OSTEOPOROSIS AND OSTEOARTHRITIS. [2 ed.] 9781071609880, 1071609882

Citation preview

Methods in Molecular Biology 2221

Andre J. van Wijnen Marina S. Ganshina Editors

Osteoporosis and Osteoarthritis Second 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.

Osteoporosis and Osteoarthritis Second Edition

Edited by

Andre J. van Wijnen and Marina S. Ganshina Stem Cell Therapy and Skeletal Regeneration Lab, Mayo Clinic, Rochester, MN, USA

Editors Andre J. van Wijnen Stem Cell Therapy and Skeletal Regeneration Lab Mayo Clinic Rochester, MN, USA

Marina S. Ganshina Stem Cell Therapy and Skeletal Regeneration Lab Mayo Clinic Rochester, MN, USA

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-0988-0 ISBN 978-1-0716-0989-7 (eBook) https://doi.org/10.1007/978-1-0716-0989-7 © Springer Science+Business Media, LLC, part of Springer Nature 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 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 Musculoskeletal regenerative medicine and orthopedic research examine molecular mechanisms, cellular pathways, and animal models for skeletal degeneration. The translational end-goal of these studies is to develop treatments for aging patients either with joint problems that restrict their mobility, or with increased fracture risk due to reduced bone mineral density bone loss. This Methods in Molecular Biology volume is intended to assist investigators concerned with research topics broadly related to osteoporosis, osteoarthritis, intervertebral disc degeneration, as well as other musculoskeletal disorders. In this second edition of Osteoporosis and Osteoarthritis, we have enlisted the expertise of investigators in the field to provide key updates on well-established methods that were previously incorporated in the earlier edition. In addition, we recruited additional experts who provided new chapters to cover recently emerging sophisticated techniques and methods that are essential for an in-depth and state-of-the-art understanding of skeletal development and homeostasis, and the pathological mechanisms that cause skeletal degeneration. The first ten chapters provide detailed methods for cell culture models for examining skeletal biology at the cellular and molecular levels, as well as methodology required for the application of powerful and novel recent techniques such as single-cell sequencing, highthroughput sequencing of novel epigenetic modifications in DNA, fluorescence recovery after photobleaching (FRAP) microscopy, and in silico modeling of signal networks and functional output at the cell level. Furthermore, we have incorporated seven chapters on powerful and informative animal models for understanding skeletal disorders and repair. Notably, we included one chapter on how to measure pain because it is a key symptom for patients with bone or joint disorders in a clinical setting. We are indebted to all authors for their willingness to share detailed protocols that have been developed, implemented, and/or tested in their research groups for use in the fields of skeletal biology and musculoskeletal regeneration. We thank John Walker for his encouragement to pursue the second edition of this volume. We trust that the contents will inspire and benefit the next generation of investigators interested in solving essential questions in skeletal biology that may eventually benefit patients with bone and joint disorders. Rochester, MN, USA

Andre J. van Wijnen Marina S. Ganshina

v

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

PART I

v ix

CELLULAR AND MOLECULAR BIOLOGY OF OSTEOARTHRITIS AND OSTEOPOROSIS

1 Isolation of Murine and Human Osteocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Matthew Prideaux, Amber Rath Stern, and Lynda F. Bonewald 2 Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived Mesenchymal Stromal Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Roberto Narcisi, Wendy J. L. M. Koevoet, and Gergo J. V. M. van Osch 3 A Novel Enzymatic Digestion Approach for Isolation and Culture of Rodent Bone Marrow Mesenchymal Progenitors. . . . . . . . . . . . . . 29 Leilei Zhong, Lutian Yao, and Ling Qin 4 Isolation of Nucleus Pulposus and Annulus Fibrosus Cells from the Intervertebral Disc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Guus G. H. van den Akker, Andy Cremers, Donatus A. M. Surtel, Willem Voncken, and Tim J. M. Welting 5 Engineering Cartilage Tissue by Co-culturing of Chondrocytes and Mesenchymal Stromal Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Yao Fu, Carlo A. Paggi, Amel Dudakovic, Andre J. van Wijnen, Janine N. Post, and Marcel Karperien 6 Generation of Induced Pluripotent Stem Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 David R. Deyle 7 Specimen Preparation for Single-Cell Sequencing Analysis of Skeletal Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Shawon Debnath and Matthew B. Greenblatt 8 Mapping 5-Hydroxymethylcytosine (5hmC) Modifications in Skeletal Tissues Using High-Throughput Sequencing . . . . . . . . . . . . . . . . . . . . . 101 Fiorella Carla Grandi and Nidhi Bhutani 9 Using FRAP to Quantify Changes in Transcription Factor Dynamics After Cell Stimulation: Cell Culture, FRAP, Data Analysis, and Visualization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Kannan Govindaraj and Janine N. Post 10 Quantitative Molecular Models for Biological Processes: Modeling of Signal Transduction Networks with ANIMO . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Sakshi Khurana, Janet Huisman, Stefano Schivo, and Janine N. Post

vii

viii

Contents

PART II 11

12 13

14

15

16

17

IN VIVO MODELS OF SKELETAL TISSUE INJURY, DEGENERATION, AND REPAIR

Generation and Characterization of Mouse Models for Skeletal Disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gabrielle E. Foxa, Ye Liu, Lisa M. Turner, Alexander G. Robling, Tao Yang, and Bart O. Williams Drill Hole Models to Investigate Bone Repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zhijun Li and Jill A. Helms Generation and Experimental Outcomes of Closed Femoral Fracture in Mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joseph L. Roberts, Christopher W. Kinter, and Hicham Drissi Mouse Models of Osteoarthritis: Surgical Model of Post-traumatic Osteoarthritis Induced by Destabilization of the Medial Meniscus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kirsty L. Culley, Purva Singh, Samantha Lessard, Mengying Wang, Brennan Rourke, Mary B. Goldring, and Miguel Otero Immunostaining of Skeletal Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Crystal Idleburg, Madelyn R. Lorenz, Elizabeth N. DeLassus, Erica L. Scheller, and Deborah J. Veis Mapping Regional Cortical Bone Responses to Local Changes in Loading and Systemic Stimuli. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sara H. Windahl, Peter J. Delisser, and Gabriel L. Galea Pain and Activity Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David H. H. Molstad and Elizabeth W. Bradley

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

165

193

205

223

261

275 291 301

Contributors NIDHI BHUTANI • Department of Orthopaedic Surgery, Stanford University, Stanford, CA, USA LYNDA F. BONEWALD • Indiana Center for Musculoskeletal Health and Department of Anatomy, Cell Biology and Physiology, Indiana University, Indianapolis, IN, USA ELIZABETH W. BRADLEY • Department of Orthopedics, University of Minnesota, Minneapolis, MN, USA; Stem Cell Institute, University of Minnesota, Minneapolis, MN, USA ANDY CREMERS • Laboratory for Experimental Orthopedics, Department of Orthopedic Surgery, Maastricht University, Maastricht, The Netherlands KIRSTY L. CULLEY • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA SHAWON DEBNATH • Department of Pathology and Laboratory Medicine, Weill Cornell Medicine, New York, NY, USA ELIZABETH N. DELASSUS • Musculoskeletal Research Center, Histology and Morphometry Core, Washington University, St. Louis, MO, USA; Department of Orthopedics, Washington University, St. Louis, MO, USA PETER J. DELISSER • Veterinary Specialist Services, Brisbane, Australia DAVID R. DEYLE • Department of Medical Genetics, Mayo Clinic, Rochester, MN, USA HICHAM DRISSI • Department of Orthopaedics, Emory University School of Medicine, Atlanta, GA, USA AMEL DUDAKOVIC • Department of Orthopedic Surgery, Mayo Clinic, Rochester, MN, USA; Department of Biochemistry and Molecular Biology, Mayo Clinic, Rochester, MN, USA GABRIELLE E. FOXA • Program for Skeletal Disease and Tumor Microenvironment, Center for Cancer and Cell Biology, Van Andel Institute, Grand Rapids, MI, USA YAO FU • Department of Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands GABRIEL L. GALEA • Developmental Biology and Cancer, UCL GOS Institute of Child Health, London, UK; Comparative Biomedical Sciences, Royal Veterinary College, London, UK MARY B. GOLDRING • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA KANNAN GOVINDARAJ • Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands FIORELLA CARLA GRANDI • Department of Orthopaedic Surgery, Stanford University, Stanford, CA, USA MATTHEW B. GREENBLATT • Department of Pathology and Laboratory Medicine, Weill Cornell Medicine, New York, NY, USA; Research Division, Hospital for Special Surgery, New York, NY, USA JILL A. HELMS • School of Medicine, Stanford University, Palo Alto, CA, USA JANET HUISMAN • Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands; Student BioMedical Engineering, University of Twente, Enschede, The Netherlands

ix

x

Contributors

CRYSTAL IDLEBURG • Musculoskeletal Research Center, Histology and Morphometry Core, Washington University, St. Louis, MO, USA; Department of Orthopedics, Washington University, St. Louis, MO, USA MARCEL KARPERIEN • Department of Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands SAKSHI KHURANA • Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands CHRISTOPHER W. KINTER • Department of Orthopaedics, Emory University School of Medicine, Atlanta, GA, USA WENDY J. L. M. KOEVOET • Department of Otorhinolaryngology, Erasmus MC, University Medical Center, Rotterdam, The Netherlands SAMANTHA LESSARD • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA YE LIU • Program for Skeletal Disease and Tumor Microenvironment, Center for Cancer and Cell Biology, Van Andel Institute, Grand Rapids, MI, USA ZHIJUN LI • School of Medicine, Stanford University, Palo Alto, CA, USA MADELYN R. LORENZ • Musculoskeletal Research Center, Histology and Morphometry Core, Washington University, St. Louis, MO, USA; Division of Bone and Mineral Diseases, Department of Medicine, Washington University, St. Louis, MO, USA DAVID H. H. MOLSTAD • Department of Orthopedics, University of Minnesota, Minneapolis, MN, USA ROBERTO NARCISI • Department of Orthopedics, Erasmus MC, University Medical Center, Rotterdam, The Netherlands MIGUEL OTERO • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA CARLO A. PAGGI • Department of Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands JANINE N. POST • Department of Developmental BioEngineering, TechMed Centre, University of Twente, Enschede, The Netherlands MATTHEW PRIDEAUX • Indiana Center for Musculoskeletal Health and Department of Anatomy, Cell Biology and Physiology, Indiana University, Indianapolis, IN, USA LING QIN • Department of Orthopaedic Surgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA JOSEPH L. ROBERTS • Department of Orthopaedics, Emory University School of Medicine, Atlanta, GA, USA ALEXANDER G. ROBLING • Department of Anatomy and Cell Biology, Indiana University School of Medicine, Indianapolis, IN, USA; Indiana Center for Musculoskeletal Health, Indianapolis, IN, USA; Richard L. Roudebush VA Medical Center, Indianapolis, IN, USA BRENNAN ROURKE • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA ERICA L. SCHELLER • Musculoskeletal Research Center, Histology and Morphometry Core, Washington University, St. Louis, MO, USA; Division of Bone and Mineral Diseases, Department of Medicine, Washington University, St. Louis, MO, USA STEFANO SCHIVO • Department of Computer Science, Open University, Heerlen, The Netherlands PURVA SINGH • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA

Contributors

xi

AMBER RATH STERN • Engineering Systems Inc., Charlotte, NC, USA DONATUS A. M. SURTEL • Laboratory for Experimental Orthopedics, Department of Orthopedic Surgery, Maastricht University, Maastricht, The Netherlands LISA M. TURNER • Pathobiology and Biorepository Team, Center for Cancer and Cell Biology, Van Andel Institute, Grand Rapids, MI, USA GUUS G. H. VAN DEN AKKER • Laboratory for Experimental Orthopedics, Department of Orthopedic Surgery, Maastricht University, Maastricht, The Netherlands GERGO J. V. M. VAN OSCH • Department of Orthopedics, Erasmus MC, University Medical Center, Rotterdam, The Netherlands; Department of Otorhinolaryngology, Erasmus MC, University Medical Center, Rotterdam, The Netherlands ANDRE J. VAN WIJNEN • Department of Orthopedic Surgery, Mayo Clinic, Rochester, MN, USA; Department of Biochemistry and Molecular Biology, Mayo Clinic, Rochester, MN, USA DEBORAH J. VEIS • Musculoskeletal Research Center, Histology and Morphometry Core, Washington University, St. Louis, MO, USA; Division of Bone and Mineral Diseases, Department of Medicine, Washington University, St. Louis, MO, USA; Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, USA WILLEM VONCKEN • Department of Molecular Genetics, Maastricht University, Maastricht, The Netherlands MENGYING WANG • Orthopedic Soft Tissue Research Program, HSS Research Institute, The Hospital for Special Surgery, New York, NY, USA TIM J. M. WELTING • Laboratory for Experimental Orthopedics, Department of Orthopedic Surgery, Maastricht University, Maastricht, The Netherlands BART O. WILLIAMS • Program for Skeletal Disease and Tumor Microenvironment, Center for Cancer and Cell Biology, Van Andel Institute, Grand Rapids, MI, USA SARA H. WINDAHL • Division of Pathology, Department of Laboratory Medicine, Karolinska University Hospital, Karolinska Institutet, Huddinge, Sweden TAO YANG • Program for Skeletal Disease and Tumor Microenvironment, Center for Cancer and Cell Biology, Van Andel Institute, Grand Rapids, MI, USA LUTIAN YAO • Department of Orthopaedic Surgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA LEILEI ZHONG • Department of Orthopaedic Surgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

Part I Cellular and Molecular Biology of Osteoarthritis and Osteoporosis

Chapter 1 Isolation of Murine and Human Osteocytes Matthew Prideaux, Amber Rath Stern, and Lynda F. Bonewald Abstract Osteocytes are thought to be the mechanosensors of bone by sensing mechanical loads imposed upon the bone and transmitting these signals to the other bone cells to initiate bone modeling and remodeling. The location of osteocytes deep within bone is ideal for their function. However, this location makes the study of osteocytes in vivo technically difficult. There are several methods for obtaining and culturing primary osteocytes for in vitro experiments and ex vivo observation. In this chapter, several proven methods are discussed including the isolation of avian osteocytes from chicks and osteocytes from calvaria and long bones of young mice. A detailed protocol for the isolation of osteocytes from hypermineralized bone of mature and aged animals is provided. In addition, a modified version of this protocol that can be used to isolate osteocytes from human trabecular bone is described. Key words Osteocyte, Isolation, Age, Culture, Collagenase, Mice

1

Introduction Osteocytes are the most abundant of the bone cells and recently found to be multifunctional [1, 2]. They serve as orchestrators of bone remodeling and regulators of mineral homeostasis. They are the mechanosensors of bone, sensing imposed bone loads, and translating these mechanical signals into biological signals of bone modeling and remodeling. They are housed in cave-like voids within the bone called lacunae. Their location deep within the mineralized bone matrix is ideal for their cellular functions, but makes their observation and study difficult. Methods to isolate these bone matrix-embedded cells have been developed throughout the years and vary by the species, state, and extent of mineralization of the bone. In 1992, the group of Peter Nijweide was the first to describe the isolation of osteocytes from 18-day-old chick embryos. Their approach yielded a relatively pure population of osteocytes based on morphology [3]. In 1995, Kumegawa and colleagues published a method for isolating primary avian osteocytes from the parietal

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021

3

4

Matthew Prideaux et al.

bones of 16-day-old chick embryos [4]. Osteocyte morphology and possible dedifferentiation into osteoblasts was noted in this study. The method proved reproducible and useful in isolating primary avian osteocytes for study by other researchers [5– 12]. The bones isolated from the chick embryos are essentially paper thin and not yet mineralized. The parietal bone is flexible and easy to digest, making the isolation of avian embryonic osteocytes rather quick and straightforward. However, the drawbacks of this initial method were that the primary osteocytes were very young themselves because they were isolated from embryonic bone, and they were avian, not mammalian. The need to develop isolation methods for osteocytes from other species was apparent. A method for isolating primary osteocytes from the calvaria of neonatal rats was described in 1995 [13]. Mikuni-Takagaki et al. characterized the osteoblast-osteocyte lineage by describing the subpopulations of isolated cells. These methods were utilized in several subsequent publications on the investigation the mechanotransduction of osteocytes [14–16]. These studies showed that the various populations of isolated bone cells responded to mechanical strain in different manners and at different magnitudes, providing insight into the highly strain responsive nature of osteocytes. Other researchers have also utilized this isolation method in their studies of primary neonatal rat calvaria osteocytes [17]. Calvaria from young chicks and neonatal rats are all very thin and easily processed using sequential collagenase digestions and calcium chelation with EDTA (ethylenediaminetetraacetic acid). The rationale for these sequential steps is that removal of mineral exposes collagen fibers that if digested will release cells embedded in mineralized tissue. Studies utilizing these primary osteocytes can provide insight into the behavior of osteocytes during early development but are not suitable for the study of osteocytes from skeletally mature bone, and do not allow the comparison between primary osteocytes isolated from animals of different ages, species, and genotypes. The calvaria are also not bones that are typically mechanically loaded longitudinally during everyday activity and regularly modeled and remodeled, such as the long bones (femurs, humeri, and tibiae). Methods for isolating calvarial osteocytes have also been adapted and applied to the isolation of osteocytes from neonatal and very young murine long bones with success, and were even utilized in the creation of several osteocyte-like cell lines from mice of 2–3 months of age [18–20]. This method was used to compare osteoblast and osteocyte function and gene expression in several studies [21–23]. To study the effects of age on osteocytes and osteocytes isolated from high bone mass mice, a method for isolating primary osteocytes from hypermineralized bone was still needed. When the methods for isolating primary osteocytes from hypomineralized bone such as young calvaria and long bones were employed for

Isolation of Murine and Human Osteocytes

5

hypermineralized bone, they produced a very low yield rate mainly yielding only the surface cells and shallowly embedded osteocytes. When characterized, the populations of cells were mixed with considerable variation from isolation to isolation. We recently published a method for isolating osteocytes from hypermineralized bone utilizing nine sequential collagenase and EDTA treatments. It is similar to previous methods, but key differences are that the periosteum was removed, and a tissue homogenizer was employed prior to the final digestion [24]. This method has been utilized by several laboratories to isolate primary osteocytes from mature (4–6 months) and aged (22–24 months) murine bone [25, 26]. Although animal models provide a valuable tool to investigate the pathobiology of human diseases, there are distinct differences between murine and human bone formation and remodeling [27]. Therefore, it is essential that discoveries made using mouse osteocytes are also confirmed in human cells. As for studies using murine osteocytes, this has proven difficult due to the location of the osteocytes buried deep within the bone matrix. In vitro studies can be performed by differentiating human osteoblast cell lines into osteocytes, but these cell lines often fail to fully recapitulate the mature osteocyte phenotype [28, 29]. We therefore describe an additional technique for the isolation of primary human osteocytes for in vitro culture, which retain similar characteristics to osteocytes in vivo. This technique will enable researchers to translate findings in murine osteocyte cell lines and primary cells into human osteocyte biology.

2

Materials This technique facilitates the isolation of osteocytes from skeletally mature bone (older than 3–4 months) to aged bone (22–24 months), and was originally published in Biotechniques [24]. A modified protocol adapting this technique for isolating osteocytes from human bone is also described [30]. The early digestions from both techniques (where noted) can also be used for obtaining primary osteoblasts. Prior to starting the isolation, several solutions and media must be prepared. Many of these solutions can be prepared the day before the isolation procedure, however the collagenase solution should be freshly prepared on the day of the procedure to ensure optimal activity of the enzyme.

