing Timeline Amplification stage

Detection stage

•••

•••

•••

Add all HCR probe sets Incubation 0

Add all HCR amplifiers Wash

Incubation 16 18 (overnight for qHCR) Experimental time (h) (45-90 min for dHCR)

Wash 34 36 20 22

Fig. 1 In situ HCR v3.0: multiplexed quantitative mRNA imaging with automatic background suppression throughout the protocol. (a) Two-stage enzyme-free protocol. Detection stage: probe sets hybridize to mRNA targets and unused probes are washed from the sample. Amplification stage: specifically bound probe pairs trigger self-assembly of a tethered fluorescent amplification polymer and unused hairpins are washed from the sample. Probes and amplifiers combine to provide automatic background suppression throughout the protocol, ensuring that reagents will not generate amplified background even if they bind nonspecifically in the sample. During the detection stage, individual probes that bind nonspecifically do not trigger HCR since each probe carries only a fraction of initiator I1. During the amplification stage, individual hairpins that bind nonspecifically do not trigger HCR since they are kinetically trapped. (b) Multiplexing timeline. The same two-stage protocol is used independent of the number of target mRNAs. HCR amplification is performed overnight for qHCR imaging experiments (to maximize the signal-to-background ratio) and for 45–90 min for dHCR imaging experiments (to resolve individual molecules as diffraction-limited dots). Adapted with permission from Development [1]

imaging mode (shorter amplification time, higher magnification) depending on the expression level observed in situ [1]. In situ HCR v3.0 protocols for qHCR imaging and dHCR imaging of mammalian cells on a slide are adapted with permission from Development [1].

146

Maayan Schwarzkopf et al.

Fig. 2 Multiplexed qHCR imaging in mammalian cells. Four-channel confocal micrograph for HEK 293T cells on a slide. GAPDH and PGK1 are predominantly localized in the cytoplasm and U6 (106 nt) and RNU48 (63 nt) are short nuclear RNAs. Confocal microscopy: 0.42  0.42  0.44 μm pixels (1 focal plane). Ch1: Target mRNA GAPDH, probe set with 10 probe pairs, amplifier B4-Alexa488, signal-to-background: 7.9  0.7 (mean  standard error, representative regions of N ¼ 6 cells). Ch2: Target mRNA PGK1, probe set with 18 probe pairs, amplifier B1-Alexa647, signal-to-background 27  5. Ch3: Target RNA U6, probe set with 2 probe pairs, amplifier B5-Alexa546, signal-to-background 32  3. Ch4: Target RNA RNU48, probe set with 1 probe pair, amplifier B3-Alexa514, signal-to-background 8.3  0.8. HCR amplification time: overnight

•• •







••

••

••

a

•••

2-channel redundant detection

b

dHCR imaging: digital single-molecule dots

3 µm

Ch1

colocalized: 0.85 ± 0.003

Ch2

Merge

colocalized: 0.82 ± 0.01

Fig. 3 Imaging single target mRNAs via dHCR imaging in mammalian cells. (a) Two-channel redundant detection of target mRNA BRAF in HEK cells using two probe sets, each initiating an orthogonal spectrally distinct HCR amplifier. (b) Confocal microscopy: 0.06  0.06  0.4 μm pixels (17 focal planes). Ch1: probe set with 23 probe pairs, amplifier B3-Alexa647. Ch2: probe set with 23 probe pairs, amplifier B4-Alexa546. HCR amplification time: 45 min. Red circles: dots detected in Ch1. Green circles: dots detected in Ch2. Yellow circles: dots detected in both channels. Colocalization represents the fraction of dots in one channel that are detected in both channels (mean  standard error, N ¼ 3; 1 field of view from each of 3 wells)

qHCR and dHCR Imaging: Mammalian Cells on a Slide

147

Table 2 Comparison of qHCR and dHCR imaging qHCR imaging

dHCR imaging

Type of signal

Analog subcellular voxel intensities

Digital single-molecule dots

Probe set size

20+ split-initiator probe pairs

30+ split-initiator probe pairs

Amplification time

Overnight

45–90 min

Magnification

Lower (20–40)

Higher (63–100)

Target expression level

Medium or high

Low or medium

2

Materials Make sure that your environment and reagents are DNase and RNase free; surfaces and pipettes should be cleaned using RNaseZAP. Prepare all solutions using Milli-Q or molecular grade water.

2.1 Sample Preparation

1. 5% CO2 incubator at 37  C. 2. RNaseZAP. 3. Tissue culture plate. 4. Milli-Q or molecular grade water. 5. Dulbecco’s PBS (DPBS) without calcium chloride or magnesium chloride (see Note 1). 6. Trypsin-EDTA (0.25%). 7. Cell growth media. 8. Phosphate-buffered saline (PBS; 10): 1.37 M NaCl, 27 mM KCl, 80 mM Na2HPO4, and 20 mM KH2PO4. 9. Fixation solution: 4% Formaldehyde in PBS. In a 15 mL conical tube combine 6.5 mL Milli-Q water, 2.5 mL of 16% formaldehyde (w/v), methanol free (see Note 2), and 1 mL 10 PBS pH 7.4 (see Note 1). Prepare fresh before use. 10. Trypan blue stain (0.4%, see Note 3). 11. Hemocytometer slide (see Note 3). 12. 0.01% Poly-D-lysine. 13. Ibidi μ-slide ibiTreat (see Note 4). 14. 70% Ethanol (ice cold).

2.2

HCR Reagents

1. HCR probes, amplifiers, and buffers for cells on a slide (probe hybridization buffer, probe wash buffer, amplification buffer) are available from Molecular Instruments (www. molecularinstruments.com) (see Note 5).

148

Maayan Schwarzkopf et al.

2. 20 Sodium chloride sodium citrate (SSC): 3 M Sodium chloride and 0.3 M sodium citrate. 3. 2 SSC. 4. 5 SSCT: 5 SSC, 0.1% Tween 20. 2.3 Additional Reagents and Equipment

1. SlowFade Gold antifade mountant with DAPI. 2. Thermocycler. 3. Hybridization oven or incubator at 37  C. 4. Confocal microscope.

3

Methods This protocol has been validated for human and mouse cell lines; additional modifications may be needed for other cell lines. The protocol describes volumes corresponding to a single chamber of an eight-chamber slide; adjust volumes according to the manufacturer’s recommendation if using other slide formats.

3.1 Sample Preparation

1. Coat the bottom of each chamber by applying 300 μL of 0.01% poly-D-lysine solution. Incubate for at least 30 min at room temperature. 2. Aspirate the coating solution and wash each chamber twice with molecular biology-grade H2O. 3. Plate desired number of cells in each chamber. (a) Aspirate growth media from culture plate and wash cells with DPBS. (b) Add 3 mL of trypsin-EDTA per 10 cm plate and incubate in a 5% CO2 incubator at 37  C for 5 min. (c) Quench trypsin-EDTA by adding 3 mL of growth media. (d) Transfer cells to a 15 mL conical tube and centrifuge for 5 min at 180  g. (e) If counting cells follow steps 3f–3g; otherwise plate desired volume. (f) Aspirate supernatant and resuspend cells in growth media or DPBS (see Note 6). (g) Mix cells with trypan blue in a 1:1 ratio and count cells (see Note 7). 4. Grow cells to desired confluency for 24–72 h. 5. Aspirate growth media and wash each chamber with 300 μL of DPBS.

qHCR and dHCR Imaging: Mammalian Cells on a Slide

149

6. Add 300 μL of fixation solution to each chamber. Caution: Use fixation solution with extreme care as it contains formaldehyde, a hazardous material. 7. Incubate for 10 min at room temperature. 8. Remove fixative (see Note 2) and wash each chamber twice with 300 μL of DPBS. 9. Aspirate DPBS and add 300 μL of ice-cold 70% ethanol. 10. Permeabilize cells overnight at 20  C. 11. Cells can be stored at 20  C or 4  C until use (see Notes 8 and 9). 3.2

Detection Stage

1. Aspirate ethanol. 2. Wash cells twice with 300 μL of 2 SSC, and remove solution by aspiration (see Note 10). 3. Pre-hybridize cells with 300 μL of probe hybridization buffer for 30 min at 37  C (see Note 11). Caution: Probe hybridization buffer contains formamide, a hazardous material. 4. Prepare a 4 nM probe solution by adding 1.2 pmol of each probe mixture (e.g., 1.2 μL of 1 μM stock) to 300 μL probe hybridization buffer at 37  C (see Note 12). 5. Remove the pre-hybridization solution and add the probe solution. 6. Incubate the sample overnight at 37  C. 7. Remove excess probes by washing 4  5 min with 300 μL of probe wash buffer at 37  C (see Note 11). Caution: Probe wash buffer contains formamide, a hazardous material. 8. Wash samples 2  5 min with 5 SSCT at room temperature. 9. Proceed to amplification stage.

3.3 Amplification Stage

1. Pre-amplify the sample in 300 μL of amplification buffer for 30 min at room temperature (see Note 13). 2. Separately prepare 18 pmol of hairpin h1 and 18 pmol of hairpin h2 by snap cooling 6 μL of 3 μM stock. Heat individual amplifier hairpins in a thermocycler to 95  C for 90 s, remove from thermocycler, and cool to room temperature in a dark drawer for 30 min (see Note 14). 3. Prepare 60 nM hairpin mixture by adding all snap-cooled hairpins to 300 μL of amplification buffer at room temperature. 4. Remove the pre-amplification solution and add the hairpin solution. 5. Incubate the sample overnight (>12 h) in the dark at room temperature (see Notes 15 and 16). 6. Remove excess hairpins by washing 5  5 min with 300 μL of 5 SSCT at room temperature.

150

Maayan Schwarzkopf et al.

7. Replace final wash with 100 μL SlowFade Gold antifade mountant with DAPI. 8. Store the sample at 4  C protected from light before imaging. 3.4 Confocal Microscopy

For qHCR imaging (medium- or high-expression targets), 20– 40 magnification provides subcellular resolution. For dHCR imaging (low- or medium-expression targets), use a highmagnification (63–100) lens with a large numerical aperture to detect individual RNA molecules.

3.5

The image analysis framework presented here was developed over a series of publications [1, 2, 4–6]. For convenience, here we provide a self-contained description of the details relevant to the present protocols.

Image Analysis

3.5.1 Raw Pixel Intensities

The total fluorescence within a pixel is a combination of signal and background. Fluorescent background (BACK) arises from three sources in each channel: l

Autofluorescence (AF): fluorescence inherent to the sample

l

Nonspecific detection (NSD): probes that bind nonspecifically in the sample and subsequently trigger HCR amplification

l

Nonspecific amplification (NSA): HCR hairpins that bind nonspecifically in the sample

Fluorescent signal (SIG) in each channel corresponds to: l

Signal (SIG): probes that hybridize specifically to the target mRNA and subsequently trigger HCR amplification

For pixel i of cell n, we denote the background: NSA AF X BACK ¼ X NSD n,i n,i þ X n,i þ X n,i

ð1Þ

X SIG n,i

ð2Þ

the signal:

and the total fluorescence (SIG + BACK): BACK ¼ X SIG X SIGþBACK n,i þ X n,i n,i

3.5.2 Measurement of Signal, Background, and Signal-to-Background for Transgenic Targets

ð3Þ

For a transgenic target mRNA, signal (SIG) and background (BACK) are characterized based on experiments of types 1a and 1b (Table 3) with SIG + BACK measured in transgenic cells containing the target and BACK measured in wild-type (WT) cells lacking the target. For the pixels in these regions, we characterize average BACK performance by calculating the mean pixel intensity (X n and Xn

SIGþBACK

for cell n). Performance across a population of cells is

characterized by calculating the sample means (X

BACK

and

qHCR and dHCR Imaging: Mammalian Cells on a Slide

151

Table 3 Experiment types for qHCR or dHCR imaging of a transgenic target mRNA Experiment type

Quantity

Probes

Hairpins

Cell type

1a

SIG + NSD + NSA + AF¼SIG + BACK

x

x

Transgenic

1b

NSD + NSA + AF ¼ BACK

x

x

WT

2

NSA + AF

x

Transgenic

3

AF

Transgenic

Adapted with permission from Development [1] SIGþBACK

X ) and standard errors (S X BACK and S X SIGþBACK ). The mean signal is then estimated as X

SIG

¼X

SIGþBACK

X

BACK

ð4Þ

with the standard error estimated via uncertainty propagation as rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi S X SIG 

S X SIGþBACK

2

þ S X BACK

2

ð5Þ

The signal-to-background ratio is estimated as X

SIG=BACK

¼X

SIG

=X

BACK

with standard error estimated via uncertainty propagation as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2ffi s s SIG BACK SIG=BACK X X s SIG=BACK  X þ SIG BACK X X

ð6Þ

ð7Þ

These upper bounds on estimated standard errors hold under the assumption that the correlation between SIG and BACK is nonnegative. If desired, additional control experiments that omit certain reagents can be used to characterize the individual components of background (AF, NSA, NSD). A type 2 experiment (no probes, NSAþAF hairpins only) yields X and a type 3 experiment (no probes, AF no hairpins) yields X (see Note 17). The background components can then be estimated via calculations analogous to (Eq. 4) and (Eq. 5). The estimated means are X

NSD

X

¼X

NSA

BACK

¼X

with estimated standard errors:

X

NSAþAF

NSAþAF

X

AF

ð8Þ ð9Þ

152

Maayan Schwarzkopf et al.

Table 4 Experiment types for qHCR or dHCR imaging of an endogenous target mRNA Experiment type

Quantity

Probes

Hairpins

Cell type

1a

SIG + NSD + NSA + AF¼SIG + BACK

x

x

WT

1b

NSD + NSA + AF ¼ BACK

x(transgenic)

x

WT

2

NSA + AF

x

WT

3

AF

WT

Adapted with permission from Development [1]

S X NSD

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2ffi  S X BACK þ S X NSAþAF

S X NSA 

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi S X NSAþAF

2

þ S X AF

2

ð10Þ ð11Þ

These upper bounds on estimated standard errors hold under the assumption that the correlations are nonnegative for the different components of background. If a type 1 experiment demonstrates SIG BACK, as is typically the case using in situ HCR v3.0, then there is little motivation to perform type 2 and type 3 experiments to characterize the individual background components (AF, NSA, NSD) as these are all bounded above by BACK. 3.5.3 Measurement of Signal, Background, and Signal-to-Background for Endogenous Targets

For an endogenous target mRNA, signal and background are characterized using experiments of types 1a and 1b (Table 4) with SIG + BACK measured using a probe set that addresses the target in WT cells and BACK measured using a probe set that addresses a different transgenic target absent from WT cells. Use of a previously validated transgenic probe set to measure background in WT cells ensures that a low measured fluorescence value does not simply indicate a dysfunctional probe set, but indeed represents low background generated by a probe set that is known to be functional if the target is present in the sample. If desired, additional control experiments of types 2 and 3 (Table 4) can be performed to characterize the components of background (AF, NSA, NSD) using the calculations of Subheading 3.5.2. In lieu of an experiment of type 1b, an experiment of type 2 can be used to obtain the partial background estimate BACK  AF + NSA.

3.5.4 Normalized Voxel Intensities for qHCR Imaging: Analog mRNA Relative Quantitation with Subcellular or Single-Cell Resolution

For qHCR imaging, precision increases with voxel size (see Subheading S2.2 of Ref. 2). To increase precision, raw voxel intensities can be calculated by averaging neighboring pixel intensities while still maintaining a subcellular voxel size. Alternatively, each cell can be treated as a single voxel using image segmentation software to identify the pixels corresponding to each cell and sum the signal per

qHCR and dHCR Imaging: Mammalian Cells on a Slide

153

cellular voxel. To facilitate relative quantitation between voxels, we estimate the normalized HCR signal of voxel j in replicate n as x n,j

X SIGþBACK  X BOT n,j X TOP  X BOT

ð12Þ

which translates and rescales the data so that the voxel intensities in each channel fall in the interval [0,1]. Here, X BOT X

BACK

ð13Þ

is the mean background across replicates and X TOP max X SIGþBACK n,j n, j

ð14Þ

is the maximum total fluorescence for a voxel across replicates. See Fig. 5 of Ref. 1 for an example of relative mRNA quantitation with subcellular resolution (2  2  2.7 μm voxels) in wholemount chicken embryos, achieving highly correlated normalized voxel intensities for two-channel redundant detection of a target mRNA [1]. In this setting, accuracy corresponds to linearity with zero intercept, and precision corresponds to scatter around the line [2]. See Ref. 2 for examples of quantification of relative expression levels and ratios (amplitudes and slopes in expression scatterplots) with subcellular resolution in whole-mountzebrafishembryos. 3.5.5 Dot Detection and Colocalization for dHCR Imaging: Digital mRNA Absolute Quantitation with Single-molecule Resolution

To validate dHCR single-molecule imaging, we perform a two-channel redundant detection experiment in which a target mRNA is detected using two independent probe sets and HCR amplifiers. Let N1 denote the number of dots detected in channel 1, N2 the number of dots detected in channel 2, and N12 the number of colocalized dots appearing in both channels. We define the colocalization fraction for each channel: C 1 ¼ N 12 =N 1 ,

ð15Þ

C 2 ¼ N 12 =N 2 :

ð16Þ

As the false-positive and false-negative rates for single-molecule detection go to zero, C1 and C2 will both approach 1 from below, providing a quantitative basis for evaluating performance. Figure 3 demonstrates two-channel redundant detection of single target mRNAs in mammalian cells on a slide, achieving approximately 82–85% colocalization. Single molecules were identified in each channel and colocalized between channels using the Dot Analysis 1.0 software package [1] available from Molecular Technologies (www.moleculartechnologies.org), a nonprofit academic resource within the Beckman Institute at Caltech.

154

4

Maayan Schwarzkopf et al.

Notes 1. Avoid using calcium chloride and magnesium chloride in PBS as this leads to increased autofluorescence in the samples. 2. Formaldehyde is a hazardous material. Use appropriate protective equipment and dispose as hazardous waste. 3. Trypan blue and a hemocytometer slide are needed only if counting cells using a hemocytometer. Cell count may also be obtained by alternative methods such as using a flow cytometer. 4. We use an eight-chamber slide with No. 1.5 coverslip bottom. Alternative slide layouts and suppliers may be used. 5. Select a different HCR amplifier (e.g., B1, B2, ...) for each target RNA that will be imaged in the same sample (for example, amplifier B1 for target 1, amplifier B2 for target 2, ...). Choose a different fluorophore label (e.g., Alexa647, Alexa594, ...) for each HCR amplifier that will be imaged in the same sample (e.g., B1-Alexa647, B2-Alexa594, ...). For each target mRNA, use a probe set with 20+ probe pairs for qHCR imaging and 30+ probe pairs for dHCR imaging, as permitted by target length. Select your optimal dye for your lowest expression target; this will depend on the light source and filters available on the imaging instrument but in general higher wavelength dyes have less sample autofluorescence (e.g., use Alexa488 for a high-expression target, use Alexa647 for a low-expression target). 6. If a fraction of the cells will be reseeded, resuspend the cells in growth media. 7. We use an automated cell counter to determine cell counts. If manually counting cells, follow recommendations for cell counting of your hemocytometer slide supplier. 8. Store samples up to a week; longer storage times should be validated experimentally. 9. Immunostaining can be combined with HCR. Immunostaining is typically performed first followed by a 4% formaldehyde (FA) fixation before proceeding to HCR. Immunostaining can also be performed after HCR. After the HCR protocol, samples can be fixed in 4% FA for 10 min at room temperature and washed 3  5 min with 1 PBS before immunostaining. 10. See Tables 3 and 4 for experiment types to characterize signal, background, signal-to-background, and/or background components for transgenic and/or endogenous target mRNAs. 11. Preheat probe hybridization and probe wash buffers to 37  C prior to use. Mix contents by swirling.

qHCR and dHCR Imaging: Mammalian Cells on a Slide

155

12. For dHCR imaging, use 10 nM (instead of 4 nM) of each probe to improve probe hybridization efficiency. 13. Equilibrate amplification buffer to room temperature before use. Mix contents by swirling. 14. HCR amplifier hairpins h1 and h2 are provided in hairpin storage buffer ready for snap cooling. h1 and h2 should be snap cooled in separate tubes. 15. We amplify the sample in a dark drawer. 16. For dHCR imaging, amplify for 45 min to ensure that singlemolecule dots are diffraction limited. 17. If a microscope generates nonnegligible fluorescence intensities in the absence of sample, this so-called instrument noise (NOISE) should be taken into consideration when calculating background and signal contributions, leading to four experiment types ((1a) SIG + BACK + NOISE, (1b) BACK + NOISE, (2) NSA + AF + NOISE, (3) AF + NOISE, (4) NOISE; cf. Tables 3 and 4).

Acknowledgments We thank M.E. Fornace for developing the Dot Analysis 1.0 software package for analyzing dHCR images. Within the Beckman Institute at Caltech, we thank the following for assistance: C.R. Calvert and G.J. Shin (Molecular Technologies), A. Collazo and S. Wilbert (Biological Imaging Facility), and J. Stegmaier and A. Cunha (Center for Advanced Methods in Image Analysis). This work was funded by the National Institutes of Health (National Institute of Biomedical Imaging and Bioengineering R01EB006192), by the Defense Advanced Research Projects Agency (HR0011-17-2-0008; the findings are those of the authors and should not be interpreted as representing the official views or policies of the US Government), by the Beckman Institute at Caltech (Programmable Molecular Technology Center, PMTC), by the Gordon and Betty Moore Foundation (GBMF2809), by the National Science Foundation Molecular Programming Project (NSF-CCF-1317694), by a Professorial Fellowship at Balliol College, University of Oxford, and by the Eastman Visiting Professorship at the University of Oxford. Competing Interests: The authors declare competing financial interests in the form of patents, pending patent applications, and a startup company (Molecular Instruments).

156

Maayan Schwarzkopf et al.

References 1. Choi HMT, Schwarzkopf M, Fornace ME et al (2018) Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust. Development 145: dev165753. https://doi.org/10.1242/dev. 165753 2. Trivedi V, Choi HMT, Fraser SE et al (2018) Multidimensional quantitative analysis of mRNA expression within intact vertebrate embryos. Development 145:dev156869. https://doi.org/10.1242/dev.156869 3. Shah S, Lubeck E, Schwarzkopf M et al (2016) Single-molecule RNA detection at depth via hybridization chain reaction and tissue hydrogel embedding and clearing. Development 143:2862–2867. https://doi.org/10.1242/ dev.138560

4. Choi HMT, Chang JY, Trinh LA et al (2010) Programmable in situ amplification for multiplexed imaging of mRNA expression. Nat Biotechnol 28:1208–1212. https://doi.org/10. 1038/nbt.1692 5. Choi HMT, Beck VA, Pierce NA (2014) Nextgeneration in situ hybridization chain reaction: higher gain, lower cost, greater durability. ACS Nano 8:4284–4294. https://doi.org/10. 1021/nn405717p 6. Choi HMT, Calvert CR, Husain N et al (2016) Mapping a multiplexed zoo of mRNA expression. Development 143:3632–3637. https:// doi.org/10.1242/dev.140137

Part IV Methods for Wholemounts and Plant Material

Chapter 10 Multiplexed Quantitative In Situ Hybridization with Subcellular or Single-Molecule Resolution Within Whole-Mount Vertebrate Embryos: qHCR and dHCR Imaging (v3.0) Harry M. T. Choi, Maayan Schwarzkopf, and Niles A. Pierce Abstract In situ hybridization based on the mechanism of hybridization chain reaction (HCR) enables multiplexed quantitative mRNA imaging in the anatomical context of whole-mount vertebrate embryos. Thirdgeneration in situ HCR (v3.0) provides automatic background suppression throughout the protocol, dramatically enhancing performance and ease of use. In situ HCR v3.0 supports two quantitative imaging modes: (1) qHCR imaging for analog mRNA relative quantitation with subcellular resolution in an anatomical context and (2) dHCR imaging for digital mRNA absolute quantitation with single-molecule resolution in an anatomical context. Here, we provide protocols for qHCR and dHCR imaging in wholemount zebrafish, chicken, and mouse embryos. Key words Accuracy, Autofluorescence, Automatic background suppression, dHCR imaging, Fluorescence in situ hybridization (FISH), Hybridization chain reaction (HCR), In situ HCR v3.0, mRNA imaging, Multiplexed, Precision, qHCR imaging, Quantitative, Signal amplification, Single-molecule imaging, Thick sample penetration, Whole-mount vertebrate embryos

1

Introduction In situ HCR v3.0 offers a unique combination of multiplexing, quantitation, sensitivity, versatility, and robustness for mRNA analysis in diverse sample types (Table 1) [1]. The same two-stage enzyme-free protocol is used independent of the number of target mRNAs (Fig. 1). In situ HCR v3.0 uses probes and amplifiers that combine to provide automatic background suppression throughout the protocol, ensuring that reagents will not generate amplified background even if they bind nonspecifically within the sample. Automatic background suppression dramatically enhances performance (signal-to-background and quantitative precision) and ease of use (no probe set optimization for new targets and organisms).

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_10, © Springer Science+Business Media, LLC, part of Springer Nature 2020

159

160

Harry M. T. Choi et al.

Table 1 mRNA analysis using in situ HCR v3.0 Property

Details

Simple

Two-stage enzyme-free protocol independent of number of targets

Amplified

Boosts signal above autofluorescence

Multiplexed

Simultaneous one-stage signal amplification for up to five target mRNAs

Quantitative

Signal scales linearly with target abundance

Penetrating

Whole-mount vertebrate embryos

Resolved

Subcellular or single-molecule resolution as desired

Sensitive

Single molecules detected in thick autofluorescent samples

Portable

Suitable for use with diverse targets in diverse organisms

Versatile

Multiple assay formats: imaging, flow cytometry, blotting

Compatible

Combine with tissue clearing as desired

Robust

Automatic background suppression throughout the protocol

Adapted with permission from Development [1]

Figure 2 illustrates multiplexed mRNA imaging in a whole-mount chicken embryo, achieving high signal-to-background without probe set optimization. With qHCR imaging, the signal is analog in the form of fluorescence voxel intensities that scale approximately linearly with the number of target molecules per voxel [1, 2]. Precision increases with voxel size so long as voxels remain smaller than the expression domains under consideration (e.g., averaging pixels to obtain 2  2  2 μm voxels enhances precision while retaining subcellular resolution) [2]. Figure 3 illustrates high accuracy and precision using qHCR imaging for analog mRNA relative quantitation with subcellular resolution in the anatomical context of whole-mount chicken embryos [1]. With dHCR imaging, the signal is digital in the form of diffraction-limited single-molecule dots representing individual target molecules [1, 3]. Figure 4 illustrates use of dHCR imaging for digital mRNA absolute quantitation with single-molecule resolution in the anatomical context of whole-mount chicken embryos [1]. qHCR and dHCR quantitative imaging modes are complementary (Table 2), with qHCR suitable for medium- and highcopy targets (where the quantitative signal dominates autofluorescent background) and dHCR suitable for medium- and low-copy targets (where the signal from individual target molecules can be spatially separated) [1]. Because the qHCR signal per imaging voxel is analog, it will naturally decrease to zero as the number of

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos

a

161

In situ HCR v3.0 Protocol Summary Detection stage

Amplification stage

H1 DNA probe set

Hybridize probe set and wash

•••

HCR initiator I1 split between pair of probes

H2 Tethered fluorescent amplification polymers

In situ HCR and wash •••

HCR initiator HCR initiator I1 I1

mRNA target

•••

b

•••

mRNA target

•••

Metastable DNA HCR hairpins

Multiplexing Timeline Amplification stage

Detection stage

•••

•••

•••

Add all HCR probe sets Incubation 0

Add all HCR amplifiers Wash

Incubation 16 18 (overnight for qHCR) Experimental time (h) (45-90 min for dHCR)

Wash 34 36 20 22

Fig. 1 In situ HCR v3.0: multiplexed quantitative mRNA imaging with automatic background suppression throughout the protocol. (a) Two-stage enzyme-free protocol. Detection stage: probe sets hybridize to mRNA targets and unused probes are washed from the sample. Amplification stage: specifically bound probe pairs trigger self-assembly of a tethered fluorescent amplification polymer and unused hairpins are washed from the sample. Probes and amplifiers combine to provide automatic background suppression throughout the protocol, ensuring that reagents will not generate amplified background even if they bind nonspecifically in the sample. During the detection stage, individual probes that bind nonspecifically do not trigger HCR since each probe carries only a fraction of initiator I1. During the amplification stage, individual hairpins that bind nonspecifically do not trigger HCR since they are kinetically trapped. (b) Multiplexing timeline. The same two-stage protocol is used independent of the number of target mRNAs. HCR amplification is performed overnight for qHCR imaging experiments (to maximize the signal-to-background ratio) and for 45–90 min for dHCR imaging experiments (to resolve individual molecules as diffraction-limited dots). Adapted with permission from Development [1]

targets per voxel decreases to zero. On the other hand, because the dHCR signal per target molecule is digital, it remains the same as the number of target molecules decreases to zero. We recommend 20+ probe pairs for qHCR imaging (signal-to-background and precision increase with probe set size) and 30+ probe pairs for dHCR imaging (fidelity increases with probe set size), as permitted by the length of the target mRNA. The same reagents can be used for either imaging mode, so imaging can be performed in qHCR mode (longer amplification time, lower magnification) or dHCR

162

Harry M. T. Choi et al.

a FoxD3

d

34 ± 3

50 µm

EphA4

Sox10

Signal-to-background 27 ± 3 59 ± 3

Dmbx1

c

b

43 ± 1

200 µm

50 µm

Fig. 2 Multiplexed mRNA imaging in whole-mount chicken embryos using in situ HCR v3.0: high signal-tobackground via automatic background suppression using unoptimized probe sets. (a) Expression schematics for four target mRNAs in the head and neural crest. (b) Four-channel confocal micrograph. (c) Zoom of depicted region of panel b. (d) Four individual channels from panel c with signal-to-background measurements (mean  standard error, representative regions of N ¼ 3 embryos). Ch1: target mRNA FoxD3, probe set with 12 probe pairs, amplifier B4-Alexa488. Ch2: target mRNA EphA4, probe set with 20 probe pairs, amplifier B2-Alexa647. Ch3: target mRNA Sox10, probe set with 20 probe pairs, amplifier B3-Alexa546. Ch4: target mRNA Dmbx1, probe set with 20 probe pairs, amplifier B1-Alexa514. Amplification time: overnight. Embryo fixed at stage HH10. Adapted with permission from Development [1]

imaging mode (shorter amplification time, higher magnification) depending on the expression level observed in situ [1]. Historically, quantitative RNA analyses for vertebrates have sacrificed anatomical context, employing some combination of microdissection, dissociation, cell sorting, and/or homogenization followed by qPCR, RNA-Seq, flow cytometry, microarray analysis, or hybridization barcoding [2]. By contrast, qHCR and dHCR imaging quantify RNA while preserving anatomical context and enable quantitative bidirectional discovery [2]: (1) read-out from anatomical space to expression space to discover quantitative co-expression relationships in selected regions of the specimen and (2) read-in from expression space to anatomical space to discover those anatomical locations in which selected quantitative

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos



••• 2-channel redundant detection

c 1

r = 0.98

Ch2 intensity

••







••

••

••

a

163

0

Ch1 intensity

1

b qHCR imaging: analog voxel intensities

50 µm

Ch1

Ch2

Merge

Fig. 3 qHCR imaging: analog mRNA relative quantitation with subcellular resolution in the anatomical context of whole-mount chicken embryos. (a) Two-channel redundant detection of target mRNA EphA4 using two probe sets, each initiating an orthogonal spectrally distinct HCR amplifier. (b) Confocal microscopy: 0.2  0.2  2.7 μm pixels (1 focal plane). Ch1: probe set with 20 probe pairs, amplifier B1-Alexa546. Ch2: probe set with 20 probe pairs, amplifier B2-Alexa647. No probe set optimization. Amplification time: overnight. Embryo fixed at stage HH10. (c) High accuracy and precision for mRNA relative quantitation in an anatomical context. Highly correlated normalized signal (Pearson correlation coefficient, r) for subcellular 2.1  2.1  2.7 μm voxels in the selected regions of panel b. Accuracy: linearity with zero intercept. Precision: scatter around the line. Adapted with permission from Development [1]

co-expression relationships occur. Figure 5 illustrates the workflow for read-out/read-in using multiplexed qHCR imaging [2]. Sample preparation protocols for whole-mount zebrafish, chicken, and mouse embryos are adapted with permission from Development [4]. In situ HCR v3.0 protocols are adapted with permission from Development [1].

2

Materials Make sure that all reagents and supplies are DNase and RNase free. Lab surfaces and tools (e.g., pipettes, tweezers) should be cleaned using RNaseZAP. Prepare all solutions using ultrapure (resistivity of 18 MΩ-cm at 25  C) or molecular grade water.

164

Harry M. T. Choi et al. •• •







••

••

••

a

••• 2-channel redundant detection

b dHCR imaging: digital single-molecule dots Ch1

Ch2

Merge

3 µm

colocalized: 0.85 ± 0.03

colocalized: 0.84 ± 0.02

Fig. 4 dHCR imaging: digital mRNA absolute quantitation with single-molecule resolution in the anatomical context of whole-mount chicken embryos. (a) Two-channel redundant detection of target mRNA Dmbx1 using two probe sets, each initiating an orthogonal spectrally distinct HCR amplifier. (b) Confocal microscopy: 0.1  0.1  0.4 μm pixels (22 focal planes). Ch1: probe set with 25 probe pairs, amplifier B1-Alexa594. Ch2: probe set with 25 probe pairs, amplifier B2-Alexa647. No probe set optimization. Amplification time: 90 min. Embryos fixed at stage HH8. Red circles: dots detected in Ch1. Green circles: dots detected in Ch2. Yellow circles: dots detected in both channels. Colocalization represents the fraction of dots in one channel that are detected in both channels (mean  standard error of the mean, representative region of N ¼ 3 embryos). Adapted with permission from Development [1] Table 2 Comparison of qHCR and dHCR imaging qHCR imaging

dHCR imaging

Type of signal

Analog subcellular voxel intensities

Digital single-molecule dots

Probe set size

20+ split-initiator probe pairs

30+ split-initiator probe pairs

Amplification time

Overnight

45–90 min

Magnification

Lower (20–40)

Higher (63–100)

Target expression level

Medium or high

Low or medium

Ch3

Ch3

Cluster 1

Cluster 2

Not selected

Image segmented by expression cluster

Ch3

Ch2

Ch1

Anatomical space

Step 2b: Display segmented images

Read-in

Discover anatomical regions

Discover expression clusters

Read-out

1

1

Ch1 Intensity

0

Ch3 Intensity

1

0

0

1

1

Cluster 2

Ch1 Intensity

Cluster 1

Voxels shaded by cluster color Cluster 1 Cluster 2 Not selected

0

1

Ch2 Intensity

Expression space

1

0

0

Step 2a: Select expression clusters of interest

1

0

Ch3 Intensity

Voxels shaded by anatomical region Region 1 Region 2

0

0

1

0

Ch2 Intensity

0 0

1

1

1

Expression space

Fig. 5 Workflow for quantitative read-out and read-in analyses using qHCR imaging. Step 0: Acquire and normalize data. Step 1: Read-out from anatomical space to expression space: select anatomical locations of interest and discover expression clusters. Step 2: Read-in from expression space to anatomical space: select expression clusters of interest and discover anatomical locations. Adapted with permission from Development [2]

Pixel intensity = signal + background

Ch3

Ch2

Ch1

Acquire multi-channel image

Step 0a: Multiplexed in situ HCR

Region 2

Region 1

Voxels outlined by region color

Ch2

Ch2

Voxel intensity = signal estimate

Ch1

Ch1

Anatomical space

Ch1 Intesnity Ch1 Intensity

Normalized signal estimate for each voxel in each channel

Step 1b: Plot pairwise expression scatter plots

Ch2 Intensity Ch2 Intensity

Step 1a: Select anatomical regions of interest

Ch3 Intesnity Ch3 Intesnity

Step 0b: Normalize data

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos 165

166

Harry M. T. Choi et al.

2.1 Sample Preparation 2.1.1 Fixation and Permeabilization (All Sample Types)

1. RNaseZAP. 2. Ultrapure (resistivity of 18 MΩ-cm at 25  C) or molecular grade water. 3. 10 Phosphate-buffered saline (PBS): 1.37 M NaCl, 27 mM KCl, 80 mM Na2HPO4, and 20 mM KH2PO4 (see Note 1). 4. PBST: 1 PBS, 0.1% Tween 20. In a 50 mL conical tube, mix 5 mL of 10 PBS, 500 μL of 10% Tween 20, and 44.5 mL of ultrapure water. 5. Fixation solution: 4% Paraformaldehyde (PFA) in 1 PBS. Microwave 200 mL of regular tap water in a 500 mL beaker for 1 min. Place beaker on a heat plate and maintain temperature right below 60  C. In a 50 mL conical tube, add 25 mL 1 PBS pH 7.4. Weigh 1 g of PFA inside a fume hood and add to conical tube (see Note 2). Close conical tube and place in water bath to allow all powder to dissolve. Do not let temperature exceed 60  C. Cool solution to 4  C before use. Use fresh fixative for each batch preparation. 6. 100% Methanol (MeOH). 7. MeOH/PBST series: 75% MeOH/25% PBST, 50% MeOH/ 50% PBST, 25% MeOH/75% PBST. Mix MeOH and PBST at corresponding volume ratios.

2.1.2 Whole-Mount Chicken Embryo Preparation

1. 3 M Filter paper. 2. Ringer’s solution: 123 mM NaCl, 1.53 mM CaCl2, 4.96 mM KCl2, 0.81 mM Na2HPO4, 0.15 mM KH2PO4, pH 7.4. Weigh 14.4 g of NaCl, 340 mg of CaCl2, 740 mg of KCl, 230 mg of Na2HPO4, and 40 mg of KH2PO4 and place them in a 2 L bottle. Bring volume up to 1.5 L with ultrapure water. Use a magnetic stir bar to help dissolve the reagents. Adjust pH to 7.5 and fill up to 2 L with ultrapure water. Filter sterilize with 0.22 μm bottle-top filter. 3. Proteinase K solution: 10 μg/mL Proteinase K in 1 PBS. Add 1 μL of 20 mg/mL proteinase K solution to 2 mL of PBST.

2.1.3 Whole-Mount Zebrafish Embryo/Larva Preparation

1. 0.003% 1-Phenyl-2-thiourea (PTU) solution: 0.003% PTU. To make a 0.3% stock solution, weigh 0.3 g of 1-phenyl-2-thiourea powder and dissolve in 100 mL of egg water. In a 50 mL conical tube, dilute 500 μL of 0.3% PTU with 49.5 mL of egg water to get 0.003% PTU solution. 2. Proteinase K solution: 30 μg/mL Proteinase K in 1 PBS. Add 1.5 μL of 20 mg/mL proteinase K solution to 1 mL of PBST.

2.1.4 Whole-Mount Mouse Embryo Preparation

1. Proteinase K solution: 10 μg/mL Proteinase K in 1 PBS. Add 1 μL of 20 mg/mL proteinase K solution to 2 mL of PBST.

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos

2.2

HCR Reagents

167

1. HCR probes, amplifiers, and buffers for whole-mount (probe hybridization buffer, probe wash buffer, amplification buffer) are available from Molecular Instruments (www. molecularinstruments.com) (see Note 3). 2. 20 Sodium chloride sodium citrate (SSC): 3 M Sodium chloride and 0.3 M sodium citrate. 3. 5 SSCT: 5 SSC, 0.1% Tween 20.

2.3

Imaging

1. SlowFade Gold antifade mountant. 2. Scotch tape. 3. No. 1 coverslip.

2.4

Equipment

1. Heat block (37  C). 2. Thermocycler (95  C). 3. Confocal microscope.

3

Methods The following sample preparation protocols have been optimized for whole-mount chicken embryos at stage HH 8–11, wholemount zebrafish embryos at 26 h postfertilization (hpf) and larvae at 5 days postfertilization (dpf), and whole-mount mouse embryos at E9.5. Additional modifications may be needed for other developmental stages.

3.1 Sample Preparation

1. Collect chick embryos on 3 M paper circles and place in a petri dish containing Ringer’s solution.

3.1.1 Whole-Mount Chicken Embryos

2. Transfer embryos into a new petri dish with fresh Ringer’s solution (see Note 4). 3. Transfer into a petri dish containing fixation solution (see Note 5). Caution: Use fixation solution with extreme care as it contains PFA, a hazardous material. 4. Fix the samples at room temperature for 1 h. 5. Transfer embryos into a petri dish containing PBST. 6. Dissect the embryos off the filter paper. Remove the fixed vitelline membrane (opaque) and cut the squares around area pellucida, without leaving excess of extraembryonic tissue. 7. Transfer embryos into a 2 mL tube containing PBST. 8. Nutate for 5 min with the tube positioned horizontally in a small ice bucket. 9. Wash embryos two additional times with 2 mL of PBST, each time nutating for 5 min on ice.

168

Harry M. T. Choi et al.

10. Dehydrate embryos with 2  5 min washes of 2 mL MeOH on ice. 11. Store embryos at 20  C overnight before use (see Note 6). 12. Transfer the required number of embryos for an experiment to a 2 mL tube (see Note 7). 13. Rehydrate with a series of graded 2 mL MeOH/PBST washes for 5 min each on ice: (a) 75% MeOH/25% PBST (b) 50% MeOH/50% PBST (c) 25% MeOH/75% PBST (d) 100% PBST (e) 100% PBST 14. Treat embryos with 2 mL of 10 μg/mL proteinase K solution for 2 min (stage HH 8) or 2.5 min (stage HH 10–11) at room temperature (see Note 8). 15. Postfix with 2 mL of fixation solution for 20 min at room temperature (see Note 9). Caution: Use fixation solution with extreme care as it contains PFA, a hazardous material. 16. Wash embryos 2  5 min with 2 mL of PBST on ice. 17. Wash embryos with 2 mL of 50% PBST/50% 5 SSCT for 5 min on ice. 18. Wash embryos with 2 mL of 5 SSCT for 5 min on ice. 19. Proceed to in situ HCR protocol (see Note 10). 3.1.2 Whole-Mount Zebrafish Embryos/Larvae

1. Collect zebrafish embryos and incubate at 28  C in a petri dish with egg water. 2. Exchange egg water with egg water containing 0.003% of PTU when embryos reach 12 hpf (see Note 11). 3. Replace with fresh egg water containing 0.003% of PTU every day until the embryos/larvae reach the desired developmental stage. 4. For young embryos/larvae, remove chorion using two pairs of sharp tweezers under a dissecting scope starting at 16 hpf. 5. When embryos/larvae reach the desired developmental stage, transfer 40 embryos/larvae to a 2 mL Eppendorf tube and remove excess egg water. 6. Fix embryos/larvae in 2 mL of fixation solution for 24 h at 4  C (see Note 5). Caution: Use fixation solution with extreme care as it contains PFA, a hazardous material. 7. Wash embryos/larvae 3  5 min with 1 mL of 1 PBS to stop the fixation. 8. Dehydrate and permeabilize with a series of MeOH washes (1 mL each):

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos

169

(a) 100% MeOH for 4  10 min (b) 100% MeOH for 1  50 min 9. Store embryos/larvae at 20  C overnight before use (see Note 6). 10. Transfer the required number of embryos/larvae for an experiment to a 2 mL Eppendorf tube. 11. Rehydrate with a series of graded 1 mL MeOH/PBST washes for 5 min each at room temperature: (a) 75% MeOH/25% PBST (b) 50% MeOH/50% PBST (c) 25% MeOH/75% PBST (d) 5  100% PBST 12. Treat embryos/larvae with 1 mL of proteinase K (30 μg/mL) for 45 min at room temperature (see Notes 8 and 12). 13. Wash embryos/larvae two times with PBST (1 mL each) without incubation. 14. Postfix with 1 mL of fixation solution for 20 min at room temperature (see Note 9). Caution: Use fixation solution with extreme care as it contains PFA, a hazardous material. 15. Wash embryos/larvae 5  5 min with 1 mL of PBST. 16. Proceed to in situ HCR protocol (see Note 10). 3.1.3 Whole-Mount Mouse Embryos

1. Wipe all dissection equipment with RNaseZAP. 2. Sacrifice a pregnant female mouse using an IACUC-approved protocol. 3. Immediately remove the uterus and submerge it in fixation solution in a fresh RNase-free petri dish (see Note 5). Caution: Use fixation solution with extreme care as it contains PFA, a hazardous material. 4. Dissect the mouse embryos from the uterus while it is submerged in fixation solution. 5. Transfer the embryos to a clean vial containing fresh fixation solution and fix them overnight or longer at 4  C (see Note 13). 6. Wash 2  5 min with PBST on ice. 7. Dehydrate embryos into MeOH with a series of graded MeOH/PBST washes for 10 min on ice: (a) 25% MeOH/75% PBST (b) 50% MeOH/50% PBST (c) 75% MeOH/25% PBST (d) 100% MeOH (e) 100% MeOH

170

Harry M. T. Choi et al.

8. Incubate embryos at 20  C overnight (>16 h) or until use (see Note 6). 9. Transfer the required number of embryos for an experiment to a 2 mL tube. 10. Rehydrate with a series of graded MeOH/PBST washes for 10 min each on ice: (a) 75% MeOH/25% PBST (b) 50% MeOH/50% PBST (c) 25% MeOH/75% PBST (d) 100% PBST 11. Wash embryos with PBST for 10 min at room temperature. 12. Immerse embryos in 10 μg/mL proteinase K solution for 15 min at room temperature (see Note 8). 13. Wash embryos 2  5 min with PBST. 14. Postfix with fixation solution for 20 min at room temperature (see Note 9). Caution: Use fixation solution with extreme care as it contains PFA, a hazardous material. 15. Wash embryos 3  5 min with PBST. 16. Proceed to in situ HCR protocol (see Note 10). 3.2 In Situ HCR Protocol 3.2.1 Detection Stage

1. For each sample in an experiment, transfer embryos to a 2 mL tube (see Note 14). 2. Incubate embryos in probe hybridization buffer for 30 min at 37  C (see Notes 15 and 16). Caution: Probe hybridization buffer contains formamide, a hazardous material. 3. Prepare probe solution with 4 nM of each probe in probe hybridization buffer at 37  C (e.g., 2 pmol of each probe mixture in 500 μL of probe hybridization buffer) (see Note 17). 4. Remove the pre-hybridization solution and add the probe solution. 5. Incubate samples overnight (12–16 h) at 37  C. 6. Remove excess probes by washing embryos 4  15 min with probe wash buffer at 37  C (see Note 15 and 16). Caution: Probe wash buffer contains formamide, a hazardous material. 7. Wash samples 2  5 min with 5 SSCT at room temperature.

3.2.2 Amplification Stage

1. Incubate embryos in amplification buffer for 5 min at room temperature (see Notes 16 and 18). 2. Prepare hairpin h1 and hairpin h2 by snap cooling them separately (heat at 95  C for 90 s and cool to room temperature in a dark drawer for 30 min). The quantity needed depends on the

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos

171

amplification volume (e.g., 30 pmol of hairpin h1 and 30 pmol of hairpin h2 for amplification volume of 500 μL) (see Note 19). 3. Prepare hairpin mixture by adding snap-cooled h1 hairpins and snap-cooled h2 hairpins to amplification buffer at room temperature. The final concentration of each hairpin should be 60 nM. 4. Remove the pre-amplification solution and add the hairpin mixture. 5. Incubate the embryos overnight (12–16 h) in the dark at room temperature (see Notes 20 and 21). 6. Remove excess hairpins by washing with 5 SSCT at room temperature (see Note 16): (a) 2  5 min (b) 2  30 min (c) 1  5 min 7. Samples can be stored at 4  C protected from light before microscopy. 3.3 Sample Mounting for Microscopy

1. Make a chamber for mounting each embryo by aligning two stacks of double-sided tape (two pieces per stack) 1 cm apart on a 25 mm  75 mm glass slide.

3.3.1 Whole-Mount Chicken Embryos

2. Place an embryo between the tape stacks on the slide and remove as much solution as possible. 3. Align the embryo for dorsal imaging and carefully touch the slide with a Kimwipe to further dry the area around the embryo. 4. Add two drops of SlowFade Gold antifade mountant on top of the embryo. 5. Place a 22 mm  30 mm No. 1 coverslip on top of the stacks to close the chamber.

3.3.2 Whole-Mount Zebrafish Embryos/Larvae

1. Make a chamber for mounting the embryos/larvae by aligning two stacks of Scotch tape (eight pieces per stack) 1 cm apart on a 25 mm  75 mm glass slide. 2. Add 200 μL of SlowFade Gold antifade mountant between the tape stacks on the slide. 3. Place embryos/larvae on the medium and orient for imaging. 4. Place a 22 mm  22 mm No. 1 coverslip on top of the stacks to close the chamber.

172

Harry M. T. Choi et al.

1. Clean a 25 mm  75 mm glass slide and a 22 mm  30 mm No. 1 coverslip with RNaseZAP. Add four beads of Vaseline to each slide to support the coverslip at the corners.

3.3.3 Whole-Mount Mouse Embryos

2. Place and orient an embryo on the slide. 3. Remove excess buffer and add 100 μL of SlowFade Gold antifade mountant over the embryo. 4. Place the coverslip over the embryo. Apply enough pressure to push the coverslip onto the embryo without flattening it. 5. Store the slides in the dark at 4  C until imaging. 3.4 Confocal Microscopy

For qHCR imaging (medium- or high-expression targets), 20– 40 magnification provides subcellular resolution. For dHCR imaging (low- or medium-expression targets), use a highmagnification (63–100) lens with a large numerical aperture to detect individual RNA molecules.

3.5

The image analysis framework presented here was developed over a series of publications [1, 2, 4–6]. For convenience, here we provide a self-contained description of the details relevant to the present protocols.

Image Analysis

3.5.1 Raw Pixel Intensities

The total fluorescence within a pixel is a combination of signal and background. Fluorescent background (BACK) arises from three sources in each channel: l

Autofluorescence (AF): fluorescence inherent to the sample

l

Nonspecific detection (NSD): probes that bind nonspecifically in the sample and subsequently trigger HCR amplification

l

Nonspecific amplification (NSA): HCR hairpins that bind nonspecifically in the sample

Fluorescent signal (SIG) in each channel corresponds to: l

Signal (SIG): probes that hybridize specifically to the target mRNA and subsequently trigger HCR amplification

For pixel i of replicate embryo n, we denote the background: NSA AF X BACK ¼ X NSD n,i n,i þ X n,i þ X n,i

ð1Þ

X SIG n,i

ð2Þ

the signal:

and the total fluorescence (SIG + BACK): BACK ¼ X SIG X SIGþBACK n,i þ X n,i n,i

ð3Þ

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos

173

Table 3 Experiment types for qHCR or dHCR imaging Experiment type

Quantity

Probes

Hairpins

Expression region

1

SIG + NSD + NSA + AF ¼ SIG + BACK

x

x

High

1

NSD + NSA + AF ¼ BACK

x

x

No/low

2

NSA + AF

x

High

3

AF

High

Adapted with permission from Development [1]

3.5.2 Measurement of Signal, Background, and Signal-to-Background

For each target mRNA, background (BACK) is characterized for pixels in a representative rectangular region of no or low expression and the combination of signal plus background (SIG + BACK) is characterized for pixels in a representative rectangular region of high expression (Table 3). The choice of representative regions depends on the type of target mRNA: l

Transgenic target: BACK voxel intensities are measured in a region of no expression in wild-type (WT) embryos lacking the target; SIG + BACK voxel intensities are measured in a region of high expression in transgenic embryos containing the target.

l

Endogenous target with local expression: BACK voxel intensities are measured in a region of no or low expression in WT embryos; SIG + BACK voxel intensities are measured in a region of high expression in the same replicate embryos.

l

Endogenous target with global expression: BACK voxel intensities are measured in a region of high expression in WT embryos using the standard in situ protocol but omitting the probes (this yields the partial background estimate BACK  AF + NSA) or alternatively using a probe set for a transgenic target absent from WT embryos (use of a previously validated transgenic probe set to measure background in WT cells ensures that a low measured fluorescence value does not simply indicate a dysfunctional probe set, but indeed represents low background generated by a probe set that is known to be functional if the target is present in the sample); SIG + BACK voxel intensities are measured in a region of high expression in a different set of WT embryo replicates (using the standard in situ protocol including probes).

For the pixels in these regions, we characterize average performance  BACK and X  SIGþBACK for by calculating the mean pixel intensity (X n n replicate embryo n). Performance across replicate embryos is charBACK SIGþBACK acterized by calculating the sample means (X and X ) and standard errors (S X BACK and S X SIGþBACK ). The mean signal is then estimated as

174

Harry M. T. Choi et al.

X

SIG

¼X

SIGþBACK

X

BACK

ð4Þ

with the standard error estimated via uncertainty propagation as rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi s X SIG 

2

s X SIGþBACK

þ s X BACK

2

ð5Þ

The signal-to-background ratio is estimated as X

SIG=BACK

¼X

SIG

=X

BACK

ð6Þ

with standard error estimated via uncertainty propagation as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2ffi s s SIG BACK SIG=BACK X X s SIG=BACK  X þ SIG BACK X X

ð7Þ

These upper bounds on estimated standard errors hold under the assumption that the correlation between SIG and BACK is nonnegative. Figure 2 illustrates multiplexed imaging of four target mRNAs in a whole-mount chicken embryo with high signal-tobackground without probe set optimization. 3.5.3 Measurement of Background Components

Calculation of the signal-to-background ratio requires only a type 1 experiment (using the terminology of Table 3), yielding the  SIGþBACK and X BACK that are needed for the calculation. values X If desired, additional control experiments that omit certain reagents can be used to characterize the individual components of background (AF, NSA, NSD). A type 2 experiment (no probes, NSAþAF hairpins only) yields X and a type 3 experiment (no probes, AF no hairpins) yields X (see Note 22). The background components can then be estimated via calculations analogous to (Eq. 4) and (Eq. 5). The estimated means are X

NSD

X

¼X

NSA

BACK

¼X

X

NSAþAF

NSAþAF

X

ð8Þ

AF

ð9Þ

with estimated standard errors: rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi S X NSD 

S X NSA 

S X BACK

2

2

þ S X NSAþAF

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi    ffi S X NSAþAF

2

þ S X AF

2

ð10Þ ð11Þ

These upper bounds on estimated standard errors hold under the assumption that the correlations are nonnegative for the different components of background. If a type 1 experiment demonstrates SIG BACK, as is typically the case using in situ HCR v3.0, then there is little motivation to perform type 2 and type 3 experiments to characterize the individual background components (AF, NSA, NSD) as these are all bounded above by BACK.

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos 3.5.4 Normalized Voxel Intensities for qHCR Imaging: Analog mRNA Relative Quantitation with Subcellular Resolution in an Anatomical Context

175

For qHCR imaging, precision increases with voxel size as long as the imaging voxels remain smaller than the features in the expression pattern (see Subheading S2.2 of Ref. 2). To increase precision, raw voxel intensities can be calculated by averaging neighboring pixel intensities while still maintaining a subcellular voxel size. To facilitate relative quantitation between voxels, we estimate the normalized HCR signal of voxel j in replicate n as x n,j

X SIGþBACK  X BOT n,j X TOP  X BOT

ð12Þ

which translates and rescales the data so that the voxel intensities in each channel fall in the interval [0,1]. Here, X BOT X

BACK

ð13Þ

is the mean background across replicates and X TOP max X SIGþBACK n,j n, j

ð14Þ

is the maximum total fluorescence for a voxel across replicates. Figure 3 demonstrates analog mRNA relative quantitation with subcellular resolution (2  2  2.7 μm voxels) in whole-mount chicken embryos, achieving highly correlated normalized voxel intensities for two-channel redundant detection of target mRNA EphA4 [1]. In this setting, accuracy corresponds to linearity with zero intercept, and precision corresponds to scatter around the line [2]. 3.5.5 Dot Detection and Colocalization for dHCR Imaging: Digital mRNA Absolute Quantitation with Single-Molecule Resolution in an Anatomical Context

To validate dHCR single-molecule imaging, we perform a two-channel redundant detection experiment in which a target mRNA is detected using two independent probe sets and HCR amplifiers. Let N1 denote the number of dots detected in channel 1, N2 the number of dots detected in channel 2, and N12 the number of colocalized dots appearing in both channels. We define the colocalization fraction for each channel: C 1 ¼ N 12 =N 1 ,

ð15Þ

C 2 ¼ N 12 =N 2 :

ð16Þ

As the false-positive and false-negative rates for single-molecule detection go to zero, C1 and C2 will both approach 1 from below, providing a quantitative basis for evaluating performance. Figure 4 demonstrates two-channel redundant detection of single target mRNAs in whole-mount chicken embryos, achieving approximately 84–85% colocalization. Single molecules were identified in each channel and colocalized between channels using the Dot Analysis 1.0 software package [1] available from Molecular Technologies (www.moleculartechnologies.org), a nonprofit academic resource within the Beckman Institute at Caltech.

176

Harry M. T. Choi et al.

3.5.6 Read-out/Read-In Analysis Framework for Quantitative RNA Discovery in an Anatomical Context

4

Pairwise expression scatterplots that each display normalized voxel intensities for two channels provide a quantitative framework for performing read-out/read-in analyses (Fig. 5) [2]. Read-out from anatomical space to expression space enables discovery of expression clusters of voxels with quantitatively related expression levels and ratios (amplitudes and slopes in the expression scatterplots), while read-in from expression space to anatomical space enables discovery of the corresponding anatomical locations of these expression clusters within the embryo. The simple and practical normalization approach of Subheading 3.5.4 translates and rescales all voxels identically within a given channel (enabling comparison of amplitudes and slopes in scatterplots between replicates), and does not attempt to remove scatter in the normalized signal estimate that is caused by scatter in the background. See Figs. 4 and 5 of Ref. 2 for examples of read-out/read-in analyses performed using the Readout/Read-in 1.0 software package [2] available from Molecular Technologies (www.moleculartechnologies.org), a nonprofit academic resource within the Beckman Institute at Caltech.

Notes 1. Avoid using calcium chloride and magnesium chloride in PBS as this leads to increased autofluorescence in the embryos/ larvae. 2. Paraformaldehyde (PFA) is a hazardous material. Use appropriate protective equipment and dispose as hazardous waste. 3. Select a different HCR amplifier (e.g., B1, B2, ...) for each target RNA that will be imaged in the same sample (for example, amplifier B1 for target 1, amplifier B2 for target 2, ...). Choose a different fluorophore label (e.g., Alexa647, Alexa594, ...) for each HCR amplifier that will be imaged in the same sample (e.g., B1-Alexa647, B2-Alexa594, ...). For each target mRNA, use a probe set with 20+ probe pairs for qHCR imaging and 30+ probe pairs for dHCR imaging, as permitted by target length. Select your optimal dye for your lowest expression target; this will depend on the light source and filters available on the imaging instrument but in general higher wavelength dyes have less sample autofluorescence (e.g., use Alexa488 for a high-expression target, use Alexa647 for a low-expression target). 4. Transferring embryos into fresh Ringer’s solution will rinse away egg yolk before fixation. 5. Always use fresh fixation solution and make sure that it is cooled to 4  C before use to avoid increased autofluorescence. 6. Embryos/larvae can be stored for 6 months at 20  C. 7. Do not place more than four embryos in each 2 mL tube.

qHCR and dHCR Imaging: Whole-Mount Vertebrate Embryos

177

8. Proteinase K concentration and treatment time should be reoptimized for each batch of proteinase K, or for samples at a different developmental stage. 9. Fixation solution does not need to be fresh for post-fixation. 10. Immunostaining can be combined with HCR. Immunostaining is typically performed first. After immunostaining, samples can be fixed in fixation solution for 20 min at room temperature and washed 3  5 min with 1 PBST before proceeding to HCR. Immunostaining can also be performed after HCR. After the HCR protocol, samples can be fixed in fixation solution for 20 min at room temperature and washed 3  5 min with 1 PBST before immunostaining. 11. PTU inhibits melanogenesis but can be toxic at high concentrations. PTU treatment must start before the initial pigmentation occurs as PTU does not remove pigment that has already formed. PTU treatment (steps 2 and 3 in Subheading 3.1.2) is not necessary for embryos fixed before 30 hpf. 12. Skip proteinase K treatment and post-fixation (steps 11–14) for zebrafish embryos 30 hpf and younger. 13. Each female mouse produces 6–9 embryos. We recommend using 2 mL of solution per group of ten embryos. 14. See Table 3 for experiment types to characterize signal, background, signal-to-background, and/or background components for target mRNAs. 15. Preheat probe hybridization and probe wash buffers to 37  C prior to use. Mix contents by swirling. 16. All hybridization, amplification, and wash buffer volumes throughout the protocol should be sufficient to ensure that all embryos/larvae are submerged in each sample: for example, 500 μL for 1–4 whole-mount chicken embryos, 500 μL for 8 whole-mount zebrafish embryos/larvae, and 1 mL for 1–4 whole-mount mouse embryos. 17. For dHCR imaging, use 10 nM (instead of 4 nM) of each probe to improve probe hybridization efficiency. 18. Equilibrate amplification buffer to room temperature before use. Mix contents by swirling. 19. HCR amplifier hairpins h1 and h2 are provided in hairpin storage buffer ready for snap cooling. h1 and h2 should be snap cooled in separate tubes. 20. For dHCR imaging, amplify for a shorter period of time (e.g., 90 min for stage HH 8 chicken embryos) to ensure that singlemolecule dots are diffraction limited. 21. Samples can be amplified in a dark drawer. 22. If a microscope generates non-negligible fluorescence intensities in the absence of sample, this so-called instrument noise (NOISE) should be taken into consideration when calculating

178

Harry M. T. Choi et al.

background and signal contributions, leading to four experiment types ((1) SIG + BACK + NOISE, (1) BACK + NOISE, (2) NSA + AF + NOISE, (3) AF + NOISE, (4) NOISE; cf. Table 3).

Acknowledgments We thank M.E. Fornace for developing the Dot Analysis 1.0 software package for analyzing dHCR images and V. Trivedi for developing the Read-out/Read-in 1.0 software package for performing read-out/read-in analyses on multiplexed qHCR images in vertebrate embryos. Within the Beckman Institute at Caltech, we thank the following for assistance: C.R. Calvert and G.J. Shin (Molecular Technologies), A. Collazo and S. Wilbert (Biological Imaging Facility), and J. Stegmaier and A. Cunha (Center for Advanced Methods in Image Analysis). This work was funded by the National Institutes of Health (National Institute of Biomedical Imaging and Bioengineering R01EB006192), by the Defense Advanced Research Projects Agency (HR0011-17-2-0008; the findings are those of the authors and should not be interpreted as representing the official views or policies of the US Government), by the Beckman Institute at Caltech (Programmable Molecular Technology Center, PMTC), by the Gordon and Betty Moore Foundation (GBMF2809), by the National Science Foundation Molecular Programming Project (NSF-CCF-1317694), by a Professorial Fellowship at Balliol College, University of Oxford, and by the Eastman Visiting Professorship at the University of Oxford. Competing Interests: The authors declare competing financial interests in the form of patents, pending patent applications, and a startup company (Molecular Instruments). References 1. Choi HMT, Schwarzkopf M, Fornace ME et al (2018) Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust. Development 145: dev165753. https://doi.org/10.1242/dev. 165753 2. Trivedi V, Choi HMT, Fraser SE et al (2018) Multidimensional quantitative analysis of mRNA expression within intact vertebrate embryos. Development 145:dev156869. https://doi.org/10.1242/dev.156869 3. Shah S, Lubeck E, Schwarzkopf M et al (2016) Single-molecule RNA detection at depth via hybridization chain reaction and tissue hydrogel embedding and clearing. Development 143:2862–2867. https://doi.org/10.1242/ dev.138560

4. Choi HMT, Calvert CR, Husain N et al (2016) Mapping a multiplexed zoo of mRNA expression. Development 143:3632–3637. https:// doi.org/10.1242/dev.140137 5. Choi HMT, Chang JY, Trinh LA et al (2010) Programmable in situ amplification for multiplexed imaging of mRNA expression. Nat Biotechnol 28:1208–1212. https://doi.org/10. 1038/nbt.1692 6. Choi HMT, Beck VA, Pierce NA (2014) Nextgeneration in situ hybridization chain reaction: higher gain, lower cost, greater durability. ACS Nano 8:4284–4294. https://doi.org/10. 1021/nn405717p

Chapter 11 Hybridization Chain Reaction for Quantitative and Multiplex Imaging of Gene Expression in Amphioxus Embryos and Adult Tissues Toby G. R. Andrews, Giacomo Gattoni, Lara Busby, Michael A. Schwimmer, and E`lia Benito-Gutie´rrez Abstract In situ hybridization (ISH) methods remain the most popular approach for profiling the expression of a gene at high spatial resolution and have been broadly used to address many biological questions. One compelling application is in the field of evo-devo, where comparing gene expression patterns has offered insight into how vertebrate development has evolved. Gene expression profiling in the invertebrate chordate amphioxus (cephalochordate) has been particularly instrumental in this context: its key phylogenetic position as sister group to all other chordates makes it an ideal model system to compare with vertebrates and for reconstructing the ancestral condition of our phylum. However, while ISH methods have been developed extensively in vertebrate model systems to fluorescently detect the expression of multiple genes simultaneously at a cellular and subcellular resolution, amphioxus gene expression profiling is still based on single-gene nonfluorescent chromogenic methods, whose spatial resolution is often compromised by diffusion of the chromogenic product. This represents a serious limitation for reconciling gene expression dynamics between amphioxus and vertebrates and for molecularly identifying cell types, defined by their combinatorial code of gene expression, that may have played pivotal roles in evolutionary innovation. Herein we overcome these problems by describing a new protocol for application of the thirdgeneration hybridization chain reaction (HCR) to the amphioxus, which permits fluorescent, multiplex, and quantitative detection of gene expression in situ, within the changing morphology of the developing embryo, and in adult tissues. A detailed protocol is herein provided for whole-mount preparations of embryos and vibratome sections of adult tissues. Key words Amphioxus, Fluorophore-labeled, Gene expression profiling, HCR, In situ hybridization, Multiplex, Single-cell resolution

1

Introduction In situ hybridization (ISH) methods have been utilized over the past several decades to describe spatial patterns of gene expression across tissues and across species. When combined with functional studies (e.g., gene expression inhibition) these methods have been

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_11, © Springer Science+Business Media, LLC, part of Springer Nature 2020

179

180

Toby G. R. Andrews et al.

particularly powerful for characterizing the roles and the regulatory interactions of specific genes in particular developmental processes. ISH methods have been, for example, key for defining genes and network interactions controlling vertebrate axial patterning [1, 2], vertebrate limb development [3, 4], and segmentation of the drosophila blastoderm [5, 6]. When combined with comparative developmental biology approaches, these methods have offered a means beyond morphological examination to define the homology of traits in different taxa, and therefore to define the origins of important evolutionary innovations. A remarkable discovery in this context was the observation that there is an evolutionarily conserved Hox code that patterns the anterior-posterior axis of most animals [7]. This sets the groundwork for comparisons between distantly related taxa, sharing homologous body parts (e.g., head, thorax, tail) that contain taxon-specific morphological specializations. Focusing on chordate evolution, gene expression profiling in the amphioxus has helped to resolve traits that are ancestral to the phylum, and those that are vertebrate innovations [8]. These include vertebrate traits, such as a complex brain and neural crest derivatives [9]. These studies have relied on the classic ISH protocol for amphioxus [10] which has been applied by researchers in the field for almost three decades [11]. While ISH methods have constantly evolved in other model systems to increase sensitivity, to target multiple genes simultaneously, and to achieve cellular and subcellular resolution, most of the gene expression profiling in amphioxus is still based on singlegene expression detection using nonfluorescent chromogenic methods. Chromogenic methods can be very sensitive, but they are not quantitative, as they depend on the nonlinear accumulation of a chromogenic product. Diffusion of this product can often compromise the spatial resolution of the signal, leading to loss of cellular and subcellular detail. Furthermore, chromogenic methods are limited in terms of multiplexing, usually restricted to threecolor reactions (color per gene), and only two in amphioxus in very exceptional cases [12]. While possible, multiplex ISH is highly technically challenging, because it demands multiple chromogenic reactions to be performed in series without cross reaction. Many technologies have emerged recently that tackle some of these technical challenges in vertebrates and offer quantitative information by virtue of fluorescent RNA detection (e.g., TSA, RNAscope, HCR). However, because these have been inconsistently applied, reconciling gene expression dynamics between model systems is becoming increasingly challenging. Therefore, there is high demand for ISH technology that can be readily applied across model systems, and offers cellular resolution, multiplex potential, and fluorescent quantitative imaging.

Hybridization Chain Reaction in Amphioxus

181

To address these challenges, we have taken advantage of thirdgeneration in situ hybridization chain reaction (HCR) technology, recently developed by Choi and collaborators [13], and adapted it to the particular requirements of amphioxus embryos and adult tissues. HCR version 3 (HCRv3) is based on the use of DNA probe pairs that bind in tandem to complementary sequences on the mRNA of interest. Split between each probe pair is an initiator sequence that triggers the focal binding and polymerization of metastable kinetically trapped DNA hairpins, each of which is conjugated to a fluorescent Alexa Fluor moiety. A complete adapter is formed where the probe pairs bind in tandem specifically to the mRNA sequence with certain separation. Only then can hairpins assemble through cooperative binding into tethered fluorescent amplification polymers, whose fluorescence intensity will scale linearly with the local density of mRNA molecules [14, 15]. Intrinsic to this design is automatic background suppression, in which the nonspecific off-site binding of a single probe will not accumulate fluorescent signal due to absence of a complete adapter sequence. This background suppression generates a significantly improved signal: noise ratio in HCRv3 imaging compared to previous versions, enabling highly sensitive imaging of even very weakly expressed genes. The specificity of adapter sequences and commercial availability of DNA hairpins conjugated to different fluorophores (Molecular Instruments) allow multiplexing probes to target up to five different genes in the same specimen per reaction [16]. The HCRv3 method is therefore multiplex, sensitive, and quantitative. Here we describe a method for application of the HCRv3 technology to multiplex and quantitatively analyze gene expression in the amphioxus, both in embryos and adult sections. This protocol can be executed in 3 days in embryos or adult sections that had been previously fixed and dehydrated in methanol or ethanol. The sample preparation is, unsurprisingly, specific for amphioxus, but the hybridization and the amplification steps are essentially as described by Choi and colleagues [13], with some minor modifications. Below we briefly outline several key developments that have facilitated optimization of our protocol in relation to that described by Choi and colleagues [13], and the classic amphioxus ISH protocol from Holland and Holland [10]: 1. We have updated the protocol for amphioxus embryo fixation to improve the quality of the starting material (Subheading 3.1), thereby increasing the specificity and sensitivity of the HCR signal. In particular, the protocol described below differs from that previously published with respect to the fixative amount and the length of fixation [17, 18]. The handling of the embryos also differs since we do not include any centrifugation steps and all pipette tips and plates are siliconized. This

182

Toby G. R. Andrews et al.

Fig. 1 Embedding and vibratome sectioning of adult amphioxus brain tissues. Once the specimen has been anesthetized (a) the head is separated from the rest of the body and fixed for 24 h as indicated in Subheading 3.3, step 3. The separated head is then washed and glued into the self-adhesive foam either vertically (b) or horizontally (c), depending on the orientation needed. For coronal sections further dissection of the ventral side is recommended to properly align brain and neural tube in a same section. The foam with the attached head is then transferred into a peel-a-way square embedding mold and filled up with liquid low-melting agarose (d). Once the agarose has solidified the block is extracted from the mold and trimmed as a pyramid to provide better grip while sectioning (e). The wider part of the pyramid is then glued to the vibratome holder and the head is sectioned with the dorsal fin facing the blade (f)

latter step is motivated by our observation that amphioxus embryos are very sticky, especially at the early steps of fixation and at the beginning of the HCR protocol, meaning that they can get damaged very easily or lost during handling. 2. We describe a completely new protocol for generating thick vibratome sections of adult tissues (Subheading 3.2), which we illustrate using brain tissue (Fig. 1). This represents an improvement on methods for analyzing gene expression in adult amphioxus tissues, which to date were mostly dependent on paraffin embedding and on a very lengthy protocol for chromogenic detection [19, 20]. The classic protocol adopted these strategies since the high hybridization temperatures of the traditional amphioxus ISH protocol (60–63  C) meant that the tissue would curl unless attached to the slide. By contrast, the milder hybridization conditions of the HCR protocol are permissive enough so the technique can be applied to thick

Hybridization Chain Reaction in Amphioxus

183

floating vibratome sections with no curling or noticeable shrinkage. Consequently, we have been able to produce a protocol that can be applied more rapidly than the previous ISH on paraffin sections that allows a finer tuning depending on the gene of interest, and that is in general more robust (Fig. 1). 3. We introduced a bleaching step at the beginning of the HCR protocol (Subheading 3.4, step 3). Bleaching is a common step in the ISH protocols of many model systems, including flatworms and most vertebrates [21–23]. In amphioxus, we have found that this step reduces the inherent autofluorescence of the tissue and completely eliminates the natural pigmentation of the eye and other photoreceptive spots. Thereafter, the tissue becomes more transparent, which permits imaging of entire embryos at single-cell resolution (Fig. 2). 4. We found that pretreatment with proteinase K is not necessary for HCR in amphioxus (Fig. 1), unlike when using DIG-labeled riboprobes, and indeed might only be useful when hybridizing very thick sections (Fig. 3). This is a major step forward, as imaging of gene expression is now possible in undigested tissues, showing a more normal morphology. Indeed, this protocol can be readily combined with immunohistochemistry for morphological landmarks. 5. We propose a novel imaging approach for fluorescent labeling in amphioxus tissues using inverted confocal microscopy. This ensures that the specimens lie flat and within the working distance of the objective throughout the imaging process, even up to 100 optical magnification. Thus, Z-stacks can be acquired of entire specimens at single-cell resolution. This configuration is compatible with tile scanning of a large field of specimens in a multi-area time lapse, but also permits adjustment of orientation to acquire the same specimen from multiple views, or to repair small deviations of position after the initial mounting process. In principle, individual specimens can be retrieved after imaging with this method, bleached, and re-stained to further increase the number of channels in the HCR.

2 2.1

Materials Reagents

1. Ethanol/methanol (for sample storage). 2. Split-initiator probe pairs (Molecular Instruments). 3. Fluorophore-labeled metastable DNA HCR hairpins (Molecular Instruments). 4. Probe hybridization buffer (Molecular Instruments). 5. Probe wash buffer (Molecular Instruments).

184

Toby G. R. Andrews et al.

Fig. 2 Triple HCR in whole-mount embryos of B. lanceolatum. HCR staining was performed on whole amphioxus embryos reared in our facility [24] to 24 hpf at 21  C with varying concentrations of probe, and in the presence and absence of proteinase K digestion. For this combination, 20-pair probe sets were designed against Soxb1c for broad labeling of the neural tube, Mnx for labeling of motor neuron progenitors, and Elav for postmitotic neurons. Exposure to hairpins in the absence of probes generates channel-specific background fluorescence profiles (a). This is conspicuous in all channels, but most severe following excitation at a 488 nm wavelength. A 4 nM probe concentration, as suggested for use on zebrafish embryos, generates specific signal but with a poor signal:noise ratio, even after extensive washing (b). In the 488 channel, signal can be difficult to distinguish from background. Signal:noise is greatly improved with expose to 20 nM (c) and 40 nM (d) probe concentrations for common incubation times. At 40 nM, signal can readily be resolved from background. Proteinase K (PK) treatment is common in amphioxus ISH protocols to enhance embryo permeability. However, treatment with PK did not enhance HCR signal beyond that achieved with only permeabilization in TritonX-100 and DMSO (e). (f) Magnified view of parasagittal section through anterior neural tube of embryo in (d), revealing regional differences in expression profiles for each gene, and a single triple-positive cell. Scale bars measure 100 μm (a–e) and 20 μm (f)

6. Amplification buffer (Molecular Instruments). 7. RNase AWAY (Ambion). 8. Proteinase K. 9. Glycerol or Aqua-Polymount (Polysciences). 2.2

Equipment

1. Siliconized or gelatinized p200 tips. 2. Sterile filter tips. 3. Siliconized or gelatinized 1.5 mL tubes. 4. Nunc untreated 4-well plates or siliconized/gelatinized 24-well plates. 5. Hybridization oven (e.g., Hybaid Shake ‘n’ Stack).

Hybridization Chain Reaction in Amphioxus

185

Fig. 3 Triple HCR in B. lanceolatum brain vibratome sections. HCR staining was performed on floating vibratome sections of adult amphioxus maintained in our amphioxus facility [24]. Sections were stained

186

Toby G. R. Andrews et al.

6. Orbital shaker. 7. Water bath. 8. Heat block. 9. Stereomicroscope. 10. Vibratome (Leica). 11. Vibratome blades. 12. Dissection Instruments (for adult tissues). 13. Peel-a-way square embedding molds (Sigma). 14. Self-adhesive foam stripes for window sealing. 15. Superglue (Loctite or Gorilla). 16. Parafilm. 17. 0.2 μm pore size filters. 18. Glass-bottom dishes. 2.3 Buffers and Solutions

1. Tricaine solution: Tricaine, double-distilled water, 1 M Tris– HCl pH 9. Weigh 400 mg of tricaine and add this to 90 mL of double-distilled water. Once the tricaine has dissolved completely adjust the pH to 8.0–8.2 with 1 M Tris–HCl pH 9. Top up the solution to 100 mL with double-distilled water, filter through a 0.2 μm pore size filter, and aliquot in 2 mL doses. Store at 20  C. 2. 2 MOPS buffer pH 7.5–7.6: 0.1 M MOPS (free acid), 2 mM MgSO4, 1 mM EGTA, 0.5 M NaCl. Weigh all in powder and dissolve in DEPC-treated water or nuclease-free water. Adjust with NaOH pellets to a pH of 7.5–7.6 (for a volume of 250 mL add approximately 8 pellets). Filter through a 0.2 μm pore size filter and store at 4  C. Discard after a month. 3. 3.7% PFA-MOPS pH 7.5–7.6: Paraformaldehyde, 2 MOPS buffer pH 7.5–7.6. Weigh 1.85 g of paraformaldehyde and reserve in a 50 mL tube. Bring some DEPC-treated water or nuclease-free water to the boil in a microwave and add 20 mL of this to the paraformaldehyde reserved in the 50 mL tube. To the paraformaldehyde in water add approximately 50 μL of

ä Fig. 3 (continued) with 20-pair probe sets designed against Pax4/6, used here as a neuronal marker, Nkx2.1, used here as a specifier of GABAergic fate and GAD, used here as a marker of GABAergic neurons. All probes were used at a concentration of 40 nM, as it was found to show the best signal:noise ratio for most of the genes in whole-mount preparations of embryos (see Fig. 2). Sections that had been preincubated in methanol after fixation (b) (see Note 1) showed a better signal:noise ratio than those that were not exposed to methanol (a). Low-expressed genes such as Nkx2.1 are for example hardly visible when sections are not preincubated in methanol (compare a and b). Additional pretreatment with proteinase K (c) improves the sharpness of the signal for all genes, including the low-level-expressing Nkx2.1, the mid-level-expressing Pax4/6, and the highly expressed GAD. Magnified views of merged images in d, e, and f show co-localization of the transcripts at a single-cell resolution

Hybridization Chain Reaction in Amphioxus

187

NaOH 10 M. Vortex until all paraformaldehyde is dissolved. If there are still particles of paraformaldehyde in suspension, leave the 50 mL tube in a water bath at 65  C until everything is completely dissolved. Thereafter, add 25 mL of 2 MOPS buffer and confirm that pH is 7.5–7.6. Adjust the pH if it is not within range. Top up with DEPC-treated water or nuclease-free water to 50 mL. Filter through a 0.22 μm pore size filter and store at 4  C. For health and safety reasons it is recommended to perform all these steps, including weighing the paraformaldehyde, in a fume hood wearing protective goggles and gloves. 4. 5 Gelatine stock for coating tips and plates: Gelatine, DEPCtreated water, formaldehyde 37%. For a volume of 50 mL of stock, weigh 0.25 g of gelatine. Dissolve by autoclaving in DEPC-treated water. Allow the solution to cool and add 250 μL of formaldehyde (37%). For health and safety reasons it is recommended to add the formaldehyde in a fume hood and wear protective goggles and gloves, as the solution might still be warm. To coat tips and plates, dilute the stock to 1 with DEPC-treated water or nuclease-free water. After coating, dry plates and tips on a 65  C oven before using them. 5. 10 NPBS: 200 mM Phosphate buffer pH 7.4, 9% NaCl. Filter through a 0.22 μm pore size filter and store at room temperature. If stocked for long, filter before use. 6. NPBST: NPBS, 0.1% Tween 20. 7. Bleaching buffer: 5% Deionized formamide, 1.5% H2O2, 0.2 SSC in DEPC-treated water or nuclease-free water. Prepare fresh when needed. For health and safety reasons it is recommended to wear protective goggles and gloves while preparing this buffer. 8. 3% LM-agarose: LM-agarose, DEPC-treated water or nuclease-free water. Weight 3 g of low-melting agarose for every 100 mL of DEPC-treated water or nuclease-free water. Add the 3 g of LM agarose to 80 mL of DEPC-treated water or nuclease-free water and boil in a microwave until completely dissolved. Add 10 mL of 10 NPBS and top up to 100 mL, if volume has decreased while boiling, with DEPC-treated water or nuclease-free water. The agarose can be kept at 4  C and remelted when needed. 9. Permeabilization solution: 1% DMSO, 1% Triton. Dissolve in NPBS. Store at room temperature. 10. 5 SSCT [13]: 5 SSC (pH 7), 0.1% Tween 20. For best results, filter just before use with a 0.22 μm pore size filter. Store at room temperature. 11. DAPI counterstain solution: 1 μg/mL DAPI, NPBST. Dilute DAPI (1 μg/mL) to 1:500 in NPBST.

188

3

Toby G. R. Andrews et al.

Methods Before starting it is recommended to clean all surfaces and equipment with RNase AWAY (Ambion). This is particularly relevant for the vibratome holder and tray, as the tissue can be especially vulnerable to RNases while sectioning. It is also recommended to autoclave all glassware and other autoclavable instruments used in the procedure. Use gloves throughout the entire procedure to protect yourself and to protect the samples from RNases.

3.1 Embryo Collection and Fixation

1. Collect the embryos in the center of the dish by concentrically moving the dish. Amphioxus embryos are very small, so it is necessary to observe this process under the stereomicroscope. 2. Pipette the embryos with a siliconized or gelatinized wide orifice p200 tip into siliconized 1.5 mL tubes. 3. Transfer the 1.5 mL tubes to an ice-cold rack and leave the embryos to pellet by gravity. When the pellet is formed, remove as much seawater as possible and fill the tube with ice-cold 3.7% PFA-MOPS buffer. 4. Leave the embryos to pellet by gravity. When the pellet is formed, remove as much liquid as possible and refill the tube with ice-cold 3.7% PFA-MOPS buffer. 5. Repeat step 4 and fix for 8–10 h (depending on the stage) at 4  C. 6. Wash the embryos with 1 MOPS buffer by pelleting the embryos by gravity, as indicated above, at least two times. 7. Wash the embryos in either ethanol or methanol, by pelleting the embryos by gravity, as indicated above, at least twice. 8. Store the embryos in either ethanol or methanol at 20  C, or proceed to the amphioxus HCR ISH protocol (Subheading 3.3, step 2).

3.2 Adult Tissue Fixation, Embedding, and Sectioning

1. Anesthetize adult amphioxus for 30 min using 2 mL of tricaine solution per 50 mL of seawater. 2. Make a series of cuts to divide the animal into four segments of equal length. Remove excess seawater from the tissues and transfer into ice-cold 3.7% PFA-MOPS buffer. 3. Fix for 24 h at 4  C. 4. Wash tissues with 1 MOPS at least twice. The samples can at this point be archived for long-term storage in either ethanol or methanol at 20  C. Otherwise, proceed to the following step to prepare samples for embedding. 5. Wash tissues with 1 NPBS at least two times.

Hybridization Chain Reaction in Amphioxus

189

6. Transfer the tissues into pre-warmed low-melting agarose in a water or dry bath at 45  C and leave to equilibrate for 30 min. 7. Embedding: Cut a small rectangle of self-adhesive foam, large enough to hold the sample in the middle. Peel off the upside of the foam and add a small drop of superglue on top. With forceps extract the tissue that was equilibrating at 45  C from the tube and place it in a suitable orientation on the top of the foam. Gently press the tissue so it gets properly attached to the foam. Transfer the foam rectangle with the attached tissue into the center of a peel-a-way square embedding mold. Fill the mold up with pre-warmed low-melting agarose (Fig. 1). Leave to cool and solidify at 4  C. 8. Sectioning: Peel away the mold and glue the agarose block to the vibratome holder plate with superglue. Place at 4  C while preparing the vibratome. Fill the tray of the vibratome with ice-cold RNase-free NPBS and the outer tray with ice. Assemble the vibratome blade following the instructions of the vibratome manufacturer. Attach the reserved holder plate containing the sample into the vibratome tray as indicated by the vibratome manufacturer. Place the hardest part of the tissue facing the blade (Fig. 1). If using a Leica VT100S the best sectioning conditions for amphioxus heads and trunks are as follows: speed 0.26 mm/s, frequency 50 Hz, and minimum thickness 50 μm. Collect the samples in gelatinized 24- or 96-well plates, to keep track of the sectioning order. 9. Sections can be stored at this point in either ethanol or methanol at 20  C (see Note 1). Otherwise, proceed to the amphioxus HCR ISH protocol (Subheading 3.3, step 2). 3.3 Amphioxus HCR In Situ Hybridization (ISH) Steps

1. Embryo/section rehydration: If the embryos or sections are stored in ethanol it is best to transfer them to methanol for the rehydration steps. Thereafter, rehydrate through a methanol/ water series, decreasing by 20% the proportion of methanol every 15 min. Perform all steps in gelatinized 4-well plates. 2. Wash embryos/sections twice in NPBST. Start the HCR protocol at this step if embryos and sections are already in NPBS or 1 MOPS buffer. 3. Bleaching: Replace NPBST with 500 μL of bleaching solution. Incubate for 30–60 min, or as long as needed to remove the pigment spots, with light and reflective foil at the base of the plate. When embryos and tissues have reached translucency, wash twice in NPBST (see Note 2). 4. Permeabilization: Replace NPBST with 500 μL of permeabilization solution. Incubate for 3 h at room temperature (see Note 3).

190

Toby G. R. Andrews et al.

5. Pre-hybridization: Remove permeabilization solution or NPBST (in case of the sections) and wash the embryos/sections in 5 SSCT for 5 min at room temperature. Wash the embryos/sections in pre-warmed hybridization buffer (without probes) at 37  C to equilibrate the tissues. Replace with fresh pre-warmed hybridization buffer and pre-hybridize for at least 1 h at 37  C (best results are obtained with longer pre-hybridization times). 6. Hybridization: Prepare the probe solution by diluting 1–10 μL of 2 μM probe mixed stock of each gene in 500 μL of pre-warmed hybridization solution. Replace the pre-hybridization buffer with the freshly made probe solution and incubate overnight at 37  C. For adult sections, a minimum of 18-h incubation is recommended to ensure homogenous probe binding. For best results wrap the plate in parafilm to prevent evaporation during the hybridization time (see Note 4). 7. Pre-amplification: Remove the probe solution and wash the embryos/sections with pre-warmed probe wash buffer at 37  C. Wash for a minimum of 20 min. In the last of the washes the plate can be transferred to an orbital shaker at room temperature, as the following steps are performed at room temperature. Replace the probe wash buffer with 5 SSCT and thoroughly wash embryos/sections three or four times, with each wash for a minimum of 1 h. Thereafter, pre-amplify by replacing the SSCT with 500 μL of pre-warmed amplification buffer at room temperature. Incubate in an orbital shaker at room temperature for a minimum of 1 h. 8. Preparation of the hairpins: Pipette 1 μL of each fluorescently labeled hairpin (of a 3 μM stock), per every 100 μL of amplification buffer to be used, into a fresh 1.5 mL tube. Heat up the hairpins by placing the 1.5 mL tube into a heat block at 95  C for 90 s. Cool the hairpins on ice, in the dark, for 30 min (see Note 5). 9. Amplification: In this step embryos and sections are transferred from the well plates to 1.5 mL tubes. In order to do this, concentrate the embryos/sections in the center of the well and use a wide-orifice siliconized p200 tip to collect them in no more of 20 μL of volume. Transfer the embryos/sections to a 1.5 mL tube containing 1 μL of each pre-cooled hairpin topped up to 80 μL with fresh amplification buffer. Protect the tubes from the light and incubate overnight in an orbital shaker at room temperature. 10. Washing: In this step embryos and sections are transferred back to a 4-well plate, so the washing volumes are bigger and therefore the excess of hairpins in the solution is removed more efficiently. To this aim, add 400 μL of 5 SSCT to the

Hybridization Chain Reaction in Amphioxus

191

1.5 mL tube containing the embryos/sections and transfer the total volume of 500 μL to a 4-well plate. Add another 500 μL of 5 SSCT to the tube, to ensure that all embryos and sections are recovered, and transfer these 500 μL to the 4-well plate. Wash three times at room temperature in 5 SSCT by replacing the buffer every 30 min. Wash a further four to five times at 4  C in 5 SSCT, by replacing the buffer every hour. Best results are obtained with overnight washes at 4  C (see Notes 6 and 7). 11. DAPI staining: Wash the embryos/sections twice in NPBST. Replace with fresh DAPI counterstain solution. Incubate overnight protected from light at 4  C (see Note 8). 12. Wash four to five times in NPBST, replacing the buffer every 15 min. 3.4 Whole-Mount Imaging

1. Replace NPBST with 100% glycerol and allow the embryos/ sections to equilibrate. 2. Fill the base of a glass-bottomed dish suitable for imaging on an inverted confocal microscope with 100% glycerol. 3. Transfer the embryos/sections to the glass-bottomed dish using a wide-orifice siliconized p200 pipette tip. Use an eyelash wand to push the specimens to the bottom of the dish, such that they lie flat in direct context with the glass in the correct orientation for imaging. Embryos can be organized into rows and columns based on variations in state, treatment, and HCR gene combination. Adult sections should be manually flattened against the glass with the eyelash wand to remove folds and curvatures (see Note 9). 4. Leave the glass-bottomed dish in the dark at 4  C for at least half an hour prior to imaging to prevent drifting during the imaging process. 5. Image whole embryos and sections using a multi-area timelapse function on an inverted confocal microscope (see Note 10). 6. After imaging, retrieve specimens from the glass-bottomed dish for long-term storage in PBS.

4

Notes 1. Methanol incubation. To maximize signal-to-background ratio in adult sections an incubation of at least 12 h in 100% MeOH is strongly recommended. This step has proved to be particularly useful for transcripts present at low levels, while patterns of highly expressed genes can be identified even in sections that are not treated with MeOH (see Fig. 3).

192

Toby G. R. Andrews et al.

2. Bleaching solution. This solution can also be used to remove the chorion of early amphioxus embryos. 3. Pretreatment with proteinase K. For very thick sections incubate with proteinase K (PK) at a concentration of 1 μg/mL, for up to 8 min at 37  C. Thereafter, thoroughly wash in NPBST and postfix with 3.7% PFA-PBS for 30 min at room temperature. Then wash thoroughly at least three times in NPBST. 4. Hybridization time. Hybridization time is critical for thick sections, as probes take longer to penetrate in the tissue. At least 18 h of incubation is recommended. 5. Amplifier label. Even with bleaching, endogenous autofluorescence in amphioxus is still particularly visible at 488 nm. We therefore advise using Alexa Fluor 488 hairpins (Molecular Instruments) for strongly expressed genes, where signal intensity is expected to exceed the autofluorescence levels. If this is not possible, background subtraction might be required during image acquisition or registration to enhance the signal of the probes. 6. Background removal. Although the V3 method ensures a high signal:noise ratio by preventing off-site binding, in amphioxus we find that background can still be an issue, particularly for weakly expressed genes where laser powers must be high to visualize the signal. However, we have found background to be reduced by extending probe washes and hairpin washes overnight at 4  C. If after an overnight wash background is still high, washes can be extended for a further 2–3 days at 4  C. 7. Combined immunohistochemistry. HCR imaging can be readily coupled to immunohistochemistry to label specific subcellular structures or morphological landmarks. For this, wash specimens out of the hairpin solution and then proceed directly to the immunoblock and subsequent steps of primary and secondary antibody incubation. All steps from here on should be in the dark at 4  C to preserve HCR fluorescence. However, this is very robust, and we find no severe quenching during the immunohistochemistry protocol. 8. Dapi staining. This step can also be done by incubating at a higher concentration (1:200) for 3 h at 4  C. 9. Vibratome sections mounting for microscopy. The use of Aqua Polymount might be more appropriate for vibratome sections of big specimens. This mounting media solidifies at application, thereby preventing the sections from moving or bend during image acquisition. This mounting media can also be used for embryos if high magnification objectives, with a shorter working distance, are used (e.g., 40, 60, 100). In this case, use a wide-orifice siliconized p200 pipette tip to transfer sections to a glass-bottom dish, then remove NPBST, and apply Aqua-

Hybridization Chain Reaction in Amphioxus

193

Polymount to fill the dish. Use tungsten needles to gently arrange sections ensuring that they are completely flat and lying at the bottom of the dish, and then leave to harden at 4  C for at least 30 min. 10. Cross-talk suppression. This method permits imaging five fluorophores simultaneously in the same specimen. However, we have found cross-talk between fluorophores when detection windows are too wide. To compensate for this, detection windows for each channel should be narrowed to sit exclusively on the emission peak of the fluorophore of interest.