2.1 Isolation of Osteocytes from Skeletally Mature Mouse Bone

1. Collagenase solution: Dissolve 300 active units/mL of collagenase type IA (Sigma-Aldrich, St. Louis, MO) in α-minimal essential medium (αMEM). 50 mL is adequate for an isolation from one or two mice (see Note 1).

6

Matthew Prideaux et al.

2. EDTA solution: Prepare the 5 mM ethylenediaminetetraacetic acid tetrasodium salt dehydrate (EDTA) solution in magnesium- and calcium-free Dulbecco’s phosphate-buffered solution (DPBS) with 1% bovine serum albumin. Bring to a neutral pH of 7.4 by adding HCl. 30 mL is adequate for osteocyte isolation from one or two mice (see Note 2). 3. Primary bone cell culture medium: On the day before the isolation, supplement α-minimal essential medium (αMEM) with 5% heat-inactivated fetal bovine serum (FBS), 5% heatinactivated calf serum (CS), and 1% penicillin and streptomycin (PS). This culture medium is chosen based on the culture of the MLO-Y4 osteocyte cell line [19]. Store at 4 ˚C. 4. Collagen-coated plates: On the day before the isolation and in a sterile tissue culture hood, dilute sterile collagen in previously filtered sterilized 0.02 M acetic acid to a final concentration of 0.15 mg/mL (see Note 3). General use, 8 mL for coating a 100 mm dish. Coat plates for 1 h at room temperature. Tilt to remove excess collagen and save. This solution can be reused approximately 6 times and should be kept at 4  C. To use plates immediately, it is best to rinse the plate with PBS to remove residual acid; otherwise dry the plates for 1 h (without rinsing with PBS) with the lids off before storing at 4  C. 5. Surgical instruments to dissect and mince bones: forceps, surgical scissors, and scalpels. 6. 25-, and/or 27-gauge needles and 1 mL syringes. 7. 100% Ethanol. 8. 70% Ethanol. 9. Hank’s balanced salt solution (HBSS) calcium and magnesiumfree. 10. α–Minimal essential medium (αMEM). 11. Heat-inactivated fetal bovine serum (FBS). 12. Penicillin and streptomycin. 13. Gentamicin (optional). 14. 6-Well petri dishes (non-TC treated). 15. 100 mm Petri dishes (non-TC treated). 16. Shaker in incubator. 17. Tissue homogenizer (Medimachine (BD Biosciences, San Jose, CA)) with a stainless steel mincing screen with a pore size of 50 μm). 2.2 Isolation of Osteocytes from Human Trabecular Bone

1. Collagenase solution: Dissolve 2 mg/mL (approx. 250–300 U) collagenase type II in α-MEM supplemented with L-glutamine, ribonucleosides and deoxyribonucleosides, 1% penicillin/streptomycin, and sterile filter. 85 mL of collagenase solution is sufficient to isolate osteocytes from 1.5 g of trabecular bone.

Isolation of Murine and Human Osteocytes

7

2. EDTA/BSA solution: Dissolve tetra sodium EDTA in calcium and magnesium-free Hank’s balanced salt solution (HBSS) to a concentration of 5 mM and add 0.1% BSA. Adjust the pH to 7.4 with concentrated HCl and sterile filter. 85 mL of EDTA solution is sufficient for 1.5 g of bone. 3. Osteoblast media preparation: Add 10% heat-inactivated FBS, 1% penicillin and streptomycin to α-MEM. 4. Osteocyte media preparation: Add 2.5% FBS, 1.8 mM potassium phosphate, 1% penicillin and streptomycin. 5. Collagen-coated plates: These are prepared the same as for the murine protocol. 6. Sterile surgical tools: Bone cutters, forceps, scalpel. These are sterilized by autoclaving prior to use. 7. 70% Ethanol. 8. Heat-inactivated FBS. 9. Calcium and magnesium-free HBSS. 10. α-MEM with L-glutamine, deoxyribonucleosides.

3

ribonucleosides

and

Methods The osteocyte isolation protocol from mouse bone takes approximately 10–12 h from the time the mice are sacrificed to the time that the bone particles are plated. The length of time depends on the number of mice used and familiarity of the researchers with the protocol. The protocol for human bone has fewer digests and takes approximately 4–5 h. We have found that preparing additional collagenase or EDTA treatments after the sixth digest does not yield a significant increase in cell number.

3.1 Isolation of Osteocytes from Skeletally Mature Mouse Bone

1. Aseptically dissect the long bones (femurs, tibiae, and humeri) from the mice using surgical scissors or scalpel. Be sure not to break any of the bones at this point and also try to keep the abdomen intact during the dissection to reduce contamination potential (see Note 4). 2. After dissection of bones and removal of as much soft tissue as possible, place them in 100 mm petri dishes containing αMEM with 1% penicillin and streptomycin (and gentamicin (25 μg/ mL)—optional) (see Note 4). 3. Remove any remaining muscle and connective tissue from the bones and scrape away the periosteum using a scalpel (see Note 4).

8

Matthew Prideaux et al.

4. Wash the bones in sequential dishes/wells of a six-well plate filled with αMEM +10% penicillin and streptomycin to remove fur and other contaminants. 5. Place bones in a 100 mm petri dishes with fresh αMEM with 1% penicillin and streptomycin (and gentamycin (25 μg/mL)— optional). 6. Cut off the bone epiphyses and flush the marrow out using a needle and syringe. 7. Wash the hollowed bone pieces again in αMEM with 1% penicillin and streptomycin (and gentamicin (25 μg/mL)— optional). 8. Cut the bones in half lengthwise and then cut into 1–2 mm lengths using a scalpel. 9. As the bone pieces are cut place in HBSS for a brief wash. 10. Collagenase Treatment 1: Incubate the bone pieces in warmed collagenase solution for 25 min (see Note 5). 11. Aspirate the solution and keep for cell plating (if interested in Digest 1 cells) (see Notes 6 and 7). 12. Wash the bone pieces with HBSS three times with 5 mL each, each time adding the HBSS rinse to the aspirated solution for cell plating. 13. Pellet, resuspend, and plate the cells on collagen-coated plates using the primary bone cell culture medium. 14. Collagenase Treatment 2: Repeat steps 10–13. 15. Collagenase Treatment 3: Repeat steps 10–13, again. Combine cells with those from step 14 (see Note 8). 16. EDTA Treatment 1: Incubate the bone pieces in warmed EDTA solution for 25 min (see Note 5). 17. Aspirate the solution and keep for cell plating (if interested in Digest 4 cells) (see Notes 6 and 7). 18. Repeat steps 12 and 13. 19. Collagenase Treatment 4: Incubate the bone pieces in warmed collagenase solution for 25 min (see Note 5). 20. Aspirate the solution and keep for cell plating (if interested in Digest 5 cells) (see Notes 6 and 7). 21. Repeat steps 12 and 13 (see Note 8). 22. EDTA Treatment 2: Incubate the bone pieces in warmed EDTA solution for 25 min (see Note 5). 23. Aspirate the solution and keep for cell plating (if interested in Digest 6 cells) (see Notes 6 and 7). 24. Repeat steps 12 and 13 (see Note 9).

Isolation of Murine and Human Osteocytes

9

25. Collagenase Treatment 5: Incubate the bone pieces in warmed collagenase solution for 25 min (see Note 5). 26. Aspirate the solution and keep for cell plating (if interested in Digest 7 cells) (see Notes 6 and 7). 27. Repeat steps 12 and 13 (see Note 10). 28. EDTA Treatment 3: Incubate the bone pieces in warmed EDTA solution for 25 min (see Note 5). 29. Aspirate the solution and keep for cell plating (if interested in Digest 8 cells) (see Notes 6 and 7). 30. Repeat steps 12 and 13 (see Note 11). 31. Collagenase Treatment 6: Incubate the bone pieces in warmed collagenase solution for 25 min (see Note 5). 32. Aspirate the solution and keep for cell plating (see Notes 6 and 7). 33. Repeat steps 12 and 13 (see Note 12). 34. Mince the bone pieces in αMEM utilizing a tissue homogenizer. 35. Directly plate the resulting suspension of bone particles in αMEM on collagen-coated plates adding additional primary bone cell culture medium if needed (see Note 13). 3.2 Isolation of Osteocytes from Human Trabecular Bone

1. Under sterile conditions, use the bone cutters to remove trabecular bone pieces from the surgical samples (see Note 14). 2. Further dissect the bone pieces into 1–2 mm fragments using the bone cutters or a scalpel. 3. Place the bone pieces (1–1.5 g, wet weight) into a 50 mL culture tube. 4. Wash the bone pieces 3 times with 20 mL HBSS containing Pen/Strep. 5. Digest the bone pieces in 20 mL pre-warmed collagenase type II (2 mg/mL in α-MEM containing Pen/Strep) for 25 min with gentle shaking at 37  C. 6. Remove the collagenase from the bone pieces. Wash the bone pieces 3 times with 10 mL HBSS (see Note 15). 7. Incubate the bone pieces in a second digest of 20 mL collagenase for 25 min. Collect the supernatant and save. Wash the bone pieces 3 times in 10 mL HBSS and add the rinsate to the saved collagenase solution. 8. Repeat step 7 (third collagenase digest) and combine the digest solutions and the rinsates. Centrifuge at 500  g for 5 min and resuspend the pelleted cells in 3 mL of osteoblast media. Divide between 2–3 wells of a collagen-coated 12-well plate (see Notes 16 and 17).

10

Matthew Prideaux et al.

9. Add 40 mL EDTA solution (5 mM in HBSS with 0.1% BSA, pH 7.4) to the bone pieces and incubate for 25 min at 37  C with gentle shaking. 10. Remove and keep the EDTA cell suspension. Rinse the bone pieces 3 times with 10 mL HBSS and add the rinsate to the EDTA. Centrifuge at 500  g and resuspend in 2 mL osteocyte media. 11. Repeat the collagenase digest as in step 7. Centrifuge the collagenase and rinsate at 500  g and resuspend the cells in 2 mL osteocyte media. 12. Repeat the EDTA treatment as in steps 9 and 10. Combine the resuspended cells with those from steps 10 and 11 and centrifuge at 500  g. Resuspend in 3 mL osteocyte media (see Note 18). 13. Split the cells between 2–3 wells of a collagen-coated 12-well plate. Culture for 4–5 days to allow the cells to attach. Replace half the culture media with fresh osteocyte media 48 h after plating (see Note 19).

4

Notes 1. The collagenase solution must be prepared fresh the morning of the isolation. 2. The EDTA solution can be prepared the day before the isolation and stored at 4 ˚C. 3. Use a chilled pipet so the collagen does not stick. 4. Steps 1 and 2 can be conducted on a lab bench. Steps 3–35 should be performed in a sterile laminar flow hood. 5. 8 mL of solution per well in a six-well plate works well for the long bones from 1–2 mice. 6. The issue of maintaining cell density is quite crucial for the cell attachment and survival of the later digests. It is recommended for an isolation using 1–2 mature mice where it is desired to plate each digest individually, one should use a 6-well plate format. If similar digests are combined together, digests 7–9 for example, one should use a 100 mm dish format. The bone particles derived from 1–2 mature mice can be split between two wells of a 6-well plate. 7. Cells will be immediately visible in digests 1–9 (for cell counting and trypan blue staining) and should attach to the plate within 24–48 h. These will be primarily surface cells such as fibroblasts and osteoblasts. 8. These will be primarily osteoblastic cells.

Isolation of Murine and Human Osteocytes

11

9. These will be a mix of osteoblastic and osteocytic cells. 10. These will be primarily osteoblastic and osteocytic cells. Each subsequent serial digest will yield a greater percentage of osteocytic cells. 11. These will be primarily osteocytic cells. 12. At this point, the bone pieces can also be used for isolation of osteocyte mRNA as described previously [31]. 13. Do not disturb the bone particle cultures for at least 48 h. Moving the dish will cause movement of the bone particles and therefore hinder the attachment of the osteocytes. It is recommended to leave the bone particles for as long as possible, adding additional primary bone cell medium to the dishes at 72 h, and changing to fresh medium at 4 or 5 days post culture. It is recommended to use the osteocytic cultures for experimental purposes before day 7 as that is when they were characterized in the BioTechniques manuscript [24]. Prolonged culture will otherwise lead to dedifferentiation/loss of phenotype or an overgrowth of the cultures by any contaminating fibro- or osteoblasts. 14. Human surgical samples may comprise femoral knee shavings or trabecular bone removed from the hip during arthroplasty. Dedicated surgical tools such as bone cutters are preferred to remove the trabecular bone pieces as scalpel blades are usually not strong enough for human bone. 15. The first digest will contain mainly fibroblasts and marrow cells. These cells can be kept and plated or discarded if only osteoblast and osteocyte fractions are required. 16. If the osteoblast fraction is not required, the collagenase solution from digests 2 and 3 can be discarded. 17. The bone pieces at this stage contain primarily osteocytes. These bone pieces can be cultured in osteocyte media without further digestions, to provide a population of osteocytes still within their native extracellular matrix. 18. Combining the cells from digests 4–6 and the additional centrifugation/resuspension step removes any remaining traces of collagenase and helps the cells to adhere to the culture plate. 19. Culturing the cells for more than 5 days may lead to loss of osteocyte morphology and phenotype. The cells from digests 4–6 may also be suspended in a collagen type I gel, which results in the cells acquiring a highly dendritic 3D morphology similar to osteocytes in vivo [30, 32].

12

Matthew Prideaux et al.

References 1. Bonewald LF (2011) The amazing osteocyte. J Bone Miner Res 26(2):229–238 2. Dallas SL, Prideaux M, Bonewald LF (2013) The osteocyte: an endocrine cell and more. Endocr Rev 34(5):658–690 3. van der Plas A, Nijweide PJ (1992) Isolation and purification of osteocytes. J Bone Miner Res 7(4):389–396 4. Tanaka K et al (1995) Time-lapse microcinematography of osteocytes. Miner Electrolyte Metab 21(1–3):189–192 5. Aarden EM et al (1996) Adhesive properties of isolated chick osteocytes in vitro. Bone 18 (4):305–313 6. Ajubi NE et al (1996) Pulsating fluid flow increases prostaglandin production by cultured chicken osteocytes—A cytoskeletondependent process. Biochem Biophys Res Commun 225(1):62–68 7. Kamioka H, Honjo T, Takano-Yamamoto T (2001) A three-dimensional distribution of osteocyte processes revealed by the combination of confocal laser scanning microscopy and differential interference contrast microscopy. Bone 28(2):145–149 8. Kamioka H et al (2007) Primary cultures of chick osteocytes retain functional gap junctions between osteocytes and between osteocytes and osteoblasts. Microsc Microanal 13 (2):108–117 9. Kamioka H et al (2006) Fluid shear stress induces less calcium response in a single primary osteocyte than in a single osteoblast: implication of different focal adhesion formation. J Bone Miner Res 21(7):1012–1021 10. Klein-Nulend J et al (1995) Pulsating fluid flow increases nitric oxide (NO) synthesis by osteocytes but not periosteal fibroblasts--correlation with prostaglandin upregulation. Biochem Biophys Res Commun 217(2):640–648 11. Klein-Nulend J et al (1995) Sensitivity of osteocytes to biomechanical stress in vitro. FASEB J 9(5):441–445 12. Westbroek I et al (2000) Differential stimulation of prostaglandin G/H synthase-2 in osteocytes and other osteogenic cells by pulsating fluid flow. Biochem Biophys Res Commun 268(2):414–419 13. Mikuni-Takagaki Y et al (1995) Matrix mineralization and the differentiation of osteocytelike cells in culture. J Bone Miner Res 10 (2):231–242 14. Kawata A, Mikuni-Takagaki Y (1998) Mechanotransduction in stretched osteocytes—Temporal expression of immediate early and other

genes. Biochem Biophys Res Commun 246 (2):404–408 15. Mikuni-Takagaki Y (1999) Mechanical responses and signal transduction pathways in stretched osteocytes. J Bone Miner Metab 17 (1):57–60 16. Mikuni-Takagaki Y et al (1996) Distinct responses of different populations of bone cells to mechanical stress. Endocrinology 137 (5):2028–2035 17. Gu G et al (2005) Estrogen protects primary osteocytes against glucocorticoid-induced apoptosis. Apoptosis 10(3):583–595 18. Kato Y et al (2001) Establishment of an osteoid preosteocyte-like cell MLO-A5 that spontaneously mineralizes in culture. J Bone Miner Res 16(9):1622–1633 19. Kato Y et al (1997) Establishment of an osteocyte-like cell line, MLO-Y4. J Bone Miner Res 12(12):2014–2023 20. Woo SM et al (2011) Cell line IDG-SW3 replicates osteoblast-to-late-osteocyte differentiation in vitro and accelerates bone formation in vivo. J Bone Miner Res 26(11):2634–2646 21. Zhao S et al (2002) MLO-Y4 osteocyte-like cells support osteoclast formation and activation. J Bone Miner Res 17(11):2068–2079 22. Kramer I et al (2010) Osteocyte Wnt/betacatenin signaling is required for normal bone homeostasis. Mol Cell Biol 30(12):3071–3085 23. Nakashima T et al (2011) Evidence for osteocyte regulation of bone homeostasis through RANKL expression. Nat Med 17 (10):1231–1234 24. Stern AR et al (2012) Isolation and culture of primary osteocytes from the long bones of skeletally mature and aged mice. BioTechniques 52 (6):361–373 25. Jahn K et al (2012) Skeletal muscle secreted factors prevent glucocorticoid-induced osteocyte apoptosis through activation of betacatenin. Eur Cell Mater 24:197–209. discussion 209-10 26. Kalajzic I et al (2013) In vitro and in vivo approaches to study osteocyte biology. Bone 54(2):296–306 27. Jilka RL (2013) The relevance of mouse models for investigating age-related bone loss in humans. J Gerontol A Biol Sci Med Sci 68 (10):1209–1217 28. Prideaux M et al (2014) SaOS2 osteosarcoma cells as an in vitro model for studying the transition of human osteoblasts to osteocytes. Calcif Tissue Int 95(2):183–193

Isolation of Murine and Human Osteocytes 29. Bodine PV, Vernon SK, Komm BS (1996) Establishment and hormonal regulation of a conditionally transformed preosteocytic cell line from adult human bone. Endocrinology 137(11):4592–4604 30. Prideaux M et al (2016) Isolation of osteocytes from human trabecular bone. Bone 88:64–72

13

31. Qing H et al (2012) Demonstration of osteocytic perilacunar/canalicular remodeling in mice during lactation. J Bone Miner Res 27 (5):1018–1029 32. Skottke J, Gelinsky M, Bernhardt A (2019) In vitro co-culture model of primary human osteoblasts and osteocytes in collagen gels. Int J Mol Sci 20(8):1998

Chapter 2 Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived Mesenchymal Stromal Cells Roberto Narcisi, Wendy J. L. M. Koevoet, and Gergo J. V. M. van Osch Abstract Human bone marrow-derived mesenchymal stem/stromal cells (BM-MSC) are adult multipotent progenitor cells that can be isolated from bone marrow. BM-MSCs have the ability to be expanded and differentiated into the chondrogenic lineage in vitro. Here we describe a standardized method to expand and chondrogenically differentiate human BM-MSCs, highlighting how to overcome technical challenges and indicating the most common readout parameters to evaluate the chondrogenic differentiation capacity. Key words Tissue engineering, Regenerative medicine, Cartilage, Chondrogenesis, Mesenchymal stem cells

1

Introduction In this manuscript we describe the materials and methods we use to isolate, expand, and chondrogenically differentiate human bone marrow-derived mesenchymal stem/stromal cells (BM-MSC). Historically, some important milestones substantially contributed to improve the understanding and the use of BM-MSC as research tool. Self-renewing bone marrow-derived stem cells have been originally identified by experiment on serial ectopic transplantation of bone marrow, thanks to their ability to recapitulate the generation of complex bone structures including the bone marrow and stroma [1, 2]. Later, the characterization and the definition of these cells evolved, and with it, the name by which those cells have been called [3–6], bringing the scientific community to define, in 2006, the minimum requirements for in vitro cultured BM-MSC [7], including surface marker expression profile. However, this is still a developing area of research and surface markers originally suggested as essential for MSC multilineage differentiation have been demonstrated to be less crucial for chondrogenesis [8]. An overview of MSC (sup)population’s surface profiles was discussed in a

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021

15

16

Roberto Narcisi et al.

previous book chapter (https://doi.org/10.1007/978-3-31953316-2_2; Chapter 2). In 1998, the first robust in vitro chondrogenic induction media for BM-MSC was developed by Johnstone et al. [9], and the next year the in vitro multilineage differentiation potential of BM-MSC was characterized [10]. Like in vivo, the heterogeneity of BM-MSC is also present in vitro. Important studies revealed the heterogeneous nature of BM-MSC in vitro [11, 12] and via in vivo transplantation-based assays [13, 14]. This heterogeneity is also linked with the limited proliferation capacity of BM-MSC that was found to be associated with loss of differentiation capacity [11, 15]. To improve proliferation and subsequent differentiation capacity of BM-MSC, the effect of different growth factor supplementations has been tested during expansion, such as WNT, EGF, PDGF but also platelet lysate or platelet-rich plasma. FGF2, however, was one of the first successfully tested factors used to stimulate expansion capacity while maintaining chondrogenic differentiation capacity of BM-MSC [16]. Based on this knowledge, in the course of the last two decades we have established in our laboratory robust protocols to efficiently expand and chondrogenically differentiate BM-MSC. Our current materials and methods, as well as all the small—yet important— details to successfully perform the assays, are here described.

2

Materials All the experiments regarding the manipulation of cells of human origin described in this manuscript are performed under sterile conditions using a Biological Safety Cabinet. All materials and solutions for cell culture should be previously sterilized according with the manufacture’s instruction for each component.

2.1 Expansion of BM-MSCs (See Table 1)

1. Complete expansion medium (CEM; see also Note 1): (a) MEM-α, nucleosides. (b) 10% Fetal calf serum (FCS; see Note 2). (c) 1 ng/mL Fibroblast growth factor-2 (FGF2). (d) 10
4 M Ascorbic acid-2-phosphate. (e) 1.5 μg/mL Fungizone (Amphotericin B). (f) 50 μg/mL Gentamicin. 2. CEM for enhanced expansion and chondrogenic differentiation capacity (CEM+): (a) CEM. (b) 250 ng/mL WNT3A or 2.5 μM CHIR99021, both canonical WNT agonists (see Note 3).

Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . .

17

Table 1 Summary of the materials for the expansion and passaging of BM-MSC Materials

Company

Catalogue

VACUETTE buis, NH (heparinized tubes)

Greiner Bio-One

455051

3% acetic acid with methylene blue

StemCell Technology

07060

Bu¨rker counting chamber

VWR International BV

631-1159

MEM-α, nucleosides

Gibco

22571038

Fetal calf serum (FCS; see Note 2)

Life technologies

Lot: 41Q2047K

Fibroblast growth factor-2 (FGF-2)

Serotec

PHP105

Ascorbic acid-2-phosphate

Sigma-Aldrich

A8960

Fungizone

Gibco

15290026

Gentamicin

Gibco

15750-037

CHIR99021

Stemgent

04-0004

WNT3A

In house production

–

Trypsin-EDTA (0.25%) phenol red

Gibco

T4049

®

Dulbecco’s phosphate-buffered saline, no calcium, no magnesium Gibco (DPBS)

14190-169

BD Falcon®

Culture flasks T175

®

353112 353136

Culture flasks T75

BD Falcon

Culture flasks T25

BD Falcon®

353108

Centrifuge tubes, 50 mL, conical bottom

Greiner Bio-One

T2318500EA

Centrifuge tubes, 15 mL, conical bottom

Westburg

P91015

3. Passaging of BM-MSCs: (a) CEM. (b) Trypsin-EDTA (0.25%) phenol red (Gibco). (c) Dulbecco’s phosphate-buffered saline, no calcium, no magnesium (DPBS). 2.2 Chondrogenic Differentiation of BMMSCs (See Table 2)

1. Complete chondrogenic medium (CCM; see Note 4): (a) DMEM-high glucose, HEPES (DMEM-HG).

GlutaMAX™

(b) Insulin transferring selenic acid (ITS+).

Supplement,

18

Roberto Narcisi et al.

Table 2 Summary of the materials for the chondrogenic differentiation of BM-MSC Materials

Company

Catalogue

Bu¨rker counting chamber

VWR International BV

631-1159

DMEM-high glucose, GlutaMAX(TM), HEPES

Gibco

32430027

Insulin transferring selenic acid (ITS+)

B&D Bioscience

354352

L-proline

Sigma-Aldrich

P5607

Sodium pyruvate

Gibco

11360-039

Transforming growth factor-β1 (TGF-β1)

R&D Systems

240-B

Ascorbic acid-2-phosphate

Sigma-Aldrich

A8960

Dexamethasone

Sigma-Aldrich

D4902

IWP2 (see Note 5)

Stemgent

04-0034

Fungizone

Gibco

15290026

Gentamicin

Gibco

15750-037

Polypropylene centrifuge tubes, screw caps (15 mL)

VWR International BV

TPPA91019

(c) 40 μg/mL L-Proline. (d) 1 mM Sodium pyruvate. (e) 10 ng/mL Transforming growth factor-β1 (TGF-β1). (f) 10
4 M Ascorbic acid-2-phosphate (Sigma-Aldrich). (g) 100 nM Dexamethasone. (h) 1.5 μg/mL Fungizone (Amphotericin B). (i) 50 μg/mL Gentamicin. 2. A medium to induce chondrogenesis with reduced hypertrophy based on the use of the WNT inhibitor IWP2 was recently developed (see Note 5).

3

Methods The most commonly used method for the isolation of BM-MSC from bone marrow aspirates is by plastic adherence. Bone marrow aspirates need to be collected in heparinized tubes and processed ideally within 6 h and anyway no later than 18 h from the harvesting time from the patient/donor. During this time the samples need to be kept at 4  C. Recently, in a comparative analysis, we showed the effect of different ways of harvesting bone marrow aspirates from patients on the chondrogenic differentiation capacity of BM-MSC [17], while maintaining the same isolation procedure.

Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . .

19

After determining the total volume of the aspirate, the cells are counted by mixing 20 μL of bone marrow aspirate with 380 μL of 3% acetic acid with methylene blue (or comparable volumes in a 1:20 ratio). After 2 min the membrane of white and red blood cells will be lysate and the remaining white blood cell nuclei can be counted, for example, with a Bu¨rker counting chamber. The cells are then seeded at a density of approximately 75,000  25,000 cells/cm2 in CEM (see Subheading 2.1) for a total volume of 20 mL in a T175 flasks or 8.5 mL in a T75 flasks (see Notes 6 and 7). In this way BM-MSCs are isolated by their ability to adhere to plastic culture flasks. After 24 h, nonadherent cells are removed by 3 washing with DPBS+2% FCS and adherent cells cultured in standard conditions (5% CO2 at 37  C) for up to 14 days. During this time, cell colonies appear and grow at a speed that depend on individual BM-MSC donor (Fig. 1a). Medium is renewed twice a week. When BM-MSCs neared confluence (Fig. 1b) or after a maximum of 14 days, they are detached and passaged for further expansion (see Subheading 2.1, step 3 and Subheading 3.1). At this point we have Passage-1 (P1) BM-MSC. 3.1 Expansion and Passaging of BMMSCs

P1 BM-MSC are 2 washed with DPBS by adding around 50% of the volume used for the expansion media, then Trypsin-EDTA is added (3 mL for a T175 flask and 1.25 mL for a T75) for 3–4 min in the culture incubator (37  C and 5%CO2) till all the cells look detached after a check with the microscope. The cell suspension is harvested in appropriate sized tubes by adding CEM (containing FCS) in order to neutralize the effect of the Trypsin-EDTA solution. For the harvesting, it is suggested to use a Trypsin-EDTA/ CEM ratio of at least a 1:4. The cell suspension is then centrifuged at 300  g for 6–8 min in 50 or 15 mL tubes, the resulting cellular pellet resuspended in CEM or CEM+ and the cell counted, for example, with a Bu¨rker counting chamber. Next, the BM-MSC are reseeded at the density of 2300 cells/cm for further expansion (see Note 6). Generally, BM-MSC take 4–6 days to reach sub-confluence (75–85% confluence; Fig. 1b) after seeding (see Note 8). In order to keep as much as possible the BM-MSC in continuous proliferation, we do not grow them till confluence (Fig. 1c). Every passage we monitor the BM-MSC for their morphology (Fig. 2) and expansion speed. These parameters are known to be influenced over time in vitro and have an impact on BM-MSC differentiation capacity, and one reason may be cellular senescence [18–20]. For example, when during expansion BM-MSC take >50% more time to reach the sub-confluence compared to the time needed to go from P1 to P2, we stop the culture and do not proceed with the differentiation assay (see Note 9). By default, we would then no longer perform surface marker analysis on our plastic-adherent-selected BM-MSC (see Note 10).

20

Roberto Narcisi et al.

Fig. 1 Representative images of BM-MSC forming a colony 2 days after isolation (a), at 75–85% confluence (b) and over-confluent (c). Scale bar ¼ 200 μm

Fig. 2 Representative images of BM-BMCs showing their classical morphology during the expansion (left panel) and a more enlarged morphology (right panel, black arrow) usually associate with aging and low expansion rate (right panel). Scale bar ¼ 100 μm 3.2 Chondrogenic Differentiation of BMMSCs

BM-MSC are trypsinized as indicated in Subheading 3.1. However, differentially compared to the passaging protocol, after centrifugation at 300  g for 6–8 min, the cells are resuspended in CCM (see Subheading 2.2). After counting, for example, with a Bu¨rker counting chamber, BM-MSC are divided in aliquots of 200,000 cell/ 0.5 mL of CCM in 15 mL polypropylene tubes (see Note 11) and centrifuged at 300  g for 6–8 min to form a pellet (Fig. 3a). The generated pellet cultures are immediately transferred in an incubator for cell culture (37  C, 5%CO2) for the next 24 h. At this stage, the pellets should roundup and look like a sphere-like cellular cluster (Fig. 3b). Now, it is suggested to loosen the pellet from the bottom of the tube by tapping it gently. From this moment-on the pellets are cultured with CCM (medium renew twice per week) for the following 3–5 weeks (see Subheading 2.2, Notes 5 and 12) depending on the research question and the history of the BM-MSC batch (see Note 13). A first positive macroscopic indication that chondrogenesis is taking place is the increasing size of the pellet over time (Fig. 3c).

Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . .

21

Fig. 3 Representative images of pellet cultures (200,000 BM-MSCs/tube) immediately after centrifugation (a), after 24 h in the presence of CCM (b), and after 10 days of chondrogenic induction in CCM (c). Arrows indicate the pellets on the bottom of the tubes 3.3 Evaluation of Chondrogenic Differentiation Capacity of BM-MSC

In this section you will find the general information about our standard assays to evaluate chondrogenic differentiation. This section is not meant to provide all the details necessary to perform the assays, but more to guide you toward the minimum set the analysis we suggest to perform to verify chondrogenic maturation of the tissue.

3.3.1 Histological Analysis

After chondrogenic differentiation we generally fixate the pellet cultures in 4% formaldehyde for 24 h (required for our Collagen type-X antibody), or for a maximum of 60 h. Then the samples are dehydrated, paraffin-embedded, cut in 6 μm thick sections, collected in microscope glass slides (VWR International BV, cat.: KN ITVS112751FEA.01), and dried for a minimum of 16 h at 37  C. We generally perform matrix analysis on the following proteins: 1. Chondrogenesis. Sulfated glycosaminoglycan by Thionin or Safranin-O, Immunostaining for Collagen type-II (Developmental Studies Hybridoma Bank; cat.: II-II6B3) 2. Hypertrophy. Immunostaining for Collagen type-X (Quartett; cat.: 1_CO097–05; or ThermoFisher, cat,: 14–9771-82). To verify cartilage stability, we usually perform an in vivo assay via the subcutaneous implantation of the chondrogenically differentiated pellets in NMRI nu/nu mice for a minimum of 6 weeks, after which we perform the histological analyses. However, in case the in vivo samples are expected to have formed calcified tissue and/or bone, after sample retrieval from the animal we first fix the samples for up to 1 week (depending on the size of the tissue) in 4% formaldehyde and then perform a decalcification step either with 10% EDTA in PBS (pH 7.4) or with formic acid 10% in water for up to 10 days. Next, we proceed with the histological analysis as mentioned above.

22

Roberto Narcisi et al.

3.3.2 Transcript Analysis (See Note 13)

After chondrogenic differentiation we extract the mRNA by first mechanically disrupting the pellet with a micro pestle (Eppendorf, cat.: 0030 120.973) in a 1.5 mL Eppendorf tube containing 350 μL of ice-cold RNA-STAT60 solution (Gentaur, cat: CS-111-200; see Note 14). At this stage the samples can be stored at
80  C till the moment of the RNA extraction (see Note 15), which is done following the manufacturer’s instructions. We generally perform RT-PCR analysis on the following genes 1. Chondrogenic genes. COL2A1 and ACAN (optional: COL2A1-IIB, SOX9) 2. Hypertrophic genes. COL10A1 and ALPL (optional: RUNX2). We have a panel of several housekeeper/reference genes and among them the most used are GAPDH, RPS27a, β-ACTIN, 18S, and HTPR1. We select our reference gene based on its stability across the conditions and/or time-points, with the intention to select one stable gene applicable for all the experiments. However, in case of high variations between the conditions of the same experiments (generally 1 Ct value) we apply the “best housekeeper index” (BHI [21]). We run the BHI by using at least 3 housekeeper/reference genes.

4

Notes If not differentially indicated, all the components need to be stored at 4  C till expiring date. For all the growth factors that need to be reconstituted, it is suggested to centrifuge the vial briefly before opening, to bring the contents to the bottom. Always check datasheet of each individual products for any additional recommendation. 1. Preparation of 500 mL of CEM medium: (a) 446.5 mL of MEM-α * (b) 50 mL FCS (10% of the total volume; see Note 2) (c) 3 mL Fungizone (167; previously aliquoted in 3 mL/ tube and stored in
20  C) (d) 500 μL Gentamicin (1000) This preparation can be stored up to 6 weeks at 4  C. In addition to this, we freshly add to each medium renew FGF2 and ascorbic acid-2-phosphate, prepared as follows: (e) FGF2 preparation: we first dissolve the powder in order to obtain a 5 μg/μL solution in TRIS pH 7.6; we make aliquots of 5 μL in 1.5 mL Eppendorf tubes and store them at
20  C. When needed, the 5 μL aliquot are

Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . .

23

thawed and 995 μL of CEM (without ascorbic acid-2phosphate) is added for a final stock concentration 5 ng/μL. This is to be considered a 5000 stock solution. These aliquots can be stored at 4  C and used for 14 days. Never refreeze the dissolved growth factors. (f) Ascorbic acid-2-phosphate preparation: dissolve the powder in MEM-α (without FCS) to obtain 14.48 mg/mL solution and store in ready-to-use aliquots at
20  C. This is to be considered a 500 stock solution. After thawing, store ascorbic acid-2-phosphate for maximum 24 h at 4  C. *Between 2010 and 20,111 we moved from DMEM to MEMα as basic media. This was based on the results of a comparative study between the two media published in 2009 [22], and on our personal experience (unpublished data). 2. Preparation of FCS: (a) Defrost overnight at 4  C. (b) Heat-inactivate by transferring the FCS at 56  C for 30 min. It is suggested to gently shake/mix the FCS bottle every 5–8 minduring the incubation at 56  C. (c) Make aliquot of 50 mL and store at
20  C until use. It is highly recommended to avoid the use of different FCS batches for experiments belonging to the same project. Different FCS batches may be significantly different in their composition, resulting in different outcomes in term of expansion and differentiation capacity of BM-MSC. It is a standard procedure in our laboratory to screen 3 to 5 different batches of serum from at least 2 different companies for their capacity to support expansion and differentiation capacity of BM-MSC. Then, we purchase the entire selected FCS batch for long-term use ( 2 year). The lot number currently used is Lot: 41Q2047K, as indicated in Table 1. 3. WNT3A and CHIR99021 are canonical WNT agonist, the former acting directly on WNT receptors on the cellular surface, the latter by inhibiting the glycogen synthase kinase 3 (GSK-3) intracellularly. (a) WNT3a preparation: WNT3A was purified from cell culture medium conditioned by Drosophila S2 cells modified with a mouse WNT3A expression vector, using affinity and gel filtration chromatography as described [23]. A stock solution of 50 μg/mL in PBS + 1% of 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) is then prepared and stored at
80  C for up to

24

Roberto Narcisi et al.

6 months. Once thawed, the stock solution can be stared at 4  C for a maximum of 24 h. Note that WNT3A is heat sensitive and in our experiment we use to renew the medium containing WNT3A every day. WNT3A is also commercially available from different companies, and although we never performed comparative studies, we are now successfully using WNT3A from R&D Systems and Sigma-Aldrich for culturing other cell types (unpublished data). (b) CHIR99021 preparation: A stock solution of 5 or 10 mM is prepared by adding dimethyl sulfoxide (DMSO) at 45 to 55  C under gentle shaking. When dissolved, aliquots are prepared according to the experimental needs. This stock solution can be stored for 1 month at
20  C. Once thawed, the stock solution can be stared at 4  C for a maximum of 24 h. We found that the effect of CHIR99021 is comparable to the effect of WNT3A on cell proliferation and subsequent chondrogenic differentiation. However, stimulation with CHIR99021 for more than 2 weeks results in changing in cellular morphology with negative effect on expansion and differentiation capacity [24]. We therefore suggest to use CHIR99021 as WNT3A substitute only for short-term experiments (10 days). 4. Preparation of 500 mL of CCM medium. (a) 485.5 mL DMEM-HG (b) 5 mL ITS+ (100) (c) 1 mL of Proline of 20 mg/mL in DMEM-HG (500; store at 4  C for 6 months) (d) 5 mL of Sodium pyruvate (100) (e) 3 mL Fungizone (167; previously aliquoted in 3 mL/ tube and stored in
20  C) (f) 500 μL Gentamicin (1000) This preparation can be stored up to 6 weeks at 4  C. In addition to this, we freshly add to medium TGF-β1*, ascorbic acid-2-phosphate and dexamethasone, prepared as follows: (g) TGF-β1 preparation: we first dissolve the powder in order to obtain a 2 μg/mL solution in 4 mM HCL +0.1% bovine serum albumin (BSA); we make aliquots of 500 μL in 1.5 mL Eppendorf tubes and store them at
20  C or
80  C up to 1 year. This is to be considered a 200 stock solution. Once thawing can be stored at 4  C and used for 14 days. Never refreeze the dissolved growth factors.

Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . .

25

(h) Ascorbic-acid preparation: dissolve the powder in DMEM-HG (without FCS) and then follow the same indication as described in Note 1. (i) Dexamethasone preparation: we prepare a 1 mM stock solution by adding 10 mL of 100% ethanol to 3.92 g of dexamethasone powder. This is a 10.000 solution. We store in aliquots of around 1 mL at
20  C for up to 1 year. *We use TGF-β1 for our chondrogenic cultures. However, TGF-β3 and sometimes TGF-β2 are used while maintaining the rest of the protocol unchanged. We and others compared the effect of different TGF-β isoforms on chondrogenic differentiation and TGF-β1 or TGF-β3 has been suggested to be the most potent [25– 27], with differences mainly related to the readout parameters taken in account for the analysis and, possibly, to the bioactivity of the recombinant protein purchased from different companies. 5. N-(6-Methyl-2-benzothiazolyl)-2-[(3, 4, 6, 7-tetrahydro-4oxo-3-phenylthieno[3, 2-dpyrimidin-2-yl)thio]-acetamide (IWP2) is a small molecule that specifically inhibits the maturation (endogenous production) of WNT proteins by blocking the Porcupine-mediated WNT palmitoylation. IWP2 is prepared as 2 mM stock solution (1000) by adding 1.072 mL of dimethyl sulfoxide (DMSO) in 1 mg of IWP2 powder. We store aliquots of 50–100 μL at
80  C for up to 2 years. During chondrogenic differentiation experiments, medium containing IWP2 is renewed every other day, and is added starting from day 10 to day 14 of chondrogenic induction [18]. Earlier administration of CCM+ media to the pellets, in our hands, reduced chondrogenic differentiation (unpublished data). 6. We use between 0.11 (T175 flasks) and 0.16 (T75) mL/cm2 of culture medium for our experiments. However, for each readout parameter (e.g., expansion, RT-PCR, Western blot), we design our experiments in order to avoid direct comparison between cells cultured with different amount of expansion media per cm2. 7. We suggest to retain a minimum of 50% of the total volume as CEM. Given that, during the first seeding, all the components of the CEM should be adjusted in order to have all the components of the CEM at the right final concentration, despite the presence of the bone marrow aspirate. 8. Do not allow the cells to reach confluence during expansion. This, in our hands, negatively influences their chondrogenic differentiation capacity (unpublished data).

26

Roberto Narcisi et al.

9. Unless specifically requested by the research question, we generally use P5 BM-MSC for our experiments. 10. Although a specific set of surface markers have been proposed in 2006 to identify BM-MSCs [7], this list is continuously updated and there is no general consensus on the minimum requirements a BM-MSC needs to have in terms of surface marker expression. Moreover, we recently showed that some of the originally selected markers, such as CD105, are not associated with the chondrogenic differentiation capacity of BM-MSC [8]. 11. The use of polypropylene tubes is necessary to prevent cell attachment to the side of the tube before and during centrifugation. Always remember to leave the lid partially open to allow air-gas exchange. It is also possible to generate 100,000 cell/ pellet. In our hands, chondrogenic differentiation progresses similarly compared to the 200,000 cell/pellets; however, the reduced amount of cells in the 100,000 cell pellets can be a limiting factor to perform subsequent analysis (e.g., RT-PCR or Western Blot). 12. Especially in the first 2 weeks of culturing, the pellets have the tendency to attach to the bottom of the tube. It is therefore suggested to gently shake/tap the tube during the medium renewal, in order to get the pellet detached. This will allow a more homogeneous diffusion of the media components and chondrogenic differentiation. 13. Although our standard chondrogenic differentiation protocol consists in 5 weeks of differentiation [9], when we use BM-MSC from a previously tested donor, or the research question we are investigating requires, for example, to investigate early chondrogenic events, we apply a 3- or 4-week differentiation protocol, which has the sole difference of being 2 or 1 week shorter, with no difference in media composition (see, for example, Cleary et al. [28]). 14. Until the end of 2019 we used RNA-Bee (Gentaur, cat.: CS-105B) for the extraction of RNA from pellet cultures and other soft tissues. Due to the fact that RNA-Bee was no longer available and after a series of comparative validation tests regarding extraction efficiency and purity among different extraction protocols, we have moved to RNA-STAT60. 15. By convention, and unless differently indicated, we harvest the samples for RT-PCR analysis 24  1 h after the last medium renewal.

Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . .

27

Acknowledgements This work is part of Medical Delta RegMed4 program. References 1. Tavassoli M, Crosby WH (1968) Transplantation of marrow to extramedullary sites. Science 161:54–56 2. Friedenstein AJ, Petrakova KV, Kurolesova AI, Frolova GP (1968) Heterotopic of bone marrow. Analysis of precursor cells for osteogenic and hematopoietic tissues. Transplantation 6:230–247 3. Pittenger MF, Discher DE, Peault BM, Phinney DG, Hare JM, Caplan AI (2019) Mesenchymal stem cell perspective: cell biology to clinical progress. NPJ Regen Med 4:22 4. Owen M, Friedenstein AJ (1988) Stromal stem cells: marrow-derived osteogenic precursors. Ciba Found Symp 136:42–60 5. Caplan AI (1991) Mesenchymal stem cells. J Orthop Res 9:641–650 6. Bianco P, Gehron RP (2000) Marrow stromal stem cells. J Clin Invest 105:1663–1668 7. Dominici M, Le Blanc K, Mueller I, SlaperCortenbach I, Marini F, Krause D et al (2006) Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy 8:315–317 8. Cleary MA, Narcisi R, Focke K, van der Linden R, Brama PA, van Osch GJ (2016) Expression of CD105 on expanded mesenchymal stem cells does not predict their chondrogenic potential. Osteoarthr Cartil 24:868–872 9. Johnstone B, Hering TM, Caplan AI, Goldberg VM, Yoo JU (1998) In vitro chondrogenesis of bone marrow-derived mesenchymal progenitor cells. Exp Cell Res 238:265–272 10. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD et al (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284:143–147 11. Muraglia A, Cancedda R, Quarto R (2000) Clonal mesenchymal progenitors from human bone marrow differentiate in vitro according to a hierarchical model. J Cell Sci 113 (Pt 7):1161–1166 12. Banfi A, Muraglia A, Dozin B, Mastrogiacomo M, Cancedda R, Quarto R (2000) Proliferation kinetics and differentiation potential of ex vivo expanded human bone marrow stromal cells: implications for

their use in cell therapy. Exp Hematol 28:707–715 13. Gronthos S, Zannettino AC, Hay SJ, Shi S, Graves SE, Kortesidis A et al (2003) Molecular and cellular characterisation of highly purified stromal stem cells derived from human bone marrow. J Cell Sci 116:1827–1835 14. Sacchetti B, Funari A, Michienzi S, Di Cesare S, Piersanti S, Saggio I et al (2007) Self-renewing osteoprogenitors in bone marrow sinusoids can organize a hematopoietic microenvironment. Cell 131:324–336 15. Digirolamo CM, Stokes D, Colter D, Phinney DG, Class R, Prockop DJ (1999) Propagation and senescence of human marrow stromal cells in culture: a simple colony-forming assay identifies samples with the greatest potential to propagate and differentiate. Br J Haematol 107:275–281 16. Martin I, Muraglia A, Campanile G, Cancedda R, Quarto R (1997) Fibroblast growth factor-2 supports ex vivo expansion and maintenance of osteogenic precursors from human bone marrow. Endocrinology 138:4456–4462 17. Sivasubramaniyan K, Ilas DC, Harichandan A, Bos PK, Santos DL, de Zwart P et al (2018) Bone marrow-harvesting technique influences functional heterogeneity of mesenchymal stem/stromal cells and cartilage regeneration. Am J Sports Med 46:3521–3531 18. Narcisi R, Cleary MA, Brama PA, Hoogduijn MJ, Tuysuz N, ten Berge D et al (2015) Longterm expansion, enhanced chondrogenic potential, and suppression of endochondral ossification of adult human MSCs via WNT signaling modulation. Stem Cell Rep 4:459–472 19. Gruber HE, Somayaji S, Riley F, Hoelscher GL, Norton HJ, Ingram J et al (2012) Human adipose-derived mesenchymal stem cells: serial passaging, doubling time and cell senescence. Biotech Histochem 87:303–311 20. Banfi A, Bianchi G, Notaro R, Luzzatto L, Cancedda R, Quarto R (2002) Replicative aging and gene expression in long-term cultures of human bone marrow stromal cells. Tissue Eng 8:901–910

28

Roberto Narcisi et al.

21. Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP (2004) Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper—Excel-based tool using pair-wise correlations. Biotechnol Lett 26:509–515 22. Chen HH, Decot V, Ouyang JP, Stoltz JF, Bensoussan D, de Isla NG (2009) In vitro initial expansion of mesenchymal stem cells is influenced by the culture parameters used in the isolation process. Biomed Mater Eng 19:301–309 23. Willert K, Brown JD, Danenberg E, Duncan AW, Weissman IL, Reya T et al (2003) Wnt proteins are lipid-modified and can act as stem cell growth factors. Nature 423:448–452 24. Narcisi R, Arikan OH, Lehmann J, Ten Berge D, van Osch GJ (2016) Differential effects of small molecule WNT agonists on the multilineage differentiation capacity of human mesenchymal stem cells. Tissue Eng Part A 22:1264–1273 25. Mueller MB, Fischer M, Zellner J, Berner A, Dienstknecht T, Prantl L et al (2010)

Hypertrophy in mesenchymal stem cell chondrogenesis: effect of TGF-beta isoforms and chondrogenic conditioning. Cells Tissues Organs 192:158–166 26. Cals FL, Hellingman CA, Koevoet W, Baatenburg de Jong RJ, van Osch GJ (2012) Effects of transforming growth factor-beta subtypes on in vitro cartilage production and mineralization of human bone marrow stromal-derived mesenchymal stem cells. J Tissue Eng Regen Med 6:68–76 27. Barry F, Boynton RE, Liu B, Murphy JM (2001) Chondrogenic differentiation of mesenchymal stem cells from bone marrow: differentiation-dependent gene expression of matrix components. Exp Cell Res 268:189–200 28. Cleary MA, Narcisi R, Albiero A, Jenner F, de Kroon LMG, Koevoet W et al (2017) Dynamic regulation of TWIST1 expression during chondrogenic differentiation of human bone marrow-derived mesenchymal stem cells. Stem Cells Dev 26:751–761

Chapter 3 A Novel Enzymatic Digestion Approach for Isolation and Culture of Rodent Bone Marrow Mesenchymal Progenitors Leilei Zhong, Lutian Yao, and Ling Qin Abstract Bone marrow mesenchymal stem cells (MSCs) are promising therapeutic tools for tissue repair and treatment of a number of human diseases. As a result, there is substantial interest in characterizing and expanding these cells to uncover their therapeutic potential. Bone marrow mesenchymal progenitors, containing both MSCs and their proliferative progeny, are commonly isolated from the central region of rodent long bones. However, challenges exist in expanding these central mesenchymal progenitors in culture. We have designed an enzymatic digestion protocol to isolate mesenchymal progenitors within rodent long bones that resides close to the bone surface, which we termed endosteal mesenchymal progenitors. These cells are more metabolically active and more responsive to external stimuli compared to central mesenchymal progenitors. Therefore, they represent a biologically important target for MSC research. This chapter describes the approach in detail how to isolate and culture endosteal mesenchymal progenitors as well as their central counterparts from rodent long bones. Key words Mesenchymal stem cells, Endosteal mesenchymal progenitors, Bone marrow, Enzymatic digestion, Colony forming unit-fibroblast

1

Introduction Almost a half century ago, Alexander Friedenstein and colleagues pioneered a flushing method to isolate bone marrow cells from the central region of rodent long bones for culturing plastic-adherent and clonogenic fibroblastoid mesenchymal progenitors [1, 2]. Since then, the flushing method has become a standard technique to isolate mesenchymal progenitors from rodents in laboratories. Mesenchymal progenitor cultures are heterogeneous and consist of mesenchymal stem cells (MSCs) and their proliferative and more differentiated offspring [3]. In addition to their multi-lineage differentiation ability, these cells are immunosuppressive and capable of homing to injured tissues and secreting a number of bioactive molecules that promote wound repair and

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_3, © Springer Science+Business Media, LLC, part of Springer Nature 2021

29

30

Leilei Zhong et al.

tissue regeneration. Therefore, they have been intensively investigated as potential therapeutic tools for tissue repair and for treatment of a number of diseases including Crohn’s disease and graftversus-host disease [4]. However, these central mesenchymal progenitors are anatomically distant from trabecular and cortical bone surfaces where constant replenishment of bone forming osteoblasts by mesenchymal progenitors is required. There are also challenges associated with the rapid expansion of these cells in culture. For example, central mesenchymal progenitor cultures, especially those from mice, are difficult to grow in vitro and have limited proliferative ability [5–8]. The endosteal bone marrow is the portion of the bone marrow that is close to the bone surface, including cells that are within the trabecular bone and close to the endocortical bone surface [9]. We previously demonstrated that, in rodents, endosteal bone marrow cells contain a much higher frequency of mesenchymal progenitors than central bone marrow cells [8]. These endosteal mesenchymal progenitors have similar cell surface marker expression (Sca-1+CD105+CD29+CD73+ CD71+CD44+ CD45 CD34 ) and multi-lineage differentiation ability to central progenitors. However, they form much larger colony forming unit-fibroblast (CFU-F) colonies due to their higher proliferative ability and can be passaged more times in culture than their central counterparts. They also exhibit greater immunosuppressive activity both in vitro and in a mouse model of inflammatory bowel disease. Moreover, aging, a major contributing factor for osteoporosis, dramatically decreases their number, while injection of parathyroid hormone, an anabolic treatment for osteoporosis, strongly increases their number. Hence, these endosteal mesenchymal progenitors are more biologically relevant to skeletal homeostasis and disease than central mesenchymal progenitors [8]. Particularly, for in vitro culturing and expansion, endosteal mesenchymal progenitors are more suitable than central mesenchymal progenitors. This chapter describes an enzymatic digestion method to isolate endosteal mesenchymal progenitors from rat and mouse long bones, along with the concomitant isolation of their central counterparts from the same bones. Methods to quantify, culture, and differentiate these cells are also presented.

2

Materials

2.1

Animals

2.2

Instruments

This protocol was developed from experiments performed with Sprague-Dawley rats and C57Bl/6 mice (see Note 1). It is also compatible with other strains of mice we have tested, such as 129. 1. Class II biological safety cabinet/cell culture hood and a horizontal laminar flow clean bench: both should be equipped with a UV light for decontamination.

A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors

31

2. Tissue culture incubator with temperature and gas composition controls. 3. Mini shaker that can be placed inside a tissue culture incubator. 4. Inverted microscope with phase-contrast ability. 5. Benchtop centrifuge with a swing-bucket rotor. 6. Sterile surgical scissors, surgical forceps, scalpel handles, and scalpel blades (#22). 7. Sterile 0.2 μm syringe filter. 8. Sterile 70 μm cell strainer. 9. Sterile 10 mL syringes with 25- and 27-gauge needles. 10. Sterile 15 and 50 mL polypropylene conical tubes. 11. Pipet-aid, sterile serological pipettes (5 and 10 mL) and gibson-type micropipettes and tips (20, 200, and 1000 μL). 12. 25 cm2 tissue culture flasks with vented seal caps, 100 mm tissue culture dishes, 100 mm Petri dishes. 13. Hemocytometer. 2.3 Reagents and Media

1. 70% Ethanol. 2. Dulbecco’s phosphate-buffered saline (PBS). 3. Flushing medium: αMEM supplemented with 1% fetal bovine serum (FBS), 100 IU/mL penicillin, and 100 μg/mL streptomycin. 4. Protease solution: 2 mg/mL collagenase A (Roche Diagnostics, Indianapolis, IN) and 2.5 mg/mL trypsin dissolved in Dulbecco’s PBS and filter sterilized using a syringe filter. This solution should be freshly prepared just before starting the isolation process. 5. Growth medium for rat mesenchymal progenitors: αMEM supplemented with 15% FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin. 6. Growth medium for mouse mesenchymal progenitors: αMEM supplemented with 15% FBS, 0.1% β-mercaptoethanol, 20 mM glutamine, 100 IU/mL penicillin, and 100 μg/mL streptomycin. 7. Osteogenic medium: αMEM containing 10% FBS, 10 nM dexamethasone, 10 mM β-glycerophosphate, and 50 μg/mL L-ascorbic acid (AA, see Note 2), 100 IU/mL penicillin, and 100 μg/mL streptomycin. 8. Adipogenic medium: αMEM with 10% FBS, 0.5 mM isobutylmethylxanthine, 10 mM indomethacin, 1 μM dexamethasone, and 10 μg/mL insulin, 100 IU/mL penicillin, and 100 μg/mL streptomycin.

32

Leilei Zhong et al.

9. Chondrogenic medium: High glucose DMEM, 0.1 μM dexamethasone, 50 μg/mL AA, 40 μg/mL L-proline, 100 μg/mL sodium pyruvate, 1  ITS+, and 10 ng/mL TGFβ3, 100 IU/ mL penicillin, and 100 μg/mL streptomycin. 10. 3% Acetic acid with methylene blue. 11. 3% Crystal violet in methanol. 12. 0.05% Trypsin/EDTA solution. 13. 0.4% Trypan blue solution.

3

Methods

3.1 Harvest of Rodent Hind Long Bones

1. Euthanize the animal by CO2 inhalation. 2. Immediately transfer the dead animal to a clean bench pre-decontaminated by UV radiation. Place the animal on a flat surface on its back and wet the pelt thoroughly with 70% ethanol. 3. Using sterile forceps and scissors, incise and peel back the skin surrounding the hind long bones. Remove the bilateral hind long bones by cutting through the hip and ankle joints using sharp surgical scissors. Cut through the knee joint to separate the tibia and femur (see Note 3). 4. Place the long bones in a 100 mm Petri dish filled with 10 mL of flushing medium. Transfer the petri dish with the bones to a tissue culture hood for bone marrow harvesting.

3.2 Isolation of Central and Endosteal Bone marrow Cells

1. Under a sterile tissue culture hood, use forceps and a scalpel to remove all of the soft tissue surrounding the bones. After removal of the soft tissue, place the long bones into a new 100 mm Petri dish filled with 10 mL of flushing medium. 2. Cut off both ends of the tibia and femur at the growth plate with a scalpel. 3. Fill a syringe with 5 (mouse) or 10 (rat) mL of flushing medium per animal (2 tibiae and 2 femurs). Attach a 25- (rat) or 27-gauge needle (mouse) to the syringe. 4. With the prefilled syringe, drill a hole at each end of the bone and press down the plunger at one end to force medium through the bone. This will flush the bone marrow out of the bone through the opposite end of the bone. 5. Reverse the bone and repeat the flushing from the other end of the bone. Use 1 (mouse) or 2 (rat) mL of flushing medium to flush out each bone. Bone marrow cells released by flushing mainly come from the central part of the diaphyseal shaft, and hence are central bone marrow cells (Fig. 1 step 1, see Note 4). The bone marrow cells that are located in close proximity to the

A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors

33

Fig. 1 Representative images of a rat femur during the isolation of endosteal bone marrow. Step 1: flush out central bone marrow cells from a bone that is free of its surrounding soft tissue and with both ends removed at the growth plates; Step 2: predigest the whole bone to remove the periosteal progenitors and then longitudinally cut the bones into two halves; Step 3: gently wash the bones to remove loosely attached bone marrow; Step 4: digest bone fragments to collect endosteal bone marrow cells. BM: bone marrow. Reprinted from Bone, 53(2), Siclari et al., Mesenchymal progenitors residing close to the bone surface are functionally distinct from those in the central bone marrow. 575–86, Copyright (2013), with permission from Elsevier

endosteum remain attached to the bone after the flushing and are only released by enzymatic digestion (For information about how to culture central bone marrow mesenchymal progenitors, see Note 5). 6. After removing the central bone marrow cells, scrape the outside surface of the bones a few times with a scalpel blade and then place the bones into a 15 mL tube containing 5 mL of protease solution (8 mouse bones or 4 rat bones per tube). 7. Place the tubes on a mini shaker in a tissue culture incubator and shake for 20 minutes at 37  C. This step removes periosteum and its associated periosteal progenitors from the long bones (Fig. 1 step 2, see Note 6). 8. After digestion, wash the bones with flushing medium twice, and then longitudinally cut the bones into two halves (Fig. 1 step 2, see Note 7). 9. Use a syringe to gently wash the inside of the bones with flushing medium to remove loosely attached bone marrow (Fig. 1 step 3, see Note 8). 10. Place the bone fragments into a 15 mL tube containing 5 mL of protease solution (8 mouse bones or 4 rat bones per tube) and perform the second digestion step for 60 minutes as described in Subheading 3.2, step 7 (Fig. 1 step 4). 11. Add 5 mL of growth medium to neutralize the protease solution, and then collect the supernatant. Wash bone fragments twice with growth medium to collect all remaining cells. 12. Pass the cells through a 70 μm cell strainer to remove debris (see Note 9).