Acknowledgments The authors would like to thank Ben Steventon for encouraging us to develop the HCR protocol in amphioxus; to Christo Christov for technical support to our lab and amphioxus facility, the latter supported by a Sir Isaac Newton Trust Research Grant (Ref. 15.07 (r)); to everybody in the histopathology and imaging facilities at the CRUK-CI; and to Matt Wayland in the imaging facilities at the Department of Zoology, which are supported by a Sir Isaac Newton Trust Research Grant (Ref. 18.07ii(c)). We also acknowledge support from CRUK (C9545/A29580) to EBG, Wellcome Trust Grant (203806/Z/16/A) to TGA, and the Claire Barnes Trust to GG. References 1. Graham A, Papalopulu N, Krumlauf R (1989) The murine and Drosophila homeobox gene complexes have common features of organization and expression. Cell 57(3):367–378 2. Dubrulle J, Pourquie´ O (2004) fgf8 mRNA decay establishes a gradient that couples axial elongation to patterning in the vertebrate embryo. Nature 427:419–422 3. Riddle RD, Johnson RL, Laufer E et al (1993) Sonic Hedgehog mediates the polarizing activity of the ZPA. Cell 75(7):1401–1416 4. Nelson C, Morgan B, Burke AC et al (1996) Analysis of Hox gene expression in the chick limb bud. Development 122:1449–1466 5. Akam M, Martinez-Arias A (1985) The distribution of Ultrabithorax transcripts in Drosophila embryos. EMBO 4:1689–1700 6. Baker NE (1987) Molecular cloning of sequences from wingless, a segment polarity gene in Drosophila: the spatial distribution of a transcript in embryos. EMBO 6:1765–1773

7. McGinnis W, Krumlauf R (1992) Homeobox genes and axial patterning. Cell 68:283–302 8. Benito-Gutie´rrez E` (2011) Amphioxus as a model for mechanisms in vertebrate development. In: eLS (ed). https://doi.org/10.1002/ 9780470015902.a0021773 9. Meuleman D, Bronner-Fraser M (2004) GeneRegulatory Interactions in neural crest evolution and development. Dev Cell 7:291–299 10. Holland ND, Holland LZ (1993) Embryos and larvae of invertebrate deuterostomes. In: Stern CD, Holland PWH (eds) Essential developmental biology, a practical approach. IRL Press, Oxford, pp 21–32 11. Holland PW, Holland LZ, Williams NA et al (1992) An amphioxus homeobox gene: sequence conservation, spatial expression during development and insights into vertebrate evolution. Development 116:653–661 ˜ eiro C, Maeso I et al (2010) 12. Irimia M, Pin Conserved developmental expression of Fezf

194

Toby G. R. Andrews et al.

in chordates and Drosophila and the origin of the Zona Limitans Intrathalamica (ZLI) brain organizer. EvoDevo 1:7 13. Choi HMT, Schwarzkopf M, Fornace ME et al (2018) Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust. Development 145: dev165753 14. Choi HMT, Chang JY, Trinh LA et al (2010) Programmable in situ amplification for multiplexed imaging of mRNA expression. Nat Biotechnol 28:1208–1212 15. Choi HMT, Beck VA, Pierce NA (2014) Nextgeneration in situ hybridization chain reaction: higher gain, lower cost, greater durability. ACS Nano 8:4284–4294 16. Choi HMT, Calvert CR, Husain N et al (2016) Mapping a multiplexed zoo of mRNA expression. Development 143:3632–3637 17. Yu JK, Holland LZ (2009) Amphioxus wholemount in situ hybridization. Cold Spring Harb Protoc 2009:pdb.prot5286 18. Yu JK, Holland LZ (2009) Amphioxus (Branchiostoma floridae) spawning and embryo collection. Cold Spring Harb Protoc 2009:pdb. prot5285 19. Benito-Gutie´rrez E, Nake C, Llovera M et al (2005) The single AmphiTrk receptor highlights increased complexity of neurotrophin

signalling in vertebrates and suggests an early role in developing sensory neuroepidermal cells. Development 132:2191–2202 20. Benito-Gutie´rrez E`, Stemmer M, Rohr SD et al (2018) Patterning of a telencephalon-like region in the adult brain of amphioxus. bioRxiv. 307629 21. Rybak-Wolf A, Solana J (2014) In: Nielsen BS (ed) Whole-Mount in situ hybridization using DIG-labelled probes in planarian in in situ hybridization protocols, methods in molecular biology, vol 1211. Springer, New York, pp 41–50 22. Fuentes R, Ferna´ndez J (2014) Fixation/permeabilization procedure for mRNA in situ hybridisation of zebrafish whole-mount oocytes, embryos and larvae. In: Nielsen BS (ed) In situ hybridization protocols, methods in molecular biology, vol 1211. Springer, New York, pp 1–12 23. Saint-Jeannet JP (2017) Whole-mount in situ hybridization of xenopus embryos. From the xenopus collection, edited by Hazel L. Sive. Cold Spring Harb Protoc. https://doi.org/ 10.1101/pdb.prot097287 24. Benito-Gutie´rrez E, Weber H, Bryant DV et al (2013) Methods for generating year-round access to amphioxus in the laboratory. PLoS One 8:e71599

Chapter 12 RNAscope™ Multiplex Detection in Zebrafish Theresa Gross-Thebing Abstract The RNAscope methodology is a powerful tool to detect RNA expression patterns with high subcellular resolution and possibility for RNA-protein colocalization studies. Presented here is a two-day protocol for robust multiplex detection of up to three different RNAs in zebrafish whole-mount embryos using the RNAscope procedure. Application of the protocol offers the simultaneous detection of multiple RNAs with a high signal-to-noise ratio in an intact embryo. Key words Zebrafish, Whole-mount embryo, RNAscope, In situ hybridization, High-resolution detection, WISH, FISH, RNA localization

1

Introduction Detecting spatiotemporal expression patterns of RNAs in intact or whole-mount embryos is important to understand the processes underlying the development of tissues and organisms. Wholemount in situ hybridization (WISH) is a major tool to detect and study RNA expression patterns in the context of development. The general principle of ISH is based on binding of one or multiple oligonucleotides to the RNA of interest [1]. Subsequently, common techniques label the RNA by chromogenic reactions allowing robust detection of RNA expression patterns [2, 3]. Nevertheless, the enzymatic reaction generates diffusible precipitates, thus reducing the spatial resolution in RNA detection, especially affecting the resolution on subcellular level. In addition, the temperatures used in standard WISH procedures likely diminish protein fluorescence interfering with simultaneous detection of RNA and proteins of interest [2]. Zebrafish is a popular model organism for studying vertebrate development. WISH in zebrafish according to the common protocol enables the simultaneous detection of two different RNAs [4]. Moreover, the development of alternative ISH techniques that are based on the use of short oligonucleotide probes as platforms for assembly of fluorescently labeled DNA offers new

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_12, © Springer Science+Business Media, LLC, part of Springer Nature 2020

195

196

Theresa Gross-Thebing

ways for multiplex detection of RNAs [5, 6]. The RNAscope technology allows the detection of multiple RNAs at high resolution, even on subcellular level. It makes use of 20-bp-long oligonucleotide probes that only hybridize in pairs to their corresponding target RNA to convey specificity. The target probes are linked with tail sequences that allow binding of a self-amplifying DNA scaffold, which provides multiple docking sites for nondiffusible fluorescent labels. The protocol was initially developed for paraffin-embedded tissue specimens [6] and later on modified to detect spatial RNA expression patterns in whole-mount zebrafish embryos [7]. In addition, employing the technique in zebrafish embryos demonstrated high subcellular resolution of RNA detection with low background staining promising a valuable approach to study intracellular RNA expression patterns [7]. Moreover, the protocol preserves the fluorescence and antigenicity of proteins, thus allowing co-detection of proteins and RNAs. The following protocol was originally published by Gross-Thebing et al. [7] and since then has been successfully used in many studies involving RNA detection in zebrafish embryos [8–12]. Beyond its application to the zebrafish model it was extended to other model systems, e.g., chicken embryos or mouse inner ear tissue [13, 14]. The protocol described is to be completed within 2 consecutive days and allows the simultaneous detection of up to three different RNAs.

2 2.1

Materials Reagents

1. Custom-made RNAscope target probes and RNAscope probe diluent (Advanced Cell Diagnostics (ACD), Hayward, CA, USA). 2. RNAscope multiplex detection reagent kit (Amp1, Amp2, Amp3 and Alt A, B, or C, DAPI) (Advanced Cell Diagnostics (ACD), Hayward, CA, USA). 3. RNAscope pretreatment reagent (Advanced Cell Diagnostics (ACD), Hayward, CA, USA). 4. Methanol. 5. Cell culture-compatible glue.

2.2 Buffers and Solutions

1. Danieau’s solution: 17.4 mM NaCl, 0.21 mM KCl, 0.12 mM MgSO4 · 7H2O, 0.18 mM Ca(NO3)2, 1.5 mM Hepes. 2. PBS: 137 mM NaCl, 2.7 mM KCL, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. 3. 0.1% PBS-Tween: 0.1% Tween-20 in PBS. 4. 0.01% PBS-Tween: 0.01% Tween-20 in PBS. 5. 4% PFA in PBS, pH 7.4.

RNAscope Multiplex Detection in Zebrafish

197

6. 0.2 0.01% SSC-Tween: 0.01% Tween-20, 15 mM NaCl, 1.5 mM TriNaCitratdihydrate, pH 7. 7. Hoechst 33342 1:10.000 in 0.2 0.01% SSC-Tween. 8. 1% Low-melting-point agarose in Danieau’s solution. 2.3

1. Glass Pasteur pipettes, pipette pump pipetting device (VWR), and glass cutter.

Equipment

2. Forceps. 3. Metal needle. 4. Water bath. 5. Plastic petri dish, compatible to match stage holders for microscopy. 6. 1.5 mL Tubes. 7. Shaker. 8. Confocal microscope with bottom-down objectives.

3 3.1

Methods Probe Design

The design and generation of specific RNAscope probes targeting a gene of interest are done by ACDBio customer service upon being provided the mRNA accession number with the prefix NM_ (see Note 1). In addition to the accession number the channel for appropriate labeling has to be selected (Table 1). There are up to three different channels available upon request that each provides a modified backbone for multiplex labeling of different channels. A number of catalogue probes for zebrafish targets are available by now. A probe targeting the bacterial encoded gene DapB can be used as negative control. 1. Raise embryos in Danieau’s solution until the developmental stage of interest (see Note 2).

3.2 Preparation of Embryos

2. Thaw 4% PFA in PBS and bring it to room temperature.

Table 1 Label probe combinations for custom-made RNAscope target probes Color module

C1

C2

C3

Amp4 Alt A

Alexa 488 (green)

Atto 550 (red)

Atto 647 (far red)

Amp4 Alt B

Atto 550 (red)

Alexa 488 (green)

Atto 647 (far red)

Amp4 Alt C

Atto 550 (red)

Atto 647 (far red)

Alexa 488 (green)

There are three different channels available that are labeled with fluorescent dyes according to the Alt labeling solution. For example using Amplification solution Alt A labels target probes in channel 2 with Atto 550

198

Theresa Gross-Thebing

Table 2 Fixation times for different developmental stages of zebrafish embryos Developmental stage

Chorion

Fixation time

4-cells to 8 hpf

Yes

4h

10 hpf

Yes

2h

12 hpf

Yes

1.5 h

12–20 hpf

Yes

1h

24 hpf to 4 dpf

No

40 min

Fixation should be carried out at room temperature with the tube positioned on its side while slowly agitating

3. Manually dechorionate embryos older than 24 hpf in a petri dish using forceps. Embryos younger than 24 hpf are dechorionated after fixation. 4. Transfer up to 25–30 embryos to a 1.5 mL tube using a glass Pasteur pipette. 5. Remove the liquid and exchange with 1 mL 4% PFA in PBS. To allow access of the PFA to the embryos, the tube should be positioned on its side aligning the embryos on the sidewall of the tube. Fixation should be performed at room temperature for the time indicated in Table 2 (see Note 2). The optimal fixation times for different stages and specific probes can be further optimized if needed. 6. Following removal of the fixation solution rinse embryos 3 in 1 mL 0.1% PBS-Tween at room temperature. Dechorionate embryos younger than 24 hpf. 7. For storage of embryos until the RNAscope assay rinse embryos 2 in 1 mL 100% methanol and store them at 20  C in 100% methanol overnight or longer. 3.3 Preparation of Wash Buffers

1. Prepare fresh 0.2 0.01% SSC-Tween as the main wash buffer. Each sample requires 18 mL of wash buffer during the assay. 2. Prepare 0.01% PBS-Tween. Each sample needs 3 mL of this buffer during the assay.

3.4 Day 1 (1 h): Probe Hybridization

1. Heat the water bath to 40  C. 2. Warm the probes at 40  C in the water bath for 10 min to dissolve precipitation. Afterwards let the probes cool down to room temperature. Spin down C2 and C3 probes to remove the content from the cap. 3. Mix the target probes of C1 (or probe diluent), C2, and C3 (both come as 50 solutions) in a 1.5 mL tube in a 50:1:1 ratio with a final volume of 50 μL per sample. It is more convenient

RNAscope Multiplex Detection in Zebrafish

199

to remove the cap of the C1 or probe diluent flask and pipette the exact amount of liquid than using the drop dispenser. The probe mix can be reused at least once. 4. Completely remove the methanol from the tube of embryos (see Note 3) and let the embryos air-dry for 10–30 min at room temperature. It is critical to sufficiently, but not completely, remove the methanol. A very thin film of methanol liquid should remain on the embryos after this step. Too much methanol left will interfere with the proteinase K treatment, whereas embryos left dry will harm the integrity of the sample and let the embryos clump together. The drying time will depend on how much methanol is initially removed. 5. Add two drops of proteinase-containing Pretreat solution (see Note 4) by carefully rinsing it down the wall of the tube and incubate the sample for 20 min at room temperature. After the drying step embryos might still stick together, but they usually separate during the following steps in detergent-containing buffer solutions. The staining is not affected by clogging. Do not try to separate the embryos by mixing or harshly tapping the tube. Only tap the tube very slightly to ensure mixing of the embryos in the solution. 6. Stop the digestion by rinsing the samples 3 with 1 mL 0.01% PBS-Tween. Again, carefully rinse the buffer down the wall of the tube. Following the last wash remove as much liquid as possible without touching the embryos. 7. Add 50 μL of probe mix solution prepared in step 2 and incubate the samples overnight at 40  C in a water bath, if protein fluorescence has to be preserved. Again, tap the tube very slightly to ensure mixing of the embryos in the solution. Otherwise, incubate the samples at 50  C. The probe mix can be recovered on the next day and reused at least one more time. 3.5 Day 2 (7 h): RNA Detection and Signal Amplification

During the assay embryo integrity is affected due to minimal fixation time and harsh buffer conditions. It is therefore critical to add all the solutions very carefully to the samples by rinsing them dropwise down the wall of the tubes. The tubes should be kept in an almost horizontal position while adding the buffers and solutions, thus avoiding whirling up the embryos. 1. Recover the probes. The probes can be reused at least one more time. 2. Thaw 4% PFA in PBS and bring it to room temperature. 3. Wash the embryos 3 15 min with 1 mL of 0.2 0.01% SSC-Tween at room temperature. During the washing steps position the tubes on their sides while slowly agitating them on

200

Theresa Gross-Thebing

a shaker to ensure homogeneous treatment of the embryos. Remove as much liquid as possible after the last wash, but do not touch the embryos with the pipette tip. 4. Add 1 mL of 4% PFA in PBS and position the tube on its side. Incubate the sample for 10 min at room temperature. This step is to enhance embryo integrity. Afterwards remove the fixation solution. 5. Repeat the washing step as described in step 3. Starting the last wash transfer Amp 1 solution from 4  C to room temperature. 6. Add two drops of Amp1 solution by carefully rinsing it down the wall of the tube (see Note 5). Tap the tube very mildly to gently mix the embryos and incubate the sample for 30 min at 40  C in a water bath. 7. Repeat the washing step as described in step 3. Starting the last wash transfer Amp 2 solution from 4  C to room temperature. 8. Add two drops of Amp2 solution by carefully rinsing it down the wall of the tube (see Note 5). Tap the tube very mildly to gently mix the embryos and incubate the sample for 15 min at 40  C in a water bath. 9. Repeat the washing step as described in step 3. Starting the last wash transfer Amp 3 solution from 4  C to room temperature. 10. Add two drops of Amp3 solution by carefully rinsing it down the wall of the tube (see Note 5). Tap the tube very mildly to gently mix the embryos and incubate the sample for 30 min at 40  C in a water bath. 11. Repeat the washing step as described in step 3. Starting the last wash transfer Amp 4 Alt solution from 4  C to room temperature. Use the appropriate Alt solution for labeling (Table 1). 12. Add two drops of Amp4 Alt solution by carefully rinsing it down the wall of the tube (see Note 5). Tap the tube very mildly to gently mix the embryos and incubate the sample for 15 min at 40  C in a water bath. 13. Carefully add two drops of DAPI or Hoechst solution. Embryos should be incubated for at least 24 h at 4  C with slow agitation to ensure labeling of all nuclei. Keep the samples in dark at 4  C. Samples can be stored for at least 2 weeks prior to imaging. Optionally, perform immunolabeling of proteins before imaging at this point. 3.6 Embryo Staging for Imaging

1. Melt 1% low-melting-point agarose in Danieau’s and keep it liquid at 37  C in a thermo block. 2. Carefully transfer the embryos from the tube to a petri dish containing 0.01% PBS-Tween using a glass pipette. If the embryos still stick together, they can be gently separated using a metal needle.

RNAscope Multiplex Detection in Zebrafish

201

3. Place a drop of low-melting-point agarose (1–2 cm in diameter) in the middle of a plastic petri dish. Try to keep the drop flat to later allow for the correct imaging distance from objective to embryo. 4. Carefully transfer the embryos into the drop of agarose with as little liquid as possible (see Note 6). Ramp the embryos into their desired position for imaging using a metal needle until the agarose solidifies and maintains the position of the embryo. 5. Fill the dish with PBS (see Note 7). The sample is ready for imaging using confocal microscopes equipped with bottomdown water objectives. In case the agarose is losing its connection to the petri dish, remove the PBS and apply few drops of a glue suitable for cell culture experiments to the border of the agarose, thus refixing it to the dish.

4

Notes 1. RNAscope probes can be designed to specifically target a defined region of the RNA, e.g., the coding region only. 2. Optionally, fix the embryos overnight at 4  C. 3. In case of fluorescent protein co-detection incubate the samples at all stages of the protocol in the dark. 4. Both Pretreat 3 and 4 solutions give the same result. 5. Alternatively, directly pipette 50 μL of the Amp solutions instead of using the solution dispenser. 6. Optionally, transfer the embryos to a tube filled with agarose before mounting them in the drop of agarose with a new glass pipette. This will further minimize the amount of leftover liquid transferred to the sample, but also increases the risk of damage to the embryos due to the extra handling step. 7. PBS-Tween might detach the polymerized drop of agarose from the dish.

Acknowledgment The author would like to thank Azadeh Paksa and Erez Raz for co-developing the protocol and Kim Joana Westerich, Zahra Labbaf, and Katsiaryna Tarbashevich for critically reading the manuscript.

202

Theresa Gross-Thebing

References 1. Femino AM (1998) Visualization of single RNA transcripts in situ. Science 280:585–590. https://doi.org/10.1126/sci ence.280.5363.585 2. Thisse C, Thisse B (2008) High-resolution in situ hybridization to whole-mount zebrafish embryos. Nat Protoc 3:59–69. https://doi. org/10.1038/nprot.2007.514 3. Hauptmann G, Lauter G, So¨ll I (2016) Detection and signal amplification in zebrafish RNA FISH. Methods 98:50–59. https://doi.org/ 10.1016/j.ymeth.2016.01.012 4. Hauptmann G (2001) One-, two-, and threecolor whole-mount in situ hybridization to Drosophila embryos. Methods 23:359–372. https://doi.org/10.1006/meth.2000.1148 5. Choi HMT, Beck VA, Pierce NA (2014) Nextgeneration in situ hybridization chain reaction: higher gain, lower cost, greater durability. ACS Nano 8:4284–4294. https://doi.org/10. 1021/nn405717p 6. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14:22–29. https://doi.org/10. 1016/j.jmoldx.2011.08.002 7. Gross-Thebing T, Paksa A, Raz E (2014) Simultaneous high-resolution detection of multiple transcripts combined with localization of proteins in whole-mount embryos. BMC Biol 12:55. https://doi.org/10.1186/ s12915-014-0055-7 8. Gross-Thebing T, Yigit S, Pfeiffer J et al (2017) The vertebrate protein dead end maintains primordial germ cell fate by inhibiting somatic differentiation. Dev Cell 43:704–715.e5.

https://doi.org/10.1016/j.devcel.2017.11. 019 9. Malhotra D, Shin J, Solnica-Krezel L, Raz E (2018) Spatio-temporal regulation of concurrent developmental processes by generic signaling downstream of chemokine receptors. eLife 7. https://doi.org/10.7554/eLife.33574 10. Reme´dio L, Gribble KD, Lee JK et al (2016) Diverging roles for Lrp4 and Wnt signaling in neuromuscular synapse development during evolution. Genes Dev 30:1058–1069. https://doi.org/10.1101/gad.279745.116 11. Yang W-J, Hu J, Uemura A et al (2015) Semaphorin-3C signals through Neuropilin-1 and PlexinD1 receptors to inhibit pathological angiogenesis. EMBO Mol Med 7:1267–1284. https://doi.org/10.15252/emmm. 201404922 12. Isaacman-Beck J, Schneider V, FranziniArmstrong C et al (2015) The lh3 Glycosyltransferase directs target-selective peripheral nerve regeneration. Neuron 88:691–703. https://doi.org/10.1016/j.neuron.2015.10. 004 13. Morrison JA, McKinney MC, Kulesa PM (2017) Resolving in vivo gene expression during collective cell migration using an integrated RNAscope, immunohistochemistry and tissue clearing method. Mech Dev 148:100–106. https://doi.org/10.1016/j.mod.2017.06. 004 14. Kersigo J, Pan N, Lederman JD et al (2018) A RNAscope whole mount approach that can be combined with immunofluorescence to quantify differential distribution of mRNA. Cell Tissue Res 374:251–262. https://doi.org/10. 1007/s00441-018-2864-4

Chapter 13 Duplex In Situ Hybridization of Virus Nucleic Acids in Plant Tissues Using RNAscope® Samar Sheat, Stephan Winter, and Paolo Margaria Abstract RNAscope has been recently introduced by Advanced Cell Diagnostics (Newark, CA, USA) for in situ hybridization (ISH) of target RNAs using a proprietary technology for probe design and hybridization assay. The method has been extensively used as a basis for sensitive diagnostic assays in the medical field, while applications of this technique in plant sciences are still rare. Here, we describe a multiplex ISH protocol for detection of two plant viruses in formalin-fixed paraffin-embedded tissue sections from cassava. The dual-color protocol described can be used as reference for virus/host interaction studies including the visualization of virus nucleic acids and plant endogenous mRNAs. RNAscope provides a specificity and sensitivity of target detection that otherwise cannot be reached. Key words Chromogenic detection, FFPE tissue, In situ hybridization, Multiplexing, Plant virus

1

Introduction In situ hybridization (ISH) is generally used to localize nucleic acids in tissues and organs and, in plant virology, to study virus invasion, movement, and spread and virus/host interactions at the cellular and subcellular level. ISH protocols to localize plant viruses with DNA or RNA genomes have been developed for several plant/virus combinations [1–8] and rapid methods with simple fixation and hybridization protocols are available to detect viruses in tissues from plants and vector insects [9]. The use of ISH techniques in plant sciences however is often limited due to the limited sensitivity and specificity of the hybridization, hampering the detection of low-abundance RNAs or the discrimination of targets with considerable sequence similarity. RNAscope can overcome these limitations, by using a hybridization assay with a unique probe design that allows simultaneous signal amplification and background suppression, to make detection of low-abundance RNA molecules down to single-molecule level [10] possible. Besides its high specificity and sensitivity, RNAscope offers the possibility of detecting

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_13, © Springer Science+Business Media, LLC, part of Springer Nature 2020

203

204

Samar Sheat et al.

multiple targets with exceptional resolution. At the DSMZ-Plant Virus Department we have been using RNAscope to localize plant virus nucleic acids in Nicotiana rustica and cassava [11], expanding the field of application from maize [12, 13] and citrus [14] to other hosts. In this chapter, we describe a multiplex chromogenic RNAscope assay for detection of two viruses (cassava brown streak virus and East African cassava mosaic virus) in cassava tissues. The duplex assay is based on the simultaneous detection of two targets in FFPE tissues using HRP- and AP-conjugated probes to generate the hybridization signals. Target detection results in sharp dots (eventually appearing as clusters) of distinctly colored chromogen precipitates that can be visualized in bright-field microscopy. The protocol presents a stepwise description of the procedure, from preparation of the plant tissues and sectioning to probe hybridization, amplification, and signal visualization, highlighting critical steps and options for optimal target detection.

2

Materials

2.1 RNAscope Assay Reagents and Samples

1. Duplex target probes in appropriate color channel (C1 and C2), designed and provided by ACD (see Note 1). 2. RNAscope 2.5 HD Duplex Reagent kit (see Note 2). 3. Formalin-fixed paraffin-embedded (FFPE) tissue samples.

2.2 Other Reagents and Materials

1. Neutral-buffered formalin (NBF), 10%. 2. PBS 1 in DEPC-treated water, pH 7.4, autoclaved. 3. Paraplast Plus paraffin. 4. SuperFrost® Plus Slides and glass coverslips. 5. Xylene. 6. Ethanol (30%, 50% 70%, 90% solutions in DEPC-treated water; 100%). 7. DEPC-treated water. 8. 5 SSC buffer. 9. Tubes (various sizes), pipettes, sterile tips (see Note 3). 10. Aluminum foil. 11. Forceps. 12. Adsorbent paper. 13. Glass beakers and bottles. 14. Digital thermometer. 15. ImmEdge™ Hydrophobic Barrier Pen. 16. Gill’s Hematoxylin, 50% solution in distilled water. 17. “Peel-A-Way” embedding blocks.

Duplex RNAscope ISH in Plant Tissues

205

18. “Histobloc” supporting blocks. 19. Glass tubes. 20. Candle and lighter. 2.3

Equipment

1. Vacuum chamber. 2. Sectioning system: Rotary microtome (e.g., Microm HM 355, Thermo Fisher Scientific). 3. Hybridization station: HybEZ™ Oven; HybEZ™ Humidity Control Tray with lid; ACD EZ-Batch™ Slide Rack; HybEZ™ Humidifying Paper (see Note 4). 4. Incubation station: Tissue-Tek® Vertical 24 Slide Rack: TissueTek® Staining Dishes for water solutions; Tissue-Tek® Clearing Agent Dish, xylene resistant. 5. Drying oven, capable of holding temperature at 60  C. 6. Standard light microscope with camera. 7. Stereomicroscope. 8. Heating plate. 9. Water bath. 10. Fume hood.

3

Methods An overview of the different steps of the duplex chromogenic RNAscope ISH assay is shown in Fig. 1.

3.1 FFPE Section Preparation

1. Cut tissue specimens into fragments of 4–5 mm, immerse tissues in fresh 10% neutral-buffered formalin fixative solution, and fix under vacuum pressure for 45 min, followed by fixative exchange and incubation for 45 min. Exchange fixation solution and incubate under vacuum pressure for 16 h at RT. For safety reasons, take care of performing all steps in a fume hood (see Note 5). 2. Wash the tissue with 1 PBS, twice for 15 min each wash, at RT. 3. Dehydrate the samples with a series of increasing ethanol concentrations (30%, 50%, 70%, 90%, 100%) for 30 min at each concentration at RT. Eventually, following incubation in 70% ethanol, samples can be stored at 4  C and processed later. 4. Liquid paraffin is needed for the next steps: melt in advance an appropriate volume of paraffin (approximately 80 g for five samples processed together) in an oven at 60  C. Prepare ethanol/xylene mixtures (2:1, 1:1, 1:2 v/v) in 50 mL glass bottles and store at RT. Prepare xylene/paraffin mixtures (2:1,

206

Samar Sheat et al.

Fig. 1 Summary of the RNAscope duplex chromogenic assay workflow for FFPE samples

1:1, 1:2 v/v) in Falcon tubes and store at 60  C; prepare each mixture separately and use within 30 min. Always use protecting tissue gloves when handling and work under fume hood with xylene solutions. 5. Infiltrate tissues with ethanol/xylene (fresh) mixtures (2:1, 1:1, 1:2 v/v), 45 min for each substitute mixture at RT, followed by incubation in pure xylene for 45 min at RT.

Duplex RNAscope ISH in Plant Tissues

207

6. Incubate samples in xylene/paraffin mixtures (2:1, 1:1, 1:2 v/ v) for 1 h at 60  C in the oven placed in the fume hood, followed by incubation in pure paraffin for 1 h at 60  C. 7. Transfer each sample into a Peel-A-Way mold filled with liquid paraffin using warm forceps and orient it into the desired position. If paraffin solidifies, incubate the mold in the oven and repeat the step after a few minutes. 8. Leave the molds at RT until paraffin solidifies and store them at RT. 9. Remove the mold and attach the paraffin sample block to the “Histobloc” supporting blocks, by quickly flaming the sample on a candle and pressing it onto the block for a few seconds to ensure attachment. Store blocks at 4  C for a couple of hours before cutting; this incubation step facilitates obtainment of thin sections with the microtome. 10. Prepare a water bath with RNase-free water at 37  C. 11. Cut 10–15 μm thick paraffin sections using a rotary microtome, transfer to heat bath at 37  C to allow expansion of the sections, and mount on SuperFrost® Plus Slides. Mounting of samples on SuperFrost® Plus Slides is recommended to ensure that sections attach to slides during the entire protocol. Make sure that sections are completely flat after attachment to the slides, to ensure optimal adhesion. More than one section can be eventually mounted per slide, for further selection of the optimal ones later in the assay. Slides with sections embedded in paraffin can be stored at RT for up to 3 months (see Note 6). 12. To begin the RNAscope assay, bake the desired amount of sections for 1 h at 60  C in a pre-warmed oven (see Note 7). 13. Wash twice in xylene for 5 min each wash, and twice in ethanol 100% for 1 min each wash. 14. Air-dry slides overnight at RT. Following deparaffinization, use slides within 1 week. Before proceeding, slides can be examined using a stereomicroscope, to select the best sections for further processing. Avoid use of slides with sections not fully attached, as they could be lost during the following steps. 3.2

Pretreatment

1. Bake slides in a dry oven at 60  C for 1 h. This baking step is highly recommended just prior to running the RNAscope assay to ensure adhesion of the sections to the slides during the whole procedure and prevent detachment. 2. Cool down slides for 5 min at RT. Create a hydrophobic barrier around the tissue sections using the ImmEdge™ Hydrophobic Barrier pen; let the barrier dry for at least 20 min before proceeding (see Note 8).

208

Samar Sheat et al.

3. Add 3–5 drops of H2O2 (provided within the kit) to entirely cover each section and incubate for 10 min at RT. This step is critical to inhibit endogenous peroxidases and prevent hazy background after detection (see Note 9). 4. Flick the slides on absorbent paper. Immediately insert the slide into a Tissue-Tek® Slide Rack submerged in a Tissue-Tek® Staining Dish filled with distilled water. Wash in distilled water by gently moving the rack up and down for 3–5 times, and repeat once in fresh water. 5. Incubate slides in target retrieval buffer maintained at boiling temperature (100–102  C) for 15 min, by gently submerging the Slide Rack using forceps. This heat-induced retrieval step is critical for breaking cross-links introduced upon formalin fixation (see Note 10). 6. Using forceps, immediately transfer the Tissue-Tek™ Slide Rack into a staining dish containing distilled H2O and wash for 15 s, followed by a washing step in 100% ethanol for 15 s. 7. Dry sections overnight at RT. 8. Check slides using a stereomicroscope, to select the best sections for further processing. 9. Equilibrate equipment by turning on the HybEZ™ oven and setting temperature to 40  C; place a humidifying paper in the HybEZ™ Humidity Control Tray, wet completely with distilled water, cover with lid, and insert into the HybEZ™ oven. Warm the tray at least for 30 min at 40  C before use. Equilibrate Amp reagents at RT, to dissolve eventual precipitates. Take care of gently mixing the reagents immediately before use. 10. Bake slides at 60  C for 30 min to improve attachments of the sections to the slides. 11. Cool down slides for 5 min at RT. Create a new hydrophobic barrier around the tissue sections using the ImmEdge™ Hydrophobic Barrier pen; let the barrier dry for at least 20 min before proceeding. 12. Place slides in the HybEZ™ Slide Rack and add 4–5 drops of Protease Plus (provided within the kit) to entirely cover each section. Place Rack in HybEZ™ Humidity Control Tray, cover with lid, and incubate in the HybEZ™ oven for 15 min at 40  C. This step is critical to permeabilize the samples and provide accessibility of the probes to the target RNA. Optimization of incubation time might be necessary depending on the sample.

Duplex RNAscope ISH in Plant Tissues

209

13. Flick one slide at a time on absorbent paper, and immediately place in Tissue-Tek™ Slide Rack submerged in a Tissue-Tek® Staining Dish filled with distilled water. Wash by gently moving the Slide Rack up and down for 3–5 times; repeat once in fresh water. 14. Flick to remove excess liquid and place slides in the HybEZ™ Slide Rack. 3.3 Duplex RNAscope Assay

1. Pre-warm target and control probes at 40  C for 10 min just before use. 2. Carefully prepare probe mix: C1-channel target probes are ready to use, while C2-channel probes are provided as a 50 concentrated stock. Mix 1:50 ratio of C2 probe to C1 probe by pipetting 1 volume of C2 probe to 50 volumes of C1 probe into a tube. Invert the tube several times to ensure complete mixing. Add an appropriate quantity of probe mix to entirely cover each section (3–4 drops). If residual probe mix is left, mixed probes can be stored at 4  C for up to 6 months, according to the manufacturer. 3. Place the HybEZ™ Slide Rack into the Humidity Control Tray, cover with the lid, and incubate in the HybEZ™ oven for 2 h at 40  C. Do not allow sections to dry during the assay, as this could result in background staining. 4. By taking one slide at a time, quickly remove excess liquid and place slides in Tissue-Tek® Slide Rack submerged in TissueTek® Staining Dish filled with 1 wash buffer. 5. Wash slides in 1 wash buffer for 5 min at RT by gently moving the Slide Rack up and down; repeat washing in fresh 1 wash buffer for 10 min with gentle agitation. The RNAscope assay procedure can be continued or conveniently split over 2 days, by keeping slides in 5 SSC buffer overnight at RT, followed by washing in 1 wash buffer 1–2 times and continuing as described below. 6. Flick slides to remove excess wash buffer and place in the HybEZ™ Slide gentle agitation. Add Amp1 to cover each section. Proceed fast between the Amp incubation steps to avoid that sections dry. 7. Place rack in the HybEZ™ Humidity Control Tray, cover with lid, and incubate in the HybEZ™ oven at 40  C for 30 min. 8. By taking one slide at a time, quickly remove excess liquid and place slides in Tissue-Tek® Slide Rack submerged in TissueTek® Staining Dish filled with 1 wash buffer. Wash slides in 1 wash buffer for 5 min at RT by gently moving the slide rack up and down; repeat once by washing in fresh 1 wash buffer for 10 min.

210

Samar Sheat et al.

9. Repeat procedures described in steps 6–8 using Amp2 and incubating for 15 min at 40  C. 10. Repeat procedures described in steps 6–8 using Amp3 and incubating for 30 min at 40  C. 11. Repeat procedures described in steps 6–8 using Amp4 and incubating for 15 min at 40  C. 12. Repeat procedures described in steps 6–8 using Amp5 and incubating for 15 min at RT. Amp5 incubation step is critical for signal detection and timing might be optimized depending on the sample in study. 13. Repeat procedures described in steps 6–8 using Amp6 and incubating for 15 min at RT. 14. Flick to remove excess wash buffer and put slides in HybEZ™ Slide Rack. 15. Pipette the RED working solution onto each tissue section, ensuring that each section is fully covered (see Note 11). Avoid touching the tissue sections with pipette tip. Place rack in the humidity control tray, close with lid, and incubate for up to 10 min at RT. 16. Flick slides one at a time and immediately insert into a TissueTek® Slide Rack submerged in staining dish filled with 1 wash buffer. Wash slides in 1 wash buffer for 5 min at RT by gently moving the slide rack up and down; repeat washing in fresh 1 wash buffer for 10 min. 17. Repeat procedures described in steps 6–8 using Amp7 and incubating for 15 min at 40  C. 18. Repeat procedures described in steps 6–8 using Amp8 and incubating for 30 min at 40  C. 19. Repeat procedures described in steps 6–8 using Amp9 and incubating for 15 min at RT. Timing of Amp9 incubation step might be optimized depending on the sample in study. 20. Repeat procedures described in steps 6–8 using Amp10 and incubating for 15 min at RT. 21. Flick to remove excess wash buffer and place slides in the HybEZ™ Slide Rack. Pipette the GREEN working solution onto each tissue section, ensuring that sections are entirely covered (see Note 12). 22. Place rack in the HybEZ® Humidity Control Tray, close with lid, and incubate for 10 min at RT. 23. Flick slides one at a time and immediately insert into a TissueTek® Slide Rack submerged in staining dish filled with distilled water. Quickly wash slides for 10–15 s at RT by gentle agitation. Proceed quickly, as green signal may fade if stored in water solution for longer time.

Duplex RNAscope ISH in Plant Tissues

3.4

Visualization

211

1. Incubate Tissue-Tek® Slide Rack in 50% hematoxylin solution for 30 s at RT (see Note 13). 2. Carefully rinse in tapped water, by moving the rack up and down two times. Repeat once in freshwater. 3. Dry slides at 60  C for 45 min in the dry oven. Alternatively, slides can be dried overnight at RT. 4. Cool slides for 5 min at RT. Mount each slide under the fume hood using 1–2 drops of Vectamount media and cover sections with coverslip, taking care to avoid trapping air bubbles. 5. Air-dry for at least 10 min and proceed to visualization using a standard bright-field microscope (see Note 14). The RNAscope protocol using chromogenic detection described above can be adapted for fluorescent detection or combined with immunohistochemistry (IHC), to obtain a more complete and robust picture of the tissues in study (see Note 15).

4

Notes 1. The bioinformatics service from ACD provides support for optimal probe design and appropriate color channel assignment, to ensure specific target detection in multiplex assays. Probes for any gene can be potentially designed. A list of target probes for a number of plantvirus species/strains is already included in the ACD catalog (https://acdbio.com/catalogprobes). Prior to running the RNAscope assays for the first time it is recommended to watch video demonstrations available at https://acdbio.com/technical-support/learn-more. ACD also offers control slides and positive/negative controlprobes for practice purposes. 2. All reagents for the assay are provided within the RNAscope 2.5 HD Duplex Reagent kit. Each kit includes RNAscope H2O2 and Protease Plus Reagents, RNAscope Target Retrieval Reagents, RNAscope Duplex Detection Reagents, and RNAscope Wash Buffer Reagents. 1 Wash buffer can be prepared ahead from the concentrated stock and stored for 1 month at RT; check for precipitates before use. Single kit components might be purchased separately. The reagents have a shelf life of 9 months from the date of bulk manufacturing, when stored as recommended by the manufacturer. Storage temperatures vary for different reagents. Use after expiring date is not recommended. 3. Most of the RNAscope assay reagents are provided in dropper bottles that are ready to use; pipettes and tubes are necessary only to prepare the probe mixture and to dispense the substrate reagents for the chromogenic reactions.

212

Samar Sheat et al.

4. Use of the HybEZ™ Hybridization System to perform the incubation steps provides the optimal conditions for humidity and temperature control to prevent sections from drying out and ensure stable temperature during the incubation steps. 5. Multiple sample types are compatible with RNAscope 2.5 Duplex HD Assays and include formalin-fixed, paraffin-embedded (FFPE) tissues (as in this study); fresh, frozen tissues; fixed, frozen tissues; tissue microarray (TMA); and cell samples. Automation is possible by using instruments for automated dehydration and infiltration and stations for temperaturecontrolled embedding. The Ventana® DISCOVERY XT or ULTRA Systems provide full process automation. 6. Cutting of sections from leaf tissues is generally straightforward; sectioning of stems might be more difficult depending on the plant (herbaceous or woody) and age, and on the amount of woody tissues. For relatively big and woody stems (5–8 mm in diameter) obtainment of such thin sections might be challenging, especially when virus infection compromises tissue integrity (for example, in case of necrotic symptoms). In our experience, sections up to 20–30 μm in thickness could be successfully used for hybridization, without compromising signal detection. Other systems for preparation of tissue sections, such as vibratome or cryostat, could be used; the optimal conditions for tissue preparation should be determined experimentally depending on the type of sample in study. 7. Each RNAscope 2.5 HD Duplex Reagent Kit provides enough reagents to stain ~20 tissue sections, each with an area of approximately 20 mm  20 mm. In case of smaller tissue sections, the number of samples can be increased, without compromising the final result. 8. Make sure not to touch the sample when drawing the barrier. Proper drying of the ink is fundamental to avoid leaking on the sample sections during the process. As ink could be rapidly released from the pen, get familiar with this step by simulation on empty slides before processing the actual samples. 9. A good practice, especially for difficult samples, consists of treating a higher number of samples than the ones expected to be used for probe hybridization, so that a proper amount of sections can be chosen later for the RNAscope assay. Single kit components that might be a limiting factor (such as hydrogen peroxide and Protease Plus) can be purchased separately, based on necessity. 10. Take care of keeping mild boiling conditions, avoiding excessive presence of bubbles, and cover the beaker with aluminum foil. Continuously monitor the temperature using a digital thermometer. It takes around 25 min to reach 100  C on the

Duplex RNAscope ISH in Plant Tissues

213

heating plate; organize timing to be ready on time, avoiding boiling of the target retrieval buffer for more than 30 min before use. Treatment time and temperature might require optimization depending on the plant sample; a decrease of both might be necessary for temperature-sensitive tissues. 11. Depending on the size of the hydrophobic barrier and number of samples, prepare an appropriate amount of RED working solution by using a 1:60 ratio of Red-B to Red-A reagents. Use working solution within 5 min, taking care of avoiding exposure to direct sunlight. Timing of substrate incubation is critical for optimal signal detection and might vary depending on the tissue in study and target concentration. Quick examination of the slides to check for eventual red-color precipitates might be done over the 10 min; in case red precipitate is visible, proceed immediately to the washing step. Incubations as short as 1 min might be sufficient for tissues with high target titer. 12. Depending on the size of the hydrophobic barrier and number of samples, prepare an appropriate volume of GREEN working solution per section by mixing Green-B and Green-A reagents in a 1:50 ratio, respectively. Use working solutions within 5 min, taking care to avoid exposure to direct sunlight. 13. Hematoxylin can be used as counterstaining reagent for better visualization of tissue context. The 50% solution can be prepared ahead and used within 1 week. Longer timing incubation at this step must be avoided to prevent fading of the green signal. 14. Signals are visible as punctate dots of two distinctly colored chromogen precipitates (red and green/blue) within cells. In case of high target concentration, dot clusters are visible. Imaging with different filters depending on microscope filter sets might help to find the optimal conditions for signal visualization. Example images of a duplex assay on cassava tissues are illustrated in Fig. 2. Slides can be stored at RT in a slide box and further imaged later. The RNAscope assay also enables a semiquantitative analysis of the target by estimating the number of punctate dots within each cell. Manual or RNAscope Spot Studio Software-guided analysis could be performed; scoring guidelines for semiquantitative assessments are provided at https://acdbio.com/technical-support/solutions. 15. Supporting information is available in the literature [15] and in the ACD online support at https://acdbio.com/science/ applications/research-solutions/dual-ish-and-ihc.

214

Samar Sheat et al.