34

Leilei Zhong et al.

13. Perform a cell count by diluting a cell aliquot 1:10 in 3% acetic acid with methylene blue to lyse the red blood cells. 14. Count the nucleated cells using a hemocytometer under a microscope. The expected cell recovery of endosteal bone marrow is 8–12  106 cells per mouse and 12–20  106 cells per rat. 3.3 CFU-F Assays of Endosteal Mesenchymal Progenitors

1. Seed 1  106 mononuclear endosteal bone marrow cells per 25 cm2 flask in the growth medium and incubate the culture at 37  C in 5% CO2 in a humidified tissue culture incubator (see Note 10). 2. Typically, after 5 (rat) or 7 (mouse) days of incubation, most colonies should contain more than 50 fibroblastic cells (Fig. 2a, b, see Note 11). At this point, remove medium and wash flasks twice with PBS. 3. Stain the colonies with 3% crystal violet in methanol for at least 1 hour at RT. 4. Rinse flasks thoroughly with tap water to remove unbound stain. 5. Air-dry the flasks completely. 6. Count the number of CFU-F colonies under an inverted microscope with a 4 objective. We recommend drawing lines on the bottom of the flask to divide the surface into 8 regions in order to facilitate counting. Only count colonies larger than 50 cells. The expected CFU-F frequency of endosteal bone marrow cells is about 80–150 CFU-Fs per 1  106 mononuclear cells from both mouse and rat (Fig. 2c, d). The size of CFU-F colonies is normally larger in endosteal bone marrow compared to central bone marrow (Fig. 2c, d, see Note 12).

3.4 Culture and Differentiation of Endosteal Mesenchymal Progenitors

1. To culture endosteal mesenchymal progenitors, seed 5  106 mononuclear endosteal bone marrow cells per 100 mm tissue culture dishes in the growth medium and incubate at 37  C in 5% CO2 in a humidified tissue culture incubator. 2. Change medium every 2–3 days. 3. When the cells reach 80–90% confluence or when individual CFU-F colonies have expanded so that they are in close proximity to each other (about 8–10 days after plating), cells should be passaged for expansion. Aspirate the medium and wash the cells with PBS. 4. Add 3 mL of 0.05% trypsin/EDTA to the cells and incubate for 2–3 min in the tissue culture incubator. Examine under the microscope to confirm that about 70–90% of the cells are detached from the plate. If not, return the plate to the incubator for another 2 min. Typically, endosteal mesenchymal progenitors require less digestion time compared to central cells.

A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors

35

Fig. 2 CFU-F assays of rodent endosteal bone marrow cells compared to their central counterparts. (a) Representative images of 25 cm2 flasks with mouse central and endosteal CFU-F colonies after staining. Note that the initial seeding densities are 3  106 and 1  106 cells per flask for central and endosteal bone marrow, respectively. (b) Representative images and morphologies of mouse CFU-F colonies at low (top) and high (bottom) magnification. (c, d) Quantification of CFU-F frequency and diameter of endosteal bone marrow cells from mouse (c) and rat (d). **: p < 0.01 vs. central

5. Neutralize the trypsin by adding 3 mL of growth medium and gently pipet up and down with a 10 mL serological pipette to obtain a single cell suspension. 6. Transfer the cells into a 15 mL tube and centrifuge at 300  g for 5 min at room temperature. 7. Resuspend the pellet in 3 mL of growth medium and count the number of cells. To count the number of live mesenchymal progenitors, dilute an aliquot of the cells in trypan blue solution to quantify live and dead cells using a hemocytometer and a light microscope. 8. Plate the cells at a density of 0.5  106/100 mm dish. The cells that grow up are passage 1 cells. 9. Change medium every 2–3 days. Normally cells reach 80–90% confluency within 5–6 days. Lift and split cells at a ratio of 1:5–1:3 for expansion (see Note 13). The cells can also be stored in liquid nitrogen from passage 2–3 for future use.

36

Leilei Zhong et al.

Fig. 3 Endosteal mesenchymal progenitors are capable of multi-lineage differentiation. (a) In vitro osteogenic differentiation as detected by von Kossa staining. (b) In vitro adipogenic differentiation as detected by oil red O staining. (c) In vitro chondrogenic pellet differentiation as detected by alcian blue staining

10. To differentiate into osteoblast or adipocyte lineages (Fig. 3a, b), endosteal mesenchymal progenitors are first expanded to confluence in the growth medium and then switched to osteogenic or adipogenic medium for 3 and 1 weeks, respectively. The differentiation media should be changed every 2–3 days. 11. To differentiate into chondrocyte pellet (Fig. 3c), endosteal mesenchymal progenitors are resuspended in chondrogenic medium at 1  106 cells/mL and seeded in a V-bottomed 96-well plate at 200 μL/well. Centrifuge the plate at 300  g for 5 min to form a cell pellet, which appears as a fuzzy cloud on the first day. Change media every 3 days for 3 weeks.

4

Notes 1. With this protocol, we have successfully isolated and cultured endosteal mesenchymal progenitors from 1- to 4-month-old Sprague-Dawley rats and 1- to 16-month-old C57Bl/6 mice. The number of endosteal mesenchymal progenitors decreases significantly with aging [8]. Therefore, young animals (1–2month-old) are the best source for obtaining these progenitors. In addition, the bones of young animals are much easier to cut while those from old animals tend to shatter and require more force during cutting (see Note 7). 2. Since AA is unstable and rapidly oxidizes in water, it should be freshly added to the medium from a frozen stock (50 mg/mL) just before a medium change.

A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors

37

3. Long bones should be harvested under sterile conditions. If a sterile bench is not available, long bones can be dissected in a tissue culture hood. Frequently dip the forceps and scissors in 70% ethanol to prevent contamination into the cultures. All procedures should be performed as quickly as possible to achieve a high yield of viable mesenchymal progenitors. 4. Be sure to hold the bone tightly with a pair of forceps during flushing to avoid dropping the bone. It is recommended to use a syringe with a screw-tip to prevent the needle from detaching from the syringe during flushing. 5. To culture central mesenchymal progenitors, flush the central bone marrow cells into a 50 mL conical tube. Gently pipet the cell suspension up and down to break up the clumps of bone marrow. Centrifuge the cells at 300  g for 5 min and resuspend the pellet in 5 mL of growth medium. Filter the cell suspension through a cell strainer and count the number of cells in the same way as described in Subheading 3.2, step 13. The expected yield of central bone marrow is 30–50  106 cells per mouse and 120–200  106 cells per rat. Seed 3  106 cells per 25 cm2 flask for CFU-F assays and 30–50  106 per 100 mm dish for expansion. CFU-F staining and counting are performed 7 (rat) and 10 (mouse) days later as described in Subheading 3.3. The expected frequency is about 20–50 CFU-Fs/1  106 mononuclear cells (Fig. 2c, d). Central mesenchymal progenitors are cultured in the same growth medium as endosteal mesenchymal progenitors but they grow much slower. Split central mesenchymal progenitors at 1:2 or 1:3 when passaging. 6. It is important to remove all soft tissue, especially the periosteum, surrounding the long bones to avoid the contamination of mesenchymal progenitors from undesired sources. Periosteum contains periosteal progenitors that have similar characteristics to bone marrow mesenchymal progenitors [10]. Several previous studies also used collagenase digestion of flushed, minced, or chopped bone fragments to increase the yield of bone marrow mesenchymal progenitors [11– 16]. However, they were likely to have contamination of periosteal progenitors because they did not remove the periosteum from the bone. We have demonstrated by histology that scraping and predigestion of the bones is sufficient to remove periosteal cells, and therefore prevent contamination of periosteal progenitors into the endosteal bone marrow [8]. 7. Bones from old animals easily shatter during cutting. To minimize the amount of shattering, while holding the bone tightly with a pair of forceps, use a scalpel blade to first mark a longitudinal line on the outside of the bone surface. Then, slowly and forcefully cut along this line.

38

Leilei Zhong et al.

8. You can choose to harvest endosteal bone marrow cells from diaphyseal and metaphyseal regions separately. To do so, use a blade to cut the longitudinally halved bones at the junctions between the metaphysis and diaphysis. Then, enzymatically digest the diaphyseal and metaphyseal bone fragments separately. We have found that the majority of endosteal mesenchymal progenitors reside in the metaphysis [8]. 9. Endosteal bone marrow cells have a tendency to clump. Always mix cell suspension well before counting and seeding. If you intend to use these cells directly for flow cytometry, pass the cells through a cell strainer immediately before flow analysis. 10. We prefer using 25 cm2 flasks over 6-well plates for CFU-F assays. Due to the concaved surface of the wells in a 6-well plate, we have found that the colonies tend to grow more at the center of the wells and become difficult to count individually. We have found the colonies to be more evenly distributed and easier to count in 25 cm2 flasks. 11. Do not change the medium of cells plated for a CFU-F assay. Overall, minimize the amount of disturbance to the flasks after plating. If possible, allow the cells to remain undisturbed in the tissue culture incubator till counting. This ensures the optimal accuracy of the CFU-F assay. 12. The adherence and proliferative ability of mesenchymal progenitors varies significantly depending on culture conditions. It is recommended to test different batches of FBS in the growth medium to select one that gives the greatest number of CFU-Fs and optimal colony morphology and to use this one batch through the entire project. 13. To maintain a healthy cell population, it is advisable to passage cells at 80–90% confluence and to avoid over-confluence. Endosteal mesenchymal progenitors grow much better and have a much shorter doubling time in culture than the commonly used central progenitors. While mouse central mesenchymal progenitors normally reach senescence and stop growth at 5–10 passages, we found that mouse endosteal mesenchymal progenitor cultures keep proliferating beyond 20 passages [8].

Acknowledgments This work was supported by NIH grants K01DK071988 and R01DK095803 (to LQ). This study was supported by the Penn Center for Musculoskeletal Disorders Histology Core (P30-AR069619).

A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors

39

References 1. Friedenstein AJ, Chailakhjan RK, Lalykina KS (1970) The development of fibroblast colonies in monolayer cultures of Guinea-pig bone marrow and spleen cells. Cell Tissue Kinet 3:393–403 2. Friedenstein AJ, Gorskaja JF, Kulagina NN (1976) Fibroblast precursors in normal and irradiated mouse hematopoietic organs. Exp Hematol 4:267–274 3. Lindner U, Kramer J, Rohwedel J et al (2010) Mesenchymal stem or stromal cells: toward a better understanding of their biology? Transfus Med Hemother 37:75–83 4. Wang S, Qu X, Zhao RC (2012) Clinical applications of mesenchymal stem cells. J Hematol Oncol 5:19.5 5. Short B, Brouard N, Occhiodoro-Scott T et al (2003) Mesenchymal stem cells. Arch Med Res 34:565–571 6. Krishnappa V, Boregowda SV, Phinney DG (2013) The peculiar biology of mouse mesenchymal stromal cells--oxygen is the key. Cytotherapy 15:536–541 7. Banfi A, Muraglia A, Dozin B et al (2000) Proliferation kinetics and differentiation potential of ex vivo expanded human bone marrow stromal cells: implications for their use in cell therapy. Exp Hematol 28:707–715 8. Siclari VA, Zhu J, Akiyama K et al (2013) Mesenchymal progenitors residing close to the bone surface are functionally distinct from those in the central bone marrow. Bone 53:575–586

9. Ellis SL, Grassinger J, Jones A et al (2011) The relationship between bone, hemopoietic stem cells, and vasculature. Blood 118:1516–1524 10. van Gastel N, Torrekens S, Roberts SJ et al (2012) Engineering vascularized bone: osteogenic and proangiogenic potential of murine periosteal cells. Stem Cells 30:2460–2471 11. Morikawa S, Mabuchi Y, Kubota Y et al (2009) Prospective identification, isolation, and systemic transplantation of multipotent mesenchymal stem cells in murine bone marrow. J Exp Med 206:2483–2496 12. Xu S, De Becker A, Van Camp B et al (2010) An improved harvest and in vitro expansion protocol for murine bone marrow-derived mesenchymal stem cells. J Biomed Biotechnol 2010:105940 13. Nakamura Y, Arai F, Iwasaki H et al (2010) Isolation and characterization of endosteal niche cell populations that regulate hematopoietic stem cells. Blood 116:1422–1432 14. Ohishi M, Ono W, Ono N et al (2012) A novel population of cells expressing both hematopoietic and mesenchymal markers is present in the normal adult bone marrow and is augmented in a murine model of marrow fibrosis. Am J Pathol 180:811–818 15. Short BJ, Brouard N, Simmons PJ (2009) Prospective isolation of mesenchymal stem cells from mouse compact bone. Methods Mol Biol 482:259–268 16. Zhu H, Guo ZK, Jiang XX et al (2010) A protocol for isolation and culture of mesenchymal stem cells from mouse compact bone. Nat Protoc 5:550–560

Chapter 4 Isolation of Nucleus Pulposus and Annulus Fibrosus Cells from the Intervertebral Disc Guus G. H. van den Akker, Andy Cremers, Donatus A. M. Surtel, Willem Voncken, and Tim J. M. Welting Abstract Cells isolated from the intervertebral disc are often used for in vitro experimentation. Correctly separating the intervertebral disc tissue in annulus fibrosus and nucleus pulposus is particularly challenging when working with surplus material from surgery or specimens from donors with an advanced age. Moreover, lineage controls are only sparsely reported to verify tissue of origin. Here we describe an approach to intervertebral disc cell isolation from human and bovine origin. Key words Intervertebral disc, Cell isolation, Nucleus pulposus, Annulus fibrosus, Collagenase, Cell characterization

1

Introduction The intervertebral disc (IVD), located between vertebra, is an avascular, non-innervated tissue of musculoskeletal origin that allows flexion and rotation of the spinal column. The IVD comprises a central cartilaginous nucleus pulposus (NP) surrounded by the fibrous annulus fibrosus (AF) with highly oriented collagen fibers. At the rostral and caudal sides, the IVD is enclosed by the end plates that connect to the vertebral body. End plate cartilage is thought to be similar to articular cartilage [1]. The end plates are important for fluid flow and nutrient supply from the vertebral body to the NP and AF. NP cells originate from the embryonic notochord in mice, and presumably this holds for other mammalians as well [2]. By contrast, the AF originates from the embryonic sclerotome [3]. NP cells constitute a chondrocyte phenotype that in part resembles articular chondrocytes [4]. Importantly, cells of the inner AF, directly bordering the NP, also acquire a chondrocyte phenotype similar to the NP [5]. The outer AF consists predominantly of fibroblastic cells.

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_4, © Springer Science+Business Media, LLC, part of Springer Nature 2021

41

42

Guus G. H. van den Akker et al.

IVD cells are isolated from tissues for cellular studies. Multiple types of (sequential) enzymatic digestions at different concentrations and durations have been reported. Digestion time, type of enzyme, and its concentration are known to affect cell viability [6]. In addition, the time post-mortem at which cells are isolated may affect cell viability and/or phenotype. For human post-mortem IVD tissue, maximum sampling times of 24–72 h have been reported [1, 7]. Surgical surplus material and animal tissues are typically obtained within hours of dissection or death of the animal. Surgical surplus material from scoliosis correction surgery often consists of smaller fragments; this demands careful selection, labeling, orientation, and dissection to obtain NP and AF cell populations. In case of surplus material from IVD herniation surgery it is recommended to verify tissue of origin by histological analyses (NP/AF). Histological scoring of IVD degeneration in human and certain animal samples can be done by various systems, e.g., Mankin [8], Thompson [9], or modified Mankin [10]. Based on our earlier work, we recommend a pre-isolation comparison and/or histological analyses of NP and AF tissue of the same donor to ensure correct isolation of the cell type of interest [11]. Following enzymatic digestion of the tissues, cells are strained, washed, and plated at a high cell density in monolayer cultures or brought into a 3D culture environment. A plethora of (pseudo) 3D culture options are available, i.a. pellets, micromasses, alginate, Matrigel, etc. This choice is mainly guided by the prevailing research question, e.g., to expand cells, limit chondrocyte dedifferentiation, or mimic the microenvironment [3, 12–14]. Importantly, different cell culture media were shown to affect marker gene expression of IVD cells [15]. DMEM or mixtures of DMEM with alpha-MEM or Ham’s F-12 are recommended for expansion of IVD cells. Finally, it is recommended to perform post-isolation measurements of IVD marker expression to confirm the successful isolation of NP or (outer) AF cell populations [11, 16, 17]. The ORS/Spine Research Interest group released a consensus definition of young healthy human NP cells [18]. The recommended markers are: stabilized expression of HIF-1, GLUT-1, aggrecan/collagen II ratio >20, Shh, Brachyury, KRT18/19, CA12, and CD24 to assess variation between NP cells. However, it may not be practical to perform all recommended measurements for each individual donor. Moreover, NP cells from aged or degenerated tissue are known to have different expression profiles [19]. Nevertheless, we advocate the systematic measurement of a limited marker subset of NP and AF phenotypic marker genes as standard practice to acquire a reference for and record on common characteristics of isolated types. This is in particular valuable when both NP and AF cells are isolated from one donor. These markers provide an internal reference for the donor and the experimenter. A subset of marker genes that we have successfully used in situ and

Isolation of Cells from the Intervertebral Disc

43

in vitro to assess correct isolation of NP and outer AF cells from bovine and human donors consists of Brachyury/T, Keratin 19 (KRT19), Secreted frizzled-related protein 2 (SFRP2), and Collagen type 12 alpha 1 (COL12A1) [11, 16, 17]. In our experience this constitutes a more sensitive and reliable method compared to Collagen type 2/Collagen type 1 mRNA ratio to characterize NP cells [11, 17]. The systematic measurement of these markers will aid in a more uniform description of IVD cell populations used in literature.

2

Materials All solution are prepared in a sterile cell culture environment. Prepared reagents are stored at 4
C unless indicated otherwise. Cell culture reagents are prewarmed at 37
C prior to use. The required waste disposal regulations should be followed.

2.1 Medical Ethical Approval for Human Tissues

2.2

Cell Culture

Prior to performing the procedures described below, it is obligatory to obtain approval from a local medical ethical committee for the use of tissues from deceased individuals or surgical leftover material for research purposes prior to performing the procedures described below. 1. Fetal calf serum (FCS): Stock is thawed overnight at 4
C. Decomplement FCS for 30 min at 56
C in a water bath. Fill out 25 mL FCS per sterile polypropylene tube. Store at 20
C in upright position. 2. Anti-Anti, Antibiotic Antimycotic (A/A), Gibco 15240-062: Stock (100) is thawed at 37
C in a water bath. Aliquot 5.5 mL per tube, store at 20
C in upright position. 3. Non-essential amino acids (NEAA), Gibco 11140-035: Stock (100) is thawed at 37
C in a water bath. Stock is stored at 4
C. 4. Sodium chloride (0.9%): Dissolve 9 g sodium chloride (suitable for cell culture, >99% purity, Mw ¼ 58.44 g/mol) in 1 L MQ-water. Sterilize the solution by autoclaving (121
C, 15 min). Store at room temperature. 5. DMEM/F12 (1:1) Glutamax, Gibco 31331-028. 6. DMEM/F12 (1:1), HEPES buffered, Gibco 31330-038. 7. Collagenase II, Invitrogen 17101-015: 10 concentrated stock ¼ 300 U/mL or 1.1%. Dissolve in HEPES-buffered DMEM/F12 with 1% A/A. Aliquot in 5 mL portions in 50 mL tubes and store at 20
C until use. Thaw and dilute to 1 (NP) or 2 (AF) in HEPES-buffered DMEM/F12 with 1% A/A.