Fig. 2 Multiplex detection of cassava brown streak virus isolate CBSV_Mo83 (red signal) and East African cassava mosaic virus Uganda variant (green/blue signal) in leaf sections prepared from mixed-infected cassava. Images were acquired using an Axioskop 2 Plus microscope (Zeiss, Jena, Germany)

Acknowledgments This work was partially funded by the “New Sources for CBSD resistance” project (OPP1113605) of the Bill & Melinda Gates Foundation (Seattle, WA, USA). References 1. Cillo F, Roberts IM, Palukaitis P (2002) In situ localization and tissue distribution of the replication-associated proteins of cucumber mosaic virus in tobacco and cucumber. J Virol 76:10654–10664

2. Gambino G, Vallania R, Gribaudo I (2010) In situ localization of grapevine fanleaf virus and phloem-restricted viruses in embryogenic callus of Vitis vinifera. Eur J Plant Pathol 27:557–570

Duplex RNAscope ISH in Plant Tissues 3. Horns T, Jeske H (1991) Localization of Abutilon mosaic-virus (AbMV) DNA within leaf tissue by in situ hybridization. Virology 181:580–588 4. Lucy AP, Boulton MI, Davies JW et al (1996) Tissue specificity of Zea mays infection by maize streak virus. Mol Plant-Microbe Interact 9:22–31 5. Shargil D, Zemach H, Belausov E et al (2015) Development of a fluorescent in situ hybridization (FISH) technique for visualizing CGMMV in plant tissues. J Virol Methods 223:55–60 6. Morilla G, Krenz B, Jeske H et al (2004) Teˆte a` teˆte of Tomato yellow leaf curl virus and Tomato yellow leaf curl Sardinia virus in single nuclei. J Virol 78:10715–10723 7. Rothenstein D, Krenz B, Selchow O et al (2007) Tissue and cell tropism of Indian cassava mosaic virus (ICMV) and its AV2 (precoat) gene product. Virology 359:137–145 8. Kliot A, Ghanim M (2016) Fluorescent in situ hybridization for the localization of viruses, bacteria and other microorganisms in insect and plant tissues. Methods 98:74–81 9. Ghanim M, Brumin M, Popovski S (2009) A simple, rapid and inexpensive method for localization of Tomato yellow leaf curl virus and Potato leafroll virus in plant and insect vectors. J Virol Methods 159:311–314

215

10. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14:22–29 11. Munganyinka E, Margaria P, Sheat S et al (2018) Localization of cassava brown streak virus in Nicotiana rustica and cassava Manihot esculenta (Crantz) using RNAscope in situ hybridization. Virol J 15:128 12. Bowling AJ, Pence HE, Church JB (2014) Application of a novel and automated branched DNA in situ hybridization method for the rapid and sensitive localization of mRNA molecules in plant tissues. Appl Plant Sci 2:1400011 13. Bowling AJ, Pence HE (2018) Use of butylmethyl methacrylate thin-sections for quantification of mRNA using branched-DNA ISH. Microsc Microanal 24(S1):1382–1383 14. Bergua M, Phelan DM, Bak A et al (2016) Simultaneous visualization of two Citrus tristeza virus genotypes provides new insights into the structure of multi-component virus populations in a host. Virology 491:10–19 15. Wang H, Su N, Wang LC et al (2015) Multiplex fluorescent RNA in situ hybridization via RNAscope. In: Hauptmann G (ed) In situ hybridization methods, vol 99. Humana Press, New York, pp 405–414

Part V Automated Methods for RNA ISH

Chapter 14 Automated ISH for Validated Histological Mapping of Lowly Expressed Genes Charles Pyke Abstract An essential part of the drug discovery and development process in the pharmaceutical industry is to provide a full characterization of cells expressing a given drug target and potential downstream markers in human tissues and in relevant preclinical animal species. This task is best solved by a combination of methods, including histological assessment of target protein and mRNA using immunohistochemistry (IHC) and in situ hybridization (ISH), respectively, as well as non-histology-based methods, such as fluorescence-activated cell sorting (FACS), and single-cell (SCS) or single-nuclei (SNS) sequencing. In reality, this work is often complicated by a combination of low target expression levels and a less than optimal availability of specific reagents for detection. In particular, the ability to specifically detect low-abundance receptor targets using IHC is notoriously difficult, due to a daunting lack of commercially available specific antibodies validated for use in IHC. In the absence of fully validated antibodies and protocols for IHC, the specific detection of target mRNA using ISH is often the only available histological method. A highly sensitive, nonradioactive, automated, and robust ISH method for use on formalin-fixed, paraffin-embedded (FFPE) tissue sections is presented for assessing histological localization of mRNA transcripts of lowly expressed genes. Key words RNAscope, Ventana Discovery Ultra, Nonoverlapping probes, Formalin-fixed paraffinembedded tissue, ZZ-pairs, GLP-1R

1

Introduction The task of mapping the cell type-specific expression of drug target genes is crucially important for understanding aspects of both efficacy and safety of any drug that is intended for human clinical use, and ideally should include, for each species, a description of all cell types expressing the target receptor as well as its relevant expression levels in each individual cell type across all organs where expression is seen. If at all possible such data should also include in vivo target engagement data, e.g., binding of I-125-labeled drug to visualize functional target in relevant organs using in situ ligand binding (ISLB) or receptor activation markers,

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_14, © Springer Science+Business Media, LLC, part of Springer Nature 2020

219

220

Charles Pyke

typically using IHC detection of phosphorylated intracellular downstream mediators of receptor activation. In many laboratories, IHC is the preferred or only method used for histological assessment of cell type-specific expression of genes of interest. This is likely due to the ease of setting up IHC protocols, and due to the abundance of antibodies with apparent specificity for almost any protein of interest. However, when the specificity of antibodies is closely investigated, we and others have discovered that many antibodies are not suited for IHC. In particular, the detection of low-abundance targets of the GPCR (G protein-coupled receptor) class of drug targets using IHC is difficult, and for many GPCRs no reliable data are available to describe their tissue localization [1–4]. Enrichment of cell types of interest combined with Western blotting and qPCR for fulllength mRNA transcript represents an alternative route to obtain spatial expression data [5]; however, for low-abundance targets this approach will often not generate sufficiently sensitive data. Even with the optimal IHC protocol established and validated for a given organ, there can be species- and organ-selective problems with unspecific signals due to cross-reactivity with nontarget epitopes. For this reason, ISH validation of IHC data is always important to consider when conducting histological expression profiling across tissues and organs. A frequently raised concern with ISH is that there is not necessarily a linear correlation between mRNA and protein expression. While this is undoubtedly a potential confounder in certain phenotypic states of some cell types and for some rare gene classes, e.g., genes that are very acutely regulated such as immediate-early genes, it is likely not a widespread phenomenon for the vast majority of genes with normal turnover rates and in steady-state conditions of transcription and translation [6]. In fact, in our own extensive work with histological expression profiling of GLP-1 receptor (GLP-1R), a GPCR-class drug target, across multiple organs and species and using a combination of IHC, ISH, and ISLB, we have only seen discordant (but specific) signal patterns in brain regions where GLP-1R IHC-positive neuronal fibers can be located at a significant distance from GLP-1R mRNA containing neuronal cell bodies [7, 8]. In all other instances, a linear relationship between GLP-1R mRNA and protein was detected (unpublished data). In the recent years, single-cell and single-nuclei transcriptomic profiling technologies have become available as a research tool in both academia and the industry, and with a potential for dramatically shifting paradigms for cell type-specific effects in normal physiology and in disease states [9–11]. Given the computational and technological challenges of this group of breakthrough technologies, there is a high need for histological validation of key findings from such studies, and for this multiplexing ISH should be considered the superior histological method.

Automated In Situ Hybridization

221

ISH protocols for tissue localization of specific mRNA have been available for 50 years [12], but these assays can be demanding to establish for low-abundance transcripts, since this typically required broad competences within molecular biology and the use of S-35 or P-33 riboprobes with strict requirements for maintaining RNase-free conditions during all protocol phases [13]. The advent of commercial methods for both manual and automated ultrasensitive ISH has markedly increased the accessibility of mRNA detection in tissue sections, and one of these platforms (RNAscope from ACD/Biotechne https://acdbio.com/) has gained widespread popularity as a robust and ultrasensitive method for specifically detecting any mRNA transcript of interest [14]. The RNAscope reagent platform consists of specific probes and sets of detection reagents, and can be used with both manual and automated protocols. Using a proprietary design of specific probe pairs (“ZZ pairs”), each creating a 28 bp binding site for preamplifier molecules, a pool of ca. 20 ZZ pairs is combined to target ca. 1000 bases of mRNA transcript region. The hybridized ZZ pairs get further amplified during multistage incubation steps, to achieve a specificity and sensitivity comparable to radioactively labeled riboprobes but with no requirement for molecular biology competences or lengthy autoradiography steps, and with a better resolution than can be achieved with radioactive probes. If optimally set up, the method can be used to specifically localize a single mRNA transcript at single-cell resolution and it can be used for multiplexed expression analysis (including using a combination of ISH and IHC). Signals can be quantitated and heatmapped using state-of-the-art image analysis software packages, e.g., from Visiopharm, Indicalabs, or Leica Biosystems. In summary, given the difficulty of validating IHC methods for detection of low-abundance targets, including many GPCRs, ISH should be considered the method of choice for histological assessment of expression profiles of genes for which no fully validated IHC assay is available, as well as for validating single-cell transcriptomic data.

2 2.1

Materials Equipment

1. Automated tissue processor, e.g., Leica APS300 (https://www. leicabiosystems.com/histology-equipment/tissue-processors/ products/leica-asp300-s/downloads/). 2. Microtome, e.g., Leica RM2235 (https://www.leicabiosystems. com/histology-equipment/microtomes/products/leicarm2235/downloads/). 3. Drying oven, capable of holding temperature at 60  C. 4. Automated slide stainer:

222

Charles Pyke

(a) Ventana Discovery Ultra or XT (https://diagnostics. roche.com/global/en/products/instruments/discov ery-ultra.html) (b) Leica Bond RX (https://www.leicabiosystems.com/ihcish-fish/fully-automated-ihc-ish-instruments/bond-rx/) 2.2

Reagents

2.2.1 FFPE Tissue Preparation

1. 10% Neutral-buffered formalin (NBF). 2. Paraffin wax. 3. 100% Ethanol, ACS grade or equivalent (prepare 99%, 96%, 70%). 4. Xylene. 5. Water bath. 6. SuperFrost® Plus slides.

2.2.2 Automated RNAscope on Ventana Discovery Ultra (See Note 1)

The RNAscope VS Universal AP Reagent Kit (Red) automated workflow requires the following: From Ventana Medical System (ventanadiscovery.com): 1. mRNA RED Detection Kit. 2. mRNA Sample Prep Kit. 3. mRNA RED Probe Amplification Kit. 4. 250 Test Probe Dispenser User-Fillables 1-25. From ACDbio/Biotechne (https://acdbio.com/rnascope% C2%AE-vs-universal-hrp-ap): 1. RNAscope 2.5 VS Target Probes (Catalog or Made-to-Order VS Probes (see Subheading 3.2 below)). 2. RNAscope 2.5 VS Control Probes (select species-specific VS positive control probes or packs of probes and VS negative control probes (see Subheading 3.2 below)). 3. RNAscope Control Slides (optional). 4. RNAscope VS Universal AP Reagent Kit (Red) (Table 1).

2.3

3

Software

Visiopharm (https://www.visiopharm.com/), HALO (http:// www.indicalab.com/), or Aperio (https://www.leicabiosystems. com/digital-pathology/analyze/ish-fish/aperio-rna-ish-algorithm/) software packages.

Methods

3.1 FFPE Tissue Preparation Using Immersion Fixation (See Note 2)

Remove sample and cut in smaller pieces (not exceeding 3–4 mm in any direction). Immersion-fix sample in freshly prepared 10% NBF for 16–32 h at room temperature. On tissue processor, dehydrate and embed samples in paraffin. Trim paraffin blocks as needed, then

Automated In Situ Hybridization

223

Table 1 Reagents needed for the full protocol Materials provided by Advanced Cell Diagnostics

Materials provided by Ventana™ Medical Systems

Other materials and equipment

RNAscope 2.5 VS Target Probe RNAscope 2.5 VS Positive Control Probe RNAscope 2.5 VS Negative Control Probe RNAscope 2.5 VS Pretreat 2–Dewax RNAscope 2.5 VS mRNA Pretreat 3-Protease RNAscope 2.5 VS Target Retrieval (Option 10) RNAscope 2.5 VS AMP1 RNAscope 2.5 VS AMP 2 RNAscope 2.5 VS AMP 3 RNAscope 2.5 VS AMP 4 RNAscope 2.5 VS AMP 5–RED RNAscope 2.5 VS AMP 6–RED RNAscope 2.5 VS AMP 7 RNAscope VS Hematoxylin RNAscope VS Bluing Reagent

DISCOVERY™ ULTRA— automated slide stainer DISCOVERY Wash Buffer 10 ULTRA LCS (Predilute) DISCOVERY SSC 5 RiboWash Buffer 10 DISCOVERY CC1 Reaction buffer (10) Probe dispensers mRNA Pretreatment Kit mRNA Probe Amplification Kit mRNA Red Detection Kit User fillable dispensers

Distilled water Dawn detergent or similar detergent Fume hood Xylene Tissue-Tek® Staining Dish (1) Tissue-Tek® Clearing Agent Dish, xylene-resistant (1) Tissue-Tek® Vertical 24 Slide Rack EcoMount/Pertex Cover Glass, 24 mm  50 mm

Reagents needed

cut 5 μm sections using a microtome, and mount onto SuperfrostPlus glass slides. Air-dry at room temperature and use within 2 weeks (see Note 3). 3.2

Probes

For each probe the company will provide information about the number of ZZ pairs and the mRNA transcript region targeted by the probe, although the exact RNAscope ZZ pair sequences are proprietary. Each probe is sufficient for staining ~30 standard slides and has a shelf life of 2 years. The optimal setup requires the following probes applied to adjacent sections from each tissue sample studied: 1. Species-specific RNAscope VS Target Probe for each gene of interest (see Note 4). These probes, if available, can be directly ordered from the suppliers catalog (https://acdbio.com/cata log-probes), or, alternatively, can be custom ordered: https:// acdbio.com/target-probes-made-order. 2. Positive control probe (see Note 5). 3. Negative control probe (see Note 5). 4. Nonoverlapping target-specific probe (optional, see Note 6).

224

Charles Pyke

5. Target-specific probe followed by IHC protocol (or only IHC protocol if running manual IHC assay), for specificity validation of antibody reactivity in this particular organ and species (optional, see Note 7). 3.3 Automated ISH (See Note 1)

Follow instructions provided by supplier (RNAscope 2.5 VS Assay For Ventana DISCOVERY™ ULTRA System (https://acdbio. com/technical-support/user-manuals)). 1. Prepare the materials: (a) Prepare the instrument. (b) Dilute bulk reagents. (c) Register new reagents. (d) Prepare instrument reagents. (e) Create instrument protocol (note: including selecting appropriate assay condition for tissue types and probes, https://acdbio.com/referenceguide) (f) Print labels. 2. Run the RNAscope 2.5 VS Assay (~10 h): (a) Load reagents. (b) Start run. (c) Prepare detergent. (d) Prepare dehydrating reagents. (e) Complete run. (f) Wash and mount slides. 3. Evaluate the results: It is important to assess each protocol by examining sections at the microscope for (a) Tissue and cell morphology (b) Positive control signal strength (c) Negative control background (d) Target probe signal

3.4 Quantitative Image Analysis

4

Follow instructions provided by supplier of image analysis software.

Notes 1. The example shown is using the Ventana Discovery Ultra equipment with the RNAscope 2.5VS-RED reagent platform. For full details of this workflow and for details of other platforms and staining options, including manual assays, user

Automated In Situ Hybridization

225

manuals are available at https://acdbio.com/technical-sup port/user-manuals. A specialized application by the same company termed BaseScope permits the detection of very short sequences and is useful, e.g., for detecting splice variants. This application is not covered here but more information can be found at the ACD’s homepage (https://acdbio.com/ basescope%E2%84%A2-vs-reagent-kit-%E2%80%93-red-0). 2. When working with rodent tissue samples, it is often a possibility to prepare perfusion-fixed (i.e., fixative solution is perfused via vascular system) tissues, and by this increase the content and quality of mRNA in the samples while at the same time ensuring that gradients of signal intensity resulting from fixative diffusion artifacts are minimal. This is particularly important when there is a need to work with whole organs, i.e., when working with samples larger than 3–4 mm in any direction. When the task is to compare organ- and cell type-specific expression patterns and mRNA levels between archival human tissue samples and preclinical animal species the use of immersion-fixed samples throughout is warranted. 3. Although other sources recommend longer storage of cut FFPE sections for ISH analysis, we typically see deterioration of signal after 2 weeks of storage, similarly to previously reported data [15]. 4. The species-specific detection of gene expression is best ensured by ordering probes that are directed to each of the species investigated, although cross-hybridization is typically seen across closely related species and genes. For instance, we consistently find mouse and human specific probes to crosshybridize with rat and monkey mRNA transcripts, respectively. As a rule of thumb, >90% homology between species will typically give rise to cross-hybridization. Note that it is often possible to order custom-designed probes that are specifically designed to hybridize with the same efficiency to homologous genes from two or more species [16]. 5. Household genes such as PPIB and POL2R are well suited as positive controls when validating protocols and for sample qualification, and the bacterial gene DapB is a much used negative control. When the goal is to ensure that the protocol does not yield genomic DNA hybridization, a probe for a pan-endothelial marker such as CD31/Pecam1 can be of value. Using this approach, a successful result is the localization of signal in endothelial cells only, with no indications of hybridization to nuclei outside of the vascular compartment (Fig. 1d, e).

226

Charles Pyke

Fig. 1 Examples of ISH and IHC results obtained when analyzing FFPE samples from dog (a–e) and cynomolgus monkey (f–o) for expression of GLP-1 receptor (GLP-1R), a lowly expressed gene belonging to the GPCR class of receptors. RNAscope signals obtained with negative control probe DapB (probe #312039, a) and with two nonoverlapping probes for dog GLP-1R (probe #428959 b) and (probe #486019 c) in serial sections from proximal duodenum from dog. Note identical signal for GLP-1R with the two nonoverlapping probes in Brunner’s gland epithelial cells, with no expression in the overlying absorptive epithelium. RNAscope signal for dog CD31/Pecam1 (probe #434439) in endothelial cells only in sample of dog prostate, with no background in any other cell type, and no evidence of genomic DNA hybridization (d). Note the total absence of GLP-1R signal in adjacent section incubated with probe #428959. f–j are adjacent sections from FFPE sample of cynomolgus monkey proximal duodenum, with IHC for GLP-1R (antibody clone 3F52 [17], f) and isotype control (g), and ISH for GLP-1R (probe #449299, h), DapB (probe #312039, i), and PPIB (probe #313909, j). Note full correlation between GLP-1R mRNA and protein in f and h, respectively. k–o are serial sections from FFPE sample from prostate from same animal as in f–j, and using same IHC and ISH protocols as for f–j: IHC for GLP-1R (k) and isotype control (l), and ISH for GLP-1R (m), DapB (n), and PPIB (o). Note the absence of GLP-1R mRNA signal (m), despite the pronounced IHC signal for GLP-1R throughout the epithelial compartment (k). Scale bars: 100 μm (a–c and f–j) and 50 μm (d–e and k–o)

Automated In Situ Hybridization

227

6. The use of a set of two nonoverlapping probes on adjacent sections represents the highest level of validation of positive signal to essentially rule out that cross-hybridization to parts of a closely related mRNA transcript has occurred. If two nonoverlapping probes containing the same number of ZZ pairs give rise to identical staining patterns and signal intensity levels, it is highly unlikely that nonspecific signals from individual ZZ pairs are contributing to the total signal. This approach should be considered when there is a concern for cross-hybridization to closely related mRNA transcripts and in situations where the observed ISH signal is biologically difficult to explain. For this approach, at least one of the probes, but often both probes, must be custom designed, so that the ZZ pairs in each probe hybridize to nonoverlapping parts of the mRNA transcript in question (see Fig. 1a–c for example). 7. As noted in Subheading 1, even with optimal IHC protocols there can be species- and organ-selective unspecific staining issues due to cross-reactivity with nontarget epitopes. ISH validation of IHC signal can be obtained by performing double labeling on one section or single labeling for ISH and IHC on two adjacent sections. As an example, Fig. 1f–j shows linear correlation between GLP-1R IHC (F) and ISH (H) signals in Brunner’s gland epithelium in monkey proximal duodenum, whereas Fig. 1k–o shows lack of correlation between GLP-1R IHC (K) and ISH (M) signals in monkey prostate, strongly indicating that the strong immunoreactivity localized in secretory epithelial cells in K is not specific for GLP-1R.

Acknowledgments The author would like to thank Pia G. Mortensen, Bettina Brandrup, and Joan H. Rasmussen at Novo Nordisk A/S for technical expertise. References 1. Pyke C, Knudsen L (2013) The glucagon-like peptide-1 receptor--or not? Endocrinology 154:4–8 2. Reubi JC (2014) Strict rules are needed for validation of G-protein-coupled receptor immunohistochemical studies in human tissues. Endocrine 47:659–661 3. Drucker DJ (2016) Never waste a good crisis: confronting reproducibility in translational research. Cell Metab 24:348–360

4. Gautron L (2019) On the necessity of validating antibodies in the immunohistochemistry literature. Front Neuroanat 13:46 5. Yusta B, Baggio LL, Koehler J et al (2015) GLP-1R agonists modulate enteric immune responses through the intestinal intraepithelial lymphocyte GLP-1R. Diabetes 64:2537–2549 6. Popovic D, Koch B, Kueblbeck M et al (2018) Multivariate control of transcript to protein variability in single mammalian cells. Cell Syst 7:398–411

228

Charles Pyke

7. Heppner KM, Kirigiti M, Secher A et al (2015) Expression and distribution of glucagon-like peptide-1 receptor mRNA, protein and binding in the male nonhuman primate (Macaca mulatta) brain. Endocrinology 156:255–267 8. Jensen CB, Pyke C, Rasch MG et al (2018) Characterization of the glucagonlike peptide1 receptor in male mouse brain using a novel antibody and in situ hybridization. Endocrinology 159:665–675 9. Campbell JN, Macosko EZ, Fenselau H et al (2017) A molecular census of arcuate hypothalamus and median eminence cell types. Nat Neurosci 20:484–496 10. Skene NG, Bryois J, Bakken TE et al (2018) Genetic identification of brain cell types underlying schizophrenia. Nat Genet 50:825–833 ˇ ilionis R, Choo-Wing R et al 11. Plasschaert LW, Z (2018) A single-cell atlas of the airway epithelium reveals the CFTR-rich pulmonary ionocyte. Nature 560:377–381 12. Gall JG, Pardue ML (1969) Formation and detection of RNA-DNA hybrid molecules in cytological preparations. Proc Natl Acad Sci U S A 63:378–383

13. Pyke C, Salo S, Ralfkiaer E et al (1995) Laminin-5 is a marker of invading cancer cells in some human carcinomas and is coexpressed with the receptor for urokinase plasminogen activator in budding cancer cells in colon adenocarcinomas. Cancer Res 55:4132–4139 14. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14:22–29 15. Lisowski AR, English ML, Opsahl AC et al (2001) Effect of the storage period of paraffin sections on the detection of mRNAs by in situ hybridization. J Histochem Cytochem 49:927–928 16. Christoffersen BØ, Skyggebjerg RB, Bugge A et al (2019) Long-acting CCK analogue NN9056 lowers food intake and body weight in obese Go¨ttingen Minipigs. Int J Obes. https://doi.org/10.1038/s41366-019-03860 17. Pyke C, Heller RS, Kirk RK et al (2014) GLP-1 receptor localization in monkey and human tissue: novel distribution revealed with extensively validated monoclonal antibody. Endocrinology 155:1280–1290

Chapter 15 Automation of Multiplexed RNAscope Single-Molecule Fluorescent In Situ Hybridization and Immunohistochemistry for Spatial Tissue Mapping Kenny Roberts and Omer Ali Bayraktar Abstract In situ transcriptomic methods hold immense promise for spatially resolved mapping of cell types across human tissues. Here, we describe a protocol for automated single-molecule fluorescent in situ hybridization (smFISH) on standard histology sections at high throughput. We focus on the RNAscope smFISH assay that combines branched DNA amplification with tyramide signal amplification (TSA) to obtain high signalto-noise ratio for tissue imaging. We describe the use of the robotic Leica BOND RX system for automation of liquid handling and the combination of the RNAscope assay with TSA-based immunohistochemistry without the need for specialized demultiplexed imaging. Key words Single-molecule fluorescent in situ hybridization, smFISH, RNAscope smFISH, Automated smFISH, Leica BOND RX

1

Introduction The remarkable cellular diversity of the human body enables our complex biology and underlies our vulnerability to disease. Our ability to identify the genes and molecular pathways that distinguish human cell types is essential for studying their roles in development, health, and illness. Transcriptomic studies can now reveal molecular signatures of human cell types and gene expression changes in disease at the single-cell resolution [1–4], but our ability to map these gene expression patterns back into the intact human organs is limited. In situ validation of transcriptomics data can indicate more precisely the level of cellular diversity, resolve spatial relationships and interactions between different cells, and reveal the functional organization of human organs [5]. Single-molecule fluorescent in situ hybridization (smFISH) allows highly sensitive quantification of gene expression at the single-cell level [6–9]. There are several approaches to amplify the

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_15, © Springer Science+Business Media, LLC, part of Springer Nature 2020

229

230

Kenny Roberts and Omer Ali Bayraktar

smFISH signal to achieve high signal-to-noise ratio (SNR) desirable for imaging tissues including branched DNA amplification [10], hybridization chain reaction [11], and probe concatemerization via the exchange reaction [12]. Here, we focus on the application and automation of the RNAscope LS Multiplex Assay (Advanced Cell Diagnostics, Bio-Techne) that combines the use of branched DNA amplification [13] with tyramide signal amplification (TSA) to achieve very high SNR for robust smFISH analysis. In the RNAscope LS Multiplex Assay, target RNAs are initially hybridized to a series of single-stranded DNA “z-probe” pairs (Fig. 1). Each z-probe is composed of (1) a 18–25-nucleotide segment complementary to the target RNA, (2) a spacer sequence, and (3) a 14-nucleotide tail sequence. These probes are labeled by branched DNA-amplification trees: double-z-probes are hybridized to oligo-preamplifiers, across their bridged tail sequences, which are then tagged by 20 oligo-amplifiers. Each oligo-amplifier is then labeled with 20 horseradish peroxidase (HRP) enzyme molecules. Most often, a 1 kilobase region on the target transcript is hybridized by 20 z-probe pairs in tandem that can yield up to 8000 HRP labels per each target. The fluorescent smFISH signal is consequently generated by the addition of tyramide-conjugated fluorophores. Tyramide is enzymatically converted into a highly oxidized intermediate by HRP that covalently binds to the proteins at or near the HRP label, depositing a large number of fluorophores for probe detection [14]. Here, we describe a protocol to automate RNAscope LS Multiplex smFISH assay combined with TSA-based immunohistochemistry on standard histology sections at high throughput [15]. We outline the steps to plan experiments using the robotic Leica BOND RX system, pretreat histology sections, and perform RNAscope smFISH and IHC on tissue sections. We present a protocol to multiplex four RNAscope targets, one antibody target, and nuclear marker staining to perform six-color imaging without the need for specialized deconvolution microscopy (Fig. 2). Automation of RNAscope and IHC not only reduces the approximately 18 h protocol to less than two hours of user preparation, but also increases throughput, reproducibility, and documentation of experiments.

2

Materials All solutions should be prepared using sterile ultrapure water (prepared by purifying deionized water, to attain a resistivity of 18 MΩ cm at 25  C, and then autoclaving at 121  C for 15 min). Prepare and store all reagents at room temperature (unless indicated otherwise). Take appropriate health and safety precautions including use of a fume cupboard and personal protective

Automated Multiplexed RNAscope smFISH

231

Fig. 1 Schematics illustrating RNAscope signal amplification. (Top) Multiple rounds of branched DNA amplification, in combination with sequential channel (gene)-specific tyramide signal amplification, are responsible for the exemplary signals of RNAscope staining. Following the building of “amplification trees”, decorated with horseradish peroxidase (HRP) molecules, fluorophores are bound to the cell infrastructure in the proximity of the RNA target molecule via tyramide signal amplification (TSA). (Bottom) In order to use diverse fluorophores, one probe channel is developed via TSA-conjugated biotin, to which streptavidin-conjugated fluorophores may then bind

equipment, especially when using paraformaldehyde. Diligently follow all waste regulations when disposing of waste materials, especially paraformaldehyde, ethanol, and BOND Dewax Solution. 2.1 Manual Pre-staining Treatment of Fixed Frozen Sections

1. SuperFrost Plus glass slides. 2. Laboratory oven, capable of reaching 65  C. 3. Vertical slide rack, such as Tissue-Tek Slide Rack.

232

Kenny Roberts and Omer Ali Bayraktar

Fig. 2 Conjugation of antibody staining to fluorescent detection via multiple platforms of amplification. Following binding of the primary and secondary antibody, the HRP conjugated to the latter conducts tyramide signal amplification via tyramide-conjugated biotin, to which streptavidin-conjugated fluorophores are recruited as above

4. Five glass Coplin-style staining jars (for 8 slides), or TissueTek Staining Dishes (for 24 slides). 5. 4% Paraformaldehyde (PFA) in PBS: Prepare an 8% stock by dissolving 80 g solid PFA in 800 mL sterile ultrapure water at 60  C with stirring and the addition of ~15 drops of concentrated sodium hydroxide; cool on ice, add 100 mL 10 PBS, and make the volume up to 1 L. Store aliquots at 20  C; thaw, dilute twofold in PBS, and chill at 4  C just prior to use (see Note 2). 6. 100% Ethanol (see Note 3). 7. 50% Ethanol: Mix 25 mL 100% ethanol with 25 mL sterile ultrapure water. Prepare fresh in a staining jar prior to use (depending on the size of the jar and position of the tissues on the slide, a larger volume may be needed). 8. 70% Ethanol: Mix 15 mL 100% ethanol with 35 mL sterile ultrapure water. Prepare fresh in a staining jar prior to use (depending on the size of the jar and position of the tissues on the slide, a larger volume may be needed). 2.2 In Situ Hybridization and Immunohistochemical Staining Using the Leica BOND RX

1. RNAscope LS Multiplex Fluorescent Reagent Kit (ACD, Bio-Techne). 2. RNAscope LS 4-Plex Ancillary Kit (ACD, Bio-Techne). 3. RNAscope BOND Wash Solution: Prepare 1 working solution by diluting RNAscope BOND Wash Solution 10 Concentrate (Leica Biosystems AR9590) tenfold in sterile ultrapure water. Store at 4  C when not in imminent (next 48 h) use. 4. 100% Ethanol. 5. Epitope retrieval solution 2 (ER2) (Leica Biosystems AR9640). Store at 4  C when not in imminent (next 48 h) use. 6. BOND Dewax Solution (Leica Biosystems AR9222). This reagent is required only for FFPE sections (see Note 4).

Automated Multiplexed RNAscope smFISH

233

7. Opal fluorophores (all Akoya Biosciences): Opal 520 (FP1487001KT); Opal 570 (FP1488001KT); Opal 620 (FP1495001KT); Opal 650 (FP1496001KT); Opal 690 (FP1497001KT). As per the manufacturer’s instructions, reconstitute each vial in 75 μL of DMSO (provided) and store at 4  C. Depending on the imaging system to be used, some of these dyes may not be required (see Note 1). 8. TSA® Plus Biotin Kit (Akoya Biosciences) and streptavidinconjugated fluorophores: Streptavidin-Atto 425 (Sigma); Streptavidin-Atto 490LS (ATTO-TEC). Reconstitute the vial of TSA-biotin in 300 μL DMSO and store at 4  C. The streptavidin-conjugated fluorophores, provided lyophilized, should be reconstituted to 1 mg/mL in sterile ultrapure water, and stored at 20  C in small aliquots. Aliquots may be stored at 4  C short term (weeks). Depending on the imaging system to be used, these reagents may not be required (see Note 1). 9. DAPI (40 ,6-diamidino-2-phenylindole, dihydrochloride): A preparation of DAPI is provided in the RNAscope LS Multiplex Fluorescent Reagent Kit, but it may be preferable to trial different dilutions to optimize the often excessively bright staining provided by DAPI: prepare a 5 mg/mL stock of DAPI (e.g., Thermo Fisher D1306) in sterile ultrapure water. Small aliquots should be stored at 20  C. Aliquots may be stored at 4  C short term (weeks). 10. Biotin and avidin (Sigma): A 50 mg/mL biotin stock solution should be produced by dissolving biotin in sterile ultrapure water, with the addition of minimal drops of concentrated ammonium hydroxide (biotin is poorly soluble in water alone). Produce a 2 mg/mL stock of avidin in sterile ultrapure water. For both reagents, store small aliquots at 20  C; aliquots may be stored at 4  C short term (weeks). These reagents are required only if two streptavidin-conjugated fluorophores will be used. This will depend upon the imaging system (see Note 1). 11. Chicken anti-NeuN primary antibody (Merck). Store at 4 C. 12. HRP-conjugated goat anti-chicken IgY secondary antibody (Thermo Fisher): Store aliquots at 20  C in small aliquots. Aliquots may be stored at 4  C short term (weeks). 13. Antibody diluent/block (Perkin Elmer): Store at 4  C. 2.3 Mounting and Imaging

1. ProLong Gold Antifade Mountant (Thermo Fisher). 2. Glass coverslips, thickness #1.5. 3. Fluorescent imaging system (see Note 1).

234

3

Kenny Roberts and Omer Ali Bayraktar

Methods The procedures described in the following Subheadings 3.1–3.4 are also summarized in checklist form in Subheading 3.5.

3.1 Leica BOND RX Setup

Please refer also to the manufacturers’ instructions for both the RNAscope kits and the Leica BOND RX. 1. Create pretreatment and staining protocols on the Leica BOND RX. Example protocols for 4-plex RNAscope smFISH utilizing TSA-biotin and fluorescent immunostaining are shown in Tables 1 and 2, respectively. 2. Register 30 mL Open Containers for each RNAscope reagent provided in the Multiplex Fluorescent Reagent Kit and 4-Plex Ancillary Kit, except Multiplex TSA Buffer, which will be used to dilute fluorophores. Clearly mark onto each Container which reagent is contained within (transferring the label from the reagent bottle will preserve maximum details, including expiry date, but do not obscure the barcode on the front or top of a Container). 3. Register Titration Containers for each remaining RNAscope reagent, including all user-supplied reagents: RNAscope probes, fluorophores, TSA-biotin, DAPI, blocking solution, antibodies, avidin, and biotin. It is recommended that these containers are given generic names to maximize flexibility of use; for example “TSA-fluorophore 1” instead of “Opal 520” or “Opal 570” allows the user to select these or other different fluorophores in alternative runs of the same protocol.

3.2 Experimental Design

1. For each experiment, calculate the required volumes of reagents (RNAscope probes, fluorophores, TSA-biotin, DAPI, blocking solution, antibodies, avidin, and biotin). For RNAscope probes, allow 500 μL per slide, plus a void volume of 300 μL. For all other reagents, allow 300 μL per slide, plus a void volume of 500 μL. Working dilutions may be found in Table 3. The number of runs and slides avilable will be dependent upon the exact protocols used (see Notes 5 and 6).

3.3 Manual Pre-staining Sample Preparation

1. FFPE sections should be cut at 5–10 μm thickness onto SuperFrost Plus slides, dried by baking at 45–50  C overnight, and stored at room temperature in a sealed slide box with desiccant. The remaining steps of this section apply only to fixed frozen tissues. No manual pretreatment steps are required for FFPE sections. This section should be intercalated, time-wise, with the steps of Subheading 3.4, such that slides and reagents are all loaded into the Leica BOND RX within around 15 min of air-drying.

Automated Multiplexed RNAscope smFISH

235

Table 1 Leica BOND RX protocol for 4-plex RNAscope smFISH with TSA-biotin RNAscope 4-plex fluorescent in situ hybridization staining Step Reagent

Time (min)

Temp Cycles ( C)

Function

1 2

Probe mix Probe mix

0 120

2 1

Ambient Ambient

Hybridization of probes to mRNA transcripts

3

BOND Wash

0

1

42

4

BOND Wash

1

1

Ambient

5

BOND Wash

5

1

Ambient

6

BOND Wash

0

5

Ambient

7

BOND Wash

1

2

Ambient

8

BOND Wash

0

1

Ambient

9

ACD Multiplex Amp1

1

1

42

Merge with cell below as in rows 16-17

10

ACD Multiplex Amp1

30

1

42

Amplification by hybridization

11

BOND Wash

0

3

Ambient

12

BOND Wash

3

2

Ambient

13

BOND Wash

0

3

Ambient

14

LS Rinse

5

2

Ambient

15

BOND Wash

0

4

Ambient

16 17

ACD Multiplex Amp2 ACD Multiplex Amp2

1 30

1 1

42 42

18

BOND Wash

0

3

Ambient

19

BOND Wash

3

2

Ambient

20

BOND Wash

0

3

Ambient

21

LS Rinse

5

2

Ambient

22

BOND Wash

0

1

Ambient

23

BOND Wash

1

3

Ambient

24

ACD Multiplex Amp3

1

1

42

Merge with cell below as in rows 16-17

25

ACD Multiplex Amp3

15

1

42

Amplification by hybridization

26

BOND Wash

0

3

Ambient

27

BOND Wash

1

5

Ambient

28 29

ACD Multiplex HRP-C1 ACD Multiplex HRP-C1

1 15

1 1

42 42

Proprietary high-stringency wash

Amplification by hybridization

Proprietary high-stringency wash

Channel-selective HRP to conduct gene-specific TSA (continued)

236

Kenny Roberts and Omer Ali Bayraktar

Table 1 (continued) RNAscope 4-plex fluorescent in situ hybridization staining Step Reagent

Time (min)

Temp Cycles ( C)

30

BOND Wash

0

3

Ambient

31

BOND Wash

1

5

Ambient

32 33

ACD Multiplex TSA-F1 ACD Multiplex TSA-F1

1 30

1 1

42 42

34

BOND Wash

0

3

Ambient

35

BOND Wash

1

5

Ambient

36

ACD Multiplex HRP Blocker ACD Multiplex HRP Blocker

1

1

42

15

1

42

37 38

BOND Wash

0

3

Ambient

39

BOND Wash

1

5

Ambient

40 41

ACD Multiplex HRP-C2 ACD Multiplex HRP-C2

1 15

1 1

42 42

42

BOND Wash

0

3

Ambient

43

BOND Wash

1

5

Ambient

44 45

ACD Multiplex TSA-F2 ACD Multiplex TSA-F2

1 30

1 1

42 42

46

BOND Wash

0

3

Ambient

47

BOND Wash

1

5

Ambient

48

ACD Multiplex HRP Blocker ACD Multiplex HRP Blocker

1

1

42

15

1

42

49 50

BOND Wash

0

3

Ambient

51

BOND Wash

1

5

Ambient

52 53

ACD Multiplex HRP-C3 ACD Multiplex HRP-C3

1 15

1 1

42 42

54

BOND Wash

0

3

Ambient

Function

Tyramide-conjugated fluorophore

Inhibits peroxidase activity of HRP and halts TSA

Channel-selective HRP to conduct gene-specific TSA

Tyramide-conjugated fluorophore

Inhibits peroxidase activity of HRP and halts TSA

Channel-selective HRP to conduct gene-specific TSA

(continued)

Automated Multiplexed RNAscope smFISH

237

Table 1 (continued) RNAscope 4-plex fluorescent in situ hybridization staining Step Reagent 55

BOND Wash

56 57

ACD Multiplex TSA-F3 ACD Multiplex TSA-F3

58

Time (min)

Temp Cycles ( C)

1

5

Ambient

1 30

1 1

42 42

BOND Wash

0

3

Ambient

59

BOND Wash

1

5

Ambient

60

ACD Multiplex HRP Blocker ACD Multiplex HRP Blocker

1

1

42

15

1

42

61 62

BOND Wash

0

3

Ambient

63

BOND Wash

1

5

Ambient

64 65

ACD Multiplex HRP-C4 ACD Multiplex HRP-C4

1 15

1 1

42 42

66

BOND Wash

0

3

Ambient

67

BOND Wash

1

5

Ambient

68 69

TSA-biotin TSA-biotin

1 30

1 1

Ambient Ambient

70

BOND Wash

0

2

Ambient

71

BOND Wash

1

4

Ambient

72 73

Streptavidin-Atto 425 Streptavidin-Atto 425

1 30

1 1

Ambient Ambient

74

BOND Wash

1

5

Ambient

75

ACD Multiplex HRP Blocker ACD Multiplex HRP Blocker

1

1

42

15

1

42

76 77

BOND Wash

0

2

Ambient

78

BOND Wash

2

2

Ambient

79

BOND Wash

10

1

Ambient

80

Deionized water

0

2

Ambient

Function

Tyramide-conjugated fluorophore

Inhibits peroxidase activity of HRP and halts TSA

Channel-selective HRP to conduct gene-specific TSA

Tyramide-conjugated biotin

Streptavidin-conjugated fluorophore

Inhibits peroxidase activity of HRP and halts TSA

238

Kenny Roberts and Omer Ali Bayraktar

Table 2 Leica BOND RX protocol for fluorescent immunohistochemistry following RNAscope (see Note 12). Fluorescent immunohistochemistry with TSA-biotin Step Reagent

Time (min)

Temp Cycles ( C)

Function

81

Block IHC

20

1

Ambient

Protein-blocking reagent

82

BOND Wash

0

3

Ambient

83 84

Avidin block Avidin block

1 20

1 1

Ambient Ambient

85

BOND Wash

0

3

Ambient

86

BOND Wash

2

3

Ambient

87 88

Biotin block Biotin block

0 30

1 1

Ambient Ambient

89

BOND Wash

0

3

Ambient

90

BOND Wash

2

3

Ambient

91 92

Primary antibody Primary antibody

1 60

1 1

Ambient Ambient

93

BOND Wash

0

3

Ambient

94

BOND Wash

2

3

Ambient

95 96

HRP secondary antibody HRP secondary antibody

1 60

1 1

Ambient Ambient

97

BOND Wash

0

3

Ambient

98

BOND Wash

2

3

Ambient

1 10

1 1

Ambient Ambient

101 BOND Wash

0

3

Ambient

102 BOND Wash

2

3

Ambient

103 Streptavidin-Atto 490LS or 1 -Alexa Fluor 700 104 Streptavidin-Atto 490LS or 30 -Alexa Fluor 700

1

Ambient

1

Ambient

105 BOND Wash

0

3

Ambient

106 BOND Wash

2

3

Ambient

107 BOND Wash

0

1

Ambient

108 DAPI

1

1

Ambient

Merge with cell below

109 DAPI

20

1

Ambient

Counterstaining of nuclei

0

4

Ambient

99 TSA-biotin 100 TSA-biotin

110 Deionized water

Avidin binds and blocks unbound TSA-biotin from RNAscope smFISH

Free biotin masks surplus avidin-binding sites

Primary antibody

HRP-conjugated secondary antibody

Tyramide-conjugated biotin

Streptavidin-conjugated fluorophore

Automated Multiplexed RNAscope smFISH

239

Table 3 Diluents and initial recommended dilution factors for reagents Reagent

Diluent

Dilution factor

Opal 520

Multiplex TSA buffer

1500

Opal 570

Multiplex TSA buffer

1500

Opal 620

Multiplex TSA buffer

1500

Opal 650

Multiplex TSA buffer

1500

Opal 690

Multiplex TSA buffer

1500

TSA-biotin

Multiplex TSA buffer

500

Streptavidin-conjugated Atto 425

Multiplex TSA buffer

400

Streptavidin-conjugated Atto 490LS

Multiplex TSA buffer

400

DAPI

BOND Wash Solution

50,000

Avidin

BOND Wash Solution

20

Biotin

BOND Wash Solution

100

Primary antibody

Antibody diluent/block

1000

Secondary antibody

Antibody diluent/block

500

2. For fixed frozen tissues, sections should be cut at 10–20 μm thickness onto SuperFrost Plus slides, and stored at 80  C until staining. 3. Remove slides from storage at 80  C and allow to thaw and dry at room temperature for around 10 min in a slide box. 4. Place slides into a vertical slide rack and bake at 65  C for 40 min. 5. Using a glass staining jar, or Tissue-Tek Slide Rack and Staining Dish, postfix the slides in pre-chilled 4% PFA in PBS for 15 min at 4  C. 6. Using four glass staining jars, or a Tissue-Tek Slide Rack and four Tissue-Tek Staining Dishes, dehydrate the sections through an ethanol series: incubate the slides at room temperature for 5 min each in 50%, 70%, 100%, and fresh 100% ethanol. 7. Air-dry the slides on a tissue for around 5 min at room temperature. 3.4 In Situ Hybridization and Immunohistochemical Staining Using the Leica BOND RX

1. Turn on the Leica BOND RX and the computer. Open the BOND software. 2. Check that all the bulk containers are present at the bottom of the machine. If any are missing, the Leica BOND RX will not

240

Kenny Roberts and Omer Ali Bayraktar

commence initialization. If necessary, refill the bulk reagent containers or empty the bulk waste containers (if any is less or more than half-full, respectively). 3. Check that all the necessary protocols are present. If not, then generate any required changes or new protocols. 4. Create a study(s). Each study should contain only one type of tissue (i.e., FFPE or fixed frozen), since the selected pretreatment (e.g., “bake and dewax”) will be applied to all slides in a given study. 5. For each study, create slides. Select the heat-induced epitope retrieval conditions and protease digestion treatments for each slide, as well as the in situ hybridization (RNAscope) and immunohistochemical staining protocols. Recommended conditions may be found in Table 4. 6. For record-keeping, export both the Slide Summary and Study Setup documents. 7. Print the label for each slide. Once the slides have been subjected to any necessary manual pretreatments, secure the correct label to each, taking care to correctly correspond previous slide references (patient or mouse IDs, for example) with the new Leica BOND RX slide ID, which incorporates a four digit incremental index (see Note 7). 8. Load all of the required 30 mL Open Containers into the Leica BOND RX (see Notes 8 and 9). 9. Scan each of the required 6 mL Titration Containers, and refill each in the software inventory. This is essential prior to every run of the machine (see Note 8). 10. Place a 6 mL Titration Container Insert into each Titration Container, and aliquot the appropriate volume of each reagent and diluent (see Note 10). 11. Place each slide into a Slide Rack, taking care to stratify the slides by protocol structure (see Note 5). Place a BOND Universal Covertile atop each slide (see Note 11).

Table 4 Recommended pretreatment conditions for examples of fixed frozen and FFPE tissue

Tissue

Recommended heat-induced epitope Preparation retrieval

Recommended protease treatment

Mouse brain

Fixed frozen

5 min, BOND ER2, 95  C

20 min, protease III

Human heart

FFPE

15 min, BOND ER2, 95  C

15 min, protease III

Automated Multiplexed RNAscope smFISH

241

Table 5 Leica BOND RX run checklist Leica BOND RX RNAscope checklist Preheat the oven to 65  C. Thaw and then dilute 8% PFA stock, and pre-chill at 4  C (fixed frozen samples only). Load and refill any missing or empty Leica BOND RX bulk reagent containers. Turn on the Leica BOND RX and open the BOND software. Generate a table listing slides (sample information, pre-staining, and staining protocols) and reagents (volumes and diluents). Check that all necessary reagents are available prior to thawing slides. The below steps (*) are required only for Generate and/or check Leica BOND RX protocols. fixed frozen samples. * Thaw slides at room temperature for ~10 min.

Generate a study(s) and slide(s) in the BOND software.

* Bake slides for 45 min at 65  C.

Check and load all 30 mL Open Containers.

* Postfix slides in pre-chilled 4% PFA at 4  C.

Refill all required 6 mL Titration Containers on the BOND software.

* Dehydrate: 5 min each in 50%, 70%, 100%, Prepare RNAscope probe mixes, fluorophores, antibodies, and 100% EtOH. and other ancillary reagents. * Air-dry slides for 5 min.

Load any remaining reagent Containers.

Print slide labels and stick onto corresponding slides, using the experiment table. Mount slides with BOND Universal Covertiles, and load Slide Racks into the Leica BOND RX. Start the run(s).

12. Load the Slide Racks into the Leica BOND RX, and start the run(s). If using multiple runs, allow one run to start fully (indicated by the appearance of a finishing time estimate) prior to starting another. 3.5 Leica BOND RX Run Checklist

3.6 Mounting and Imaging

Table 5 provides a suggested scheme for intercalating the steps of pre-staining sample preparation, Leica BOND RX preparation, and run assembly, and is designed as a prompt for novice users and those of intermediate experience alike. It has been formatted in the form of a table that might be stored in the vicinity of the Leica BOND RX. 1. Remove the slides from the Leica BOND RX by disengaging the Slide Rack(s). 2. Wipe the underside of each slide with a tissue moistened with 70% ethanol. 3. Flick or tap the slides hard over a tissue on the lab bench to remove any excess liquid.

242

Kenny Roberts and Omer Ali Bayraktar

Fig. 3 A mouse brain tissue section processed for 4-plex RNAscope smFISH, antibody staining against the neuronal transcription factor NEUN, and the nuclear marker DAPI. RNAscope targets are housekeeping genes (italicized). TSA was used to boost both the smFISH and IHC signal. The left panel shows an image from the mouse cerebellum while the right panel shows a close-up of the area indicated with the white box. Scalebars: (left) 100 μm, (right) 10 μm

4. Swiftly add a minimal number of drops of ProLong Gold Antifade Mountant required to cover the section. Do not allow the stained section to dry out. 5. Mount with a glass coverslip, endeavoring to avoid introducing air bubbles. 6. Store slides in the dark prior to imaging. Store at room temperature if imaging is imminent (24 h); otherwise, store at 4  C. Slides should be imaged within several weeks for optimal results, though signals may remain strong for months at 4  C. See Fig. 3 for an example image from the mouse brain.

4

Notes 1. The imaging system must be capable of simultaneous imaging of six fluorophores, including DAPI (or another nuclear stain). Ultimately, the choice of microscope should be coordinated with the choice of fluorophores. This protocol is designed with the Operetta CLS High-Content Analysis System or Opera Phenix High-Content Screening System (both Perkin Elmer) in mind. Alternatives include the Vectra Polaris Automated Quantitative Pathology Imaging System (Perkin Elmer), which allows multiplexing of up to six Opal dyes alongside DAPI—this would remove the necessity of using TSA-biotin and streptavidin-conjugated fluorophores.

Automated Multiplexed RNAscope smFISH

243

2. Paraformaldehyde is very hazardous, especially in powder form. It is a respiratory and dermal irritant, sensitizer, and toxin; it causes damage to organs and is a category 2 suspected carcinogen. It must only be used inside a ducted fume cabinet using appropriate PPE. 3. Ethanol is highly flammable. It should be stored in a flammable storage cabinet, and used in a well-ventilated area, ideally in a ducted fume cabinet or on a down-flow table. 4. BOND Dewax Solution is an alternative to the commonly used deparaffinizing agent xylene, but is still hazardous and should be handled with care, according to the safety data sheet. 5. The Leica BOND RX can operate three differently structured protocols simultaneously. That is, if protocols are to be run alongside one another on the same Slide Rack, each component step should be of the same length, though different reagents may be dispensed. This should be borne in mind when designing an experiment, which should be done with care prior to any slide preparation. Each slide with a different protocol structure should be loaded onto a separate Slide Rack. 6. For the 4-plex RNAscope smFISH protocol, a maximum number of 30 slides could be processed against 10 different probe mixtures in a single run (e.g., 40 different genes screened across 3 sets of biological replicates). The combined multiplexed RNAscope smFISH and IHC protocol for 20 slides can be run overnight on the Leica BOND RX lasting ~18 h. 7. While each Leica BOND RX will generate unique slide IDs for each slide stained on that machine, each Leica BOND RX uses the same sequence of incremental identifiers (though the additional three digits are non-incremental). If you will routinely be using multiple machines, take care during record-keeping to identify which machine was used, in case of duplicate slide IDs. 8. Each 30 mL Open Container and 6 mL Titration Container has a refill limit volume of 40 mL. That is, 40 mL may be dispensed from each Container before another must be registered to replace it. 9. The location of each Container within the Reagent Racks within the Leica BOND RX is not specified—its location will be identified by the machine by barcode scanning—with the exception of the BOND Research Detection System rack, which must always be present within the machine during a run, and must contain a “Detection Reagent”—see the Leica BOND RX and RNAscope product literature for further details.

244

Kenny Roberts and Omer Ali Bayraktar

10. Many of the reagents and diluents contain detergents and are thus prone to air bubbles, which may disrupt the automated liquid handling. Limit bubble generation by pipetting liquids down the side of the Titration Tubes, and by mixing by gentle vortexing or swirling, not pipetting. 11. The Leica BOND RX does not verify whether each slide is mounted with a BOND Universal Covertile, so take care to ensure that this is carried out. 12. TSA is used to amplify the signal in both RNAscope and the subsequent IHC staining (Fig. 2). In the scheme used here, TSA-biotin-streptavidin-fluorophore conjugation is used to detect both RNA and protein signals, necessitating the inclusion of an avidin-biotin blocking step between the assays. References 1. Nowakowski TJ, Bhaduri A, Pollen AA et al (2017) Spatiotemporal gene expression trajectories reveal developmental hierarchies of the human cortex. Science 358:1318–1323. https://doi.org/10.1126/science.aap8809 2. Hodge RD, Bakken TE, Miller JA et al (2019) Conserved cell types with divergent features in human versus mouse cortex. Nature:1–8. https://doi.org/10.1038/s41586-019-15067 3. Vento-Tormo R, Efremova M, Botting RA et al (2018) Single-cell reconstruction of the early maternal-fetal interface in humans. Nature 563:347–353. https://doi.org/10.1038/ s41586-018-0698-6 4. Young MD, Mitchell TJ, Vieira Braga FA et al (2018) Single-cell transcriptomes from human kidneys reveal the cellular identity of renal tumors. Science 361:594–599. https://doi. org/10.1126/science.aat1699 5. Lein E, Borm LE, Linnarsson S (2017) The promise of spatial transcriptomics for neuroscience in the era of molecular cell typing. Science 358:64–69. https://doi.org/10.1126/sci ence.aan6827 6. Raj A, van den Bogaard P, Rifkin SA et al (2008) Imaging individual mRNA molecules using multiple singly labeled probes. Nat Methods 5:877–879. https://doi.org/10. 1038/nmeth.1253 7. Moffitt JR, Bambah-Mukku D, Eichhorn SW et al (2018) Molecular, spatial and functional single-cell profiling of the hypothalamic preoptic region. Science 15:eaau5324. https://doi. org/10.1126/science.aau5324 8. Eng C-HL, Lawson M, Zhu Q et al (2019) Transcriptome-scale super-resolved imaging in

tissues by RNA seqFISH. Nature:1. https:// doi.org/10.1038/s41586-019-1049-y 9. Qian X, Harris KD, Hauling T et al (2018) A spatial atlas of inhibitory cell types in mouse hippocampus. bioRxiv 431957. https://doi. org/10.1101/431957 10. Battich N, Stoeger T, Pelkmans L (2013) Image-based transcriptomics in thousands of single human cells at single-molecule resolution. Nat Methods 10:1127–1133. https:// doi.org/10.1038/nmeth.2657 11. Shah S, Lubeck E, Schwarzkopf M et al (2016) Single-molecule RNA detection at depth by hybridization chain reaction and tissue hydrogel embedding and clearing. Development 143:2862–2867. https://doi.org/10.1242/ dev.138560 12. Kishi JY, Lapan SW, Beliveau BJ et al (2019) SABER amplifies FISH: enhanced multiplexed imaging of RNA and DNA in cells and tissues. Nat Methods 16:533. https://doi.org/10. 1038/s41592-019-0404-0 13. Wang F, Flanagan J, Su N et al (2012) RNAscope. J Mol Diagn 14:22–29. https://doi. org/10.1016/j.jmoldx.2011.08.002 14. Kerstens HM, Poddighe PJ, Hanselaar AG (1995) A novel in situ hybridization signal amplification method based on the deposition of biotinylated tyramine. J Histochem Cytochem 43:347–352. https://doi.org/10. 1177/43.4.7897179 15. Bayraktar OA, Bartels T, Holmqvist S et al (2019) Single-cell in situ transcriptomic map of astrocyte cortical layer diversity. Nat Neurosci (in press) https://doi.org/10.1038/ s41593-020-0602-1

Chapter 16 Automated Co-in Situ Hybridization and Immunofluorescence Using Archival Tumor Tissue Leah K. Officer, Kleopatra E. Andreou, Ana V. Teodo´sio, Zhangyi He, and John P. Le Quesne Abstract In situ hybridization (ISH) and immunohistochemistry (IHC) are valuable tools for molecular pathology and cancer research. Recent advances in multiplex technology, assay automation, and digital image analysis have enabled the development of co-ISH IHC or immunofluorescence (IF) methods, which allow researchers to simultaneously view and quantify expression of mRNA and protein within the preserved tissue spatial context. These data are vital to the study of the control of gene expression in the complex tumor microenvironment. Key words Autostainer, Digital pathology, Image analysis, Immunofluorescence, Immunohistochemistry, In situ hybridization, Multiplex, Pathology informatics, Quantitative analysis, Translational control

1

Introduction Control of mRNA translation is fundamental to gene regulation and is highly dysregulated in cancer cells [1]. As evidence surrounding dysregulated mRNA translation continues to grow, it is emerging as an additional hallmark of cancer [2]. Until recently, however, methods of mRNA and protein analysis have relied on single-cell methods such as flow cytometry, or the in situ detection of a single protein or mRNA target on a glass slide. These methods, although highly quantitative, either contain limited information about tissue spatial context or are limited to the quantification of a single molecule, respectively. Recent developments in multiplex in situ methods have the potential to address the control of gene expression in intact tissues, generating quantitative single-cell data, and this type of data is essential for the understanding of translational control in the complex tumor microenvironment. Multiplex immunohistochemistry methods have advanced rapidly over the past decade with the

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_16, © Springer Science+Business Media, LLC, part of Springer Nature 2020

245

246

Leah K. Officer et al.

development of multispectral deconvolutional microscopy, enabling the quantitative unmixing of multiple fluorophores [3]. The multiple-layer images generated are an ideal substrate for digital image analysis methods. These technologies have successfully been used to identify previously unknown tumor subtypes and phenotypic relationships [4–7]. These discoveries underline the concept of the tumor as a complex neo-organ in which malignant cells are able to manipulate the stroma to optimize their own local environment, favoring tumor growth and metastasis [8, 9]. Multiplex immunohistochemical approaches have been accompanied by the introduction of readily applicable nonradioactive in situ hybridization methods for the detection of nucleic acids. In particular, mRNA detection using multiple pairs of oligonucleotide probes has resulted in high-quality in situ data on gene expression [10]. A method to combine these technologies, giving simultaneous immunofluorescent detection of protein and mRNA, would be highly desirable to discern translational control in situ. The laborious nature of the methods when applied manually is a major limiting factor, with large multiplex assays containing hundreds of individual steps taking up to several days to complete. However, the recent automation of these methods on several histological staining platforms has opened the door to high-throughput quantitative staining [11–14]. We have developed a method to generate data of this type by combining proprietary technologies for mRNA ISH (Advanced Cell Diagnostics, RNAscope) and multiplex immunofluorescence (Akoya Biosciences Inc. OPAL) on an autostainer platform (Roche Tissue Diagnostics, Ventana Discovery Ultra), followed by digital slide scanning and deconvolution (Akoya Biosciences Inc. Vectra, Phenochart) and extraction of quantitative single-cell data (Akoya Biosciences Inc. Inform®). In this way we can visualize a snapshot of mRNA and protein levels within the preserved spatial context of fixed archival tumor tissue. With appropriate image analysis methods, we can identify cells expressing a given gene at the mRNA and/or protein levels, and directly visualize the consequences of altered transcriptional and translational control in situ.

2

Materials

2.1 Advanced Cell Diagnostics

1. RNAscope 2.5 VS Target Probes. 2. RNAscope 2.5 VS Positive Control Probe_Hs-UBC. 3. RNAscope 2.5 VS Negative Control Probe_dapB. 4. RNAscope VS Universal HRP Reagent Kit.

2.2 Roche Tissue Diagnostics

1. Ventana Discovery Ultra Autostainer with mRNA Universal procedure (v0.00.0213).

Co-in Situ Hybridization and Immunofluorescence

247

2. DISCOVERY ChromoMap DAB Kit (RUO). 3. DISCOVERY FITC Kit (RUO). 4. DISCOVERY Wash (RUO). 5. mRNA Amplification, Pretreatment & DAB Kit. 6. ULTRA LCS (Predilute). 7. SSC (10). 8. Reaction buffer (10). 9. Cell conditioning 2 (CC2). 10. Protease 2. 11. DISCOVERY QD DAPI (RUO). 12. Bluing reagent. 13. Hematoxylin II. 14. 250 Test detection dispensers. 15. DISCOVERY UltraMap anti-Ms HRP (RUO). 16. DISCOVERY Inhibitor. 2.3 Target IHC Antibodies

1. Multi-Cytokeratin (AE1, AE3) (Leica Biosystems). 2. Anti-Ornithine Decarboxylase/ODC antibody (ODC1/485) (Abcam). 3. EnVision FLEX Antibody Diluent (Dako Agilent).

2.4 Akoya Biosciences Inc.

1. Vectra 3.0 Automated Quantitative Pathology Imaging System, 200 Slide. 2. OPAL fluorophores. 3. 1 Plus Amplification Diluent. 4. inForm® software. 5. Phenochart software.

2.5 Multiplex Assay Validation Slides

2.6

Mounting

FFPE sections of target tissue, 4.5 μm thick, prepared no more than 48 h in advance. Our method uses approximately 12 slides per marker. These slides are not for validation of antibody specificity and/or sensitivity. This should be done prior to multiplexing with positive control tissue and/or cell lines. 1. Molecular Probes™ ProLong™ Diamond Antifade Mountant. 2. Glass coverslips, #1.5.

3

Methods

3.1 Multiplex Assay Construction and Validation

Our co-ISH IF method is validated in three stages, beginning with well-validated bright-field immunohistochemical and RNAscope assays. Each assay is then optimized in monoplex fluorescence

248

Leah K. Officer et al.

incorporating digital image deconvolution and using the brightfield images/slides as references for staining accuracy. Finally, the assay is optimized in multiplex and the resulting images are analyzed with inForm® software, and H-scores generated using this software are validated against H-scores obtained manually from bright-field images. 1. Consecutive sections of a validation TMA block containing archival FFPE lung adenocarcinoma samples were used for all validation steps. The TMA block was constructed specifically for assay testing and contained donor cores from 50 archival cases. Sections were 4.5 μm thick and were freshly prepared 24–48 h prior to staining. A positive and negative control probe section was used on every RNAscope run. The following steps are carried out for each new assay (Fig. 1). Once the assay has been validated it is applied to test tissues. Each individual stain was validated chromogenically using DAB and a hematoxylin counterstain. The DAB-stained slide is used as a reference slide throughout the validation process. The pretreatment used for each protein was CC2 for 24 min at 97  C followed by an 8-min incubation with protease 2 at 37  C (see Note 1). 2. Each protein was assigned an OPAL fluorophore. Fluorophores were selected based on colocalization of markers, spectrally separating markers expected to be expressed in the same compartments of the same cells as much as possible to avoid spectral cross talk. The mRNA probe was assigned the Ventana FITC detection as an “open detection” step was not available for mRNA detection. 3. The DAB visualization kit was deselected and replaced with the chosen fluorophore for each single-plex assay and the counterstain is deselected. A fluorescent single-plex optimization slide was stained. 4. For OPAL fluorophores the starting concentration of 1:100 was used with an incubation time of 8 min. 5. The concentration was optimized using the multispectral image (MSI) exposure time on the Vectra imaging system. An exposure time of below 50 ms was considered too high, and the OPAL fluorophore was diluted (see Note 2). An exposure time of 200 ms was considered too low, and the fluorophore was concentrated (see Note 3). This process was repeated until an optimal exposure time of between 50 ms and 200 ms was achieved on the brightest part of the slide (Fig. 2). Stain intensity is verified using inForm® software. 6. Staining specificity was checked using the reference slide and unmixed snapshot images of single-plex optimization slides in inForm®.

Co-in Situ Hybridization and Immunofluorescence

249

Fig. 1 Flowchart showing outline of multiplex assay development and validation. Single-plex DAB-stained images are used for validation of immunofluorescence images at two stages. Firstly, while the multiplex assay is being constructed, the immunofluorescent signal from single-plex staining is validated qualitatively against DAB. Secondly, following construction of the multiplex assay, a quantitative validation is conducted (Fig. 6)

7. For ready to use Ventana fluorophores (FITC) the above process was repeated altering incubation times to achieve optimal stain intensity. 8. A multiplex protocol was built using the mRNA Universal procedure. 9. Sequence 1 is automatically assigned to mRNA detection by the software. 10. After each detection sequence a denaturation step was selected, incubating CC2 at 100  C for 8 min. 11. Each protein was tested in every available protocol sequence with its optimal fluorophore counterpart to check for loss of

250

Leah K. Officer et al.

Fig. 2 Exposure times (ms) for a 4-plex protocol using Vectra software within the acceptable range for MSI images

Fig. 3 Library snapshots of lung adenocarcinoma taken using Vectra: A, Autofluorescence, B, Cytokeratin OPAL 650, C, Protein OPAL 570, D, mRNA FITC

signal; no counterstain was used. Dewax and retrieval steps for each sequence were active. The resulting slide was checked as in step 5. 12. The RNAscope assay was tested in Sequence 1 with subsequent denaturation steps selected and tested as in step 5.

Co-in Situ Hybridization and Immunofluorescence

251

13. A snapshot of each single stain in its highest performing sequence was taken for use in the spectral control library (Fig. 3). 14. Autofluorescence from collagen and blood cells can be a problem with fluorescent staining, particularly in lung tissue. An autofluorescent control slide was treated with the dewax and retrieval steps for the multiplex protocol. A snapshot of the area with the highest autofluorescent signal was taken for the project library. 15. Each protein was placed into the multiplex protocol in its highest performing sequence; an 8-min DAPI counterstain was added. 16. The exposure times for the multiplex assay were checked using Vectra and the final protocol was saved (see Notes 4 and 5). A whole-slide image was captured at 10 magnification. 17. Regions of the 10 image were selected for high-power scanning using Phenochart. MSIs were captured at 40 for analysis. 18. A spectral library was built using inForm® software and the snapshots of each single-plex stain (Fig. 4). 19. The autofluorescence control and 40 images were imported into inForm® and spectrally unmixed using the spectral library (see Note 6) (Fig. 5). 20. Images were checked for cross talk and nonspecific staining using “pathology views,” pixel intensity tool, and reference DAB validation slides (see Notes 7 and 8). 21. Both fluorescent and DAB slides/images were reviewed by a pathologist. 22. An image analysis project was prepared using inForm® to segment and phenotype cells; the data from this project was then cross-referenced with manual H-Scores of the DAB-stained images. 23. In this example, a multiplex fluorescent approach was used to detect ODC1 (ornithine decarboxylase 1) mRNA and protein expression in lung adenocarcinoma tissues. Inform® software generated automated histological H-scores for each channel of the multiplex assay in the cytoplasm of tumor cells in each TMA tissue core. These scores were correlated against conventional histological H-scores of monoplex DAB-stained TMAs. Analysis was performed using an R script (R version 3.5.3) and confirms the positive correlation between automated and manual histological scores (Fig. 6). 3.2 Validated Protocol

The following steps were selected for the final protocol using Ventana mRNA Universal procedure (v0.00.0213) on the Ventana Discovery Ultra autostainer. Stringency washes using reaction

252

Leah K. Officer et al.

Fig. 4 Spectral library for unmixing 4-plex protocol taken from single-stain images

buffer or SSC are predetermined by the protocol and do not require manual entry. Steps marked ∗ are predetermined but included for clarification: 1. Sequence 1. 2. Deparaffinization. 3. mRNA dewax solution, 60  C∗. 4. Cell conditioning. 5. CC2, 97  C, 24 min.

Co-in Situ Hybridization and Immunofluorescence

253

Fig. 5 Channels of an unmixed lung adenocarcinoma image obtained using inForm® software. (a) Original mixed image. (b) Spectrally unmixed composite image. (c) Cytokeratin OPAL 650. (d) Target protein OPAL 570. (e) Target mRNA FITC. (f) Co-ISH IF expression

Fig. 6 Quantitative validation of automated scoring of multiplex co-ISH IF assay against manually scored single-plex DAB. (a) Representative multiplex image of a lung adenocarcinoma TMA core immunostained for ODC1 mRNA (red), ODC1 protein (yellow), and cytokeratin (green). (b) Monoplex DAB ISH for ODC1 mRNA. (c) Monoplex DAB IHC for ODC1 protein. (d) Spearman correlation plot of fluorescent (automated) and monoplex (manual) H scores of individual TMA cores for ODC1 mRNA (n ¼ 9, rho 0.9790881, p-value ¼ 4.266e-06). (e) Spearman correlation plot of fluorescent (automated) and monoplex (manual) H scores for ODC1 protein (n ¼ 17, rho 0.6216815, p-value ¼ 0.007716). Scale bar ¼ 100 μ

254

Leah K. Officer et al.

6. mRNA HRP detection. 7. Discovery inhibitor, 37  C, 12 min∗. 8. HRP detection third pretreatment. 9. (mRNA) Protease, 37  C, 16 min. 10. Apply two drops of target probe, 43  C, 2 h. 11. Amp 1, 39  C, 32 min. 12. Amp 2, 39  C, 32 min. 13. Amp 3, 37  C, 12 min∗. 14. Amp 4, 37  C, 32 min∗. 15. Amp 5, RT, 16 min. 16. Amp 6, RT, 12 min∗. 17. Amp 7, RT, 4 min∗. 18. Apply two drops of Disc FITC, RT, 4 min∗. 19. Apply one drop of Disc FITC H2O2, RT, 1 h 20 min. 20. Dual sequence (DS). 21. Antibody denaturation, antibody denature CC2, CC2, 100  C, 8 min. 22. DS inhibitor. 23. Neutralize, 40  C, 20 min∗. 24. DS antibody. 25. Target primary antibody, 37  C, 28 min. 26. DS second antibody. 27. Secondary antibody in detection dispenser (UltraMap antimouse HRP), 37  C, 12 min. 28. DS third antibody. 29. OPAL fluorophore in detection dispenser, 37  C, 8 min. 30. Triple stain (TS). 31. Antibody denaturation, antibody denature CC2, CC2, 100  C, 8 min. 32. TS inhibitor. 33. Neutralize, 40  C, 20 min∗. 34. TS antibody. 35. Target primary antibody, 37  C, 60 min. 36. TS second antibody. 37. Secondary antibody in detection dispenser (UltraMap antimouse HRP), 37  C, 20 min. 38. TS third antibody. 39. OPAL fluorophore in detection dispenser, 37  C, 8 min. 40. Counterstain.

Co-in Situ Hybridization and Immunofluorescence

255

41. Apply one drop of QD DAPI, 37  C, 8 min. 42. Slides are removed from the machine and washed manually in two changes of discovery wash for 2 min. 43. Slides are washed in running water for 2 min. 44. Slides are mounted with Diamond Prolong anti-fade mountant and a glass coverslip and left to dry at room temperature prior to imaging.

4

Notes 1. We have found that the antigen retrieval method used for RNAscope is sometimes incompatible with target antigens. In instances where incorporation of a marker of interest into the co-ISH IF assay is not possible, these markers may be stained on consecutive sections. Data from the multiplex and singleplex assays is then combined. 2. Where staining is too intense (exposure time lower than 50 ms) fluorophores should be diluted or incubation times decreased until the optimum signal is achieved. Failing to do this may result in an umbrella effect where signals from other markers of interest or counterstains may be masked. 3. Where staining is too weak (exposure time higher than 200 ms) fluorophores should be concentrated or incubation times increased to achieve the optimum signal intensity. If a signal is too weak images may be pixelated or appear to have falsepositive staining where the background is amplified by software attempting to detect a weak signal. Where two signals appear to be the same or there appears to be cross talk between two separate channels: 4. The spectral library should be checked, each fluorophore within the multiplex assay should be different, and each separate fluorophore should have separate spectral wavelengths with different excitation and emission peaks. If the same fluorophore has been used twice, or a fluorophore is contaminated the emission spectra will overlap. 5. The “show cube info” tool should be used to check if the exposure times for the MSI are in range (50–200 ms). If they are out of range and signals appear to be masked or non-specific background fluorophore dilutions or incubation times may require re-optimization. 6. The autofluorescence image should be checked for signal; if a signal appears to be present there may be a lack of specific staining in the visible channel across test sections within the project.

256

Leah K. Officer et al.

7. The pixel intensity tool should be used to check the signal in the affected area; if the visible signal has a very low pixel intensity it may be nonspecific background and a threshold can be applied to analysis. 8. If cross talk appears to be true and none of the above apply, a “dropout” staining run should be carried out. In this run the multiplex protocol should be applied on consecutive sections omitting one of the problematic primary antibodies per slide with all other steps remaining, and the resulting slides should be assessed for cross talk. If no issues are visible but the problem remains with the multiplex assay an additional denaturation reagent may be required. References 1. Topisirovic I, Siddiqui N, Orolicki S et al (2009) Stability of eukaryotic translation initiation factor 4E mRNA is regulated by HuR, and this activity is dysregulated in cancer. Mol Cell Biol 29:1152–1162. https://doi.org/10. 1128/MCB.01532-08 2. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144:646–674. https://doi.org/10.1016/j. cell.2011.02.013 3. Perkin Elmer Phenoptics ™ Tissue Biomarker Detection Solutions. http://image.info.per kinelmer.com/Web/PerkinElmer/% 7Bf032f380-a59e-4d6d-bca2-e906160c10e8% 7D_PDF-File_LST-Phenoptics-GLO-2018Q1-BannerAd-Cell-ebook_013782_01_STL. pdf?_ga¼2.93522525.1901772079. 1532254929-1250708859.1531136125&_gac ¼1.137101572.1532255082.E. Accessed 11 May 2019 4. Kawashima R, Chang YH, Tsujikawa T et al (2017) Quantitative multiplex immunohistochemistry reveals myeloid-inflamed tumorimmune complexity associated with poor prognosis. Cell Rep 19:203–217. https://doi.org/ 10.1016/j.celrep.2017.03.037 5. Ying L, Yan F, Meng Q et al (2017) Understanding immune phenotypes in human gastric disease tissues by multiplexed immunohistochemistry. J Transl Med 15:1–11. https://doi. org/10.1186/s12967-017-1311-8 6. Parra E, Francisco-Cruz A, Wistuba I (2019) State-of-the-art of profiling immune contexture in the era of multiplexed staining and digital analysis to study paraffin tumor tissues. Cancers (Basel) 11:247. https://doi.org/10. 3390/cancers11020247 7. Heindl A, Khan AM, Rodrigues DN et al (2018) Microenvironmental niche divergence

shapes BRCA1-dysregulated ovarian cancer morphological plasticity. Nat Commun 9:3917. https://doi.org/10.1038/s41467018-06130-3 8. Mittal V, El Rayes T, Narula N et al (2016) The Microenvironment of Lung Cancer and Therapeutic Implications. In: Ahmad A, Gadgeel SM (eds) Lung cancer and personalized medicine: novel therapies and clinical management. Springer, Cham, pp 75–110 9. Egeblad M, Nakasone ES, Werb Z (2010) Tumors as organs: complex tissues that interface with the entire organism. Dev Cell 18:884–901. https://doi.org/10.1016/j. devcel.2010.05.012 10. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagnostics 14:22–29. https://doi.org/ 10.1016/j.jmoldx.2011.08.002 11. Leica Biosystems Inc. (2019) BOND RX Fully Automated Research Stainer. https://www. leicabiosystems.com/ihc-ish-fish/fullyautomated-ihc-ish-instruments/bond-rx/. Accessed 02 July 2019 12. Roche Diagnostics (2016) Roche launches DISCOVERY 5-Plex procedure for cancer research applications. https://diagnostics. roche.com/global/en/news-listing/2016/ roche-launches-discovery-5-plex-procedurefor-cancer-research-ap.html. Accessed 02 July 2019 13. Advanced Cell Diagnostics Inc. (2019) Automated assay Ventana. https://acdbio.com/ automated-assay-ventana. Accessed 02 July 2019 14. Advanced Cell Diagnostics Inc. (2019) ACD automated assay Leica. https://acdbio.com/ automated-assay-leica. Accessed 02 July 2019

Chapter 17 Automated Five-Color Multiplex Co-detection of MicroRNA and Protein Expression in Fixed Tissue Specimens Lorenzo F. Sempere, Erin Zaluzec, Elizabeth Kenyon, Matti Kiupel, and Anna Moore Abstract microRNAs are an important class of noncoding regulatory RNAs with functional roles in development, physiology, and disease. Visualization of microRNA expression at a single-cell level has contributed to a better understanding of their biological function in animal models and their etiological contribution to human diseases. In addition, several microRNAs have been highlighted as potential biomarkers carrying diagnostic and prognostic information. Co-detection of microRNA expression with that of cell-typespecific proteins can enhance the interpretative power of expression changes during development or altered expression in pathological conditions. Here, we describe an automated fluorescence-based five-color multiplex assay for co-detection of microRNA (e.g., miR-10b, miR-21, miR-205), noncoding RNA (e.g., snRNA U6, 18S rRNA), and protein expression (e.g., cytokeratin 19, vimentin, collagen I) in paraffin-embedded formalin-fixed tissue slides on a Leica Bond Rx staining station. While this protocol uses mainly mouse tissues to demonstrate multiplex detection, it can be generally applied to single-cell expression analysis of other animal models and clinical specimens. Key words Animal models, Automated staining, Breast cancer, In situ hybridization, Locked nucleic acid (LNA), microRNA (miR, miRNA), Molecular pathology, Multiplex detection, Pancreatic cancer

1

Introduction microRNAs are an evolutionarily conserved class of short noncoding regulatory RNA genes in animals [1]. The biologically active ~22-nucleotide-long RNA guides an Argonaute-containing multiprotein silencing complex to partially complementary sites on the 30 UTR of target mRNAs [2]. This interaction between the microRNA-induced silencing complex and target mRNA leads to mRNA degradation and/or translational inhibition and consequent decrease in protein output. This posttranscriptional regulatory mechanism enables a single microRNA to modulate the protein output of tens to hundreds of target mRNAs depending on cell type and context. Altered expression and dysregulated

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_17, © Springer Science+Business Media, LLC, part of Springer Nature 2020

257

258

Lorenzo F. Sempere et al.

activity of microRNAs have been associated with many human diseases, including cancer [1]. Functional studies in cell lines and animal models have shed light on molecular mechanisms of microRNA action [2]. Over the last decade, there have been many methodological and technological advances to sensitively and robustly detect microRNA expression at a single-cell resolution. Among these methodologies, in situ hybridization (ISH) protocols have been established for chromogenic or fluorescent detection of microRNAs in frozen and fixed tissues from animal models of human disease and human subjects. While perhaps not as sensitive as other techniques, unique features of ISH make it a powerful technique for detection of microRNA expression in both a research environment and clinical setting [3–5]. Characterization of microRNA expression by ISH on diseased tissue from animal models and in clinical specimens has also provided valuable and actionable information to uncover the etiological contribution of microRNAs in complexed human diseases and to interrogate contextual changes of microRNA expression for diagnostic and prognostic applications [4]. This book chapter is an updated and improved version of our previous published protocol for automated multiplex co-detection of microRNA and protein markers in formalin-fixed paraffin-embedded tissues [5]. We highlight recent advances and applications of microRNA ISH detection within the last 5 years, and refer readers to previous review and protocol publications for a broader historical perspective [4–9]. With few exceptions, microRNA detection protocols [10– 15], including ours, use locked nucleic acid (LNA)-modified DNA probes [16]. Recent protocols describe different approaches for co-detection of microRNA and protein markers [10, 13, 14], and co-detection of microRNAs and mRNAs either combining LNA and RNAscope technology [15] or exclusively using RNAscope technology for both microRNA and mRNA species [17]. These or similar protocols have been implemented to evaluate microRNA expression changes in animal models and/or human specimens of different neurological conditions [18], skin conditions [19, 20], and cancer types [15, 21–25] as well as for tissue slide-based microRNA biomarker discovery studies in bladder cancer [26], breast cancer [27], colorectal cancer [28], hepatocellular carcinoma [29], melanoma [30], and oral squamous cell carcinoma [31]. In this book chapter, we describe a detailed step-by-step protocol for fluorescence-based co-detection of a microRNA, a reference abundant noncoding RNA, and two protein markers on a Leica Bond Rx automated staining station. Our previous protocol [5] described the use of FDA-approved Leica Bond MAX for co-detection of microRNA and protein expression in formalinfixed paraffin-embedded (FFPE) clinical specimens by combined ISH and immunohistochemistry (IHC) staining techniques. The Leica Bond Rx is an open platform with more flexibility and

Automated 5-Plex ISH/IHC Assay

259

Fig. 1 Workflow of the multiplex staining procedure. Pre-run preparatory steps are manual tasks (green boxes) required for protocol programming and preparation of materials and reagents. Automated steps (red boxes) of Bond Rx with description of protocol subsections. Post-run steps include manual tasks for slide counterstaining and slide loading on Versa imaging system, automated slide scanning, and computer-assisted (orange box) tasks for image analysis and figure composition. Estimated times for each step are for the preparation, staining, and analysis of 30 tissue slides with 5 colors per slide

protocol versatility in the research environment and biomarker development space. This new protocol is programmed as a continuous IHC protocol for streamlined co-detection of microRNA and protein markers. We have adapted for use in an open and continuous protocol (see Note 1) Leica protocols for proteinase K digestion, probe hybridization, and probe removal, which are only readily available in a sequential IHC and ISH staining protocol. We optimized and demonstrate on normal and tumor tissues from genetically engineered mouse models of breast and pancreatic cancer the robustness of this automated four-color multiplex IHC protocol with nuclear counterstaining with DAPI (Figs. 1, 2, 3 and 4). The same or slightly modified protocol should be generally applicable for co-detection of microRNA, noncoding RNA, and protein expression in a broader spectrum of organs and animal species. We have used this protocol for co-detection of microRNA and protein expression in FFPE cat (Fig. 5) and baboon (Sempere and Fazleabas, unpublished observations) tissue specimens. Some of the antibodies that we describe, including anti-collagen I, anticytokeratin 19, and anti-vimentin, can detect cognate protein in multiple animal species, including mouse, cat, baboon, and humans.

Fig. 2 Co-detection of RNA and protein markers on tumor tissue from a mouse model of K-Ras-driven pancreatic cancer. Cartoon representation of post-hybridization steps for automated co-detection of miR-21, U6 snRNA, tubulin, and vimentin. Horseradish peroxidase (HRP) activates the tyramide moiety conjugated to the fluorochrome, resulting in covalent deposition of the signal in close proximity of the RNA probe or target protein. Sequential rounds of HRP inactivation with H2O2 followed by a new protein or antibody conjugated to HRP can be used to reveal the expression of multiple markers. Representative cumulative staining of RNA and protein markers as well as DAPI. Scale bar is 200 μm

Fig. 3 Setting up a new multiplex protocol. Screen snapshot shows how to mark selection for a single continuous IHC staining protocol for co-detection of RNA and protein markers. Protocol details and position of the covertile are shown for key heated steps of the proteinase K digestion and probe hybridization

Automated 5-Plex ISH/IHC Assay

261

A key step of the ISH detection is enzymatic digestion with proteinase K to expose probing (micro)RNAs; automation of this step increases consistency and reproducibility of the stain across samples. While proteinase K digestion is a sufficient treatment to enable detection of several protein markers, heat-induced epitope retrieval (HIER) is required for other protein markers. Alternative ISH/IHC protocols exist in which a HIER step [32, 33], instead of proteinase K digestion, is used to expose microRNA and protein molecules. Our protocol allows for the addition of a HIER step after probe detection to maximize co-detection with other proteins of interest. However, this HIER may need to be of reduced time or temperature to minimize loss of signal from microRNA and reference RNA. We have not investigated this HIER modification in sufficient detail to provide general guidelines for its use (but see [14, 34]). Multiplex detection of protein markers has flourished in recent years thanks to its application in the field of immunotherapy with immune checkpoint regulators [35–38]. Our protocol could be applied in a similar fashion for the diagnostic characterization of multiple microRNAs and other noncoding RNAs in cancer. We also describe a general companion protocol for image acquisition and analysis using Aperio Versa imaging system (Figs. 1 and 5). This image acquisition protocol uses DAPI nuclear counterstaining for tissue location and focusing. This acquisition protocol generates a multilayer whole-tissue image file for DAPI, FITC, rhodamine, Dylight 594, and Dylight 650 fluorochrome channels. This general protocol should be compatible without modifications for use with other filter cube-based microscopy system, but it may need modification if multispectral or confocal microscopy is used.

2

Materials This method assumes that the user has access to and is proficient at operating the Leica Bond Rx staining station (referred to hereafter as the machine). Use RNase-free reagents and chemicals to prepare solutions and buffers, and work in an RNase-free laboratory station. In this example protocol, volumes of buffers and solutions as well as quantity of reagents are calculated to co-detect miR-21a, U6 snRNA, tubulin, and vimentin (Fig. 2) on 30 FFPE mouse normal and cancerous tissue slides, which is the maximum slide capacity per run. With the exception of the “Buffer” container that is linked to the Open Research Kit, all other reagents can be designated as ancillary reagents and need to be provided by the researcher. The only reagents that need to be purchased from Leica are bulk solutions with a specific containers in the machine (e.g., Bond Dewax solution, Bond Wash buffer). For each reagent, we indicate the

262

Lorenzo F. Sempere et al.

Fig. 4 Adding a slide to a current study. Screen shows how to add a slide (slide 1 MMTV-PyMT) to a current study (breast mouse models) and to mark selections for the ISH4plex co-detection protocol

reagent label as it will be entered in the protocol program and container size (6 or 30 mL). 2.1 Consumables and Bulk Leica Reagents

1. Leica ancillary reagents: Novocastra Bond Dewax Solution (AR9222), Novocastra Bond Wash Solution 10 Concentrate (AR9590). 2. Leica consumables: Novocastra Bond Universal Covertile (S21.4611), Novocastra Bond Titration Kit with 6 mL containers (OPT9049), and Novocastra Bond Open 30 mL (OP309700). 3. Open Research Kit: There is no need for any Leica reagent to be dispensed in this open configuration. A “Buffer” container filled with PBST is the only reagent linked to the Open Research Kit (DS9777) and it is dispensed once per slide to validate the protocol program. Linking the Open Research Kit to a wash step after probe removal washes provides flexibility to configure different protocol variants with the same kit. The system is configured to run as single IHC staining protocol (Figs. 3 and 4). This standard protocol and most variants do not require the use of Novocastra Bond Epitope Retrieval

Automated 5-Plex ISH/IHC Assay

263

Fig. 5 Cell classification based on the expression of a microRNA or reference noncoding RNA on feline breast cancer tissues. Representative image of co-detection of indicated microRNA and reference noncoding RNAs in cat breast cancer tissue showing expression in normal adjacent skin. DAPI signal was used as nuclear counterstaining to segment cells and intensity of each RNA marker was used to classify cells as positive or negative. Basal cells of skin layer were classified as miR-21 negative and miR-205 positive, whereas cancerassociated fibroblasts were miR-21 positive and miR-205 negative. Most tumor and normal adjacent cells were classified as snRNA6 or 18S rRNA positive indicating appropriate RNA quality of these tissues. Scale bar is 500 μm

Solution 1 or Solution 2; within Bond Admin program, navigate to Hardware configuration tab, and deselect the containers for Novocastra Bond Epitope Retrieval Solution 1 and Solution 2. 2.2 General Solution and Reagents

We provide specific products that we used for this protocol. Comparable reagents from other companies should provide similar results. 1. 100% Ethanol. 2. RNase-free water.

264

Lorenzo F. Sempere et al.

3. 30% Hydrogen peroxide (H2O2). 4. 10 Phosphate-buffered saline (PBS). 5. 1 M Tris–Cl pH ¼ 8. 6. 0.5 M EDTA pH ¼ 8. 7. 5 M NaCl. 8. Proteinase K (PK). 9. Glycine. 10. 20% Paraformaldehyde (PFA). 11. Hydrochloric acid. 12. Acetic anhydride (AA). 13. Triethanolamine (TEA). 14. Triton X-100. 15. Formamide. 16. Tween∗20. 17. Yeast tRNA solution. 18. 50 Denhardt’s solution. 19. 20 Saline-sodium citrate. 20. Bovine serum albumin (BSA). 21. 40 ,6-Diamidino-2-phenylindole, dilactate (DAPI). 22. Prolong Gold. 23. Coverslips. 2.3 Buffers and Reagent Solutions (See Also Table 1 for Probes, Table 2 for Antibodies, Table 3 for Dyes)

1. PK solution (2 x 6 mL container): Prepare fresh 10 μg/mL proteinase K (see Note 2) in 10 mM Tris–Cl pH 8, 5 mM EDTA pH 8, and 50 mM NaCl. Add 100 μL each of 1 M Tris–Cl pH ¼ 8, 0.5 M EDTA pH ¼ 8, and 5 M NaCl to 9.3 mL of RNase-free water in a 15 mL tube. Mix well and add 5 μL of proteinase K enzyme (20 mg/mL) to complete proteinase K digestion solution. Mix well and transfer 5 mL to each of the two “PK” 6 mL containers. Please note that PK solution will be dispensed twice (300 μL per slide) and hence the need to prepare double the volume amount of this reagent. 2. Glycine 2% (6 mL container): 2% Glycine (w/v) in PBS. Add 1 g of glycine to 50 mL of PBS (this solution keeps well for 1 month at 4  C). Transfer 5 mL to the “Glycine 2%” container. 3. 4% PFA solution (6 mL container): Prepare fresh 4% PFA in PBS (v/v). Add 10 mL of 20% PFA to 40 mL of PBS (this solution keeps well for a week at 4  C). Transfer 5 mL to the “4% PFA” solution container.

Automated 5-Plex ISH/IHC Assay

265

Table 1 Probe information and general hybridization conditions Probename Probe sequence

Probe concentration

Let-7a

AA+CTA+TA+CAA+CC+TA+CTA+CCT+CA

50 nM (25–100 nM)

miR-10b

CA+CAA+ATT+CGG+TT+CTA+CA+GG+GTA

50 nM (25–100 nM)

miR-21

T+CAA+CAT+CA+GT+CTG+ATA+AG+CTA

50 nM (25–100 nM)

miR-34a

A+CAA+CCA+GCT+AAG+ACA+CTG+CCA

50 nM (25–100 nM)

miR-126

CG+CAT+TAT+TAC+TCA+CGG+TAC+GA

100 nM (100–200 nM)

miR-145

AG+GGA+TTC+CTG+GGA+AAA+CTG+GAC

50 nM (25–100 nM)

miR-155

T+TA+AT+GCT+AAT+CGT+GAT+AG+GG+GT

100 nM (100–200 nM)

miR-205

CA+GAC+TCC+GGT+GGA+ATG+AAG+GA

50 nM (25–100 nM)

miR-210

TCA+GCC+GCT+GTC+ACA+CGC+ACAG

25 nM (25–100 nM)

miR-216

T+CA+CAG+TTGC+CAG+CTGA+GAT+TA

50 nM (25–100 nM)

miR-451

AAAA+CT+CAG+TA+AT+GG+TAA+CG+GT+TTA

100 nM (50–100 nM)

U6

CGTGTCATCCTTGCGCAGGGGCCATGCTAATCTTCTC TGT

250 (50–250 nM)

18S rRNA

GGGCAGACGTTCGAATGGGTCGTCGCCGCCACGGG

100 (50–250 nM)

28 S rRNA ACGAACGTGCGGTGCG TGACGGGCGAGGGGGCGGCCG

100 (50–250 nM)

Probe sequence complementary to indicated microRNA or reference noncoding RNA. The use of multiple terminal hapten-tagged probes enhances sensitivity of detection. FAM and DIG are commonly used haptens for microRNA probes and biotin for reference RNAs. Ready-to-use or custom-designed LNA-modified probes can be purchased from Qiagen (formerly Exiqon; Germantown, Maryland, USA). We custom-designed LNA-modified probes with 50 and 30 terminally FAM moieties for microRNA detection; +N indicates the LNA-modified nucleotide. We custom-designed DNA probes with 50 and 30 terminally biotin moieties for reference RNA detection. For FAM2X or Biotin2X probes, these moieties were attached to an extra terminal “T” nucleotide on both 50 and 30 ends that were not complementary to the microRNA or reference sequence (not shown in table). For FAM4X probes, an internal modification consisting of a 6-FAM dT was inserted next to the terminal extra Ts. The LNA-modified probes were purchased from Integrated DNA Technologies (IDT; Coralville, Iowa, USA) or from Eurogentec (Seraing, Belgium); DNA probes were purchased from IDT. Probe concentration indicates best working concentration for each microRNA on mouse breast and pancreas tissues with conditions used in the main protocol. These concentrations may need to be adjusted for microRNA detection in other tissue types and/or detection conditions could be further optimized by conducting more thorough analysis of combining different probe concentration (10–200 nM), hybridization temperature, and/or wash temperature (45–55  C) as well as time and/or concentration of proteinase K digestion

4. AA/TEA acetylation solution (6 mL container): 66 mM HCl, 0.66% acetic anhydride (v/v), and 1.5% triethanolamine (v/v) in RNase-free water. Prepare fresh just before run. For 5 mL, add 33 μL of acetic anhydride and 75 μL of triethanolamine to 4.892 mL of 66 mM HCl. Shake well to mix thoroughly. Transfer 5 mL to the “AA/TEA” container.

266

Lorenzo F. Sempere et al.

Table 2 Antibody information and general detection conditions Host species

Stock (mg/mL)a

Final (μg/ Vendor Dilutiona mL)

Cat #

Anti-chicken/ HRP

Goat

0.4

1:500

0.8

Santa Cruz Biotechnology

sc-2901

Anti-collagen I

Rabbit

1

1:300

3.33

Abcam

ab34710

Anti-cytokeratin 19

Rat

0.288

1:200

1.44

Hybridoma

Troma-III

Anti-FITCb

Rabbit

Unknown

1:200

Unknown

Dako

P5100

Anti-mouse/HRP Goat

0.5

1:500

1

Biorad

170-6516

Anti-rabbit/HRP Goat

0.5

1:500

1

Biorad

170-6515

Anti-rat/HRP

Goat

1

1:500

2

Abcam

ab7097

Streptavidin/ polyHRPc

n/a

0.5

1:2000

0. 25

Thermo Scientific

21140

Anti-tubulin

Rat

1

1:300

3.33

Abcam

ab6160

Anti-vimentin

Chicken

Unknown

1:300

Unknown

Lifespan Biosciences LS-B291100

Antibody name

a

This table indicates stock concentration of in-use antibodies in our laboratory; these concentrations may vary between lots and batches. It may be required to accordingly calculate dilutions for new antibodies to achieve indicated assay concentration. Antibody dilutions may also need to be adjusted if a different dye is used for staining b Rabbit anti-fluorescein antibody (P5100, DAKO) is already conjugated to HRP. However, the use of a secondary antibody conjugated to HRP provides further enhancement of signal amplification via antibody sandwich c Streptavidin binds directly to biotin; there is no antibody sandwich application as for detection of other indicated markers. The use of poly-HRP-conjugated streptavidin increases stain intensity and also allows to use a more diluted concentration to minimize background stain (see Note 5) Antibody names and details for primary and secondary antibodies used in 4ISHplex protocol described in main protocol and variants. The primary antibodies should recognize target protein in mouse, human, baboon, cat, and other mammalian species

5. 0.5% Triton X (6 mL container): 0.5% Triton X-100 (v/v) in PBS. Add 250 mL of Triton X-100 to 50 mL of PBS (this solution keeps well for 1 month at 4  C). Transfer 5 mL to the “0.5% Triton X” container. 6. PBST solution (50 mL tube preparatory reagent): 0.1% Tween∗20 (v/v) in PBS. Add 50 μL of Tween∗20 to 50 mL of PBS. Mix well and set aside. 7. 3% H2O2 solution (30 mL container): Prepare fresh 3% H2O2 (v/v) in PBST. Add 2 mL of 30% H2O2 to 18 mL of PBST. Mix well. Transfer 30 mL to the “3% H2O2” container. 8. 5% BSA blocking solution (6 mL container): 5% BSA (w/v) in PBST. Add 2.5 g of BSA to 50 mL of PBST (this solution keeps well for a month at 4  C). Mix well. Transfer 5 mL to the “5% BSA” container.