44

Guus G. H. van den Akker et al.

8. IVD expansion medium: DMEM/F12, Glutamax (440 mL), 10% decomplemented FCS (50 mL), 1 A/A (5 mL of 100 stock), 1 NEAA (5 mL of 100 stock), Limited shelf life, store at 4
C. 9. 0.5% Trypsin-EDTA (TE), Gibco 15400-054: Stock (100) is thawed for 30 min at 37
C in a water bath. Dilute 10 with 0.9% sodium chloride and aliquot 5 mL per 50 mL tube (10 stock), dilute again 10 in Sodium Chloride (0.9%) to obtain 1 stock. Store at 20
C in upright position. Keep 1 stock at 4
C after thawing. 10. Cell strainer 70 μm, Greiner, 542070. 11. Freeze-down medium (10 mL, 2 concentrated stock): DMEM/F12, Glutamax (5 mL), FCS (3 mL ¼ 30%), DMSO (2 mL ¼ 20%). This solution can be aliquoted and stored at 20
C. Avoid repeated freeze/thaw cycles. 12. Mr. Frosty, Nalgene, VWR, 479-3200. Fill with isopropanol until the mark on the side and refrigerate to 4
C. 13. Isopropanol, p.a Merck 1.09634.1011. 14. DMSO, cell culture grade, Santa Cruz, SC-358801. 15. Cryovials (2 mL), Greiner, 126277. 2.3

Cell Counting

1. Bu¨rker Turk cell counter: Depth 0.1 mm, surface per square 0.04 mm2. 2. Trypan blue (0.4%), Sigma-Aldrich Chemie T8154. 3. Phosphate-buffered saline (PBS, 20 concentrated stock): 87.52 g Sodium chloride (suitable for cell culture, >99% purity, Mw ¼ 58.44 g/mol), 14.16 g disodium hydrogen phosphate dihydrate (>99.5% purity, Mw ¼ 177.99 g/mol), 2.15 g potassium dihydrogen phosphate (>99.5% purity, Mw ¼ 136.09 g/mol), fill up to 500 mL. pH is between the required range of 7.2–7.4, no further pH-correction is needed. Stock is diluted 20 in MQ water to obtain 1 stock. Sterilize by autoclaving (121
C, 15 min).

2.4 IVD Marker Gene Analyses by RT-qPCR

RNA was isolated according to standard protocols and cDNA synthesis was done using random hexameres. RT-qPCR reactions were set up in Takyon™ No Rox SYBR Master Mix dTTP blue (Eurogentec) in a CFX96 Real-Time PCR Detection System (Biorad) according to the MIQE guideline [20]. Primer pair sequences for human and bovine were described before and are provided in Tables 1 and 2 [11, 16, 17].

Isolation of Cells from the Intervertebral Disc

45

Table 1 RT-qPCR primer sequences for human NP (T & KRT19) and AF (SFRP2 & COL12A1) marker genes Human Gene

FW primer (50 -30 )

RV primer (50 -30 )

Brachyury/T

CCACCTGCAAATCCTCATCCT

TTGGAGAATTGTTCCGATGAGC

KRT19

GCAGTCACAGCTGAGCATGAA

TCCGTTTCTGCCAGTGTGTCT

SFRP2

TGTGCCACGGCATCGA

TCGTGGCCCAGCAGGTT

COL12A1

TGACAACCCTTTCCGACACA

CTCCTCACGGTTCTAAAATTTGC

ACTB

CCTGGCACCCAGCACAAT

GCCGATCCACACGGAGTACT

PPIB

CCTGCTTCCACCGGATCAT

CGTTGTGGCGCGTAAAGTC

Beta Actin (ACTB) and Peptidyl-Prolyl Isomerase B (PPIB) are recommended as stable reference genes

Table 2 RT-qPCR primer sequences for bovine NP (T & Krt19) and AF (Sfrp2 & Col12a1) marker genes Bovine Gene

FW primer (50 -30 )

RV primer (50 -30 )

bBrachyury/T

CACACGGCTGCGAAAGGTA

TGAACTGTCGGAATAGGTTGGA

bKrt19

GACCTGCGGGACCAGATTCTC

GTCAGCCTCCACACTCATGC

bSfrp2

CAGGACAACGACCTTTGCAT

TCACATACCTTTGGAGCTTCCT

bCol12a1

ACCGGCTACACTGTGACCTA

TCCAGGCGCATCTCTTTGG

bGapdh

CACCCACGGCAAGTTCAAC

TCTCGCTCCTGGAAGATGGT

bRps14

CATCACTGCCCTCCACATCA

TTCCAATCCGCCCAATCTTCA

Bovine glyceraldehyde-3-phosphate dehydrogenase (bGapdh) and bovine ribosomal protein S14 (bRps14) are recommended as stable reference genes

3

Methods Procedures are carried out at room temperature in a laminar flow cabinet unless indicated otherwise.

3.1 Tissue Dissection and Cell Isolation

1. Tissues are obtained as fresh as possible from human or animal donors (90%). 5. Add 5–10 mL cell culture medium to the trypsinized cells to neutralize the Trypsin-EDTA. 6. Transfer the solution to a sterile 50 mL tube and centrifuge for 5 min at 316 RCF at RT. 7. Remove the supernatant and resuspend the pellet in an appropriate amount of IVD expansion medium (4–8 mL is convenient for passaging 1:4). Carefully pipet up and down 10 after the last clump has disappeared to make a single-cell solution. 8. Count cells and replate for further expansion, an experiment or for freezing down. Write down the number of seeded cells at the start and end of each passage to calculate population doubling level (see Note 11). 3.5 Freezing Down IVD Cells

1. Write down the donor number, cell type, passage number, and date on the cryotubes. 2. Trypsinize and count the cells according to “Subheading 3.4.” 3. Centrifuge for 5 min at 316 RCF at room temperature (RT). 4. Remove supernatant and resuspend the IVD cells at a density of 2. 106 cells/mL. 5. Add cold 2 freeze down medium 1:1 drop-by-drop and mix at the same time. Note: be quick, DMSO at this concentration is toxic to the cells. 6. Aliquot 1 mL cell suspension (1. 106 cells) per cryotube and close the cap. Put cryotubes on ice immediately when freezing multiple donors. 7. Put the vials in a refrigerated Mr. Frosty and store overnight at 80
C to freeze the cells. Note: A maximum of 18 cryotubes fits in one Mr. Frosty. 8. The next day, transfer the cryotubes to a liquid nitrogen container for long-term storage.

3.6 Thawing IVD Cells from the Liquid Nitrogen

1. Transport tubes from the liquid nitrogen on ice to the cell culture laboratory. 2. Release any liquid nitrogen that might have gotten into the tube by slowly unscrewing the cap in the flow cabinet.

50

Guus G. H. van den Akker et al.

3. Rapidly thaw the cells in a 37
C water bath. Important: stop when a small clump of ice is still visible. 4. Transfer the cells to a 50 mL tube and add 9 mL RT IVD expansion medium dropwise and mix at the same time. 5. Centrifuge for 5 min at 316 RCF at RT. 6. Remove supernatant and plate the cells in a T75 culture flask with 12 mL IVD expansion medium. 7. The next day, refresh medium to remove DMSO that might have been released from the cells overnight. The far majority (>80%) should attach and survive the freeze/thawing procedure. Cells should be allowed to recover for a day or two prior to use in experiments. 3.7 RT-qPCR Analyses of NP and AF Marker Genes

4

In brief, RT-qPCR reactions are set up with 15 ng cDNA, 300 nM forward and 300 nM reverse primers. The program was: 50
C for 2 min, denaturation at 95
C for 10 min, 40 cycles of amplification (15 s 95
C and 1 min 60
C) and followed by a melting curve (95–50
C, 1
C/min). The standard curve method was used to calculate gene expression and this was normalized to stable reference genes indicated in Tables 1 and 2.

Notes 1. It can be challenging to distinguish the NP from AF in human aged/herniation surgery or bovine tail specimens. Cut transversally into the AF toward the NP using a scalpel blade (see Fig. 1a). It should become clear from the separation between the lamellar AF tissue when the AF transitions into the interzone and NP (see Fig. 1b). Store representative samples for histological confirmation and grading (standard formalin fixation, paraffin embedding, and histology, Fig. 1c). 2. For preadolescence tissue samples, the separation between NP and AF should be easily discernable by eye. Store representative samples for histological grading. 3. Do not leave tissue pieces for too long outside the sodium chloride container in which it was collected to avoid drying. 4. Mincing can be challenging for the AF. It is sometimes better to limit the dissection time and digest slightly larger AF tissue pieces. 5. Human outer AF tissue pieces tend to be tougher and it is advisable to use 60 U/mL (instead of 30 U/mL) collagenase (0.22%) solution and digestion usually takes 4–6 h. 6. For certain animal tissues, such as bovine tail AF tissue, it is recommended to perform an overnight (16 h) digestion.

Isolation of Cells from the Intervertebral Disc

51

7. The 70 μm cell strainer is thought to prevent isolation of notochordal cells that can have a diameter of up to 100 μm. 8. If a minority of small AF tissue pieces have not digested completely within the indicated time periods, proceed with cell isolation. 9. Pellet will be very unstable and can detach easily. 10. Not all cells will attach within this time frame, but the majority should attach if the initial cell isolate had a good viability (>90%). 11. The formula for PDL calculation ¼ 3.32(log (total viable cells at harvest/total viable cells at seed)) [21]. The PDL of each passage should be plotted in a cumulative figure.

Acknowledgements This work was supported by grants from ReumaNederland and Stichting de Weijerhorst. References 1. Power KA, Grad S, Rutges JPHJ, Creemers LB, van Rijen MHP, O’Gaora P, Wall JG, Alini M, Pandit A, Gallagher WM (2011) Identification of cell surface–specific markers to target human nucleus pulposus cells: expression of carbonic anhydrase XII varies with age and degeneration. Arthritis Rheum 63 (12):3876–3886. https://doi.org/10.1002/ art.30607 2. McCann MR, Se´guin CA (2016) Notochord cells in intervertebral disc development and degeneration. J Dev Biol 4(1):3. https://doi. org/10.3390/jdb4010003 3. Pattappa G, Li Z, Peroglio M, Wismer N, Alini M, Grad S (2012) Diversity of intervertebral disc cells: phenotype and function. J Anat 221(6):480–496. https://doi.org/10.1111/j. 1469-7580.2012.01521.x 4. Minogue BM, Richardson SM, Zeef LA, Freemont AJ, Hoyland JA (2010) Transcriptional profiling of bovine intervertebral disc cells: implications for identification of normal and degenerate human intervertebral disc cell phenotypes. Arthritis Res Ther 12(1):R22–R22. https://doi.org/10.1186/ar2929 5. Hayes AJ, Isaacs MD, Hughes C, Caterson B, Ralphs JR (2011) Collagen fibrillogenesis in the development of the annulus fibrosus of the intervertebral disc. Eur Cell Mater 22:226–241

6. Molinos M, Almeida CR, Goncalves RM, Barbosa MA (2015) Improvement of bovine nucleus pulposus cells isolation leads to identification of three phenotypically distinct cell subpopulations. Tissue Eng Part A 21 (15–16):2216–2227. https://doi.org/10. 1089/ten.tea.2014.0461 7. Thorpe AA, Binch ALA, Creemers LB, Sammon C, Le Maitre CL (2016) Nucleus pulposus phenotypic markers to determine stem cell differentiation: fact or fiction? Oncotarget 7(3):2189–2200. https://doi.org/10. 18632/oncotarget.6782 8. Mankin HJ, Dorfman H, Lippiello L, Zarins A (1971) Biochemical and metabolic abnormalities in articular cartilage from osteo-arthritic human hips. II Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 53(3):523–537 9. Thompson JP, Pearce RH, Schechter MT, Adams ME, Tsang IK, Bishop PB (1990) Preliminary evaluation of a scheme for grading the gross morphology of the human intervertebral disc. Spine (Phila Pa 1976) 15(5):411–415. https://doi.org/10.1097/00007632199005000-00012 10. Thomas CM, Fuller CJ, Whittles CE, Sharif M (2007) Chondrocyte death by apoptosis is associated with cartilage matrix degradation. Osteoarthr Cartil 15(1):27–34. https://doi. org/10.1016/j.joca.2006.06.012

52

Guus G. H. van den Akker et al.

11. van den Akker GGH, Koenders MI, van de Loo FAJ, van Lent PLEM, Blaney Davidson E, van der Kraan PM (2017) Transcriptional profiling distinguishes inner and outer annulus fibrosus from nucleus pulposus in the bovine intervertebral disc. Eur Spine J 26(8):2053–2062. https://doi. org/10.1007/s00586-017-5150-3 12. Tang X, Jing L, Richardson WJ, Isaacs RE, Fitch RD, Brown CR, Erickson MM, Setton LA, Chen J (2016) Identifying molecular phenotype of nucleus pulposus cells in human intervertebral disc with aging and degeneration. J Orthop Res 34(8):1316–1326. https://doi.org/10.1002/jor.23244 13. Huang Y-C, Hu Y, Li Z, Luk KDK (2018) Biomaterials for intervertebral disc regeneration: current status and looming challenges. J Tissue Eng Regen Med 12(11):2188–2202. https://doi.org/10.1002/term.2750 14. Johnson WEB, Stephan S, Roberts S (2008) The influence of serum, glucose and oxygen on intervertebral disc cell growth in vitro: implications for degenerative disc disease. Arthritis Res Ther 10(2):R46–R46. https://doi.org/10. 1186/ar2405 15. Schubert AK, Smink JJ, Pumberger M, Putzier M, Sittinger M, Ringe J (2018) Standardisation of basal medium for reproducible culture of human annulus fibrosus and nucleus pulposus cells. J Orthop Surg Res 13:14. https://doi.org/ 10.1186/s13018-018-0914-y 16. van den Akker GGH, Surtel DAM, Cremers A, Richardson SM, Hoyland JA, van Rhijn LW, Voncken JW, Welting TJM (2016) Novel immortal cell lines support cellular heterogeneity in the human annulus Fibrosus. PLoS One

11(1):e0144497–e0144497. https://doi.org/ 10.1371/journal.pone.0144497 17. van den Akker GGH, Surtel DAM, Cremers A, Rodrigues-Pinto R, Richardson SM, Hoyland JA, van Rhijn LW, Welting TJM, Voncken JW (2014) Novel immortal human cell lines reveal subpopulations in the nucleus pulposus. Arthritis Res Ther 16(3):R135. https://doi. org/10.1186/ar4597 18. Risbud MV, Schoepflin ZR, Mwale F, Kandel RA, Grad S, Iatridis JC, Sakai D, Hoyland JA (2015) Defining the phenotype of young healthy nucleus pulposus cells: recommendations of the spine research interest group at the 2014 annual ORS meeting. J Orthop Res 33(3):283–293. https://doi.org/10.1002/ jor.22789 19. Rutges J, Creemers LB, Dhert W, Milz S, Sakai D, Mochida J, Alini M, Grad S (2010) Variations in gene and protein expression in human nucleus pulposus in comparison with annulus fibrosus and cartilage cells: potential associations with aging and degeneration. Osteoarthr Cartil 18(3):416–423. https:// doi.org/10.1016/j.joca.2009.09.009 20. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wittwer CT (2009) The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55(4):611. https://doi.org/10.1373/ clinchem.2008.112797 21. Roth S (1974) Tissue Culture. Methods and Applications. Paul F. Kruse, Jr. , M. K. Patterson, Jr. Q Rev Biol 49(2):150–151. https:// doi.org/10.1086/408024

Chapter 5 Engineering Cartilage Tissue by Co-culturing of Chondrocytes and Mesenchymal Stromal Cells Yao Fu, Carlo A. Paggi, Amel Dudakovic, Andre J. van Wijnen, Janine N. Post, and Marcel Karperien Abstract Co-culture of chondrocytes and mesenchymal stromal cells (MSCs) has been shown to be beneficial in engineering cartilage tissue in vitro. In these co-cultures, MSCs increase the proliferation and matrix deposition of chondrocytes. The MSCs accomplish this beneficial effect by so-called trophic actions. Thus, large cartilage constructs can be made with a relatively small number of chondrocytes. In this chapter, we describe different methods for making co-cultures of MSCs and chondrocytes. We also provide detailed protocols for analyzing MSC-chondrocyte co-cultures with cell tracking, proliferation assays, speciesspecific polymerase chain reactions (PCR), rheological analysis, compression analysis, RNA-sequencing analysis, short tandem repeats analysis, and biochemical examination. Key words Chondrocytes, Mesenchymal stromal cells, Co-culture, Trophic effects, Cartilage engineering, Matrix deposition

1

Introduction Partial replacement of chondrocytes by alternative cell sources can reduce the number of chondrocytes needed to engineering cartilage constructs in vitro [1–3]. Hendriks et al., co-cultured bovine primary chondrocytes with human expanded chondrocytes, human dermal fibroblasts, mouse embryonic stem cells, mouse-3T3 feeder cells, or human mesenchymal stromal cells (MSCs) in cell pellets [4]. Their data indicated that cartilage matrix deposition increased in co-culture pellets. Replacement of 80% of the chondrocytes with other cell types resulted in similar amounts of GAG production when compared to pure chondrocyte pellets. This beneficial effect on cartilage formation is most prominent in co-cultures of chondrocytes with mesenchymal stromal cells [5]. In a more recent study, we used a xenogeneic co-culture model of human MSCs and bovine chondrocytes to study the contribution of each cell type to cartilage matrix formation [6, 7]. Our data showed a

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_5, © Springer Science+Business Media, LLC, part of Springer Nature 2021

53

54

Yao Fu et al.

significant decrease in MSCs in co-culture pellets over time, resulting in an almost homogeneous cartilage tissue predominantly derived from the initially seeded chondrocytes. Our data showed that the beneficial effect of co-culture is largely due to increased chondrocyte proliferation and matrix formation, while chondrogenic differentiation of the MSCs only marginally contributed to cartilage formation. We also demonstrated that these observations present in co-culture pellets of chondrocytes and MSCs are independent of donor variation and culture conditions [8]. Subsequent experiments indicated that increased secretion of fibroblast growth factor 1 (FGF1) in co-culture of MSCs and chondrocytes is responsible for increased chondrocyte proliferation in pellet co-cultures [9]. Thrombospondin-2 has also been reported to be secreted by MSCs to promote chondrogenic differentiation both in vitro and in vivo [10]. To facilitate the engineering of large cartilage constructs it is beneficial to combine the co-culture with a support matrix. Many different scaffolds have been described in literature. Of these, hydrogels are particularly promising for cartilage tissue engineering. We showed that the beneficial co-culture effect on cartilage matrix formation is preserved in a hydrogel matrix consisting of agarose or tyramine-conjugated hyaluronic acid and tyramineconjugated dextran [11]. The tyramine-conjugated polymers cross-link in situ by a biocompatible enzymatic cross-linking reaction. The method is compatible with incorporation of chondrocytes and mesenchymal stem cells which can be mixed with the polymer conjugates prior hydrogel gelation. The cross-linked hydrogel supports long-term cell survival and matrix deposition in vitro [11]. This matrix production can even be further accelerated by incorporation of cellular micro-aggregates of 50–250 cells that can be uniformly produced using a high-throughput microwell platform [12]. Our data showed that micro-aggregation of the expanded monoculture cells prior to incorporation in the hydrogel boosted the chondrogenic potential and cartilage matrix formation both in vitro and in vivo [12, 13]. Different co-culture systems were subsequently performed in the injectable hydrogels or as micro-aggregates. The beneficial effects of MSCs in co-cultures were also observed in these studies. These reports revealed that the beneficial effects in co-cultures were largely due to the stimulation of proliferation and the matrix formation of chondrocytes induced by a trophic effect of the MSCs [14]. There is a dynamic regulation and interplay of stem cellderived cytokines that influence tissue survival, repair, and regeneration, including the activation of resident and circulating stem cells [15]. These studies point to a dominant role of the MSCs in stimulating resident or co-implanted chondrocytes to initiate a regenerative response. These findings starting from observations in vitro have been confirmed in animal studies and recently as well as a clinical trial [16, 17].

Methods for Co-culturing of Chondrocytes and MSCs

55

Moreover, while chondrogenic differentiation of MSCs can be triggered by chondrocytes, it also induces death of the MSCs [6]. It was postulated that the underlying mechanism is likely related to one of the deliberate “suicide” programs present within cells [14]. The programmed suicide death has usually two main forms, apoptosis and autophagy [18–20]. These suicide programs are usually induced intrinsically or extrinsically by external stimuli such as mechanical stress, oxidative processes, and drug treatments. Further studies need to be performed to investigate the mechanism behind the trophic effects of MSCs in co-cultures.

2 2.1

Materials Cell Sources

1. Bovine primary chondrocytes (bPCs) are isolated from fullthickness cartilage knee biopsies of female calves that are approximately 6 months old. Cartilage is separated and digested to extract primary chondrocytes (see Subheading 3.1). 2. Human primary chondrocytes (hPCs) are obtained from fullthickness cartilage dissected from knee biopsies of a patient undergoing total knee replacement (see Subheading 3.2). 3. Human MSCs (hMSCs) are isolated from bone marrow aspirates of healthy donors (see Note 1).

2.2 Media, Solutions, Chemicals, and Kits

1. Chondrocyte proliferation medium contains DMEM supplemented with 10% FBS, 1
nonessential amino acids, 0.2 mM ascorbic acid 2-phosphate (AsAP), 0.4 mM proline, 100 U penicillin /mL, and 100 μg/mL streptomycin. 2. Chondrogenic differentiation medium contains DMEM supplemented with 40 μg/mL of proline, 50 μg/mL ITS-premix, 50 μg/mL of AsAP, 100 ug/mL of sodium pyruvate, 10 ng/ mL of transforming growth factor beta 3 (TGFβ3, R&D system), 10–7 M of dexamethasone, 500 ng/mL of bone morphogenetic protein 6 (BMP6, R&D system), 100 U penicillin/ mL, and 100 μg/mL streptomycin. 3. MSC proliferation medium contains α-MEM plus 10% fetal bovine serum, 1% L-glutamine, 0.2 mM ascorbic acid, 100 U/mL penicillin, 10 μg/mL streptomycin, and 1 ng/ mL basic fibroblast growth factor (bFGF, R&D system). 4. Proteinase K digestion buffer: 1 mg/mL proteinase K (Sigma) in Tris/EDTA buffer (pH 7.6), 18.5 μg/mL iodoacetamide, and 1 μg/mL pepstatin A. The proteinase K solution can be stored in aliquots at 20  C for several weeks. After one thaw, do not freeze again. Tris/EDTA buffer: Dissolve 6.055 g Tris and 0.372 g EDTA · 2 H2O in 1000 mL of H2O. Adjust pH to 7.6. 5. PBE buffer: 14.2 g/L Na2HPO4 and 3.72 g/L Na2EDTA, pH 6.5.

56

Yao Fu et al.

6. GAG stock solution: 50 mg/mL, 17.5 mg of cysteine-HCl was dissolved in 10 mL of PBE buffer; GAG stock solution can be aliquoted and stored in 20  C freezer. A GAG working solution is 250 times dilution of GAG stock solution in PBE buffer, which contains 200 μg/mL of GAG. 7. DMMB solution: add 9.5 mL of 0.1 M HCl solution to 90.5 mL of d2H2O plus 0.304 g of glycine and 0.237 g of NaCl; adjust to pH 3 before adding 1.6 mg of DMMB to the buffer. When stored in the dark at RT, the solution is stable for 3 months; filter to get rid of precipitates before use. 8. Tyramine-conjugated polymers: dextran/hyaluronic acidtyramine conjugates dissolved in appropriate medium (Hy2Care). To obtain stable hydrogel constructs the end concentration of the polymers should be between 5 and 10 wt/v %. 9. Organic fluorescent dye (CM-DiI), Click-iT® EdU Imaging Kit, and the CyQuant DNA Kit (Molecular Probes). 10. QIAamp DNA Mini Kit and RNeasy Mini Kit (Qiagen). 11. iScript cDNA Synthesis kit and iQ SYBR Green Supermix (Bio-Rad). 12. Direct-zolTM RNA kit (Zymo Research). 13. TruSeq RNA library preparation kit (Illumina). 14. PowerPlex 16 System (Promega). 15. Collagenase type II (Worthington). 16. Click-iT® EdU Imaging Kit (Invitrogen). 17. A round-bottom ultra-low attachment 96 wells plate (Corning). 18. Cryomatrix (Shandon). 19. DMMB (1, 9-Dimethyl-Methylene Blue, Sigma). 20. DMTMM (4-(4,6-Dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium Chloride, Fluorochem Ltd. UK). 21. Milli-Q water (Milli-Q Advantage A10 system with 0.22 μm Millipak®-40 Express filter). 22. Horseradish peroxidase (HRP, Sigma-Aldrich). 23. Hydrogen peroxide (H2O2, 30%, Sigma-Aldrich). 24. Ultrapure™ agarose (Invitrogen). 2.3

Equipment

1. BD pathway 435 confocal microscope (BD Biosciences). 2. ELISA reader (TECAN). 3. MyiQ2 Two-Color Real-Time PCR Detection System (Bio-Rad). 4. PTEE mold for hydrogel formation (see Fig. 1a, b).

Methods for Co-culturing of Chondrocytes and MSCs

57

Fig. 1 Schematic illustration of PTEE mold for hydrogel formation (a and b) and PDMS convex mold for microaggregation (c)

5. PDMS convex mold for micro-aggregation (see Fig. 1c). 6. Rheometer (MCR 302, Anton Paar). 7. Texture Analyser TA-HD plus (StableMicro Systems Ltd.).

3

Methods

3.1 Isolation of Human Articular Chondrocytes

1. Human cartilage tissue was obtained from total knee or hip joint replacement. 2. Cartilage tissue is cut from underlying bone and connective tissue with scalpels and chopped into pieces of approximately 2
2 mm. 3. Digest cartilage pieces for 20–22 h in collagenase type II (0.15%) in DMEM supplemented with penicillin (100 U/ mL) and streptomycin (100 mg/mL).

3.2 Isolation of Human Bone marrow Mesenchymal Stromal Cells

1. Collect bone marrow aspirates in sterile heparin tubes. 2. Pour aspirate into 50 mL Falcon tubes. 3. Remove red blood cells by incubating 100 μL aliquots of aspirate with 900 μL red blood cell lysing buffer for 5–10 min on ice or until transparent. 4. Count cell numbers with trypan blue staining. Plate cells at 50,000/cm2 in T75 in MSC proliferation medium plus 1% heparin.

3.3 Cell Tracking of Cell Populations in Pellet Co-cultures with Organic Fluorescent Dyes CM-DiI

1. Trypsinize bovine or human chondrocytes and resuspend in PBS at a concentration of 2
106 cells/mL. 2. Incubate the cells with the fluorescent dye CM-DiI (final concentration of 4 μM) at 37  C for 5 min followed by incubation at 4  C for 15 min. 3. Wash cells two times by suspending cells in PBS followed by collecting cells by centrifuging at 300
g for 3 min.

58

Yao Fu et al.

3.4 Co-culture of bPCs and hMSCs in Pellets

1. Trypsinize hMSCs and suspend in chondrocyte proliferation medium at a concentration of 1
106 cells/mL. Resuspend labeled bPCs or hPCs from Subheading 3.1 at the same concentration as hMSCs in chondrocyte proliferation medium. 2. Mix hMSCs with bPCs or hPCs at ratios of 80/20% and 50/50%. Seed a total of 200,000 cells in one well of a roundbottom ultra-low attachment 96 wells plate in chondrocyte proliferation medium. 3. Use monoculture of hMSCs only or bPCs only or hPCs only as controls. Cell numbers per well are the same as in co-culture pellets. 4. Make pellets by centrifugation of the plate at 500
g for 5 min. 5. Xenogeneic co-cultures (bPCs and hMSCs), including corresponding controls, are cultured in chondrocyte proliferation medium at all times. 6. For allogenic co-cultures (hPCs and hMSCs), including corresponding controls, medium is changed to chondrogenic differentiation medium (see Subheading 2.2) on the second day after seeding.

3.5 Co-culture of bPCs and hMSCs in Injectable Hydrogels

1. Dissolve the tyramine-conjugated polymers in sterile PBS and incubate the polymer solution with HRP overnight at 4  C. 2. Trypsinize hMSCs and bPCs and resuspend in chondrocyte proliferation medium. Mix hMSCs with bPCs at ratios of 80/20%. 3. Combine the cell mixture with polymers solution mentioned above at concentration of ten million cells / mL. Pure polymers groups without cells are also prepared as control groups. 4. Add the freshly prepared 0.3% hydrogen peroxide (H2O2) to the mixture and immediately transfer to the mold (see Subheading 2.3) using a 1 mL pipette after brief vortex (see Note 2). 5. After gelation, transfer the gels to six-well plates and incubate in chondrogenic differentiation medium at all times.

3.6 Co-culture of bPCs and hMSCs in Micro-aggregates

1. Agarose microwell chips containing ~5000 microwells with a diameter of 200 μm were fabricated by pouring a 3% UltrapureTM agarose solution on convex molds of PDMS (see Subheading 2.3). 2. After agarose solidification, separate the agarose chips from the PDMS molds and transfer into 12-well culture plates covering with PBS (see Note 3). 3. Trypsinize hMSCs and bPCs and resuspend in chondrocyte proliferation medium. Mix hMSCs with bPCs at ratios of 80/20%.

Methods for Co-culturing of Chondrocytes and MSCs

59

4. Seed the cell mixture onto the agarose chips resulting in aggregates consisting of 100 cells. Immediately after seeding, centrifuge the agarose chips at 300
g for 3 min to allow the cells to settle down in the microwells and then incubate at 37  C for 24 h to form micro-aggregates. 5. The micro-aggregates can be obtained by flushing the well plates with medium and centrifugation. 3.7 Examination of Cell Proliferation in Pellets by EdU Labeling and Staining

1. 2 or 3 days after making pellets, add EdU (5-ethynyl-20 -deoxyuridine, provided in Click-iT® EdU Imaging Kit) to the culture medium of pellets at a concentration of 10 μM. 2. Harvest samples for analysis, 24 h later by transferring pellets to Eppendorf tubes. 3. Wash cell pellets with PBS and fix with 10% formalin for 15 min. 4. Embed samples in cryomatrix and cut 10 μm sections with a cryotome. 5. Permeabilize sections and stain for EdU with Click-iT® EdU Imaging Kit according to the manufacturer’s protocol. In this kit, nuclei are counterstained with LP435 (Hoechst 33342, provided in Click-iT® EdU Imaging Kit).

3.8 Image Acquisition and Analysis by Fluorescent Microscopy

1. Make fluorescent images with a BD pathway 435 confocal microscope (see Note 4). 2. Capture three separate images for each pellet section, using BP536/40 (Alexa 488), BP593/40 (DiI), and LP435 (Hoechst 33342) and pseudo color green, red, and blue, respectively. 3. Open blue image of one pellet section with ImageJ software [21]. 4. Set threshold by clicking drop-down menu via Image – > Adjust –> Threshold (see Note 5). 5. Open particle analyzer via Analyze –> analyze particles. 6. Set area restrictions: 100-infinite; choose Display results, Exclude on edges, Include holes; click OK to count NUMBER of total cell (see Note 6). 7. Open red image of the same pellet section; set threshold as described above (see Note 7). 8. Open image calculator via Process- > Image calculator. 9. Select blue image in the box of Image 1; select red image in the box of Image 2; select “AND” in the box of Operation; then click OK to generate a new image named “result of blue.” 10. Run “Analyze particles” on new image “result of blue” with the same setting as above to count NUMBER of red cell.

60

Yao Fu et al.

11. Open green image of the same pellet; set threshold and area restriction (see step 6) to count NUMBER of green cell. 12. Run “Image calculator” by selecting green image in Image 1 box and red image in Image 2 box, with AND in Operation box to generate new image named “result of green.” 13. Run “Analyze particles” on new image “result of green” with same setting as above to count NUMBER of green plus red cell. 14. Input all NUMBERs into an Excel spreadsheet and perform the following calculations: Rate of EdU positive Chondrocyte¼ NUMBER of green plus red cell  NUMBER of red cell
100%; Rate of EdU positive MSCs¼(NUMBER of green cell – NUMBER of green plus red cell)(NUMBER of total cell – N UMBER of red cell)
100%; Labeling efficiency ¼ NUMBER of red cell NUMBER of total cell
100% (see Note 8). 3.9 Quantitative GAG and DNA Assay

1. Perform glycosaminoglycan (GAG) and DNA assay at the end of co-culture (i.e., 4 weeks). 2. Wash cell pellets (n ¼ 6) with PBS and freeze pellets overnight at 80  C. 3. Digest pellets in 500 μL of proteinase K digestion buffer (see Note 9) for more than 16 h at 56  C. 4. To prepare a standard curve, make dilution series of cysteineHCl, according to Table 1. 5. Add 5 μL of a 2.3 M NaCl solution and 25 μL of the samples or the standard in one well of a 96-well non-tissue culture treated plate. 6. Add 150 μL of the DMMB (1, 9-Dimethyl-Methylene Blue) solution (see Subheading 2.2) and read the absorbance at 520 nm on an ELISA reader. Figure 2 gives an example of a standard curve (see Subheading 2.2). 7. Determine cell number by quantification of total DNA using a CyQuant DNA Kit, according to the manufacturer’s instructions.

Table 1 Series dilution of GAG standards GAG amount

Blank

0.5 μg

1 μg

1.5 μg

2 μg

2.5 μg

GAG working solution (see Note 9)

0 μL

10 μL

20 μL

30 μL

40 μL

50 μL

PBE buffer (see Note 10)

100 μL

90 μL

80 μL

70 μL

60 μL

50 μL

Methods for Co-culturing of Chondrocytes and MSCs

61

Fig. 2 An example of standard curve for GAG quantification. The blank (25 μL PBE, 5 μL 2.3 M NaCl and 150 μL DMMB solution) has an O.D. value of 0.18  0.03

3.10 Cell Tracking with Species-Specific PCR

1. Perform species-specific PCR to determine the ratio of MSCs and chondrocytes in xenogeneic co-culture (hMSCs and bPCs) pellets at the end of culture (i.e., 4 weeks). 2. Isolate DNA samples of pellets with a QIAamp DNA Mini Kit according to the manufacturer’s protocols. 3. Extract RNA samples of pellets with an RNeasy Mini Kit (see Note 10). 4. Reverse-transcribe one microgram of total RNA into cDNA using the iScript cDNA Synthesis kit. 5. Perform species-specific quantitative PCR (qPCR) on genomic DNA or cDNA samples by using the iQ SYBR Green Supermix. 6. Carry out PCR Reactions on MyiQ2 Two-Color Real-Time PCR Detection System under the following conditions: cDNA is denatured for 5 min at 95  C, followed by 45 cycles consisting of 15 s at 95  C, 15 s at 60  C, and 30s at 72  C. 7. Generate a melting curve for each reaction to test primer dimer formation and nonspecific priming. 8. The primers for real-time PCR, either species specific or crossspecies specific, are listed in Tables 2 and 3. 9. For each gene, standard curves are obtained by serial dilutions of cDNA (see Note 11). Figure 3 gives an example of standard curve for qPCR.

62

Yao Fu et al.

Table 2 Forward (F) and Reverse (R) primers used for quantitative PCR on genomic DNA

Gene name

Primer sequence 0

0

Product size

Gene bank no.

Cross-species GAPDH

F: 5 GCATTGCCCTCAACGACCA 3 R: 50 CACCACCCTGTTGCTGTAGCC 30

179 OR 171a

NC_000012 & NC_007303

Human-specific GAPDH

F: 50 TTCCACCCATGGCAAATTCC 30 R: 50 TTGCCTCCCCAAAGCACATT 30

131

NC_000012

Bovine-specific GAPDH

F: 50 AGCCGCATCCCTGAGACAAG 30 R: 50 CAGAGACCCGCTAGCGCAAT 30

132

NC_007303

a

Product size of human genomic GAPDH is 179, of bovine genomic GAPDH is 171

Table 3 Forward (F) and Reverse (R) primers used for quantitative RT-PCR Product size

Gene name

Primer sequence

Gene bank no.

Cross-species β-actin

F: 50 GCGCAAGTACTCCGTGTGGA 30 R: 50 AAGCATTTGCGGTGGACGAT 30

123

NM_001101 & NM_173979

Cross-species GAPDH

F: 50 AGCTCACTGGCATGGCCTTC 30 R: 50 CGCCTGCTTCACCACCTTCT 30

116

NM_002046& NM_001034034

Human-specific GAPDH

F: 50 CGCTCTCTGCTCCTCCTGTT 30 R: 50 CCATGGTGTCTGAGCGATGT 30

82

Bovine-specific GAPDH

F: 50 GCCAT CACTG CCACC CAGAA 30 R: 50 GCGGCAGGTCAGATCCACAA 30

207

NM_001034034

Human-specific aggrecan

F: 50 TTCCCATCGTGCCTTTCCA 30 R: 50 AACCAACGATTGCACTGCTCTT 30

121

NM_013227

Bovine-specific aggrecan

F: 50 CCAAGCTCTGGGGAGGTGTC 30 R: 50 GAGGGCTGCCCACTGAAGTC 30

98

NM_173981

Human-specific collagen II

F: 50 GGCGGGGAGAAGACGCAGAG 30 R: 50 CGCAGCGAAACGGCAGGA 30

129

NM_001844

Bovine-specific collagen II

F: 50 AGGTCTGACTGGCCCCATTG 30 R: 50 CTCGAGCACCAGCAGTTCCA 30

101

NM_001001135

Human-specific collagen IX

F: 50 GGCAGAAATGGCCGAGACG 30 R: 50 CCCTTTGTTAAATGCTCGCTGA 30

150

NM_001851

Bovine-specific collagen IX

F: 50 GGACTCAACACGGGTCCACA 30 R: 50 ACAGGTCCAGCAGGGCTTTG 30

102

XM_601325

NM_002046

Methods for Co-culturing of Chondrocytes and MSCs Standard

63

Unknown

35

Threshold Cycle

30

25

20

15

10

-0,5

0

0,5

1

1,5

2

2,5

Log Starting Quantity,copy number SYBR

E=118, 1% R^2=0, 996 slope=-2, 953 y-int=22, 150

SYBR1 E=110, 3% R^2=0, 998 slope=-3, 097 y-int=27, 560 SYBR2 E=110, 4% R^2=0, 993 slope=-3, 095 y-int=30, 151

Fig. 3 An example of standard curve for qPCR. SYBR, SYBR1, and SYBR2 stand for three different primer sets

10. Use Bio-Rad iQ5 optical system software (version 2.0) to calculate copy numbers for each condition using the standard curve as reference. 11. Ratio of bovine or human cells in the xenogenic co-culture pellets are defined as the proportion of human or bovine GAPDH copy numbers as percentage of the total copy numbers of both human and bovine genes determined by speciesspecific PCR using genomic DNA as a template. 12. The relative mRNA expression level of bovine or human genes in xenogenic co-cultures is determined by normalizing the values using cross-species-specific GAPDH and β-actin primers. 3.11 Rheological Analysis

1. Perform Rheological measurements to determine the storage modulus of hydrogels incorporated with MSC-Chondrocytes co-cultures pellets at the end of culture (i.e., 4 weeks) (see Note 12). 2. Wash the hydrogels with PBS (see Note 13). 3. Start the RheoCompass™ software. Load the related parallel plate measuring system. 4. Adjust the measuring chamber temperature at 20  C. 5. Load one sample on the chamber and select the measuring protocol. This protocol set parameters with a 0.05 N normal force in the oscillatory mode with 0.5% strain and 1 Hz, which was in the LVE range according to measured frequency and strain sweeps. Figure 4a gives an example of a standard curve.