Automated 5-Plex ISH/IHC Assay

267

Table 3 Dye information and general detection conditions Dye substrate

Stock Fluorochrome (mM)a

Dilution

Final (μM)

Green dye

Fluorescein

77

1:200 (1:200–300)

Orange dye

Rhodamine

71

1:400 (1:300–500)

Red dye

Dylight 594

44

1:1000 (1:750–1500)

Near-IR dye

Dylight 650

46.9

1:300 (1:200–400)

Vendor #

Cat #

385–257

Thermo Scientific

46410

237–142

Thermo Scientific

46406

59–30

Thermo Scientific

46412

234–110

Thermo Scientific

62265

a

Stock concentration of in-house fluorochrome-conjugated tyramide substrate generated as described [39]. Briefly, 10 mg/mL of fluorochrome-NHS ester in dimethylformamide (DMF) is allowed to react with an equal molarity of 10 mg/mL tyramine in 1% of triethylamine (v/v)/DMF for 2 h at room temperature and then reaction is stopped with an equal volume of 100% ethanol In a typical five-color multiplex assay, green dye is used to reveal microRNA probe, orange dye for reference RNA, red and near-IR dyes for protein markers, and blue dye for nuclear counterstaining with DAPI. Preferably, anti-marker antibody that yields more intense stain should be used in the last round of HRP-mediated fluorochrome deposition with red dye

9. 5 SSC pre- and hybridization solution (15 mL tube preparatory reagent): 50% Deionized formamide, 5 SSC, 500 μg/ mL yeast tRNA, 1 Denhardt’s solution, 0.1% Tween∗20. For 10 mL, add 2.5 mL of 20 SSC, 1.79 mL of RNase-free water, 500 μL of yeast tRNA (10 mg/mL), 200 μL of 50 Denhardt’s solution, and 10 μL Tween∗20 to 5 mL of freshly thawed formamide. 10. Probe 1 ISH (2  6 mL containers): 50 nM FAM4X-tagged LNA-modified DNA probe against miR-21 and 250 nM biotin 2-tagged DNA probe against U6 snRNA. Dispense 5 μL of 100 μM miR-21 probe and 25 μL of 100 μM U6 probe into a 1.5 mL tube. Boil at 94  C for 1 min and immediately place on ice for 5 min. Dispense denatured probes into 10 mL of 5 SSC hybridization solution in a 15 mL tube. Mix well. Transfer 4.8 mL to each of the two “Probe 1 ISH” 6 mL containers. Please note that Probe 1 ISH solution will be dispensed twice (300 μL per slide) and hence the need to prepare double the volume amount of this reagent (see Note 1). For steps 12–18 and 20–23 concentrated reagents are dispensed directly in the titration tube inside the 6 mL container already with appropriate amount of diluent. 11. PBT solution (50 mL tube preparatory reagent): 1% BSA (w/v), 0.1% Tween∗20 (v/v) in PBS. Add 10 mL of 5% BSA and 50 μL of Tween∗20 to 50 mL of PBS (this solution keeps well for a month at 4  C). Mix well and set aside.

268

Lorenzo F. Sempere et al.

12. Anti-FITC antibody (6 mL container): 1:200 dilution of rabbit anti-FITC/FAM antibody in PBT. Add 25 μL of stock antibody to 5 mL of PBT. 13. Anti-rabbit/HRP antibody (6 mL container): 1:500 Dilution of goat anti-rabbit secondary antibody conjugated to horseradish peroxidase (HRP) in PBT. Add 10 μL of stock antibody to 5 mL of PBT. 14. Streptavidin/HRP (6 mL container): 1:2000 Dilution of streptavidin conjugated to poly-HRP in PBT. Add 2.5 μL of stock streptavidin/poly-HRP to 5 mL of PBT. 15. Marker 1 antibody (6 mL container): 1:300 Dilution of rat anti-tubulin in PBT. Add 16.66 μL of stock antibody to 5 mL of PBT. 16. Anti-Rat/HRP antibody (6 mL container): 1:500 Dilution of goat anti-rat secondary antibody conjugated to HRP in PBT. Add 10 μL of stock antibody to 5 mL of PBT. 17. Marker 2 (6 mL container): 1:300 Dilution of chicken antivimentin in PBT. Add 16.7 μL of stock antibody to 5 mL of PBT. 18. Anti-chicken/HRP antibody (6 mL container): 1:500 Dilution of stock goat anti-chicken antibody conjugated to HRP in PBT. Add 10 μL of stock antibody to 5 mL of PBT. 19. HRP-mediated dye deposition solution (50 mL tube preparatory): 0.015% H2O2 (v/v) in PBST. Add 1 μL of 30% H2O2 to 20 mL of PBST. Prepare fresh just before run. 20. Green dye solution (6 mL container): 1:200 Dilution of stock FITC-conjugated tyramide substrate in HRP-mediated dye deposition solution. Add 25 μL of stock FITC-conjugated tyramide to 5 mL of HRP-mediated dye deposition solution. 21. Orange dye solution (6 mL container): 1:400 Dilution of stock rhodamine-conjugated tyramide substrate in HRP-mediated dye deposition solution. Add 12.5 μL of stock rhodamineconjugated tyramide to 5 mL of HRP-mediated dye deposition solution. 22. Red dye solution (6 mL container): 1:1000 Dilution of stock Dylight 594-conjugated tyramide substrate in HRP-mediated dye deposition solution. Add 16.7 μL of stock Dylight 594-conjugated tyramide to 5 mL of HRP-mediated dye deposition solution. 23. Near-IR dye solution (6 mL container): 1:300 Dilution of stock Dylight 650-conjugated tyramide substrate in HRP-mediated dye deposition solution. Add 16.7 μL of stock Dylight 650-conjugated tyramide to 5 mL of HRP-mediated dye deposition solution.

Automated 5-Plex ISH/IHC Assay

269

2.4 Tissue Processing and Slide Preparation

Normal and tumor breast, pancreas, or other mouse organs were sectioned in 2–3 mm slices and fixed in 10% formalin for up to 24 h and then processed in a Leica PELORIS II Premium Tissue Processing System using a standard 8-h xylene-free isopropanol schedule for paraffin embedding. Briefly, tissues were incubated at 55  C sequentially in formalin for 30 min, in 85% ethanol for 20 min, in 85% ethanol for 30 min, in 80% ethanol for 30 min, in 20% isopropyl alcohol for 60 min, and in isopropyl alcohol for 20 min, 40 min, and 80 min. Then, tissues were incubated at 85  C in wax for 60 min and 50 min and finally at 65  C in wax for 40 min. Fourmicron-thick tissue sections were mounted on Leica Microsystems Bond Plus Slides.

2.5 Equipment for Microscopy

An Aperio Versa 8 Brightfield&Fluorescence imaging system (Leica Biosystems, Buffalo Grove, IL, USA) with customized narrowwidth band excitation and emission filter cubes (Chroma Technology Corp, Bellows Falls, VT, USA): Standard set (49000) for DAPI, standard set (49020) for fluorescein, custom set (ET546/ 10x,T555lpxr,ET570/20x) for rhodamine, custom set (t620lpxr; et630/20) for Dylight 594, and custom set (t665lpxr, et667/ 30 m) for Dylight 650 were used for image acquisition.

3

Methods This method describes fully automated co-detection of miR-21, U6 snRNA, tubulin, and vimentin, which is followed by manual counterstaining with DAPI by slide immersion (Figs. 1 and 2). This four-color fluorescence-based ISH/IHC assay is performed using a single-IHC staining protocol with a customized Open Research Kit (Figs. 3 and 4). We separate in different subsections the preparatory, hybridization, post-hybridization, and detection steps of this single continuous protocol. One hundred and fifty microliters are dispensed per slide at each step of the procedure (see Note 1). The covertile position in all steps is set to close position (dispense type: 150 μL), with the exception of proteinase digestion and hybridization steps set to intermediate position (dispense type: Intermediate), and some of the probe removal steps set to open position (dispense type: Open); see specific sections below for more details.

3.1 Programming 4-Plex ISH/IHC Assay

1. Enter reagents in the machine with names indicated in Subheading 2. Link reagents to appropriate size container (6 mL or 30 mL) and the 30 mL container labeled “Buffer” to the Open Research Kit. 2. Program the protocol with staining method set to single, protocol type set to IHC staining, and preferred detection system set to “Open Research Kit” (see specifications below and

270

Lorenzo F. Sempere et al.

Fig. 3). For this example, we named the protocol ISH4plex (microRNA, snRNA, 2proteins). 3. Add study. Write a descriptive name for study ID; in this example, we used “breast mouse models” (Fig. 4). 4. Set slide preparation to “∗Bake and Dewax.” 5. Add slide. Set run parameters as shown in Fig. 4. 6. Print labels, stick labels to slides, place slides on trays, and load trays in the machine. 7. Prepare reagents and dispense in appropriate containers as described in Subheading 2. 8. Load reagent racks in machine. Start run without delay. Alternatively, run start can be delayed up to 4 h without significant impact on staining results. 3.2 Pre-hybridization Steps

There are three Bond Washes in between steps 2–7; the first wash is set for 2 min and the others for 0 min. 1. Slides are heated at 60  C for 20 min as part of the “∗Bake and Dewax” for slide preparation. 2. Slides are incubated with dewaxing solution as part of the “∗Bake and Dewax” for slide preparation. 3. Slides are incubated with proteinase K digestion solution at 37  C for 15 min. The proteinase K digestion solution is dispensed twice: 0 min at 37  C with 150 μL for dispense type and 15 min at 37  C with intermediate position for dispense type. 4. Slides are incubated with 2% glycine solution for 2 min. 5. Slides are incubated with 4% PFA solution for 10 min. 6. Slides are incubated with AA/TEA acetylation solution for 5 min. 7. Slides are incubated with 0.5% Triton X-100 permeabilization solution for 5 min.

3.3 ISH Hybridization (See Table 1 for Probes)

1. Slides are incubated with hybridization solution containing probes against microRNA and reference RNA for 60 min at 45  C. Please note that 300 μL is required per slide since the probe solution is dispensed in two consecutive steps of 30 min at 45  C with intermediate position for dispense type (see Notes 1 and 3). 2. Slides are washed with Bond Wash at the same temperature of hybridization to remove excess probe and provide stringency washes (see Note 4). Please note the times and selection of dispense type for Bond Wash series: 0 min at RT with 150 μL for dispense type; 0 min at 45  C with 150 μL for dispense type;

Automated 5-Plex ISH/IHC Assay

271

1 min at 45  C with open position for dispense type; 5 min at 45  C with 150 μL for dispense type; 0 min at RT with 150 μL for dispense type; 0 min at RT with 150 μL for dispense type; and 1 min at RT with open position for dispense type. 3. Slides are incubated with “Buffer” linked to Open Research Kit for 2 min. This is a short PBST wash incorporated as part of the probe removal steps. 3.4 Sequential Detection of ISH Probes (See Table 2 for Antibodies and Table 3 for Dyes)

There are three Bond Washes in between steps; the first wash is set for 2 min and the others for 0 min. 1. Slides are incubated with 3% H2O2 solution for 15 min. 2. Slides are incubated with 5% BSA blocking solution for 30 min. Detection of FAM-Tagged microRNA Probe by HRP-Mediated Deposition of Fluorescein-Tyramide (Steps 3–5)

3. Slides are incubated with rabbit anti-FITC 1 antibody solution for 30 min. 4. Slides are incubated with goat anti-rabbit HRP-conjugated 2 antibody solution for 30 min. 5. Slides are incubated with Green dye for 20 min. 6. Slides are incubated with 3% H2O2 solution for 15 min. Detection of Biotin-Tagged Reference RNA Probe by HRP-Mediated Deposition of Rhodamine-Tyramide (Steps 7–8)

7. Slides are incubated with HRP-conjugated streptavidin solution for 30 min. 8. Slides are incubated with Orange dye for 20 min. 3.5 Sequential Detection of Protein Markers (See Table 2 for Antibodies and Table 3 for Dyes)

There are three Bond Washes in between steps; the first wash is set for 2 min and the others for 0 min. 1. Slides are incubated with 3% H2O2 solution for 15 min to inactive HRP from previous steps. Detection of Protein 1 by HRP-Mediated Deposition of Dylight 594-Tyramide (Steps 2–4)

2. Slides are incubated with rat anti-tubulin 1 antibody solution (marker 1) for 30 min. 3. Slides are incubated with goat anti-rat HRP-conjugated 2 antibody solution for 30 min. 4. Slides are incubated with Red dye for 20 min. 5. Slides are incubated with 3% H2O2 solution for 15 min. Detection of Protein 2 by HRP-Mediated Deposition of Dylight 650-Tyramide (Steps 6–8)

272

Lorenzo F. Sempere et al.

6. Slides are incubated with chicken anti-Vimentin 1 antibody solution (“marker 2”) for 30 min. 7. Slides are incubated with goat anti-chicken HRP-conjugated 2 antibody solution for 30 min. 8. Slides are incubated with near-IR dye for 20 min. For detection of other protein markers or marker combinations (see Table 2), add appropriate amount of anti-marker of interest (marker 1 and marker 2) and accordingly make modifications to the following secondary antibody steps (see Note 5). Please see Table 3 for brief description of fluorochrome-conjugated tyramide substrate production. 3.6 Slide Mounting (Manually)

1. Remove slides from machine. Transfer to vertical tray with PBS. The slides can be stored at 4  C for 1 h to overnight to remove excess unbound dye(s) and decrease background signal. 2. Counterstain with DAPI. Immerse slides in 350 nM DAPI in PBS (see Note 6). Incubate for 15 min at room temperature. 3. Wash three times for 5 min with PBS. 4. Dispense one drop of (~20 μL) Prolong Gold (Invitrogen) on top of each tissue section. Seal with 1-mm-thick glass coverslips. Let slides cure overnight before image analysis under fluorescence microscopy.

3.7 Image Acquisition and Computer-Assisted Image Analysis

4

Fluorescent images of each tissue sample were acquired and analyzed with an Aperio Versa 8 Brightfield&Fluorescence imaging system. We followed vendor’s recommendations for image acquisition and analysis. For best imaging results, primary antibody and dye concentrations may need to be adjusted to obtain stains in the same range of signal intensity using similar exposure time depending on the animal species and tissue type (see Note 7). For optimal results, signal of each fluorochrome should be balanced and have a similar intensity with a 50–150 ms exposure using the 20 objective. We used Aperio ImageScope tools for tissue browsing and annotation (Figs. 1 and 5), and Aperio Cellular IF algorithm for cell segmentation and cell classification based on expression levels of selected markers (Fig. 5).

Notes 1. Leica Bond Rx is an open research instrument with more versatility and programmable options than counterparts clinical locked-down system, Bond-MAX. Our protocol is in a complete open configuration programmed as a single continuous IHC protocol. With this configuration, a single container

Automated 5-Plex ISH/IHC Assay

273

linked to the open research kit from which a single dispense per slide is needed is sufficient to validate this IHC protocol. However, pre-programmed Leica protocols for proteinase K digestion, probe hybridization, and probe removal are needed to be adapted for this continuous run. The operator has some restrictions in the volume that can be dispensed per slide; the only option is 150 μL, and the position of the covertile: close (standard dispense type: 150 μL), intermediate, or open. Proteinase K and probe removal protocols we describe are virtually identical to the original Leica protocols. However, dispensing for probe hybridization is slightly different due to these operator restrictions. The Leica pre-programmed protocol dispenses first 150 μL of the probe solution and at a later time forcefully dispenses an additional 75 μL. Our protocol dispenses 150 μL of the probe solution at the beginning of the hybridization step and then dispenses another 150 μL at 30 min of hybridization incubation, replacing the initial 150 μL. 2. Concentration and activity of PK may vary between batches; accordingly adjust dispense volume of enzyme stock. It is advisable to test each new batch between 5 and 20 μg/mL to select best concentration in 10–20-min incubation time at 37  C. 3. Leica recommends to limit heated incubation steps to 60 min. An alternative protocol of single 60–90-min hybridization step can be used to minimize the amount of stock probe needed, but some tissue dehydration may start to occur after 30 min. 4. This series of Bond Wash buffer recapitulates Leica ∗probe removal program, with minor modifications due to operator restriction on the position of the covertile. We also include a wash at RT immediately after the probe hybridization to allow for changing the temperature of probe removal steps. Efficient probe removal is required for optimal staining results. SSC washes do not provide better stringency than Bond Wash buffer, and we do not recommend them. Changes in temperature between 37 and 55  C for hybridization and probe removal steps can be used to decrease or increase stringency. Changes in SSC concentration of hybridization solution (add 1 mL of 20 SSC and 2.36 mL of RNase-free water for 2.5 SSC hybridization solution, and 0.4 mL of 20 SSC and 2.96 mL of RNase-free water for 1 SSC hybridization solution while maintaining constant the amount of other reagents) can also be considered, but for the workflow of a Bond Rx run changing the hybridization temperature is a more convenient modification. Each slide can be incubated at a different temperature as long as the incubation time is kept identical among slides. Thus, 30 different hybridization and washing conditions can be tested in a single run.

274

Lorenzo F. Sempere et al.

5. Detection of FAM-tagged microRNA probe requires an antirabbit HRP secondary antibody to detect the anti-FITC/FAM rabbit antibody in the first round of staining. We can detect a strong protein marker with an anti-rabbit HRP secondary in the fourth round of staining, typically without carryforward spurious stain from microRNA probe. If needed, blocking of previous primary rabbit and secondary anti-rabbit antibodies can be achieved with sequential incubation with 5% rabbit serum (011-000-120, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) and 20 μg/mL of anti-Rabbit IgG goat Fab fragment (111-007-003, Jackson ImmunoResearch Laboratories, Inc). Detection of three or more protein markers can be achieved by using secondary antibodies raised in different host species or using isotype-specific antibodies. Ideally, detection of the most abundant protein, expected to have the strongest signal, should be revealed in the last round of staining. 6. DAPI staining may vary depending on the animal species, tissue type, and cellularity of the tissue (e.g., normal vs. cancerous tissue). The DAPI concentration may need to be adjusted within the range of 200–500 nM for optimal results. An alternative approach to slide immersion, especially if only a few slides 1,400 targets in a given ROI. This readout also allows for the resolution of individual tiles along a transcript, allowing multiple independent measurements per RNA transcript [2]. Preparing FFPE tissues for this assay uses conventional ISH methods and common equipment. Slides are dewaxed and treated with Tris-EDTA pH 9.0 to expose RNA targets. A mild proteinase K digestion is then performed to remove protein bound to RNA and further expose transcripts [9]. To preserve tissue integrity, a brief post-fixation is performed. Tissues are incubated with RNA detection probes overnight at 37  C in buffer-containing blocking agents, formamide, and dextran sulfate (Buffer R, NanoString Technologies). Stringent washes are performed to remove nonspecific probe binding and fluorescently labeled antibodies are added to enable imaging of the tissue. The GeoMx DSP is an automated microscope equipped to image tissue sections and cleave and collect photo-releasable indexing oligos. Once slides are loaded on the GeoMx DSP, a whole-slide fluorescence image is generated using the GeoMx DSP software package. ROIs for DSP profiling are then selected by the user. ROIs can further be segmented into areas of illumination (AOI) based on the fluorescence image (e.g., tumor-enriched compartment or immune cell-enriched compartment). After ROI selection, photocleaving light (385 nm) is projected onto the first user-defined AOI using an automated, programmable, digital micromirror device (DMD). Photo-released indexing oligos are then collected with a microcapillary and deposited in a 96-well PCR plate. This process is repeated for all user-selected ROIs, with several automated sample washing steps between each collection to eliminate AOI-to-AOI cross-contamination. Indexing oligos are then quantified using nCounter Technology to generate digital quantification of spatial RNA expression (Fig. 1).

334

Daniel R. Zollinger et al.

Here, we describe the slide preparation and nCounter readout of the GeoMx RNA assay. For additional details on the GeoMx DSP, the GeoMx Protein assay, and the GeoMx high-plex RNA assay with NGS readout, please see Merritt et al. 2020 [2] or contact NanoString Technologies directly.

2 2.1

Materials Equipment

1. Pipettes for 5–1000 μL. 2. 12-Channel P20 multichannel pipetter. 3. Filter tips (RNase/DNase free). 4. Microtubes. 5. Kimwipes (8.4  4.4 in.). 6. Large Kimwipes (16.6  14.7 in.). 7. Vortex mixer. 8. Benchtop centrifuge or Picofuge. 9. Heated water bath. 10. TintoRetriever Pressure Cooker (Bio SB). 11. In situ hybridization oven with hybridization chamber. 12. Humidity chamber. 13. GeoMx Digital Spatial Profiler (DSP). 14. AeraSeal film (Sigma). 15. Eppendorf 96-Well twin.tec PCR plates, semi-skirted, 250 μL (Fisher Scientific). 16. ALPS 50 plate sealer (Thermo Fisher) with seals. 17. Heat block or thermal cycler.

2.2 Materials for Slide Preparation

1. SuperFrost Plus Slides (Thermo Fisher). 2. Opaque staining jars (Tissue-Tek; see Note 1) or Coplin jars. 3. RNase AWAY (Thermo Fisher). 4. Citr Solv (Decon Labs, Inc.) or xylene. 5. 100% Ethanol 6. 95% Ethanol: Prepare 500 mL of 95% ethanol by adding 25 mL of DEPC-treated water to 475 mL of 100% ethanol. Prepare fresh daily. 7. 1 Phosphate-buffered saline pH 7.4 (PBS): Prepare 1 L of 1 PBS by combining 100 mL of 10 PBS pH 7.4, RNase free (Invitrogen), and 900 mL of DEPC-treated water. 8. 1 Tris-EDTA pH 9.0 buffer (Sigma-Aldrich).

GeoMx RNA Assay

335

9. Proteinase K solution: Prepare proteinase K in 1 PBS. Dilute to 0.1 or 1 μg/mL as noted in Table 2. 10. 10% Neutral-buffered formalin Scientific).

(NBF;

Thermo

Fisher

11. NBF stop buffer: Prepare 24.5 g Tris base (0.1 M) and 15 g glycine (0.1 M). Bring to 2 L with DEPC-treated water. This can be stored at RT for 1 month. 12. 2 Saline sodium citrate (SSC): Prepare 1 L of 2 SSC by combining 100 mL of 20 SSC and 900 mL of DEPC-treated water. 13. 2 SSC–T: Prepare 1 L of 2 SSC by combining 100 mL of 20 SSC, 10 mL of 10% Tween 20, and 890 mL of DEPCtreated water. 14. 4 SSC: Prepare 1 L of 4 SSC by combining 200 mL of 20 SSC and 800 mL of DEPC-treated water. 15. Stringent wash solution: Mix equal parts 4 SSC and 100% formamide. 16. Hybridization solution (NanoString Technologies): GeoMx RNA Probe Mix (75 nM per target), 1 Buffer R, DEPCtreated water (see Table 3). 17. GeoMx Morphology Kit for RNA (NanoString Technologies): 500 nM SYTO13, 1 anti-PanCK, 1 anti-CD45, or applicable markers (see Table 4). 18. GeoMx RNA Slide Prep Kit for FFPE: Buffer S, Buffer W, Buffer R (NanoString Technologies). 19. Buffer S (NanoString Technologies). 20. HybriSlip Hybridization Covers (Grace Bio-Labs). 2.3 Materials for nCounter Readout

1. Adhesive seals for 96-well PCR plates. 2. GeoMx Hyb Code Pack, RNA A–H reagent tubes, in situ capture probe (ICP), nCounter Hybridization Buffer (NanoString Technologies). 3. DEPC-treated water. 4. NanoString nCounter® MAX/FLEX system or SPRINT system.

3

Methods

3.1 Preparing Reagents and Equipment

Take care to maintain nuclease-free conditions as RNase contamination can drastically reduce signal. Oligo contamination is a risk with the GeoMx RNA assay due to the high sensitivity of the nCounter readout. We recommend the use of RNase AWAY (Thermo Scientific), as it will remove oligos, GeoMx RNA detection probes, and nucleases.

336

Daniel R. Zollinger et al.

Fig. 2 Slide tray dimension: Gasket shown in gray. Tissues must be placed in the area shown in green or they will be covered by the gasket

Spray down all Coplin jars, hybridization chamber, humidity chamber, pipettes, and your work area before beginning. Rinse away excess with DEPC-treated water or let air-dry. 3.2 Preparing Tissues

1. Criteria for selecting and sectioning FFPE blocks are discussed in Subheading 4 (see Notes 2 and 3). Unstained tissue sections should be 5 μm thick and mounted on Superfrost Plus slides. Tissue sections should be placed in the center of the slide and be no larger than 36.2 mm long by 14.6 mm wide. If sections are larger than this size and/or placed off-center, it is likely that the tissue will be covered by the gasket on the instrument and not be measured by the GeoMx DSP (Fig. 2). 2. Bake sections on slides in a 60  C drying oven for 30–60 min immediately prior to starting the assay.

3.3

Slide Preparation

1. Preheat the pressure cooker to the temperature specified below in step 3 using the low-pressure setting. Before turning on the pressure cooker, ensure that the water level is above four cups. 2. Deparaffinize and rehydrate FFPE tissue sections. Gently place the slides in a rack and perform the following washes using staining jars. Submerge slides in either CitrSolv or xylene three times for 5 min, followed by 100% ethanol twice for 5 min, 95% ethanol for 5 min, and then 1 PBS for 1 min. Slides can be stored in the final PBS wash for up to 1 h at room temperature. 3. Incubate slides in 1 Tris-EDTA pH 9.0 in the pressure cooker at low pressure. Incubation times and temperatures differ by tissue and may need to be empirically determined (see Table 1 and Note 4). 4. Wash in PBS for 5 min at room temperature. Slides can be stored at this step for up to 1 h at room temperature. 5. Dilute proteinase K to the desired concentration in 1 PBS. Proteinase K concentration and incubation times differ by tissue and may need to be empirically determined (see Table 2

GeoMx RNA Assay

337

Table 1 Target retrieval conditions for various tissue types when using TintoRetriever Pressure cooker Tissue type

Target retrieval

Tonsil

Tris-EDTA for 10 min at 100  C

Colorectal tumor

Tris-EDTA for 20 min at 100  C

Breast tumor

Tris-EDTA for 10 min at 100  C

NSCLC

Tris-EDTA for 20 min at 100  C

Melanoma

Tris-EDTA for 20 min at 100  C

Prostate tumor

Tris-EDTA for 20 min at 100  C

Cell pellets

Tris-EDTA for 10 min at 85  C

Table 2 Proteinase K digestion conditions for various tissue types Tissue type

Proteinase K digest

Tonsil

1 μg/mL for 15 min

Colorectal tumor

1 μg/mL for 15 min

Breast tumor

0.1 μg/mL for 15 min

NSCLC

1 μg/mL for 15 min

Melanoma

1 μg/mL for 15 min

Prostate tumor

1 μg/mL for 15 min

Cell pellets

1 μg/mL for 5 min

and Note 4). Place proteinase K in a Coplin jar and prewarm to 37  C in water bath. Incubate slides in proteinase K at 37  C for 15 min (or modified time according to Table 2). Wash in 1 PBS for 5 min. 6. Postfix the tissue by submerging in 10% NBF for 5 min. Wash two times in NBF stop buffer for 5 min, and then in PBS for 5 min (see Notes 5 and 6). 3.4 Overnight In Situ Hybridization

1. Thaw GeoMx RNA Probe Mix on ice and make the hybridization solution in a microtube (Table 3). Mix thoroughly by pipetting. Avoid all air bubbles (see Note 7). 2. Clean all equipment with RNase AWAY. The hybridization chamber is a key source of contamination by RNA detection probes. Arrange a fresh large Kimwipe on the bottom of the hybridization chamber.

338

Daniel R. Zollinger et al.

Table 3 Hybridization solution Buffer R

RNA Probe Mix

DEPC-treated water

Final volume

(200 μL  n)

(37.5 μL  n)

(12.5 μL  n)

(250 μL  n)

_____

+ _____

+ _____

¼ _____

3. Wet Kimwipe with 2 SSC. Take care that the Kimwipe and 2 SSC do not contact the slides to avoid wicking of the hybridization solution. 4. One at a time, remove slides from PBS, wipe away excess liquid with a fresh Kimwipe, and set in hybridization chamber. Take care not to let the slide dry out. Add 200 μL hybridization solution to each slide. Take care not to introduce any bubbles. Gently apply a Grace Biolabs HybriSlip. Lay down gradually starting from one edge to avoid the formation of air bubbles. 5. Repeat with remaining slides. 6. Close hybridization chamber, place it into the oven, and clamp into place. Incubate at 37  C overnight (16–24 h) (see Note 8). 3.5 Perform Stringent Washes

1. Warm 100% formamide to room temperature before opening. 2. Prepare stringent wash solution by mixing equal parts 4 SSC and 100% formamide to fill two staining jars. Prewarm to 37  C in water bath. 3. Place the slides in 2 SSC-T and allow the coverslips to slide off by themselves. Take care that the coverslips slide off freely without hitting the slides of the Coplin jar containing 2 SSC-T. Once coverslips are removed, submerge in prewarmed stringent wash solution (see Note 9). 4. Wash two times for 25 min in stringent wash solution at 37  C. Wash two times for 2 min in 2 SSC at room temperature. Slides can be stored in the final wash of 2 SSC for up to 1 h.

3.6 Apply Morphology Markers

1. Add 2 SSC to the bottom of the humidity chamber and move one slide at a time to the chamber for antibody staining. Cover tissue with up to 200 μL buffer W and leave at RT for 30 min. Ensure that the humidity chamber is protected from light (see Note 10). 2. Dilute morphology markers and SYTO13 in buffer W according to Table 4. Spin morphology marker solution at max speed in a benchtop centrifuge to remove any morphology reagents that may precipitate. 3. Remove buffer W by tapping the edge of the slide onto a Kimwipe. Add 200 μL of morphology marker mix to each

GeoMx RNA Assay

339

Table 4 Morphology marker mix SYTO 13

Morphology marker 1

Morphology marker 2

Buffer W

Total volume

(22 μL  n)

(5.5 μL  n)

(5.5 μL  n)

(187 μL  n)∗

(220 μL  n)

_____

+ _____

+ _____

+ _____

¼ _____

*If a different number of detection or visualization antibody tubes are used, Buffer W amount needs to be adjusted to bring total volume up to 220 μL per slide

slide. Stain for 1 h in the humidity chamber at room temperature. Ensure that the humidity chamber is protected from light. 4. Wash slides two times for 5 min in 2 SSC. Transfer slides to the GeoMx DSP slide holder and wet with 3 mL of buffer S. Slides can be stored for up to 6 h prior to running on the GeoMx DSP, but must be submerged in 2 SSC, stored at 4  C, and protected from light exposure. 3.7 Using the GeoMx DSP

Operation of the GeoMx DSP will not be covered in depth in this chapter. Please consult the GeoMx DSP Instrument Manual and Merritt et al. (2020) [2] for details. In brief, slides will be scanned and ROIs will be defined. Oligos will be photo-released and collected in a 96-well PCR plate for downstream processing.

3.8 GeoMx Hyb Code Setup

1. An overview of the nCounter readout is shown in Fig. 3 (see Note 11). 2. To ensure that equal volumes of indexing oligos are hybridized for each well, aspirates are dried down and rehydrated in 7 μL of DEPC-treated water (see Note 12). When rehydrating samples, add 7 μL of DEPC-treated water and mix by pipetting. Cover plate with an adhesive seal and let sit for 10–20 min at room temperature. 3. Thaw ICP and the GeoMx Hyb Code reagents required for the number of samples to be hybridized. Dilute ICP in DEPCtreated water (Table 5). Set up the hybridization readout in a separate area than you prepared your RNA probe mix to avoid contamination by the RNA detection probes. 4. Pipette 80 μL nCounter hybridization buffer per GeoMx Hyb Code to be used into a fresh tube. Add 16 μL of ICP per GeoMx Hyb Code to be used into the nCounter hybridization buffer (Table 6). Vortex and spin down in a picofuge. 5. After the GeoMx Hyb Code tubes have been completely thawed, spin them down briefly in a picofuge to bring the contents to the bottoms of the tubes. Add 84 μL of buffer/ ICP mix to each aliquot of GeoMx Hyb Code. Mix by flicking the tubes, NOT vortexing. Spin briefly in the picofuge. These are the individual GeoMx Hyb Code Master Mixes.

340

Daniel R. Zollinger et al.

Fig. 3 Overview of GeoMx RNA Assay for nCounter Hybridization Setup

GeoMx RNA Assay

341

Table 5 ICP working dilution #hyb codes

ICP master stock

Nuclease-free water

Total volume

1

4 μL

29 μL

33 μL

2–3

8 μL

58 μL

66 μL

4–6

14 μL

102 μL

116 μL

7–8

20 μL

145 μL

165 μL

Table 6 ICP/hyb buffer calculation #hyb codes n ¼ ___

ICP working pool

Hybridization buffer

(n  16 μL)

(n  80 μL)

___ μL

___ μL

6. Set up the hybridization reactions in a new 96-well plate with a tight seal that does not allow for evaporation, and incubate overnight at 65  C. Using a heat sealer is preferred (see Note 13). Prior to adding reagents, confirm that this plate seats completely in the heat block/thermocycler that will be used. The heat block/thermocycler should have a heated lid to prevent condensation on the plate seal, preferably floating at +5  C above the block temperature, although a heated lid with a fixed temperature is also acceptable. 7. Pipet 8 μL of each GeoMx Hyb Code Master Mix across the hybridization plate into each of the 12 wells of the appropriate row, matching GeoMx Hyb Code A–H to the respective plate row. Transfer 7 μL of aspirate from GeoMx DSP (see Note 14) to the matching well on the hybridization plate (one sample per well, in the same plate location order as collected). 8. Seal the plate carefully; as stated above, a heat sealer is preferable to prevent evaporation. Quick spin the hybridization plate, spinning just long enough to reach 2000  g. Incubate the plate at 65  C for 16–24 h in a thermocycler with a heated lid. If the thermocycler is programmable, it can be set to ramp down to 4  C indefinitely after the hybridization. 9. Following the overnight nCounter hybridization, quick spin the plate, spinning just long enough to reach 2000  g. Pool the samples by columns down the plate and into a strip tube (Fig. 3). Mix the final pool by gently pipetting up and down five times. Cap the strip tube, and quickly spin down.

342

Daniel R. Zollinger et al.

10. Load the strip tube with the pooled samples on the MAX/FLEX Prep Station or SPRINT workstation. For the MAX/FLEX Prep Station, use the entire pool. 11. For the SPRINT, load 30 μL from each sample well into the corresponding lane on a SPRINT cartridge. If running only one Hyb Code, add 15 μL of nuclease-free water to the hybridizations before loading the entire volume into the SPRINT cartridge. If there is remaining pooled material (i.e., if using three or more Hyb Codes), the hybridized samples can be frozen and saved at 80  C in case a rerun is required. 12. Transfer the GeoMx DSP Reporter Library File (RLF) “DSP_v1.0.rlf” (or current version) to the platform for the nCounter run. 13. If using the MAX/FLEX, transfer the cartridge to the Digital Analyzer and scan as per the manufacturer’s instructions using a CDF generated by the GeoMx DSP software. 14. Data analysis will also be performed in the GeoMx DSP software and data can be exported for analysis in other programs.

4

Notes 1. Clear glass Coplin jars should not be used unless wrapped in foil to occlude light. We recommend the use of opaque plastic staining dishes such as the Tissue-Tek staining dish or plastic Coplin jars. 2. FFPE blocks should meet the following criteria for the best performance with the GeoMx RNA assay. (a) Prior to embedding, tissues should be fixed in 10% neutral-buffered formalin (NBF) for 18–24 h at room temperature. This applies to tissues less than 0.5 cm in thickness. Larger tissues have not been tested by NanoString. (b) Tissues should be fixed immediately after excision for best results. Up to 1-h post-excision is acceptable. (c) Tissues should be thoroughly dehydrated in ethanol gradients prior to embedding in paraffin. (d) FFPE blocks should be stored at room temperature in a desiccator. (e) NanoString does not recommend the use of FFPE blocks that are greater than 10 years old. 3. It is important to avoid any scratches or folds in the section. These scratches and folds can be exacerbated by the subsequent slide washes on the GeoMx DSP, resulting in tissue loss. The following are general guidelines for sectioning FFPE blocks for

GeoMx RNA Assay

343

optimal GeoMx RNA assay performance. This is not meant to be an all-inclusive guide on sectioning. Please refer to your local pathologist or core for training on sectioning. (a) Sections should be cut at 5 μm thickness on a calibrated microtome. (b) Always discard the first few sections from the block face. (c) Sections should be mounted in the center of the slide to allow room for the gasket on the GeoMx DSP slide holder (Fig. 2). (d) NanoString recommends the use of SuperFrost Plus slides. (e) If mounting multiple sections per slide, ensure that all tissues are at least 5 mm from the edge of the gasket and 2–3 mm apart. (f) Mounted slides should be allowed to air-dry overnight. Store slides in a vertical position such that any remaining water can drain away from the tissue section. (g) Any water trapped under the wax or tissue section should be removed by gently touching a folded Kimwipe onto the corner of the wax section. The Kimwipe should not contact the tissue. (h) Sections should be used within 2 weeks of sectioning for optimal signal. Slides should be stored at room temperature or 4 ºC in a desiccator prior to processing. 4. Baking, target retrieval, and proteinase K conditions were determined based on blocks meeting the constraints outlined in Note 1. These conditions may vary by sample, the tumor content of the sample, and other factors. These conditions were optimized for large tumor sections and may not apply to arrayed tissues, cored tissues, and needle biopsies. If conditions have not been empirically determined, use 60-min baking at 60  C, 15-min incubation at 100  C in Tris-EDTA pH 9.0, and 15-min incubation with 1 μg/mL proteinase K. 5. Before proceeding to the overnight in situ hybridization, visually inspect tissues throughout the assay to verify that the tissue is intact. Tissue detachment is a high risk for the GeoMx RNA assay, as tissues are subject to multiple washes on the GeoMx DSP. 6. We have validated the automation of Slide Preparation steps 4 through 9 on a Leica BOND RXm. The protocol is available through NanoString. 7. RNA detection probes should not undergo more than two freeze-thaw cycles. Probes arrive frozen and should be stored at 20  C until the initial use. After thawing, store at 4  C for

344

Daniel R. Zollinger et al.

several months. Probe mixes should be assembled in an area separate from any nCounter work, NGS library prep, or other GeoMx DSP workflows. GeoMx detection reagents can contaminate results. Areas should be cleaned thoroughly with RNase AWAY after probe mix formulation. Alternately, mixes can be made in PCR workstations and decontaminated with UV light. Gloves should also be changed after handling any probe mixes to avoid contamination. 8. Prepare the hybridization chamber according to product instructions. If your chamber is light permeable, minimize light exposure (e.g., by wrapping the lid in aluminum foil). NanoString has tested and recommends the following hybridization ovens: Boekel Scientific InSlide Out Slide Hybridizer, ACD HybEZ II Hybridization System, and the Boekel Scientific RapidFISH Slide Hybridizer. 9. Forcibly removing coverslips will damage the tissue. Place slides in SSC-T to help loosen the coverslips. Soak coverslips for up to 5 min or until they freely fall off the slide. If coverslips still do not fall off freely, move to stringent wash at 37  C. If coverslips still are not removed after the first stringent wash, discard the slide. 10. When staining with morphology markers, ensure that tissues are adequately covered with buffer W. Ensure that the edges of tissues will remain covered throughout the 1-h incubation. Hydrophobic pen can be applied to the slide but MUST be removed with a razor blade before placing on the GeoMx DSP. 11. This section only applies to the nCounter readout of GeoMx RNA Assays. Protocols for the GeoMx Protein Assay readout are available from NanoString Technologies. Protocols for the NGS-based readout will be made available at the time of commercialization. 12. There are multiple options for drying down GeoMx DSP aspirates. The simplest is to cover the plate with an AeraSeal and leave it at room temperature overnight. A quicker method is to place the plate in a thermocycler set to 65  C with the lid open until the wells are dry (30–60 min). An AeraSeal is recommended to avoid introducing contamination. 13. Heat sealers are the only method validated by NanoString to avoid evaporation during the overnight incubation at 65  C. If a heat sealer is not available, we advise testing other seals for evaporation before using with the GeoMx Assay. 14. For the highest quality data in the GeoMx RNA assay, we recommend loading the full 7 μL volume of the GeoMx DSP aspirate.

GeoMx RNA Assay

345

References 1. Chung JY, Braunschweig T, Williams R et al (2008) Factors in tissue handling and processing that impact RNA obtained from formalin-fixed, paraffin-embedded tissue. J Histochem Cytochem 56(11):1033–1042. https://doi.org/10. 1369/jhc.2008.951863 2. Merritt CR, Ong GT, Church S et al (2020) Multiplex digital spatial profiling of proteins and RNA in fixed tissue. Nature Biotechnology (in press) https://doi.org/10.1101/559021 3. Toki MI, Merritt CR, Wong PF et al (2019) High-plex predictive marker discovery for melanoma immunotherapy treated patients using Digital Spatial Profiling. Clin Can Res. https:// doi.org/10.1158/1078-0432.CCR-19-0104 4. Nanostring Technologies, Inc. MAN-1008703, GeoMx Digital Spatial Profiler Online User Manual 5. nCounter® Expression CodeSet Design Manual MAN-C0002-03. https://www.nanostring.

com/download_file/view/247/3779. Accessed 25 June 2019 6. Raj A, van den Bogaard P, Rifkin SA et al (2008) Imaging individual mRNA molecules using multiple singly labeled probes. Nat Methods 5 (10):877–879. https://doi.org/10.1038/ nmeth.1253 7. Geiss GK, Bumgarner RE, Birditt B et al (2008) Direct multiplexed measurement of gene expression with color-coded probe pairs. Nat Biotechnol 26(3):317–325 8. Nanostring User manual: PlexSet™ Reagents for Gene Expression MAN-10040-05. https:// www.nanostring.com/download_file/view/ 972/3778. Accessed 25 June 2019 9. Nuovo GJ, Elton TS, Nana-Sinkam P et al (2009) A methodology for the combined in situ analyses of the precursor and mature forms of microRNAs and correlation with their putative targets. Nat Protoc 4(1):107–115. https:// doi.org/10.1038/nprot.2008.215

Part VII Target Selective Methods and Single Molecule Detection

Chapter 22 In Situ Point Mutation Detection in FFPE Colorectal Cancers Using the BaseScope Assay Ann-Marie Baker and Trevor A. Graham Abstract In situ mutation detection (ISMD) is a powerful tool for the characterization of tumor heterogeneity at cellular resolution while preserving tissue morphology and spatial context. The BaseScope assay is a novel approach to ISMD, offering excellent specificity and sensitivity, with little requirement for assay optimization or technical expertise. Here we describe the validation and application of BaseScope ISMD probe sets to human formalin-fixed paraffin-embedded (FFPE) samples, firstly by testing the probes in wellcharacterized cell lines of known mutational status, and then by applying the assay to archival FFPE colorectal cancers. Key words BaseScope, Colorectal cancer, In situ hybridization, In situ mutation detection, Tumor heterogeneity

1

Introduction Genetic intra-tumor heterogeneity (ITH) is a major underlying cause of therapy resistance and tumor recurrence in the clinic [1], yet many ITH studies have been limited to the analysis of “bulk” tumor extractions or flow cytometry-sorted single cells. These methods destroy the tumor morphology and architecture; therefore analysis of genetic heterogeneity within the native spatial and microenvironmental context is no longer possible. Direct in situ visualization of point mutations can circumvent these issues, offering a unique opportunity to study ITH and subclonal architecture while preserving tumor morphology. BaseScope is a commercially available RNA-in situ hybridization (RNA-ISH) technique that can be used for in situ mutation detection (ISMD) in human tumors [2]. The novel BaseScope technology (developed by Advanced Cell Diagnostics, ACD, see Fig. 1 for workflow) is based upon the well-established RNAscope method, which is extensively used for in situ mRNA expression studies [3]. However, whereas RNAscope uses up to 20 pairs of

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_22, © Springer Science+Business Media, LLC, part of Springer Nature 2020

349

350

Ann-Marie Baker and Trevor A. Graham

Fig. 1 Flowchart outlining the main steps involved in performing the BaseScope assay on FFPE samples

BaseScope in FFPE Samples

351

Fig. 2 Schematic diagram showing the binding of ZZ probes to a point mutation on an mRNA transcript, allowing the sequential hybridization of a signal amplification tree. Detection is performed using Fast Red chromogenic reagent

DNA oligonucleotide “Z” probes, usually designed to hybridize along the length of an mRNA of interest, BaseScope relies upon the selective binding of a single pair of “Z” probes to wild-type or mutated mRNA transcripts. It is crucial that the wild-type ZZ probes and mutant ZZ probes bind to their respective mRNA transcripts (that differ by only one nucleotide) with absolute specificity. Upon the binding of a pair of ZZ probes to target mRNA, a series of further complementary DNA oligonucleotides are hybridized, creating a large signal amplification “tree” (Fig. 2). In this chapter we outline the steps involved in the validation of BaseScope point mutation-specific probes for the purpose of ISMD in human colorectal cancers. Firstly, we describe the generation of FFPE blocks for cell lines of known mutational status, and then the application of BaseScope to these cell lines. If the probes are found to be sensitive and specific in the cell line setting, then they are next tested in human tumors of known mutational status before finally being approved for the analysis of tumors of unknown mutational status. We highlight the key methodological differences when running the BaseScope assay on cell lines versus tumor samples, and also describe analysis and interpretation of BaseScope results derived from experimental controls and clinical samples.

2 2.1

Materials Equipment

1. Tissue culture incubator. 2. Centrifuge (capable of holding 15 mL and 50 mL conical tubes). 3. Roller mixer. 4. Microwave. 5. Microtome. 6. Fume hood.

352

Ann-Marie Baker and Trevor A. Graham

7. Hot plate. 8. HybEZ oven (ACD). 9. Dry oven. 10. 175 cm2 Tissue culture flasks. 11. 15 mL Conical tubes. 12. 50 mL Conical tubes. 13. Parafilm. 14. Ice pack. 15. Slide holders and staining racks. 16. HybEZ tray (ACD). 17. EZ-batch slide holder and wash tray (ACD). 18. Kimwipes. 19. Superfrost plus slides. 20. ImmEdge hydrophobic barrier pen (ACD). 21. HybEZ humidifying sheets (ACD). 22. Coverslips. 2.2

Reagents

1. Cell growth media. 2. Sterile phosphate-buffered saline (PBS). 3. Trypsin-EDTA 0.25%. 4. 10% Neutral-buffered formalin (NBF) 5. Histogel (Fisher Scientific). 6. RNase-free water. 7. RNaseZap. 8. Xylene. 9. Ethanol. 10. BaseScope v2 reagent kit (ACD, containing the following: ready-to-use hydrogen peroxide, 10 target retrieval solution, ready-to-use protease IV, 20 wash buffer, detection reagents [AMP1-8 and Fast Red A + B]). 11. Target retrieval solution (dilute the 10 target retrieval solution in ddH2O). 12. Wash buffer (dilute the 20 wash buffer in ddH2O). 13. Fast Red solution (add Fast Red B to Fast Red A at a 1:60 ratio). 14. BaseScope 1ZZ positive control probe (ACD, human POLR2A). 15. BaseScope 1ZZ negative control probe (ACD, dapB).

BaseScope in FFPE Samples

353

16. BaseScope probes of interest (ACD, e.g., KRAS G12 V and KRAS nt35 wild type). 17. Hematoxylin (Gill No. 2 [Sigma-Aldrich] at a 1:5 ratio in ddH2O). 18. Ammonia water (0.02% ammonium hydroxide in ddH2O). 19. Vectamount (Vector Labs).

3

Methods

3.1 Preparation of FFPE Cell Pellets

1. Identify a cell line carrying the mutation of interest, and a cell line that is wild type at that locus. This information can be found on the ATCC (American Type Culture Collection) website, on the COSMIC (Catalogue of Somatic Mutations in Cancer) database, or by performing a literature search (see Note 1). 2. Thaw the cell lines and culture in appropriate growth conditions. Follow the cell line-specific guidelines that can be found on the ATCC website. 3. Once the cells have expanded to cover a 175 cm2 flask at approximately 70–80% confluency, remove growth media by aspirating, wash cells in 20 mL sterile PBS, and then trypsinize cells by adding 5 mL of trypsin to cover the base of the flask. Transfer the flasks to a 37  C incubator until the cells have detached (approximately 5 min, time required is dependent on the cell line). 4. Neutralize the trypsin by adding 15 mL cell culture media and mix by pipetting up and down. 5. Transfer the suspension of cells to a 50 mL conical tube, and then centrifuge at 250  g for 10 min. Ensure that the centrifuge is balanced for all centrifugation steps. 6. Carefully aspirate the media without disturbing the cell pellet. Resuspend the cell pellet in 30 mL sterile PBS and centrifuge again at 250  g for 10 min. 7. Carefully aspirate the PBS without disturbing the cell pellet. Resuspend cells in 30 mL 10% NBF. 8. Place the tubes on a roller mixer for 16 h at room temperature. 9. Centrifuge at 250  g for 10 min. Carefully aspirate the NBF without disturbing the cell pellet, and then resuspend cells in 15 mL PBS. At this point transfer the cell suspension to a 15 mL conical tube. 10. Centrifuge at 250  g for 10 min. Carefully aspirate the PBS without disturbing the cell pellet and place the 15 mL conical tube in a glass beaker of warm water.

354

Ann-Marie Baker and Trevor A. Graham

11. Place a vial of Histogel in a glass beaker of room temperature water, and then heat in the microwave until the gel has melted (approximately 1–2 min at full power). 12. Resuspend the cell pellet in 200 μL of melted Histogel. Place the 15 mL conical tube in a 50 mL conical tube containing 10 mL warm water. 13. Centrifuge at 250  g for 5 min. Carefully remove the upper layer of Histogel and discard. 14. Pipette the lower layer onto a piece of parafilm placed on top of a cold ice pack, taking care not to introduce bubbles. 15. Leave the pellet on the ice pack for 10 min, or until the Histogel has solidified. Then carefully transfer the solidified pellet to PBS and process into an FFPE block using standard histological procedures (see Note 2). 3.2 Preparation of FFPE Tumor Samples

1. Place a piece of freshly dissected tissue into a large volume (e.g., 20 mL) of 10% NBF for 24 h at room temperature. 2. Remove NBF and place the fixed tissue into PBS. 3. Process the sample into an FFPE block using standard histological procedures.

3.3 Running the BaseScope Assay

This protocol can be followed for FFPE cell pellets and for FFPE colorectal tumors, with the appropriate modifications at steps 14 and 15. 1. Prepare sections by trimming the FFPE block using a standard microtome and a fresh blade. Discard the first few sections, and then cut the required number of sections at 5 μm thickness (for comments on section thickness see Note 3). 2. Float sections in a bath containing RNase-free water at 40  C, then collect sections onto Superfrost Plus slides, and air-dry overnight at room temperature (see Note 4). 3. Bake slides in a preheated dry oven at 60  C for 1 h. During this time, switch on the HybEZ oven and set to 40  C. Place the HybEZ tray inside the oven with a damp HybEZ humidifying sheet. Prepare 1 target retrieval solution (dilute the 10 target retrieval solution from the BaseScope v2 reagent kit with RNase-free water) and preheat to 100  C on a hot plate. 4. Place slides in a staining container. Dewax for 5 min by immersing the slides in fresh xylene with gentle agitation. Remove the xylene, replace with fresh xylene, then incubate for a further 5 min with gentle agitation. 5. Place slides in a staining container containing fresh 100% ethanol and incubate for 2 min with gentle agitation. Remove ethanol, replace with fresh 100% ethanol, and then incubate at room temperature for a further 2 min with gentle agitation.

BaseScope in FFPE Samples

355

6. Remove slides from ethanol and place in a preheated dry oven at 60  C for 2 min. 7. Allow slides to cool to room temperature on the bench, then apply hydrogen peroxide reagent (ready-to-use from the BaseScope v2 reagent kit) to completely cover the tissue section, and incubate for 10 min at room temperature. 8. Place slides in a staining container, and then wash in RNase-free water for 1 min with gentle agitation. Remove water, then replace with fresh RNase-free water, and wash for a further 1 min with gentle agitation. 9. Carefully add the slides to the preheated 1 target retrieval solution and maintain temperature at 100  C for 15 min (for comments on target retrieval see Note 5). 10. Remove the slides from the target retrieval and immediately place in a staining container with RNase-free water for 1 min with gentle agitation. Remove water, then replace with fresh RNase-free water, and wash for a further 1 min with gentle agitation. 11. Immerse the slides in 100% ethanol for 30 s with gentle agitation. 12. Remove slides from ethanol and place in a dry oven at 60  C for 2 min. 13. Use the ImmEdge hydrophobic barrier pen to draw a barrier around each section, ensuring that the pen does not touch the tissue section. Allow the barrier to dry for 5 min at room temperature. At this point, remove the BaseScope probes from storage at 4  C and allow to equilibrate to room temperature for 30 min. 14. Apply protease III (for cell pellets, ready-to-use from the BaseScope v2 reagent kit) or protease IV (for tumors, readyto-use from the BaseScope v2 reagent kit) to completely cover the tissue section. 15. Place the slides onto the preheated HybEZ tray within the HybEZ oven at 40  C for 15 min (for cell pellets) or for 30 min (for tumors, see Note 6). 16. Wash the slides in RNase-free water for 1 min with gentle agitation. Remove water, then replace with fresh RNase-free water, and wash for a further 1 min with gentle agitation. 17. Without touching the tissue sections, use a Kimwipe to remove excess wash buffer from tilted slides. Work quickly to ensure that the tissue sections do not dry out. Apply the BaseScope probe of interest (e.g., KRAS G12V, KRAS nt35 wild type, POLR2A, or dapB) to completely cover the tissue section, and

356

Ann-Marie Baker and Trevor A. Graham

place slides onto the preheated HybEZ tray within the HybEZ oven. Incubate for 2 h at 40  C. At this point, remove the AMP 1–8 reagents from storage at 4  C and allow to equilibrate to room temperature. 18. Wash the slides in 1 wash buffer for 2 min with gentle agitation. Remove wash buffer, then replace with fresh 1 wash buffer, and incubate for a further 2 min with gentle agitation. 19. Without touching the tissue sections, use a Kimwipe to remove excess wash buffer from tilted slides. Apply AMP 1 reagent (from the BaseScope v2 reagent kit) to completely cover the tissue section and place slides into the HybEZ oven. Incubate at 40  C for 30 min. 20. Repeat steps 18 and 19, working through AMP 2–8 with the following conditions: AMP 2 for 30 min at 40  C, AMP 3 for 15 min at 40  C, AMP 4 for 30 min at 40  C, AMP 5 for 30 min at 40  C, AMP 6 for 15 min at 40  C, AMP 7 for 30 min at room temperature, and AMP 8 for 15 min at room temperature. 21. After the AMP 8 incubation, perform the final washes in 1 wash buffer as described in step 18. 22. Prepare the appropriate volume of Fast Red reagent. Approximately 100 μL ( 50 μL) is required per section, depending on the area within the hydrophobic barrier. Remove any remaining wash buffer with a Kimwipe and detect the BaseScope signal by adding Fast Red reagent to completely cover the tissue section. 23. Allow signal to develop by incubating the slides for 10 min at room temperature in the dark. 24. Remove Fast Red reagent by tapping, and then wash slides three times in ddH2O (1 min for each wash). 25. Immerse the slides in diluted hematoxylin for 20 s with gentle agitation. Remove the counterstain by washing slides three times in ddH2O (1 min each, see Note 7). 26. To blue the counterstain, place slides in ammonia water for 20 s with gentle agitation. When preparing the ammonia water, concentrated ammonium hydroxide should be used only in a fume hood. 27. Perform final three washes in ddH2O (1 min each) with gentle agitation. 28. Remove the slides from ddH2O and place in a dry oven at 60  C for 15 min.

BaseScope in FFPE Samples

357

29. Once the slides have cooled to room temperature, add 1–3 drops of Vectamount and carefully lower a single coverslip over the cell pellet or tumor section. Press down firmly to remove any air bubbles. 30. Allow the Vectamount to dry at room temperature before imaging and quantifying the signal. 3.4

Analysis

1. Quantification of the number of positive signals per cell can be performed manually, or by using a software package such as ImageJ, HALO (Indica Labs), or Visiopharm. 2. ACD guidelines recommend that the number of signals per cell in sections incubated with 1ZZ negative control probe (dapB) should be less than 1 dot for every 20 cells. In practice we believe that it should be much lower than this: on average we found a signal in 0.29% (0.20%, n ¼ 11) of cells in cell pellets, and 0.18% (0.21%, n ¼ 5) of cells in human tumors [2]. If there is considerable signal on the dapB negative control slides then further analysis should be curtailed and optimization of the pretreatment conditions may have to be performed. 3. The number of signals from the 1ZZ positive control probe (POLR2A) should be high, particularly in cell pellets, where many cells will contain dot clusters. In FFPE cell pellets we found that 99.1% of cells (0.4%, n ¼ 11) contain positive signal, and in FFPE tumors there was 66.0% (24.8%, n ¼ 4) of cells containing positive signal for POLR2A [2]. If there is very low signal from the POLR2A control slide then the pretreatment conditions may need further optimization. However, it is also possible that low POLR2A signal may indicate that the quality of the FFPE block is very poor and therefore the sample is not suitable for BaseScope analysis. 4. If initial analysis reveals the BaseScope signal on the positive and negative control slides to be as expected, then the signal from the wild-type and mutant probes of interest can be considered. We recommend beginning with analysis of cell lines of known mutational status (see Fig. 3). We expect the wild-type cell line to contain signal only for the wild-type probe. A very small amount of signal from the mutant probe may be acceptable: we found that 0.08% (0.08%, n ¼ 9) of wild-type cells contained such nonspecific signal [2]. The mutant cell line should contain signal from only the mutant probe (if a homozygous cell line), or from both the wild-type and mutant probes (if a heterozygous cell line). We found that homozygous mutant cell lines contain a nonspecific wild-type signal in 0.19% (0.27%, n ¼ 4) of cells (see Note 8) [2].

358

Ann-Marie Baker and Trevor A. Graham

Fig. 3 Representative images of the validation of the KRAS G12 V probe set in FFPE wild-type (a) and homozygous mutant (b) cell lines, and archival FFPE wild-type (c) and heterozygous mutant (d) tumors. Each sample was stained using a negative control probe (dapB), a positive control probe (POLR2A), the KRAS wildtype probe, and the KRAS G12V mutant probe. Probe binding is visualized as punctate red dots. Scale bars represent 50 μm and 10 μm (inset)

5. If the specificity and sensitivity of the probes in cell lines are acceptable then we progress to testing the probes in FFPE tumors. Ideally, we begin by analyzing tumors that are known to carry the mutation of interest, and tumors known to be wild type at the locus of interest (see Fig. 3). Finally, if the specificity of the probes is acceptable in human tumors of known mutation status, we then begin to analyze tumors of unknown mutational status (see Note 9).

4

Notes 1. Cell line selection. It is desirable to use a homozygous mutant cell line rather than a heterozygous mutant cell line, so that the specificity of the wild-type probe can also be ascertained.

BaseScope in FFPE Samples

359

2. Storage of fixed cell pellets. It is preferable to process the cell pellet immediately; however if it is necessary to store the pellet for >24 h before processing, the pellet can be kept in 70% ethanol. 3. Thickness of tissue sections. In our experience, FFPE sections as thin as 3 μm can be processed. This may be advantageous for tumor samples as there will be fewer overlapping nuclei in the stained section; however probe hybridization will be reduced as there will be fewer target sites available. 4. Number of sections per slide. For small samples it may be desirable to place more than one tissue section per slide, in order to hybridize the probe of interest and the control probes on serial sections placed on one slide. 5. Target retrieval. In our experience, exceeding the recommended temperature of the target retrieval solution results in nonspecific probe hybridization, and reducing the temperature results in a loss of specific probe hybridization. Therefore, careful monitoring is required to maintain the optimal temperature of the target retrieval solution. Small bubbles should be visible, but vigorous boiling should be avoided. 6. The EZ-Batch system. If using the EZ-Batch system then the slides can be transferred to the EZ-Batch slide holder at this stage. The slides can then remain in the EZ-Batch slide holder, and all incubations and washes are performed in the holder. We recommend the use of this system to facilitate the washing steps. 7. Counterstaining. At this point the counterstain can be examined under a light microscope and if a stronger counterstain is required then the slides can be reimmersed for a further 10–20 s, followed by washing in ddH2O. 8. BaseScope signal in cell lines. The amount of wild-type and mutant signal is highly variable between target genes of interest, and in cell line pellets it is likely to be largely dependent on the amount of that transcript present in the cells. 9. BaseScope signal in tumors. As with cell lines, there is a lot of variability in the BaseScope signal; however it can be even more pronounced in archival tumors as there are further factors to consider, such as poor adherence to tissue processing protocols, suboptimal storage of FFPE blocks, and degradation of mRNA over time.

360

Ann-Marie Baker and Trevor A. Graham

Acknowledgments We thank Advanced Cell Diagnostics for providing ongoing BaseScope technical support, and the BCI histopathology team for expert tissue handling and processing. References 1. Gay L, Baker AM, Graham TA (2016) Tumour cell heterogeneity. F1000Res:5. https://doi. org/10.12688/f1000research.7210.1 2. Baker AM, Huang W, Wang XM et al (2017) Robust RNA-based in situ mutation detection delineates colorectal cancer subclonal evolution.

Nat Commun 8(1):1998. https://doi.org/10. 1038/s41467-017-02295-5 3. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14(1):22–29. https://doi.org/10.1016/ j.jmoldx.2011.08.002

Chapter 23 Using In Situ Padlock Probe Technology to Detect mRNA Splice Variants in Tumor Cells Lilli Hofmann, Thomas Kroneis, and Amin El-Heliebi Abstract Advanced prostate cancer (PC) patients commonly receive anti-hormonal drugs targeting the androgen receptor (AR) signaling pathways. However, almost all patients acquire therapy resistance that can be caused by AR amplification or expression of AR splice variant 7 (AR-V7). Therefore, AR-V7 and AR expression are potential biomarkers for early detection of therapy resistance. Here, we present our padlock probe (PLP)based approach for the in situ detection of AR full length, AR-V7, and prostate-specific transcripts in PC cell lines, which is applicable for circulating tumor cells (CTCs) isolated from cancer patients. First, PC cell lines are seeded on glass slides. Then, cDNA is created using target-specific reverse transcription primers. PLPs are hybridized to the cDNA and ligated to form circular single-stranded DNA molecules. The PLP sequence is ligated and amplified by rolling circle amplification and the resulting rolling circle products can be detected using fluorescently labeled probes. Quantification can be automated using the image analysis software CellProfiler. Key words Androgen receptor, AR-V7, CTC, In situ, mRNA, Padlock probe, Prostate cancer, Single cell, Splice variant

1

Introduction Anti-hormonal drugs targeting the androgen receptor (AR) are the standard-of-care treatment for advanced prostate cancer (PC) [1]. Despite a good initial response, most patients acquire resistance. The resistance to anti-hormonal therapy has been linked to androgen receptor alterations, such as AR amplification or expression of AR splice variant 7 (AR-V7) [2]. Prostate-specific antigen (PSA) has been used as a prognostic biomarker in prostate cancer for decades [3]. In recent years, the potential of new biomarkers has been investigated extensively. The detection of AR amplification or expression of AR-V7 might uncover an evolving resistance to anti-hormonal therapies [2, 4, 5]. Moreover, as chemotherapy seems to be more effective than anti-hormonal drugs in

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_23, © Springer Science+Business Media, LLC, part of Springer Nature 2020

361

362

Lilli Hofmann et al.

AR-V7-positive patients, AR-V7 might prove useful as a treatment selection biomarker [6]. Here, we present a detailed protocol for the in situ detection of PSA, AR full length (AR-FL), and AR-V7 transcripts in PC cell lines. Notably, this protocol can easily be adapted to be used on circulating tumor cells (CTCs) [7]. Our approach is based on in situ padlock probe (PLP) technology, a robust and specific method for the in situ detection of mRNA molecules, that has been described in depth by Krzywkowski and Nilsson [8–10]. An overview of the whole in situ padlock procedure is outlined in Fig. 1b.

Fig. 1 In situ padlock probe (PLP) technology for the detection of mRNA molecules. (a) PLPs consist of two target complementary hybridization arms (15–20 nucleotides) at the 30 - and 50 -ends. The phosphorylation at the 50 -end is necessary for ligation. The central linker region (approximately 40 nucleotides) contains a reporter sequence that can be targeted by detection probes. (b) For the in situ detection of mRNA, cDNA is synthesized during reverse transcription. RNase H degrades the mRNA template. PLPs are hybridized to the cDNA and circularly closed by enzymatic ligation. The PLP acts as template for rolling circle amplification (RCA). Fluorescently labeled detection probes are hybridized to the resulting rolling circle product (RCP) and RCPs can be detected as bright fluorescent spots

Padlock Probes to Detect mRNA Splice Variants

363

CTCs can be isolated from patients and transferred to glass slides by cytocentrifugation [7]. Here, we focus on PC cell lines as they are a suitable model system and should also be used as a control when analyzing CTCs. Cultured cells can be seeded directly onto glass slides. After formaldehyde fixation of the cells, targetspecific primers are used for reverse transcription (RT). Part of these RT primers contain locked nucleic acid (LNA) modifications. These modifications cause enhanced binding affinity and stabilize RNA/DNA hybrids, thereby improving the efficiency of cDNA synthesis [8, 10]. The newly synthesized cDNA molecules are chemically cross-linked to their surroundings by formaldehyde postfixation. Then, mRNA templates are degraded by RNase H, an endonuclease specifically hydrolyzing RNA phosphodiester bonds in RNA/DNA duplexes. However, the part of the mRNA hybridized to LNA-modified primers is protected from degradation and the cDNA remains physically linked to the template mRNA [8, 10]. Single-stranded cDNA is now accessible for the hybridization of PLPs. PLPs are 50 -phosphorylated oligonucleotides consisting of two 15–20-nucleotide-long target complementary hybridization arms at the 30 - and 50 -ends, and a central linker region of approximately 40 nucleotides (see Fig. 1a). Upon hybridization to their target cDNA, the hybridization arms are threaded around the cDNA in a typical helical structure. PLPs form a loop and the 30 - and 50 -ends of PLPs bind juxtaposed to each other [10]. Only after a perfect hybridization will the PLPs be ligated by Ampligase DNA ligase to become circularly closed molecules [8]. Then, the circularized PLPs are amplified by isothermal, target-primed rolling circle amplification (RCA). Phi29 DNA polymerase recognizes and binds to free 30 -ends of the cDNA. With its 30 –50 exonuclease activity, Phi29 DNA polymerase degrades the single-stranded cDNA until it reaches the hybridized PLP. Then, the remaining cDNA acts as a primer and the PLP as a template for RCA [8, 11]. The resulting rolling circle products (RCPs) contain up to a thousand tandem copies of the PLP complementary sequence and spontaneously coil up into compact balls of single-stranded DNA [10, 12]. Fluorescently labeled detection probes are hybridized to a reporter sequence that is included in the central linker region of PLPs. Using a fluorescence microscope, the RCPs can be visualized as bright spots. Finally, an image analysis software such as CellProfiler can be used for automated quantification of RCPs [7, 10, 13]. In our multiplexed approach, PLPs contain a reporter sequence that is unique for each target. Thus, it is possible to distinguish between the targets by hybridizing detection probes with different fluorescent labels. The AR is encoded by eight exons (exons 1–8). AR-V7 lacks exons 4–8 and additionally contains a cryptic exon (cryptic exon 3, CE 3). By targeting RT primers and PLPs to CE 3 and the 30 -untranslated region (30 -UTR) of AR-V7, it can be

364

Lilli Hofmann et al.

Fig. 2 Target-specific RT primers and PLPs. Schematic illustration of the exon structure of AR-FL, AR-V7, and PSA mRNA. AR-FL is encoded by 8 exons (E 1–8). AR-V7 lacks E 4–8 and additionally contains a cryptic exon 3 (CE 3). AR-FL-specific RT primers and PLPs target exon-exon junctions (EEJs) 4, 6, and 7 that are not present in AR-V7. In contrast, AR-V7 is targeted at EEJ 3 and CE 3 including the 30 -untranslated region (30 -UTR). PSA is encoded by 5 exons and is targeted at E 3 and EEJs 1, 2, 3, and 4

distinguished from AR-FL, which in turn is targeted at exon-exon junctions (EEJs) 4, 6, and 7 [7] (see Fig. 2). To increase the detection efficiencies we apply more than one RT primer and PLP for each target. Using droplet digital PCR data as reference, we determined a detection limit as low as one transcript per cell for AR-V7 [7].

2

Materials

2.1 RT Primers, PLPs, and Detection Probes

The sequences of RT primers, PLPs, and detection probes are listed in Table 1. Sequences of all targets were retrieved from the National Center for Biotechnology Information (GenBank accession numbers FJ235916.1 (AR-V7), NM 000044.3 (AR-FL), and NM 001030047 (PSA)). The oligonucleotides were designed according to the guidelines published by Weibrecht et al. [9], using the CLC Main Workbench software (CLC Bio Workbench version 7.6, Qiagen). Oligonucleotides were checked for the formation of secondary structures or primer dimers using the OligoAnalyzer Tool by Integrated DNA Technologies (IDT). The specificity of RT primers and PLPs was confirmed using Primer-BLAST (National Institute of Health, NIH) [14]. For each target, a mix of 4–10 RT primers, with one of them carrying LNA modifications for enhanced hybridization affinity, and a mix of 4–10 PLPs are used (see Notes 1 and 2). AR-V7 is targeted at EEJ 3, CE 3, and 30 -UTR. AR-FL is

Padlock Probes to Detect mRNA Splice Variants

Table 1 Oligonucleotide sequences

reverse transcription primer mixes

PSA

AR-FL

AR-V7

target

name

sequence 5’  3’

RV_AR-V7_4

AAGCCACATTACAGGAAACA

RV_AR-V7_3

ACAAATAAAGATGGCCACAG

RV_AR-V7_2

AAGGCTAGATGTAAGAGG

RV_AR-V7_1

GACTCCACTTCTCCACTA

pAR-V7_CE3_3’LNA +GG+GT+CT+GGTCATTTTGAGATGC AR-FL_LNA_1

C+CA+TC+TG+GT+CG+TCCACGTGTAAGTT

RV_AR-FL_1

TGATTTTTCAGCCCA

RV_AR-FL_2

GAGTTCCTTGATGTAGT

RV_AR-FL_4

GATTAGCAGGTCAAAAGTG

PSA_LNA_1

G+AG+GT+CCA+CAC+ACT+GAAGTTT

RV_PSA_1

AGACAGGATGAGGGGTG

RV_PSA_2

GGGTTGGGAATGCTTCT

RV_PSA_3

TTCAGGATGAAACAGG

RV_PSA_4

TATCGTAGAGCGGGTGTG

RV_PSA_5

GGAGCAGCATGAGGT

RV_PSA_6

TGACCTTCACAGCATCC

RV_PSA_7

TAGGGAGCCATGGAGGA

RV_PSA_8

ACTAGACACCTCCTCTCCA

padlock probe mixes plp_AR-V7

/5Phos/AAAAATTCCGGGTTGTTCCTTTTACGACCTCAA TGCTGCTGCTGTACTACTCTTCGGATGACTCTGGGAG

plp_AR-V7_2

/5Phos/GGCTGACTTGCCTCATTAACCTCAATGCTGCTG CTGTACTACAATGCGTCTATTTAGTGGAGCCGGCTATC ACCAGACCCTGAAGAAA

plp_AR-V7_3

/5Phos/GGTTACCACTCATGTAGAACCTCAATGCTGCTG CTGTACTACAATGCGTCTATTTAGTGGAGCCGCCTATC GGGCTGTAGAAGTAATAGT

plp_AR-V7_4

/5Phos/TGCTTTTCGTGGTGTAACCTCAATGCTGCTGCTG TACTACAATGCGTCTATTTAGTGGAGCCGTCTATCTCT GGCTCAGTCGCT

plp_AR-V7_5

/5Phos/TGTCTGTCTGAGGTTCCTAACCTCAATGCTGCT GCTGTACTACAATGCGTCTATTTAGTGGAGCCGACTAT CTTGGACAAGAAGCAACTG

plp_AR-V7_6

/5Phos/CTAGCCTTCTGGATCCCAACCTCAATGCTGCTG CTGTACTACAATGCGTCTATTTAGTGGAGCCCGCTATC TATTTCCCCCTTAGGTT

plp_AR-V7_7

/5Phos/AGCTGATCCACAGAAGTTAACCTCAATGCTGCT GCTGTACTACAATGCGTCTATTTAGTGGAGCCATCTAT CACTGAGCTGAAGGTAGT

plp_AR-V7_8

/5Phos/GTCCGACTTTCCCTCTTAACCTCAATGCTGCTGC TGTACTACAATGCGTCTATTTAGTGGAGCCCTCTATCC ACACTGAGAGACTACA

plp_AR-V7_9

/5Phos/TGCTTCTCTTCATCAGTAACCTCAATGCTGCTGC TGTACTACAATGCGTCTATTTAGTGGAGCCCACTATCT GCTCTCTCACATCACA

(continued)

365

366

Lilli Hofmann et al.

plp_AR-V7_10

/5Phos/CCATAGCTTCCATATTGACAACCTCAATGCTGC TGCTGTACTACAATGCGTCTATTTAGTGGAGCCAGCTA TCATCTTGTTCTCTCTCTGCT

plp_AR-FL1

/5Phos/GGGCCAAGGCCTTGCCTGGCCTCAATGCACATG TTTGGCTCCTAAAGTCGGAAGTACTACTCTCTCTTGTA CACGTGGTCAAGT

plp_AR-FL_5

/5Phos/AGTGGATGGGCTGAAAACCTCAATGCACATGTT TGGCTCCAATGCGTCTATTTAGTGGAGCCAACTATCCT CTTCAGCATTATTCC

plp_AR-FL_6

/5Phos/TCTACCAGCTCACCAAACCTCAATGCACATGTT TGGCTCCAATGCGTCTATTTAGTGGAGCCTACTATCCC TGCTCAAGACGCT

plp_AR-FL_7

/5Phos/GCGAGAGAGCTGCATAACCTCAATGCACATGTT TGGCTCCAATGCGTCTATTTAGTGGAGCCACCTATCTC CGTGCAGCCTATT

plp_PSA_1

/5Phos/ACCAGAGGAGTTCTTGTCCTAGTAATCAGTAGC CGTGACTATCGACTGGTTCAAAGCTGGGGCAGCATTGA

plp_PSA_4

/5Phos/CGTGACGTGGATTGGAAAGTAGCCGTGACTATC GACTAATGCGTCTATTTAGTGGAGCCACACTATCTCTT CCTCACCCTGTC

plp_PSA_5

/5Phos/CGTGATCTTGCTGGGAAAGTAGCCGTGACTATC GACTAATGCGTCTATTTAGTGGAGCCAGACTATCGCA TCAGGAACAAAAG

plp_PSA_6

/5Phos/CCACAGCTTCCCACAAAAGTAGCCGTGACTATC GACTAATGCGTCTATTTAGTGGAGCCACTCTATCAGGT ATTTCAGGTCAG

plp_PSA_7

/5Phos/TCGATTCCTCAGGCCAAAAGTAGCCGTGACTAT CGACTATGCGTCTATTTAGTGGAGCCAGTCTAATCGA GCCTCCTGAAGAA

plp_PSA_8

/5Phos/GATCACCGAACTGACAAAGTAGCCGTGACTATC GACTAATGCGTCTATTTAGTGGAGCCAGGCTATCGCT CGTGGGTCATTCT

PSA

AR-FL

AR-V7

Table 1 (continued)

detection probes AR-V7 Lin16 AR-V7 AR-FL Lin33 AR-FL PSA

B2_DO PSA

Cy3 - CCTCAATGCTGCTGCTGTACTAC Cy5 - CCTCAATGCACATGTTTGGCTCC FAM - AGTAGCCGTGACTATCGACT

+ indicates that the following base is a locked nucleic acid (LNA) base with increased hybridization affinity; /5Phos/ indicates phosphorylation on the 50 -end of PLPs; target complementary hybridization arms of PLPs are underlined; reporter sequences (Lin16, Lin33, and B2_DO) for detection probes are highlighted in yellow, red, and green color; 50 -ends of detection probes are Cy3, Cy5, or FAM labeled

targeted at EEJs 4, 6, and 7. PSA is targeted at exon 3 and at EEJs 1, 2, 3, and 4. LNA-modified primers can be purchased from Exiqon, 50 -phosphorylated PLPs from IDT, and fluorescently labeled detection probes from Biomers. 1. Oligonucleotides are delivered as lyophilisates and can be stored at room temperature or 4  C (see Note 3). 2. Use nuclease-free water (NF-H2O) to prepare 100 μM stock solutions of oligonucleotides for long-time storage at 20  C (see Note 4).

Padlock Probes to Detect mRNA Splice Variants

367

3. Prepare 10 μM dilutions of detection probes in NF-H2O and store them at 20  C for short-term use. Protect the detection probes from light at all times. 4. For each target, prepare an RT primer mix containing 10 μM of each primer. Store them at 20  C for short-term use. 5. For each target, prepare a PLP mix containing 10 μM of each PLP. Store them at 20  C for short-term use. 2.2 Cell Lines and Cell Culture Media

1. VCaP cells. 2. LNCaP cells. 3. PC-3 (see Note 5). 4. Complete growth medium VCaP cell line: Dulbecco’s modified Eagle medium (DMEM) high glucose (4500 mg/L glucose, 4 mM glutamine, 1 mM sodium pyruvate, and 1500 mg/ L sodium bicarbonate) supplemented with 10% fetal bovine serum (FBS), and 1% penicillin/streptomycin. 5. Complete growth medium LNCaP cell line: RPMI1640 supplemented with 10% FBS, and 1% penicillin/streptomycin. 6. Complete growth medium PC-3 cell line: Kaighn’s modification of Ham’s F-12 medium (F-12K) supplemented with 10% FBS, 1% penicillin/streptomycin, and 1.5 g/L sodium bicarbonate.

2.3 Reagents, Buffers, and Solutions

1. Hanks’ balanced salt solution (HBSS). 2. Trypsin-EDTA (0.05%/0.02%). 3. Diethylpyrocarbonate (DEPC, Sigma-Aldrich)-treated H2O: Add 1 mL DEPC to 1 L ultrapure H2O (e.g., PURELAB flex 2, ELGA LabWater), stir overnight at room temperature, and then autoclave to inactivate DEPC. 4. 100%, 85%, and 70% ethanol in DEPC-treated H2O. 5. 10 DEPC-treated phosphate-buffered saline (PBS): To prepare a 10 stock solution, dissolve 36 g disodium-hydrogenphosphate-dodecahydrate, 80 g sodium chloride, and 4.1 g potassium dihydrogen phosphate in 700 mL DEPC-treated H2O; use NaOH to adjust to pH 7.2; fill to 1000 mL; and autoclave. 6. 1 DEPC-PBS-Tween (0.05%): 200 mL DEPC-treated PBS (1), 100 μL Tween-20. 7. 0.1 M Hydrochloric acid (HCl): 200 mL DEPC-treated H2O, 1.66 mL HCl (37%). 8. 1 M Potassium chloride (KCl): 14.91 g KCl, 200 mL DEPCtreated H2O.

368

Lilli Hofmann et al.

9. 20 Saline-sodium citrate buffer (SSC): To prepare a 20 stock solution, dissolve 175.3 g sodium chloride and 88.2 g trisodium citrate in 800 mL DEPC-treated H2O, use HCl to adjust to pH 7.0, fill to 1000 mL, and autoclave. 10. 2 Hybridization buffer: 20 mL Formamide (99.5%), 80 mL 4 SSC buffer. Formamide is a hazardous chemical. Wear nitrile rubber gloves and safety glasses, work under fume hood, and follow local lab regulations for disposal. 11. 2 SSC-Tween (0.05%) buffer: 20 mL 20 SSC buffer, 180 mL DEPC-treated H2O, 100 μL Tween-20. 12. 37% Formaldehyde solution. Formaldehyde is toxic/carcinogenic. Wear nitrile rubber gloves and safety glasses, work under fume hood, and follow local lab regulations for disposal. 13. 3.7% Formaldehyde: 900 mL 1 DEPC-treated PBS, 100 mL 37% formaldehyde. 14. 3.0% Formaldehyde: 919 μL 1 DEPC-treated PBS, 81 μL 37% formaldehyde. 15. Antifade mounting medium (e.g., SlowFade Gold Antifade Mountant, Thermo Fisher Scientific). 16. Reverse transcription mix (for 50 μL per reaction, all units are final concetrations): 5 μL of 10 RT reaction buffer (DNA-Gdansk), 1 μM RT primer mix AR-V7, 1 μM RT primer mix AR-FL, 1 μM RT primer mix PSA, 0.5 mM dNTPs mix, 0.4 μg/μL bovine serum albumin (BSA), 20 U/μL TranscriptMe reverse transcriptase (DNA-Gdansk), 1 U/μL RiboLock RNase inhibitor (Thermo Fisher Scientific). Use DEPCtreated H2O to fill to 50 μL (see Note 6). 17. PLP hybridization/ligation mix (for 50 μL per reaction, all units are final concetrations): 5 μL of 10x Ampligase reaction buffer (Biozym Biotech Trading), 0.05 M KCl, 20% formamide, 0.1 μM PLP mix AR-V7, 0.1 μM PLP mix AR-FL, 0.1 μM PLP mix PSA, 0.4 μg/μL BSA, 0.5 U/μL Ampligase DNA ligase (Biozym Biotech Trading), and 0.4 U/μL RNase H (Thermo Fisher Scientific). Use DEPC-treated H2O to fill to 50 μL (see Note 6). 18. Rolling circle amplification mix (for 50 μL per reaction, all units are final concetrations): 5 μL of 10 Phi29 DNA polymerase reaction buffer (Thermo Fisher Scientific), 0.25 mM dNTPs mix, 5% glycerol, 0.4 μg/μL BSA, and 1 U/μL Phi29 DNA polymerase. Use DEPC-treated H2O to fill to 50 μL (see Note 6).

Padlock Probes to Detect mRNA Splice Variants

369

19. Detection probe hybridization mix (for 50 μL per reaction, all units are final concetrations): 25 μL of 2 hybridization buffer, 0.1 μM detection probe AR-V7, 0.1 μM detection probe AR-FL, 0.1 μM detection probe PSA, and 2 μg/mL DAPI. Use DEPC-treated H2O to fill to 50 μL. 2.4 Consumables and Equipment

1. Cell culture flasks (e.g., Nunc EasYFlask 25 cm2). 2. SuperFrost Plus microscope slides (Thermo Fisher Scientific). 3. Coverslips. 4. 50 μL Hybridization chamber gasket, 9 mm diameter, 0.8 mm deep (Secure-Seal, Thermo Fisher Scientific). 5. PCR plate seals. 6. Forceps. 7. Diamond pen. 8. Humidity chamber. 9. Incubator for cell culture (37  C, 5% CO2). 10. Incubator for in situ PLP experiment (37  C and 45  C). 11. Fluorescence microscope with appropriate filter sets and 10, 20, and 40 objectives. 12. CellProfiler image analysis software [13].

3

Methods

3.1 Seeding PC Cell Lines

1. Aseptic cell culture techniques are required for seeding PC cell lines. To avoid contaminations, work in a laminar flow safety cabinet, spray your gloves with 70% ethanol, wipe the safety cabinet with 70% ethanol, and spray all equipment and reagents with 70% ethanol before placing them in the safety cabinet. 2. Culture VCaP, LNCaP, and PC-3 cells in T25 cell culture flasks at 37  C and 5% CO2 in the appropriate complete growth media (see Subheading 2.2, items 4–6). 3. Upon reaching a high cell density (see Note 7), harvest the cells for seeding. 4. Use a pencil to label the glass slides (see Note 8). 5. Sterilize the glass slides for 1 min in a glass cuvette filled with 70% ethanol and wipe a humidity chamber with 70% ethanol. Place the glass slides in the humidity chamber and air-dry them in the laminar flow safety cabinet. 6. Remove the growth medium from the cell culture flask and rinse the cell layer with 5 mL 1 HBSS. 7. Cover the cell layer with 1.5 mL trypsin/EDTA and incubate at 37  C for 10–15 min until the cells detach from the culture flask.

370

Lilli Hofmann et al.

8. Resuspend the cells in 3.5 mL complete growth medium to stop trypsin digestion. 9. Remove an aliquot of the cell suspension to determine the cell number. 10. Centrifuge the cell suspension at 300  g for 5 min at room temperature. 11. Remove the supernatant and resuspend the cell pellet in complete growth medium to achieve a cell density of 1  105 cells/ mL (see Note 9). 12. Transfer 500 μL of the cell suspension (i.e., 5  104 cells per slide) directly to each glass slide (see Note 10). 13. Add approximately 40 mL HBSS to the humidity chamber. 14. Incubate at 37  C and 5% CO2 for 24 h or up to 48 h to allow the cells to adhere to the slides (see Note 11). 15. Discard the growth medium and transfer the glass slides to a glass cuvette filled with 1 DEPC-treated PBS to wash them for 1 min at room temperature. Repeat the washing step one more time. 16. Fix the slides in 3.7% formaldehyde for 15 min at room temperature (see Note 12). 17. Wash twice in 1 DEPC-treated PBS for 1 min each. 18. Dehydrate the slides in ascending ethanol series (70%, 85%, and 100% for 1 min each) and air-dry the slides. 19. The slides are now ready to use and you can directly proceed to sample preparation, Subheading 3.2, step 3. Alternatively, keep the slides at 20  C for short-term use or 80  C for long-time storage. 3.2 Sample Preparation

1. Thaw slides containing seeded cells at room temperature. 2. Dehydrate the slides in ascending ethanol series (70%, 85%, and 100% for 2 min each) and air-dry the slides. 3. Mount 50 μL hybridization chambers on the slides (see Note 13 and Fig. 3a–c). 4. Fill the hybridization chambers with 50 μL 1 DEPC-PBSTween (0.05%) (see Note 14 and Fig. 3c) to rehydrate the cells for 5 min at room temperature. 5. Remove DEPC-PBS-Tween (0.05%) from the hybridization chamber using a pipet and permeabilize the cells by filling the hybridization chamber with 50 μL 0.1 M HCl for 5 min at room temperature. 6. Wash twice with 50 μL 1 DEPC-PBS-Tween (0.05%) at room temperature for 5 min each (see Notes 15 and 16).

Padlock Probes to Detect mRNA Splice Variants

371

Fig. 3 Handling of glass slides and hybridization chambers. (a) The 50 μL hybridization chamber has a diameter of 9 mm and a depth of 0.8 mm. The rubber wall (red) has an adhesive surface to mount the hybridization chamber on a glass slide. We recommend to trim the rubber wall to a thickness of approximately 1 mm. This enables easy removal of the hybridization chamber from the glass slide and minimizes the risk of breaking the slide, while leaving enough adhesive surface to prevent leakage. Through two chamber ports, the hybridization chamber can be charged and discharged using a pipette. (b) To prevent evaporation of the reaction mix, the chamber ports are covered with thin strips of PCR plate seals for all incubation steps exceeding 15 min. (c) Note the orientation of the hybridization chamber on the glass slide. To charge and discharge the hybridization chamber without the formation of air bubbles, tilt the glass slide as indicated by the arrow, and use the lower chamber port for pipetting, so that air can flow through the upper chamber port 3.3 Reverse Transcription

1. Prepare the RT mix (50 μL per reaction) directly before use, add enzymes last, and keep it on ice. 2. Fill the hybridization chambers with 48 μL RT mix (see Note 17). 3. Use thin strips of PCR plate seals to cover the access ports of the hybridization chambers (see Fig. 3b), place the slides in a humidity chamber (see Note 18), and transfer the humidity chamber to an incubator. 4. Incubate for 3 h at 45  C. 5. Remove slides from the humidity chamber and carefully detach the strips of PCR plate seals to access the chamber ports of the hybridization chambers (see Note 19). 6. Apply 50 μL 3.0% formaldehyde to fix the samples for 10 min at room temperature. 7. Wash twice with 50 μL 1 DEPC-PBS-Tween (0.05%) at room temperature for 2 min each (see Notes 15 and 16).

372

Lilli Hofmann et al.

3.4 Hybridization and Ligation of PLPs

1. Prepare the PLP hybridization/ligation mix (50 μL per reaction) directly before use, add enzymes last, and keep it on ice. 2. Fill the hybridization chambers with 48 μL PLP hybridization/ ligation mix. 3. Use thin strips of PCR plate seals to cover the access ports of the hybridization chambers, place the samples in a humidity chamber, and transfer to the incubator. 4. Incubate for 30 min at 37  C. Then, increase the temperature to 45  C and incubate for another 45 min (see Note 20). 5. Remove slides from the humidity chamber and carefully detach the strips of PCR plate seals to access the chamber ports of the hybridization chambers. 6. Wash with 50 μL prewarmed 2 SSC-Tween (0.05%) buffer for 5 min at 37  C. 7. Wash twice with 50 μL 1 DEPC-PBS-Tween (0.05%) at room temperature for 2 min each (see Notes 15 and 16).

3.5 Rolling Circle Amplification

1. Prepare the RCA mix, add the Phi29 DNA Polymerase last, and keep it on ice. 2. Fill the hybridization chambers with 48 μL RCA mix. 3. Cover the access ports of hybridization chambers and place the slides in a humidity chamber. 4. Incubate overnight at room temperature (see Note 21). 5. Carefully remove the strips of PCR plate seals to access the chamber ports of the hybridization chambers. 6. Wash twice with 50 μL 1 DEPC-PBS-Tween (0.05%) at room temperature for 2 min each (see Notes 15 and 16).

3.5.1 Hybridization of Detection Probes

1. From now on, work protected from light to avoid bleaching of the fluorophores. 2. Prepare the detection probe hybridization mix (50 μL per reaction). 3. Fill the hybridization chambers with 48 μL detection probe hybridization mix. 4. Cover the access ports of hybridization chambers, place the slides in a humidity chamber, and transfer to an incubator. 5. Incubate for 30 min at 37  C. 6. Remove the slides from the humidity chamber and carefully detach the strips of PCR plate seals to access the chamber ports of the hybridization chambers. 7. Wash twice with 50 μL 1 DEPC-PBS-Tween (0.05%) at room temperature for 2 min each.

Padlock Probes to Detect mRNA Splice Variants

373

8. Use a diamond pen to mark the position of the hybridization chambers at the back of the glass slides. 9. Carefully remove the hybridization chambers using forceps (see Note 22). 10. Dehydrate the slides in ascending ethanol series (70%, 85%, and 100%) for 2 min each. Air-dry the slides. 11. Apply antifade mounting medium to a coverslip and place it on the slide. 12. The samples can be stored protected from light at 4  C (see Note 23). 3.5.2 Imaging

1. Use a fluorescence microscope and 20 or 40 objectives to obtain images for image analysis and quantification of RCPs. 2. Adjust the exposure time for each channel to avoid overexposure of RCPs. 3. To detect RCPs in different focal planes throughout the cells, acquire z-stacks at range and intervals allowing full focused images (see Note 24). 4. Perform a maximum intensity projection to combine the signals of all focal planes in a single 2D image. 5. We recommend scanning several areas of the sample, so that the number of detected cells will be sufficient for statistical analysis (see Note 25).

3.5.3 Image Analysis and Quantification

1. Semiautomated quantification of RCPs can be done using the image analysis software CellProfiler. 2. Use grayscale images of individual channels in TIFF file format as input for analysis. 3. Employ the module “identify primary objects” to identify nuclei from DAPI staining. 4. Identify cell borders using the module “identify secondary objects” (see Note 26). 5. Identify RCPs using “enhance and suppress features” followed by “identify primary objects.” 6. Exclude unspecific signals, i.e., RCPs that are detected in more than one channel, using the module “mask objects.” 7. Employ “relate objects” to assign the RCPs to the respective cell. The software will then determine the number of RCPs per cell and the results can be exported to a spreadsheet (see Fig. 4).

374

Lilli Hofmann et al.

Fig. 4 Representative in situ results and image quantification by CellProfiler. (a) Representative microscopy image of VCaP cells; RCPs are visualized in red, yellow, and green, DAPI-stained nuclei in gray. (b) The image analysis software CellProfiler identifies nuclei (gray), cell borders (blue), and RCPs (yellow ¼ AR-V7, red ¼ ARFL, green ¼ PSA), and determines the number of RCPs per cell. (c) Comparison of expression levels in VCaP, LNCaP, and PC-3 cells; median and interquartile range are plotted in the bar chart; sample size n ¼ 184–193

4

Notes 1. The number of detected RCPs increases by targeting each transcript with an array of several RT primers and PLPs. 2. To minimize the formation of stable primer dimers in the primer mixes, only one LNA-modified primer is included per target. According to Weibrecht et al. [9], an increased number of LNA-modified primers negatively affects the efficiency of in situ detection of transcripts. 3. According to IDT, lyophilized oligonucleotides are stable for 40 weeks when stored at 37  C, and >60 weeks at 4  C. 4. Briefly centrifuge the tubes containing lyophilized oligonucleotides to avoid any loss when opening the cap of the tube. Add the appropriate amount of NF-H2O to prepare a 100 μM stock solution. Place the tubes on a thermoshaker and resuspend the oligonucleotides for 1–5 min at 55  C and 200 rpm. Briefly centrifuge to collect the whole content at the bottom of the tube. 5. We chose VCaP, LNCaP, and PC-3 cell lines to compare different mRNA expression levels [7, 15]. VCaP cells have high expression of AR-V7 and AR-FL, and medium expression of PSA. LNCaP cells have medium expression of AR-V7 and

Padlock Probes to Detect mRNA Splice Variants

375

AR-FL, and high expression of PSA. PC-3 cells serve as negative control, as they do not express AR-V7, AR-FL, or PSA (see Fig. 4). 6. The protocol can easily be adapted for blood samples. As blood samples can have an inhibitory effect on PCR reactions, we recommend doubling the enzyme concentrations in the reaction mixes for RT, hybridization/ligation of PLPs, and RCA [7, 16]. 7. PC-3 cells form a monolayer, while LNCaP cells tend to form cell clusters. Both cell lines should be harvested for seeding before reaching confluence. VCaP cells form large cell clusters and may take weeks to reach confluence. They can be harvested for seeding at ~50% confluence. 8. Permanent markers can dissolve upon contact with ethanol and might contaminate the sample, resulting in high fluorescence background. Thus, it is recommended to use a pencil instead. 9. If the cell density on the slides is too low, large areas will have to be imaged in order to analyze an acceptable number of cells. If the cell density is too high, the segmentation of cell clusters during image analysis and consequently the quantification of signals/cell can be problematic. A cell density of 1  105 cells/ mL (i.e., 5  104 cells per slide) worked well in our hands. 10. To ensure uniform cell density on all slides, gently resuspend the cells frequently. When the cell suspension is transferred to the slides, surface tension will keep it from spilling over the edges of the slides; however, gentle pipetting is required. 11. PC-3 and LNCaP cells usually adhere to the slide within 24 h, while VCaP cells require longer and should be incubated for up to 48 h. The density of adherent cells can be checked under the microscope before moving on to washing and fixation steps. 12. The use of the cross-linking fixative formaldehyde is sometimes considered problematic for downstream RNA analysis. RNA might be degraded and covalent modifications of RNA might block the extraction of RNA from tissues and cells and might interfere with base pairing during in situ hybridization [17]. However, Yan et al. [18] have described formaldehyde as a suitable fixative for RNA in situ hybridization. In our in situ PLP approach, formaldehyde fixation delivers good results [7, 19], while the number of detectable signals is significantly decreased in acetone-fixed cells. Omitting the fixation step will cause cell rupture, leading to low cell morphology and low quality of in situ results. 13. You can use more than one hybridization chamber per slide. In this case, we recommend separating the hybridization chambers to minimize the risk of cross-contamination between leaky

376

Lilli Hofmann et al.

hybridization chambers. Glass slides can break upon removing the hybridization chambers! To prevent this, it can be helpful to trim the adhesive surface of the hybridization chambers before mounting them on the slide. However, leave enough area for attachment of the hybridization chamber to prevent leakage during incubation. We noticed leakage if the rubber wall diameter of the hybridization chamber fell below 1 mm (see Fig. 3a). 14. For all incubation and washing steps carried out in the hybridization chambers, a pipette is used to fill the chamber with reagent/washing buffer. After incubation, a pipette is used to remove the reagent/washing buffer from the hybridization chamber before moving on to the next step. To avoid the formation of bubbles, always tilt the slide for filling or emptying the hybridization chambers (see Fig. 3c) and do not press the pipette knob to the second stop. 15. Samples should never run dry. Washing steps with 1 DEPCPBS-Tween (0.05%) can be prolonged if necessary, as this has no impact on the efficiency of in situ detection in our experience. 16. It is possible to pause the experiment at this point. The samples can be stored in 1 DEPC-PBS-Tween (0.05%) overnight at 4  C. 17. 48 μL will be sufficient to fill the hybridization chamber. A larger volume of reaction mix will likely spill upon mounting the strips of PCR seal to cover the chamber ports of the hybridization chamber. 18. Cover the chamber ports of the hybridization chamber for all incubation steps exceeding 15 min (see Fig. 3b) and place the slides in a humidity chamber to avoid evaporation of the reaction mix. 19. When you remove the strips of PCR plate seals, we recommend inserting the tip of your forceps below the strip to loosen the adhesive, and stabilizing the hybridization chamber with another pair of forceps when you peel off the strip. Otherwise, the hybridization chamber might lose adhesion to the glass slide and become leaky. 20. Incubation at 37  C facilitates the digestion of RNA in RNA/cDNA hybrids by RNase H. Subsequent incubation at 45  C facilitates the ligation of hybridized PLPs by Ampligase DNA ligase. High temperature, as well as the addition of formamide to the reaction mix, supports specific hybridization of PLPs [8, 10].

Padlock Probes to Detect mRNA Splice Variants

377

21. Incubation overnight at room temperature will result in large RCPs. Incubating overnight at higher temperatures might cause long RCPs to break into smaller fragments [10]. 22. To avoid breaking of glass slides when removing the hybridization chambers, place the slides flat on the working surface. Insert the tip of the forceps below the rubber wall of the hybridization chamber to loosen the adhesive and slowly lift the hybridization chamber. If necessary, repeat this from several directions. Take care not to scratch the cells inside the hybridized area. 23. In our experience, the signals are stable for >1 year, when the slides are stored protected from light at 4  C. 24. We usually acquire z-stacks covering a thickness of 6 μm. The optimal interval for the acquisition of z-stacks depends on the used objective and is determined according to the Nyquist criterion. The interval should be 0.5 the axial optical resolution. For example, the optimal interval for z-stacks determined by the Zeiss imaging software Zen is 0.86 μm for the Zeiss objective LD Achroplan 40x/0.60 Korr, and 1.94 μm for the Zeiss objective LD plan-Neofluar 20x/0.4 Korr. 25. We observed high variations in expression levels between individual cells. Furthermore, the formation of air bubbles inside the hybridization chambers cannot always be avoided. These local variations can be compensated by scanning several areas of each sample and analyzing an appropriate number of cells. We recommend not to scan close to the area where the hybridization chamber was attached, as these bubbles appear more often at the walls of the hybridization chamber and the hybridization chamber ports. 26. To identify cell borders, the previously identified nuclei act as “seeds.” You can define the cell borders by expanding the nuclei by a fixed number of pixels. Alternatively, if the cytoplasm of cells has detectable background fluorescence (most likely in the FITC channel), you can identify the cell borders based on this background fluorescence. References 1. Cornford P, Bellmunt J, Bolla M et al (2017) EAU-ESTRO-SIOG guidelines on prostate cancer. Part II: Treatment of relapsing, metastatic, and castration-resistant prostate cancer. Eur Urol 71:630–642. https://doi.org/10. 1016/j.eururo.2016.08.002 2. Prekovic S, Van den Broeck T, Moris L et al (2018) Treatment-induced changes in the androgen receptor axis: liquid biopsies as diagnostic/prognostic tools for prostate cancer.

Mol Cell Endocrinol 462:56–63. https://doi. org/10.1016/j.mce.2017.08.020 3. Stamey TA, Yang N, Hay AR et al (1987) Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med 317:909–916. https://doi.org/10. 1056/NEJM198710083171501 4. Antonarakis ES, Lu C, Wang H et al (2014) AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N Engl J Med

378

Lilli Hofmann et al.

371:1028–1038. https://doi.org/10.1056/ NEJMoa1315815 5. Scher HI, Lu D, Schreiber NA et al (2016) Association of AR-V7 on circulating tumor cells as a treatment-specific biomarker with outcomes and survival in castration-resistant prostate cancer. JAMA Oncol 2:1441. https://doi.org/10.1001/jamaoncol.2016. 1828 6. Antonarakis ES, Lu C, Luber B et al (2015) Androgen receptor splice variant 7 and efficacy of taxane chemotherapy in patients with metastatic castration-resistant prostate cancer. JAMA Oncol 1:582. https://doi.org/10. 1001/jamaoncol.2015.1341 7. El-Heliebi A, Hille C, Laxman N et al (2018) In situ detection and quantification of AR-V7, AR-FL, PSA, and KRAS point mutations in circulating tumor cells. Clin Chem 64:536–546. https://doi.org/10.1373/ clinchem.2017.281295 8. Larsson C, Grundberg I, So¨derberg O et al (2010) In situ detection and genotyping of individual mRNA molecules. Nat Methods 7:395–397. https://doi.org/10.1038/ nmeth.1448 9. Weibrecht I, Lundin E, Kiflemariam S et al (2013) In situ detection of individual mRNA molecules and protein complexes or posttranslational modifications using padlock probes combined with the in situ proximity ligation assay. Nat Protoc 8:355–372. https://doi.org/10.1038/nprot.2013.006 10. Krzywkowski T, Nilsson M (2018) Padlock probes to detect single nucleotide polymorphisms. In: Gaspar I (ed) RNA detection. Springer, New York, NY, pp 209–229 11. Larsson C, Koch J, Nygren A et al (2004) In situ genotyping individual DNA molecules by target-primed rolling-circle amplification of padlock probes. Nat Methods 1:227–232. https://doi.org/10.1038/nmeth723

12. Baner J, Nilsson M, Mendel-Hartvig M et al (1998) Signal amplification of padlock probes by rolling circle replication. Nucleic Acids Res 26:5073–5078. https://doi.org/10.1093/ nar/26.22.5073 13. Carpenter AE, Jones TR, Lamprecht MR et al (2006) CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol 7:R100. https://doi.org/10. 1186/gb-2006-7-10-r100 14. Ye J, Coulouris G, Zaretskaya I et al (2012) Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinformatics 13:134. https://doi.org/10. 1186/1471-2105-13-134 15. Ma Y, Luk A, Young F et al (2016) Droplet digital PCR based androgen receptor variant 7 (AR-V7) detection from prostate cancer patient blood biopsies. Int J Mol Sci 17:1264. https://doi.org/10.3390/ijms17081264 16. Al-Soud WA, Radstrom P (2001) Purification and characterization of PCR-inhibitory components in blood cells. J Clin Microbiol 39:485–493. https://doi.org/10.1128/JCM. 39.2.485-493.2001 17. Evers DL, Fowler CB, Cunningham BR et al (2011) The effect of formaldehyde fixation on RNA. J Mol Diagn 13:282–288. https://doi. org/10.1016/j.jmoldx.2011.01.010 18. Yan F, Wu X, Crawford M et al (2010) The search for an optimal DNA, RNA, and protein detection by in situ hybridization, immunohistochemistry, and solution-based methods. Methods 52:281–286. https://doi.org/10. 1016/j.ymeth.2010.09.005 19. El-Heliebi A, Kashofer K, Fuchs J et al (2017) Visualization of tumor heterogeneity by in situ padlock probe technology in colorectal cancer. Histochem Cell Biol 148:105–115. https:// doi.org/10.1007/s00418-017-1557-5

Chapter 24 Automated One-Double-Z Pair BaseScope™ for CircRNA In Situ Hybridization Boye Schnack Nielsen, Trine Møller, and Jørgen Kjems Abstract Circular RNAs (circRNAs) are single-stranded RNA, typically exons, connected head to tail by backsplicing. The functions of circRNAs include binding of microRNA, regulation of transcription, regulation of alternative splicing, and modulation of immune response. As for other RNA transcripts their levels vary during development and may also become deregulated during disease progression. Different from linear RNAs, the circRNAs are not susceptible to traditional exonuclease activity and therefore more stable in tissues and blood. This makes the circRNAs an attractive new group of potential biomarkers. Specific detection of circRNAs in situ is challenged by the need to discriminate bona fide circRNAs from the linear precursor forms and splice variants that contain largely overlapping sequences. Knowing the sequence around the splice junction site makes the branched DNA probe technology, BaseScope, suitable for selective detection of unique circRNAs. Here, we present the automated application of BaseScope with a onedouble-Z pair probe set designed for the junction of circHIPK3. Key words Automation, BaseScope, circHIPK3, circRNA, In situ hybridization, Ventana

1

Introduction Circular RNAs (circRNAs) constitute a relatively new group of RNAs that previously were thought to be an insignificant by-product of RNA splicing. Today, thousands of circRNA have been reported, many of which have been shown to be functionally important. CircRNAs are formed by a process called head-to-tail back-splicing where the 50 and 30 ends of one or more exons are connected by the spliceosome to form a novel unique splice site. One of the best described circRNAs is derived from the cerebellar degeneration-related protein 1 antisense transcript (CDR1AS), a gene located on the X chromosome [1]. This circRNA is also known as a circular RNA sponge for miR-7 (and therefore also named ciRS-7) as it was found to contain more than 60 binding sites for miR-7 [2, 3]. Since the identification of ciRS-7’s sponging activity, several other circRNAs have been found to contain

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_24, © Springer Science+Business Media, LLC, part of Springer Nature 2020

379

380

Boye Schnack Nielsen et al.

microRNA-binding sites, including circHIPK3 [4, 5], which is formed by back-splicing of the Exon 2 of the homeodomain-interacting protein kinase 3 (HIPK3) transcript. CircHIPK3 has been reported to bind various microRNAs, including miR-124, miR-558, miR-29b-3p, and miR-193a-3p [4–7]. However, the microRNA sponging activity appears not to be a general trend among circRNAs. Other general functions of circRNA include participation in protein translation, regulation of transcription and alternative splicing, and controlling of immune response [8– 11]. There is mounting evidence that circRNAs could play a direct role in disease development or serve as biomarkers as reported for cancer, cardiovascular diseases, dementia, and autoimmune diseases [11–14]. Since the single-stranded RNA molecules are connected head-to-tail, they are not accessible to exonucleases and are therefore more stable than the linear forms, making circRNAs an attractive group of biomarkers. The circles are, on the other hand, cleaved by endo-ribonucleases such as RNase L [11] or as a consequence of strong complementarity between a microRNA and circRNA [1]. Their extensive resemblance to the linear form set high demands for probe design and method choice for in situ hybridization (ISH) analysis, which needs to be highly specific and with significant signal amplification. Branched DNA (bDNA) signal amplification [15] is the basis of RNAscope probe technology resulting in high specificity and high sensitivity [16]. The high specificity is obtained by designing two antisense DNA oligonucleotides, also called double-Z probes, that bind to adjacent sequences as pairs on the target sequence. The high sensitivity of RNAscope probes is based partly on the antisense design of typically up to 20 pairs for individual mRNA targets, and partly on the bDNA oligo amplification [16]. In the recently developed BaseScope assay, only one double-Z pair can be used and an additional signal amplification step ensures sufficient sensitivity for detection of single molecules [17]. For specific detection of circRNA, the back-splice junction (BSJ) sequence is used as target. The BaseScope probe technology has been found useful in several other applications including detection of specific exon junctions [18], mRNA splice variants [19–21], and single-point mutation RNA ISH [17]. In addition to the BaseScope technology, other techniques have been used for circRNA ISH, including Padlock probes that involve rolling circle amplification in situ [22], and locked nucleic acid (LNA):DNA chimeric probes [2] that have found broad utility in microRNA ISH [23, 24]. In this method chapter, we introduce the use of automated BaseScope in situ hybridization for detection of circRNAs in paraffin-embedded samples. The specific junction sequence covering approximately 30 nucleotides on each side of the junction (see

CircRNA In Situ Hybridization

381

Note 1) was submitted for probe design by ACD/Bio-Techne resulting in a one-double-Z probe named BA-Hs-HIPK3-E2-circRNA-Junc. Our application is based on ACD’s detailed manual for running BaseScope on the Ventana instrument (VS BaseScope, see www.acdbio.com).

2

Materials

2.1

Tissue Samples

2.2

Reagents

Colon cancer paraffin-embedded samples (BioIVT/Asterand, W Sussex, UK) and cervix cancer FFPE sample (ProteoGenex, Inglewood, CA). 1. BaseScope probes: All probes were 1 x ZZ. HIPK3: BA-HsHIPK3-E2-circRNA-Junc targeting 1629-579 of NM_005734.4. One double-Z positive control probe: ubiquitin C (UBC), and a one-double-Z negative control: dihydrodipicolinate reductase (dapB) mRNA (see Note 2). 2. BaseScope VS Reagent Kit (Red) (ACD/Bio-Techne). 3. BaseScope VS AMP 1–8 (Roche). These eight reagents contain the probe detection reagents (ACD/Bio-Techne). 4. RNAscope VS Protease (ACD/Bio-Techne). 5. RNAscope VS Target Retrieval solution (ACD/Bio-Techne). 6. RNAscope VS Dewax (ACD/Bio-Techne). 7. Discovery Wash 10 (Roche Diagnostics). 8. Discovery CC1 (Roche Diagnostics). 9. SSC buffer (Roche Diagnostics). 10. Reaction buffer (Roche Diagnostics). 11. ULTRA LCS (Pre-diluted liquid cover-slipping oil, Roche Diagnostics). 12. mRNA AMP 1–8 dispensers—for BaseScope VS AMP 1–7 (Roche Diagnostics). Amp8 uses an option dispenser. 13. Test Probe #1–20 dispensers (Roche Diagnostics). 14. mRNA Target Retrieval dispenser (Roche Diagnostics). 15. mRNA Dewax dispenser (Roche Diagnostics). 16. mRNA Protease dispenser (Roche Diagnostics). 17. mRNA Fast Red dispenser—prefilled (Roche Diagnostics). 18. Hematoxylin kit for Ventana (Roche Diagnostics). 19. Bluing reagent for Ventana (Roche Diagnostics). 20. EcoMount (Biocare Medical, Pacheco, CA). 21. Detergent (regular dish washing product).

382

2.3

Boye Schnack Nielsen et al.

Equipment

1. Microtome. 2. SuperFrost Plus Glass slides (Thermo Fisher). 3. Cover glass 24  50 mm. 4. Control slides with FFPE HeLa cells (ACD, Bio-Techne) (see Note 3). 5. Water baths. 6. Glass jars (Ziehl-Nielsen type). 7. Ventana DISCOVERY™ ULTRA instrument (Roche, Basel, Switzerland). 8. Software ACD, mRNA Universal Procedure v2. 9. Microscope or slide scanner, for bright field or fluorescence.

3

Methods

3.1 Preparation of Tissue Sections

1. Prepare a water bath with room-temperature water and another water bath with 45  C water. Place RNase-free Milli-Q water in clean Ziehl-Nielsen jars placed inside the water baths (see Note 4). 2. Place paraffin-embedded samples on a cooling plate, cooled to 14  C, for 5–10 min. 3. Cut 5 μm thick tissue sections using a microtome, one section for each probe. 4. Place tissue sections in the RNase-free Milli-Q water within the chilled (RT) water bath. 5. Using a Superfrost Plus glass slide, move tissue sections to the jar with Milli-Q water in the warm (45  C) water bath, let the tissue sections unfold for a few seconds, and then place the sections centrally on individual Superfrost Plus glass slides. 6. Leave the glass slides upright and let the tissue sections dry for 30–60 min at room temperature. 7. Melt the paraffin in an oven for 1 h at 60  C.

3.2 Preparation of the Ventana DISCOVERY ULTRA Instrument

1. In the lower liquid containers, place CC1, SSC, reaction buffer, wash buffer, and LCS in the appropriate positions. 2. In the upper chamber with the carrousel, place liquid dispensers with dewax, protease, target retrieval solution, hematoxylin, and bluing reagent. 3. Fill Roche dispensers AMP 1–8 with the appropriate ACD AMP 1–8 and place the dispensers in the upper chamber. 4. Fill Roche dispensers with the 1-ZZ probes and place the dispensers in the upper chamber.

CircRNA In Situ Hybridization

383

Fig. 1 Screenshot of the Ventana software. The screenshot shows the first steps in the BaseScope protocol (mRNA Universal Procedure v2). Some of the steps allow modification of the experimental conditions by using the pull-down menus. Not all steps in the BaseScope protocol are shown in this example 3.3 Programming the Ventana DISCOVERY ULTRA Instrument

1. Ensure that the Ventana System has an updated software package installed (mRNA Universal Procedure v2 or similar). When designing the experimental setup, and programming the software for a BaseScope run, note that not all incubation steps are adjustable. 2. Open the Ventana program, mRNA Universal Procedure v2. 3. Set Dewax time to 16 min using the pull-down menu (see Fig. 1). 4. Set Pretreatment 2 (heat-induced target retrieval) to 16 min at 97  C. 5. Set Pretreatment 3 (protease step) to 16 or 24 min at 37  C (see Note 5). 6. Set Amp 1–6 according to the standard conditions. 7. Set Amp 7 to 32 or 60 min (see Note 6). 8. The Amp 8 and Fast Red incubation steps cannot be modified.

3.4 Running BaseScope

1. A protocol is prepared for the individual slides using the software menu above. 2. Print adhesive barcode labels and place them on the respective slides. 3. Insert the slides with tissue sections and barcode labels into individual drawers of the Ventana.

384

Boye Schnack Nielsen et al.

4. Run the program; the expected duration is 12–14 h. 5. Remove the glass from the individual drawers of the Ventana and place the slides into a Ziehl-Nielsen jar filled with lukewarm tap water with detergent. 6. Wash carefully by moving slides up and down in the solution to remove excess oil from the slide. 7. Wash three times with tap water without detergent. 8. Dry slides by heating in an oven for 30 min at 60–65  C (slides are not dehydrated in alcohol). In case of fluorescence application see Note 7. 9. Soak slides shortly in xylene and mount slides using EcoMount mounting media. 10. Inspect the slides in a microscope. The UBC mRNA ISH signal should be very intense, whereas there should be no red staining with the dapB probe (see Fig. 2). The UBC mRNA signal is typically located in the cytoplasm (see Fig. 3). The circHIPK3 ISH signal (see Note 8) can be inspected using a 20 or 40 objective in a standard bright-field microscope. We used a Zeiss AxioScan Slide scanner with a 20 objective to obtain digital slides (examples in Fig. 2), from which digital zooming allows immediate evaluation of the ISH signal spots (Figs. 3 and 4).

Fig. 2 Test of 1-ZZ reference probes in paraffin samples. Examples of UBC mRNA ISH using a 1-ZZ probe in a colon cancer, two cervix cancers, and paraffin-embedded HeLa cells. The colon cancer and the HeLa cells show abundant UBC mRNA ISH signal (red color), whereas the two cervix cancers have lower levels, probably due to partly degraded RNA. The 1-ZZ probe to bacterial dapB mRNA shows no ISH signal in the four paraffinembedded samples

CircRNA In Situ Hybridization

385

Fig. 3 Test of 1-ZZ probe to circHIPK3 in HeLa cells. Three adjacent tissue sections were incubated with 1-ZZ probes targeted to UBC mRNA, circHIPK3, and dapB mRNA. Intense ISH signal is, as expected, seen with the UBC mRNA probe. CircHIPK3 ISH signal spots are most frequently seen in the cytoplasm. Note that the number of dots per cell varies. No signal spots are seen with the dapB probe

Fig. 4 Localization of circHIPK3 in colon cancer. Three adjacent tissue sections were incubated with 1-ZZ probes targeted to circHIPK3, UBC mRNA, and dapB mRNA. The circHIPK3 ISH signal spots are frequently seen over the tumor cells, but also some spots are noted in the adjacent stroma (upper left panel). Note that, in the tumor cells, the circHIPK3 ISH spots vary in size (upper right panel). Intense ISH signal is seen with the UBC mRNA probe and virtually no signal spots are seen with the dapB probe

386

4

Boye Schnack Nielsen et al.

Notes 1. Probe design: The specific sequence at the junction site can be obtained from www.circbase.org. For circHIPK3, the ID sequence was hsa_circ_0000284, targeting 1629-579 of NM_005734.4 (HIPK3), thus exon 2: 569..1667 of HIPK3. The gene ID is then submitted to ACD (www.ACDbio.com) with the notion that the head-to-tail splice junction is needed for the probe design. 2. One-double-Z reference probes: In the current experimental setup, we used one-double-Z probes for detection of linear targets as positive and negative controls. Future studies may reveal if a circRNA may be better applicable as reference probe in circular RNA ISH studies. The positive control probe to UBC mRNA shows a positive signal in the HeLa cell sample and in both colon and cervix cancer samples (Fig. 2). UBC mRNA is considered a highly expressed transcript. Alternative positive control probes are available from ACDbio. The reduced staining intensity in some FFPE samples may reflect partly degraded RNA. The negative control probe, directed against bacterial dapB, was overall negative and only occasionally resulted in ISH signal spots or background stain. 3. BaseScope on Cell lines: FFPE cell lines provided by ACD are reference slides for optimizing the experimental conditions including pretreatment and amplification times and are considered for tests with the reference probes. In circBase, it was noted that circHIPK3 is relatively highly expressed in HeLa cells [25]. We therefore also submitted a HeLa cell slide to circHIPK3 ISH, and found a decent ISH signal identified as spots located in the perinuclear cytoplasmic space (see Fig. 3). 4. RNase-free environment: Although tissue sections from FFPE samples aimed for RNA ISH with RNAscope and BaseScope can be prepared without requirements of RNase-free reagents and tools, we systematically apply conditions that can reduce RNase contamination, including using RNase-free Milli-Q water for handling of tissue sections, using rubber gloves for slide and reagent handling and RNaseZAP for cleaning surfaces. 5. Pretreatments: We tested if prolongation of pretreatment step 3 (the protease step) could improve the circHIP3 signal output; however, this was without significant effect on the prevalence of the circHIPK3 signal spots. 6. Amplifier steps: The amplifier step 7 is the Fast Red substrate incubation step and can be prolonged for better visualization of the signal dots. We tested incubation at 32 min and 60 min, which as expected resulted in overall larger signal dots.

CircRNA In Situ Hybridization

387

7. Fluorescence application: The Fast Red chromogen will also fluoresce under a fluorescence microscope. If it is necessary to inspect the ISH signal as fluorescence signal, then omit the dehydration steps and mount the slides directly with a DAPIcontaining mounting medium, such as ProLong™ Gold Antifade from Thermo Fisher Scientific. 8. CircHIPK3 in cancer samples: CircHIPK3 expression has been reported in various cancer types, including colorectal cancer [26]. We tested three different colorectal cancer samples and found that the circHIPK3 ISH signal dots were primarily found in cancer cells and tend to gather toward the lumen of the crypt-like structures (Fig. 4). Lasda and Parker reported that circRNA including circHIPK3 is enriched in extracellular vesicles [27], which may help to explain the apical localization of the circHIPK3 ISH signal. We also noted that the BaseScope ISH signal dots varied in size. We speculate that the reason for this variation could be that if circRNAs are packed in vesicles, the varying number of circRNA copies within a vesicle may cause variation in dot size.

Acknowledgments This study was supported by The Danish Agency for Science and Higher Education. References 1. Hansen TB, Wiklund ED, Bramsen JB et al (2011) miRNA-dependent gene silencing involving Ago2-mediated cleavage of a circular antisense RNA. EMBO J 30:4414–4422 2. Hansen TB, Jensen TI, Clausen BH et al (2013) Natural RNA circles function as efficient microRNA sponges. Nature 495:384–388 3. Memczak S, Jens M, Elefsinioti A et al (2013) Circular RNAs are a large class of animal RNAs with regulatory potency. Nature 495:333–338 4. Zheng Q, Bao C, Guo W et al (2016) Circular RNA profiling reveals an abundant circHIPK3 that regulates cell growth by sponging multiple miRNAs. Nat Commun 7:11215 5. Li Y, Zheng F, Xiao X et al (2017) CircHIPK3 sponges miR-558 to suppress heparanase expression in bladder cancer cells. EMBO Rep 18:1646–1659 6. Liu X, Liu B, Zhou M et al (2018) Circular RNA HIPK3 regulates human lens epithelial cells proliferation and apoptosis by targeting

the miR-193a/CRYAA axis. Biochem Biophys Res Commun 503:2277–2285 7. Ni H, Li W, Zhuge Y et al (2019) Inhibition of circHIPK3 prevents angiotensin II-induced cardiac fibrosis by sponging miR-29b-3p. Int J Cardiol. https://doi.org/10.1016/j.ijcard. 2019.04.006 8. Ebbesen KK, Hansen TB, Kjems J (2017) Insights into circular RNA biology. RNA Biol 14:1035–1045 9. de Fraipont F, Gazzeri S, Cho WC et al (2019) Circular RNAs and RNA splice variants as biomarkers for prognosis and therapeutic response in the liquid biopsies of lung cancer patients. Front Genet 10:390 10. Shen B, Wang Z, Li Z et al (2019) Circular RNAs: an emerging landscape in tumor metastasis. Am J Cancer Res 9:630–643 11. Liu CX, Li X, Nan F et al (2019) Structure and degradation of circular RNAs regulate PKR activation in innate immunity. Cell 177:865–880 e821

388

Boye Schnack Nielsen et al.

12. Kristensen LS, Hansen TB, Veno MT et al (2018) Circular RNAs in cancer: opportunities and challenges in the field. Oncogene 37:555–565 13. Floris G, Zhang L, Follesa P et al (2017) Regulatory role of circular RNAs and neurological disorders. Mol Neurobiol 54:5156–5165 14. Aufiero S, Reckman YJ, Pinto YM et al (2019) Circular RNAs open a new chapter in cardiovascular biology. Nat Rev Cardiol. https://doi. org/10.1038/s41569-019-0185-2 15. Player AN, Shen LP, Kenny D et al (2001) Single-copy gene detection using branched DNA (bDNA) in situ hybridization. J Histochem Cytochem 49:603–612 16. Wang F, Flanagan J, Su N et al (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14:22–29 17. Baker AM, Huang W, Wang XM et al (2017) Robust RNA-based in situ mutation detection delineates colorectal cancer subclonal evolution. Nat Commun 8:1998 18. Erben L, He MX, Laeremans A et al (2018) A novel ultrasensitive in situ hybridization approach to detect short sequences and splice variants with cellular resolution. Mol Neurobiol 55:6169–6181 19. Wang H, Zhang CZ, Lu SX et al (2019) A coiled-coil domain containing 50 splice variant is modulated by serine/arginine-rich splicing factor 3 and promotes hepatocellular carcinoma in mice by the Ras signaling pathway. Hepatology 69:179–195 20. Guo X, Zhao Y, Nguyen H et al (2018) Quantitative analysis of alternative pre-mRNA

splicing in mouse brain sections using RNA in situ hybridization assay. J Vis Exp. https://doi. org/10.3791/57889 21. Volpi CC, Pietrantonio F, Gloghini A et al (2019) The landscape of d16HER2 splice variant expression across HER2-positive cancers. Sci Rep 9:3545 22. Zaghlool A, Ameur A, Wu C et al (2018) Expression profiling and in situ screening of circular RNAs in human tissues. Sci Rep 8:16953 23. Jorgensen S, Baker A, Moller S et al (2010) Robust one-day in situ hybridization protocol for detection of microRNAs in paraffin samples using LNA probes. Methods 52:375–381 24. Nielsen BS, Holmstrom K (2013) Combined microRNA in situ hybridization and immunohistochemical detection of protein markers. In: Moll J, Colombo R (eds) Target identification and validation in drug discovery, Methods in molecular biology, vol 986. Humana Press, New York, pp 353–365 25. Salzman J, Chen RE, Olsen MN et al (2013) Cell-type specific features of circular RNA expression. PLoS Genet e1003777:9 26. Barbagallo C, Brex D, Caponnetto A et al (2018) LncRNA UCA1, upregulated in CRC biopsies and downregulated in serum exosomes, controls mRNA expression by RNA-RNA interactions. Mol Ther Nucleic Acids 12:229–241 27. Lasda E, Parker R (2016) Circular RNAs co-precipitate with extracellular vesicles: a possible mechanism for circRNA clearance. PLoS One 11:e0148407

Correction to: In Situ Sequencing: A High-Throughput, Multi-Targeted Gene Expression Profiling Technique for Cell Typing in Tissue Sections Markus M. Hilscher, Daniel Gyllborg, Chika Yokota, and Mats Nilsson

Correction to: Chapter 20 in: Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_20 As per the editor’s request, figure 2 in chapter 20 has been revised as shown below: A)

5’ arm

Anchor Seq

Linker

CTGTCCACCTTCCAGCAGATTCTACGATTTTACCAGTGGCTGCGTCTATTTAGTGGAGCCTCGTCTATCTTCTTTATGGCTCCAGCCTGGCCTCA

‘5th’ base

B)

CGA TC T A

PLP

TTT

5’-

Base Library

U G C G U C U U UU A G U G G A G C C T N N Detection Probe NC TGCGTCTATTTAGTGGAG TA C G G CC -3’ TC GT TC A C G T AC

TC CT

3’-

Anchor Probe

3’ arm

AT CT

TT TT

5’-

Barcode

5’ 3’

TAGACGACCTTCCACCTGTC ACTCCGGTCCGACCTCGGTA

Detection

-5’ RCP RCA

-3’ cDNA RT

3’-

UAGACGACCUUCCACCUGUCACUCCGGUCCGACCUCGGUA

-5’ mRNA

The updated online version of this chapter can be found at: https://doi.org/10.1007/978-1-0716-0623-0_20 Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0_25, © Springer Science+Business Media, LLC, part of Springer Nature 2021

C1

INDEX A Alkaline phosphatase (AP) .................................. 4, 10, 21, 222, 279, 295 Amphioxus............................................................ 179–193 Androgen receptor (AR) ..................................... 361–363 Animal models............................................................... 258 Antibody elution ................................................. 279, 280, 284, 297 AP, see Alkaline phosphatase (AP) AR splice variant 7 (AR-V7)............................... 361–364, 368, 369, 374, 375 Autofluorescence (AF) ....................................19–31, 132, 133, 136–138, 140, 150–152, 155, 172, 173, 178 Automation ........................................................... 23, 212, 229–244, 246, 261, 278, 280, 343 Autostainer ........................................................... 246, 248

B Background suppression ..................................... 127, 129, 143, 145, 159, 161, 162, 181, 203 Bacteria ...................................................... 15, 64, 87, 139 BaseScope .............................................................. 20, 225, 349–360, 379–387 Bouins fixative ............................................. 279, 281, 282 Brain cortex ................................................................... 308 Branched DNA amplification ..................... 230, 231, 278 Break-apart ...................................................49–51, 53, 56 Breast cancer...................................................65, 258, 263 Brightfield.................................................... 265, 272, 274

C CaCo-2 cells ......................................................... 108, 109 Cassava brown streak virus .................................. 204, 214 CD3 ............................................................................... 309 Cervical cancer ..................................................... 101, 309 Chromogenic detection....................................... 182, 210 Chromosome identification......................................71, 72 Chromosome painting .................................................... 20 CircHIPK3 ...................................................380, 384–387 CircRNA, see Circular RNAs (CircRNAs) Circular RNAs (CircRNAs) ........................ 379, 380, 387 Circulating tumor cells (CTCs) .......................... 362, 363 Colorectal cancer.................................258, 349–359, 387

Controls...........................................................5, 6, 12, 14, 15, 50–52, 56, 67, 75, 85, 93, 105, 108, 113–116, 122, 123, 137, 151, 173, 197, 205, 208, 210, 212, 222–226, 245–248, 251, 303, 305, 306, 310, 351, 352, 355, 358, 359, 363, 375, 381, 382, 386 CP2 mRNA .......................................................... 279, 281 CTCs, see Circulating tumor cells (CTCs)

D Dark-field......................................................................... 11 Deconvolve ...................................................................... 93 Digital pathology .............................................22, 23, 222 Dihydrodipicolinate reductase (DapB) mRNA ........................................... 15, 381 DNA probes ......................................................36, 37, 42, 43, 60, 78–80, 181, 258, 265, 267 Dual ISH-IF .................................................302, 308–310 Dual ISH-IHC .............................................................. 310

E East African cassava mosaic virus ........................ 204, 214 Embryos .................................................................. 9, 153, 159–177, 179–193, 195–201 Embryo staging .................................................... 200–201

F Fluorescence in situ hybridisation (FISH)...................................................... 5, 20, 21, 23, 24, 26, 31, 35–40, 42, 43, 45, 46, 48, 49, 53–68, 71, 72, 74–75, 80–81, 85–94, 313 Fluorophore-labelled .......................................... 139, 154, 176, 183, 311 Formalin fixation................................................. 5, 22, 55, 63, 208 Formalin-fixed paraffin-embedded (FFPE) ....................................................... 8, 9, 21, 22, 37, 40, 42, 44–45, 50, 53, 56, 63–65, 68, 103, 107, 204–207, 212, 222, 225, 226, 232, 234, 240, 247, 248, 258, 259, 261, 279, 302–305, 308–310, 327, 331–345, 349–359, 381, 382, 386

Boye Schnack Nielsen and Julia Jones (eds.), In Situ Hybridization Protocols, Methods in Molecular Biology, vol. 2148, https://doi.org/10.1007/978-1-0716-0623-0, © Springer Science+Business Media, LLC, part of Springer Nature 2020

389

IN SITU HYBRIDIZATION PROTOCOLS

390 Index G

K

GAPDH mRNA ............................................................ 133 Gene arrangements ......................................................... 86 Gene expression profiling ............................180, 313–329 Genomics ...........................................................36, 54, 60, 64, 72, 86, 87, 225, 226 GeoMX DSP ....................................................... 333, 334, 336, 339, 341–344 GLP-1 receptor (GLP-1R) ......................... 220, 226, 227

KRAS ........................................................... 353, 355, 358 KRAS G12V ......................................................... 355, 358

H HALO® ............................................................. 23–26, 28, 30, 31, 222, 355 HCR, see Hybridization chain reaction (HCR) Heat induced epitope retrieval (HIER)............................................ 261, 278–281, 296, 297 HEK cells............................................................. 130, 131, 133, 135 HeLa cells .....................................................382, 384–386 HIER, see Heat induced epitope retrieval (HIER) High-plex..................................................... 277, 280, 331 High throughput ................................................ 111, 112, 313–329 Horse radish peroxidase (HRP) ......................... 4, 10, 21, 113–116, 122, 230–232, 235–237, 246, 247, 254, 260, 266, 268, 271, 274, 279, 280, 284, 306, 307, 310 HRP, see Horse radish peroxidase (HRP) HT-29 cells........................................................... 102, 109 Hybridization chain reaction (HCR)129, 130, 133, 136, 138–140, 145–147, 150, 151, 153, 154, 161, 163, 164, 172, 175–177, 180–185, 188–193

I Image analysis.................................................... 15, 19–31, 93, 100, 115, 119, 150–153, 172–176, 221, 224, 246, 251, 259, 272, 277, 323, 363, 368, 373–375 Immunofluorescence (IF).......................... 247, 252, 253, 272, 285, 301, 302, 309–311 In situ HCR......................................................... 127, 128, 137, 143–145, 152, 159–163, 168–171, 173 In situ hybridization (ISH) ..........................3–12, 14, 15, 19–22, 24, 100, 101, 105, 107–109, 124, 179–184, 188–191, 195, 203, 205, 219–227, 246, 253, 258, 259, 261, 265, 267, 270, 271, 285, 294–297, 301, 302, 305–311, 331, 333, 380, 384–387 In situ sequencing ................................................ 313–329

L Leica BOND ................................................ 20, 222, 230, 232–236, 238, 240, 241, 243, 244, 258, 261, 272, 280, 282, 306, 343 LNCaP cells................................................. 366, 374, 375 Locked nucleic acids (LNA) ............................... 4, 8, 100, 101, 104, 108, 109, 258, 363, 364, 366, 380 Long non-coding RNA (lncRNA).................... 8, 20, 124

M Mammalian .................................................................3, 22, 127–140, 143–155, 266 Messenger RNA (mRNA) .................................. 3, 5, 7, 8, 14, 15, 20, 127–131, 133, 136–139, 143–146, 150–154, 159–164, 172, 175, 176, 181, 197, 220–223, 225–227, 235, 245–254, 257, 258, 277–279, 308, 309, 314–316, 331, 349, 351, 359, 361–377, 380–386 MicroRNA (miRNA) ............................................ 4, 8, 20, 100, 105, 107 Mir-15a ............................................................................ 13 Mir-17.......................................................... 101, 108, 109 Mir-21......................................................... 100, 101, 105, 107–109, 260, 263, 265, 267 miRNA, see MicroRNA (miRNA) Molecular cytogenetics ................................................... 56 Morphology.....................................................6, 9, 10, 36, 37, 40, 56, 118, 122, 183, 224, 280, 302, 327, 335, 338, 339, 344, 349, 375 mRNA flow cytometry ............................... 127, 129, 130 Multiplex .......................................................8–10, 14, 24, 113, 114, 121, 122, 179–193, 195–201, 204, 210, 214, 230, 232–237, 239, 245–249, 251–253, 256–274, 277–297, 302, 305–308, 310, 311, 331–344 Mutation detection ........................................20, 349–359

N NeuN ........................................................... 242, 308, 309 Nonoverlapping probes .............................. 223, 226, 227 Numerical analysis.....................................................49, 53

IN SITU HYBRIDIZATION PROTOCOLS Index 391 O

Q

Olfm4 mRNA.................................................................. 11 Oligonucleotide probes ....................................9, 20, 195, 196, 246, 333 Operetta....................................................... 113, 115, 242

qHCR .................................................................. 127–140, 143–155, 159–178 Quality assurance (QA)................................................... 61

P Padlock probes (PLPs) ....................................... 314–318, 322, 326, 362–364, 366, 372, 374–376 Pancreatic cancer ......................................... 259, 260, 302 Paraffin embedding ..........................................5, 182, 265 PC-3 cells..................................................... 368, 374, 375 PD1 ................................................................................ 309 PD-L1 ................................................................... 101, 309 PGK1 mRNA ....................................................... 131, 146 Plants ................................................................... 3, 71–73, 79, 203–213 PLPs, see Padlock probes (PLPs) POLR2A.................................................................. 12, 14, 352, 355, 358 PPIB (peptidyl-prolyl cis-trans isomerase B) mRNA ............................................................14, 15 Precision ....................................................... 37, 101, 127, 128, 133, 138, 143, 144, 151, 153, 159–161, 163, 175 Pretreatment...................................................... 6, 7, 9–10, 13, 36, 37, 41, 43, 44–45, 60, 62, 64, 65, 68, 80, 106, 107, 183, 186, 192, 196, 207–208, 234, 240, 247, 248, 264, 268, 280, 281, 296, 304, 305, 310, 311, 357, 383, 386 Probes ............................................................ 4–15, 19–24, 26, 28–31, 36–38, 42–46, 48–54, 56–68, 71, 72, 79–82, 86, 87, 89–94, 101, 104, 105, 107–109, 112–114, 122, 127–131, 133, 134, 136, 137, 139, 143–147, 149–151, 153, 154, 159, 161–164, 170, 172–177, 181, 183, 184, 186, 190, 192, 196–199, 201, 203, 204, 208, 210, 212, 221–227, 230, 231, 234, 235, 241, 243, 246–248, 254, 259–262, 265, 267, 270, 271, 273, 274, 278, 279, 282, 285, 295, 297, 302, 303, 305, 310, 314–316, 318, 321, 322, 324–326, 328, 329, 333, 335, 337, 339, 343, 344, 351–353, 355, 358, 359, 361–377, 380–382, 384–386 Profiling ............................................................... 220, 315, 331, 333 Proteases ..........................................................9, 107, 113, 114, 122, 208, 210, 212, 240, 247, 248, 254, 279–281, 296, 297, 302, 305, 310, 352, 355, 381–383, 386

R RCA, see Rolling circle amplification (RCA) Reproducibility.................................................5, 230, 261 RNA-FISH .................................................................... 111 RNAscope............................................................ 5, 20, 21, 23, 24, 100, 111–122, 180, 195–201, 203–213, 221–224, 226, 229–244, 246–248, 250, 252, 277–297, 301–311, 349, 380, 381, 386 Rolling circle amplification (RCA)....................................................... 314, 316, 321, 322, 362, 363, 372, 375

S Signal amplification ...........................................8, 10, 101, 199–200, 203, 231, 232, 266, 302, 351, 380 Single-cell resolution .......................................... 183, 186, 301, 313, 315 Single-molecule ................................................... 112, 144, 153, 155, 159–178, 203, 245, 315 Single-molecule fluorescent in situ hybridization (smFISH) ........................................ 229, 230, 234, 235, 238, 242, 243 snRNA-U6 (U6) ......................................... 106, 107, 108 Smooth muscle actin............................................ 281, 282 Spatial transcriptomics .................................................. 313 Splice variants ......................................225, 361–377, 380 Stringency wash.................................................36, 38, 41, 46–48, 60, 81, 248, 270

T Tissue micro array (TMA) ................................ 23–26, 28, 29, 212, 248, 251, 253, 306, 308 TMA, see Tissue micro array (TMA) TP2 mRNA .......................................................... 279, 281 Translational control ............................................ 245, 246 TSA, see Tyramide signal amplification (TSA) Tumor heterogeneity .................................................... 349 Tyramides ................................................... 230–232, 260, 267, 268, 272, 278, 284, 285, 289, 290, 292–294 Tyramide signal amplification (TSA)....................................................4, 8, 10, 24, 101, 108, 113, 115, 116, 122, 180, 230, 231, 233–237, 239, 242, 244, 278, 285, 297, 302, 303, 306, 310, 311

IN SITU HYBRIDIZATION PROTOCOLS

392 Index U

UBC (ubiquitin) mRNA ...................................... 14, 384, 385, 386

V Validation.......................................................3–12, 36, 61, 220, 224, 227, 229, 247–249, 251, 253, 351, 358 VCaP cells.................................................... 366, 374, 375 Ventana ....................................................... 212, 222–224, 246, 248, 249, 381–384 Vgat mRNA................................................................... 308 Vglut1 mRNA ............................................................... 308 Vglut2 mRNA ............................................................... 308

Vimentin .............................................................. 260, 261, 265, 281, 282 Virus...................................................................... 203–214

W Whole-mount ..................................................3–5, 9, 153, 159–177, 195, 196

Z Zebrafish .............................................................. 153, 163, 164, 167–169, 171, 177, 184, 195–201 ZZ-pairs ....................................................... 221, 223, 227