64

Yao Fu et al.

Fig. 4 Example figures of mechanical properties analysis. (a) Storage modulus curve for rheological analysis. (b) Stress–strain curve for compressive test

3.12 Test

Compression

1. Perform texture analysis to determine the compressive modulus of hydrogels prepared and equilibrated for the rheological testing. 2. Load the sample on the loadcell at room temperature. The sample undergo three compression cycles with a maximum strain of 50% using a compression speed of 0.05 mm/s. Figure 4b gives an example of a stress strain curve. 3. The initial compressive modulus was calculated from the stress– strain curves using a linear slope at a strain ranging from 0% to 2.5%. The high strain compressive modulus was calculated from the stress–strain curves using a linear slope at a strain ranging from 40% to 49.5%. The percentage of energy dissipated during a compression cycle was calculated by dividing the surface of the hysteresis loop by the surface under the compression trace.

Methods for Co-culturing of Chondrocytes and MSCs

3.13 RNASequencing Data Analysis

65

1. Perform RNA-sequencing analysis to determine the gene expression of sex-mismatched MSCs and human chondrocytes in pellet co-cultures (see Note 14). 2. Isolate total RNA samples from pellets using the Direct-zolTM RNA kit. Perform the RNA-sequencing procedure with the RNA samples. The values obtained are normalized and expressed as reads per kilobase pair per million mapped reads (RPKM). 3. To remove the values related to noise or highly influenced by noise, consider an RPKM average value above 0.3. For average value is considered the average obtained by combining the value of a gene for all the samples. RPKMaverage ¼

n 1X RPKMn > 0:3 n i¼1

where n is the number of samples considered in the analysis (see Note 15). 4. The sex-mismatch of the MSC and chondrocyte donors can be used to attribute gene expression to either one of the donors. To determine the RPKM level of the two donors in the co-culture, the ratio between the monoculture (male donor) pellets and the co-culture (male + female donors) is calculated. %RPKMmale ¼

RPKMmonoculture RPKMcoculture

%RPKMfemale ¼ 100  %RPKMmale The percentage obtained is the expected value of the gene expression of the male and female donor in the co-culture. Moreover, the same value can be used to determine the expected value of the co-culture with the assumption of cells not interacting with each other. RPKM Cocultureexpected ¼ %RPKMmale ∗RPKM donor1mono þ %RPKMfemale ∗RPKM donor2mono where RPKM donor1mono and RPKM donor2mono are the real value of the individual donors cultured in monoculture and the %RPKMmale/female is the % obtained in step 3.

66

Yao Fu et al.

5. To determine the fold change, compare the real values with the expected values: Fc ¼

RPKM Coculturereal RPKM Cocultureexpected

6. Values above 2 are considered highly upregulated and values below 0.5 downregulated (see Note 16). Fc < 0:5&Fc > 2 ðhigh sensitivity and specificityÞ 7. Add the upregulated gene names into STRING software to determine gene-interconnections and GO-pathways. Perform the same procedure for the downregulated (as shown in Fig. 5a). 8. Add the entire data set of raw RPKM data to MORPEUS software. The system gives as output the heatmap of all the genome and allow samples comparison (as shown in Fig. 5b). 9. Up-load the upregulated or downregulated genes in PAN THER. This software determines the function of the individual genes and creates clusters of genes involved in the same pathway (as shown in Fig. 5c).

Fig. 5 Example figures of RNA-sequencing data analysis. (a) Heatmap of collagen-related genes comparison in different conditions. (b) Inter-connecting cluster showing collagen-related genes

Methods for Co-culturing of Chondrocytes and MSCs

3.14 Short Tandem Repeats (STR) Analysis

67

1. Perform STR analysis to determine the ratio of MSCs and chondrocytes in allogeneic co-cultures (hMSCs and hPCs) pellets. 2. Extracts genomic DNA samples from pellets (n ¼ 6) with the QIAamp DNA Mini Kit. 3. Amplify the sixteen loci of the kit PowerPlex 16 System, type “sequence,” and analyze all loci according to manufacturer’s protocol. 4. Compare monocultures of hMSCs or hPCs to find informative alleles only present in either the hMSCs or the hPCs donor (see Note 17). 5. Make electropherograms of the informative loci. 6. In the resulting electropherogram, calculate the area under the peaks, which reflects the abundance of the alleles. 7. The sum of the area under the peak for the two donor-specific alleles represents a relative amount of DNA for this donor. 8. Calculate the relative DNA amount for both the hMSC and the hPC donor. 9. Calculate the ratio of hMSCs and hPCs in the pellet by dividing through the total amount of relative DNA present in the pellet.

4

Notes 1. We define the “primary” cells (bPCs, hPCs and hMSCs) in this manuscript as cells with low passage number (90%) and minimization of the procedure duration, particularly enzymatic digestion steps, are easily accessible basic proxies for the overall impact of the procedure on cells. We here focus solely on specimen preparation for scRNA-seq studies; however, equally important is the selection of scRNA-seq methodology and subsequent methods for data analysis. Discussion of considerations in selection of these methods in the context of skeletal biology have been recently published [8]. Each scRNA-seq methodology has different requirements and limitations in terms of numbers of input cells that must be considered when planning specimen preparation. In particular, methods such as CEL-Seq2 and SMART-seq will require that specimens undergo index sorting prior to analysis, aiming to place single cells in as many wells as possible in a microtiter plate. In contrast, most other methods utilize cells in suspension as the input, and for these methods FACS is optional, though it can have utility in helping to remove cell doublets, acellular debris from the specimen in addition to allowing the scRNA-seq study to focus on a specific population of interest through the use of cell surface or genetically encoded markers. Thus, optional methods for both index sorting and FACS are included here. Therefore, the following sections focuses on preparation of the bone samples for scRNA seq studies. However, this protocol has also been successfully used to sort different mesenchymal cell populations by flow cytometry to determine their cellular hierarchy and physiologic contribution in bone using various types of transplantation studies.

Isolation of Skeletal Cells for scRNA-seq

2

91

Materials The digestion buffer and DNase I solution must be prepared fresh each time prior to use. Tissue suspension media, FACS buffer, and wash buffer should be made regularly and can be stored up to 2 months at 4
C. The FACS staining and blocking buffers and primary and secondary antibody dilutions must all be prepared fresh each time prior to use. Any unused reagents should be discarded according to disposal regulations of the individual’s institution.

2.1

Reagents

1. Tissue digestion buffer: Collagenase P (Roche, cat. 11213857001) and Dispase II (Roche, cat. 04942078001) should always be stored at 4
C when not in use. To prepare the digestion buffer, weigh required amount of Collagenase P and Dispase II using a scale capable of detecting 0.1 mg. Dissolve both reagents in α-MEM media containing 2% fetal bovine serum (FBS) using 15 or 30 mL falcon tubes. It is important to prepare the digestion mix in media with 2% FBS, since presence of serum greatly improves the viability of mesenchymal cells including the rare cell types which otherwise can be lost in media without serum. Gently swirl the reagents in the falcon tube until dissolved. Do not dissolve reagents by vortexing. The final digestion buffer should contain Collagenase P at 1 mg/mL and Dispase II at 2 mg/mL. Store the digestion buffer on ice before adding it to the tissue sample. 2. DNase I solution: Add DNase I at 2 units/mL in α-MEM basal media without serum. Presence of serum will interfere with DNase activity. Mix gently and do not vortex. Keep the DNase I solution on ice before use. 3. Tissue suspension media: Prepare tissue suspension media in bulk using α-MEM basal media and 2% FBS and store it at 4
C. 4. FACS buffer: Prepare FACS buffer in bulk using distilled phosphate-buffered saline (DPBS), 2% FBS, 1 mM ethylene diamine tetra acetic acid (EDTA). Store at 4
C. 5. Blocking buffer: Prepare fresh blocking buffer by adding rat anti-mouse CD16/CD32 (clone 2.4G2, BD bioscience) antibody to FACS buffer at a dilution of 1:50. Keep on ice. 6. Primary antibody solution: Prepare fresh primary antibody solution by adding the fluorophore-conjugated antibodies at a dilution of 1:100 in brilliant stain buffer (BD bioscience). Keep on ice and protect it from light (see Note 1). 7. Secondary antibody solution: If an unconjugated primary antibody is used, a fluorophore-conjugated secondary antibody directed against the primary antibody needs to be used. Prepare

92

Shawon Debnath and Matthew B. Greenblatt

fresh secondary antibody solution at a dilution of 1:500 in brilliant stain buffer. Keep on ice and protect it from light (see Note 1). 8. DAPI solution: Prepare DAPI (40 , 6-diamidino-2-phenylindole) stock solution in PBS at a concentration of 1 μg/mL. Keep it cold and protect it from light.

3

Methods Carry out all procedures at room temperature until specified.

3.1 Tissue Sample Preparation

1. Euthanize the animals according to institutional guidelines. Make a cut on the back of the mouse with fine scissor and pull out the skin exposing the mouse femur and tibia. Carefully remove the skin exposing the mouse knee joint and hip joint/ pelvic girdle. Carefully pull the mouse femur from the pelvic girdle separating the femur head. It is easy to separate mouse femur head above 3 weeks of age. However, it may be difficult to do so in young mice up to 2 weeks of age. In that case, we first recommend to make a cut with fine scissors taking a portion of the pelvic joint and the femur head and then carefully separating out the femur head with fine forceps. After isolating the entire femur, strip away as much muscle as possible. Keep the mouse femur in tissue suspension media in a 12-well plate. Please refer to Fig. 1 (see Notes 2–5). 2. Pick up the femur with forceps and place it on a glass slab such as the one used in casting SDS-PAGE gels. Use two razor blades to chop the femur into small pieces until a finely minced tissue is generated. Thorough mincing of femur section is very critical to ensure good yield of mesenchymal cells. While mincing, make sure the chopped tissue is never dry, since it will affect cell viability. It should always be moist and

Fig. 1 Digestion of mouse femurs. Isolated 15-day-old mouse femur suspended in media in a 12-well plate (a). Mouse femurs placed on a glass slab to be chopped by razor blades (b). Texture of a thoroughly minced mouse femur sample (c). Slurry texture of an appropriately digested mouse femur that can be easily pulled up by pipetting (d)

Isolation of Skeletal Cells for scRNA-seq

93

Table 1 Suggested digestion conditions based on mouse age and specimen site Digestion buffer volume (mL)

DNasel solution volume (mL)

Falcon tube size (mL)

Temerature (
C)

Shaking speed Time (rpm) (min)

Mouse femur (7-day-old)

1

1

15

37

200

15–20

Mouse femur (15-day-old)

1

1

15

37

200

30–40

Mouse femur (1-month-old)

1.5

1.5

15

37

200

50–60

Mouse femur (2-month-old)

2

2

30

37

225

60

Mouse femur (6-month-old)

3

3

30

37

225

60

Bone organoid in mammary fat pad

1

1

15

37

200

15

Source

surrounded by tiny amount of suspension medium. Please refer to Fig. 1 (see Notes 6 and 7). 3. Transfer the minced tissue homogenate into a 15 mL falcon tube. Add 1 mL of tissue digestion buffer to the tube. We recommend to add 1 mL of digestion buffer to digest two femurs within 2 weeks of age. Quickly spin the tube for a minute at 524  g to pull down any minced tissue that is attached to the surface. Incubate the tube at 37
C, gently shaking at 200 rpm in a shaker. The digestion process will approximately take 30–40 min for a 2-week-old mouse femur. Resuspend the tissue digest several times with 1 mL pipet to ensure that the homogenate is slurry. At this point, the tissue digestion step is complete (see Notes 8, 9 and Table 1). 4. Add 10 mL of tissue suspension media to the falcon tube. Centrifuge the tube at 524  g for 10 min at 4
C. 5. Discard the media by gently turning the falcon tube upside down without disturbing the pellet. We recommend not to use vacuum to remove excess liquid, since it can accidentally dislodge the pellet. The media can also be discarded using manual pipetting method. 6. Resuspend the tissue pellet thoroughly in 1 mL of DNase I solution. Incubate the tube at 37
C, gently shaking at 200 rpm inside the shaker for 10 min. Please refer to Table 1.

94

Shawon Debnath and Matthew B. Greenblatt

7. Add 8–9 mL of tissue suspension media and thoroughly suspend the tissue homogenate by vigorously pipetting the sample for several times. This step is highly critical. Thorough suspension will cause physical dissociation of tissue, and thereby minimize cell loss eventually leading to better single-cell suspension. 8. Once the tissue is thoroughly suspended, filter it through a 70 μm nylon mesh into a 30 mL falcon tube. Make sure, you filter the tissue suspension using a 5 mL or 10 mL pipet. During the filtration, the pipetting force should be strong enough to push the tissue filtrate through the nylon mesh without causing any splash. This will cause physical dissociation of tissue giving rise to single-cell suspension. It will need some practice to obtain the suitable pipetting force. To increase cell recovery, collect the remaining filtrate attached to the other side of the filter using a 1 mL pipet. 9. Transfer the filtered cell suspension from 30 mL falcon tube to a 15 mL falcon tube. We have observed that using 15 mL falcon tube for the later stages of sample preparation for FACS staining minimizes cell loss during the washing steps. 10. Centrifuge the tubes at 524  g for 10 min at 4
C. Discard the liquid by gently turning the falcon tube upside down without disturbing the pellet. 11. Gently add 5–10 mL of ice-cold FACS buffer to the pellet. Do not suspend the cell pellet. This will minimize cell loss. 12. Centrifuge the tubes at 524  g for 5 min at 4
C. Discard the liquid as much as possible without disturbing the pellet. Keep the tube on ice. The samples are now ready for FACS staining. 3.2

FACS Staining

Skeletal mesenchymal cells can be subjected to single-cell sequencing with or without sorting. However, sorting provides efficient elimination of contaminating cell population such as hematopoietic and endothelial cells and additionally facilitates focus on specific mesenchymal populations if desired. The following steps need to be carried out on ice (see Note 10). 1. Add 200 μL of freshly prepared blocking buffer to the samples. Suspend the pellet thoroughly by using 1 mL pipet. Do not vortex. Incubate the tubes on ice for 30 min (see Note 11). 2. Add 200 μL of freshly prepared primary antibody solution to the tubes. Suspend the pellet thoroughly by using 1 mL pipet. Do not vortex. Incubate the tubes on ice for 45 min to 1 h. Protect the tubes from light. After adding equal volumes of blocking buffer and primary antibody solution, the final working solution of the blocking buffer is 1:100 and primary antibody solution is 1:200 (see Note 12).

Isolation of Skeletal Cells for scRNA-seq

95

Fig. 2 Schematic representation of isolation of mesenchymal cells by FACS. DAPI-negative live cells are analyzed for cell size and singlets are isolated from the doublets. Mesenchymal cells (Ter119, CD45, CD31) can be sorted by exclusion of mature red blood cells (Ter119+), hematopoietic cells (CD45+), and endothelial cells (CD31+)

3. For washing, add 10 mL of ice-cold FACS buffer to the tubes. Centrifuge the tubes at 524  g for 10 min at 4
C. 4. Gently decant the supernatant without resuspending or otherwise disturbing the cell pellet. This will minimize cell loss. 5. Repeat steps 3 and 4 twice. The centrifugation time can be cut to 5 min during the second and third wash. 6. Add 200 μL of freshly prepared secondary antibody solution to the tubes. Thoroughly suspend the pellet by pipetting. Do not vortex. Incubate the tubes on ice for 20–30 min. Protect the tubes from light. 7. Wash the samples twice following steps 3 and 4. Do not resuspend the cell pellet. 8. Add 500 μL of ice-cold FACS buffer to the tubes. Suspend the pellet thoroughly and transfer it to appropriate FACS tubes. Filter the cell suspension through a 40 μm nylon mesh during the transfer. This will remove any undesirable cell aggregates that may clog the nozzle of the sorter. 9. Add DAPI to the cell suspension at a dilution of 1:100 and gently mix by tapping the tubes. Do not vortex. The samples are ready for sorting (see Note 13). 10. A template strategy for sorting mesenchymal cells is shown in Fig. 2. Generally, DAPI negative live cells are chosen. Forward versus side scatter gating (FSC-A versus SSC-A) is used to choose the cell population of interest based on size and cytoplasmic complexity. FSC-H versus FSC-W and SSC-H versus SSC-W gates allow exclusion of doublets. Mesenchymal cells can be sorted by eliminating Ter119+ erythroid cells, CD45+ hematopoietic cells, and CD31+ endothelial cells. After gating on mesenchymal cells, additional cell surface or genetic markers can then be applied to specify subpopulations of mesenchymal cells if desired.

96

Shawon Debnath and Matthew B. Greenblatt

3.3 Sample Collection for Single-Cell Sequencing

During FACS or after bone tissue digestion without further FACS isolation, samples must be collected or suspended using a method that both preserves cell viability and is compatible with the downstream single-cell sequencing methodology. The approach will vary depending on whether a droplet based (e.g., Drop-seq, 10 Genomics) or index sorting based (e.g., SMART-seq, MARS-seq, CEL-seq) methodology is subsequently employed [8]. 1. For droplet based assays using Drop Seq or 10 Genomics platforms, mesenchymal cells are sorted in bulk to conduct single-cell sequencing. Sorted mesenchymal cells should be collected in collection media using 1.5 mL Eppendorf tubes. Collection media should be free of serum, and basal media containing 0.5% bovine serum albumin is suitable for many applications. Sorting cells into PBS should be avoided, since it will cause significant loss of sensitive cell types such as mesenchymal stem cells. We recommended to follow the guidelines for the specific sequencing platform utilized. 2. Index-based techniques (such as CEL-seq, Smart-seq) require sorting single cells into individual wells of 96-well or 384-well plates. In many cases, the collection plate will be provided by the sequencing facility and are pre-loaded with unique barcodes, primers, and lysis buffer. Following sorting, the plates are sealed and snap frozen on dry ice before submitting the samples for sequencing. Validation experiments must be conducted to determine that sorting produces as few empty wells as possible while avoiding doublets. This validation can take the form of sorting a test cell line with direct microscopic visualization of cell capture after colonies are grown from the sorted cells.

4

Notes 1. We recommend that the primary antibody solution should be made at a 1:100 dilution. However, the appropriate dilution may need to be adjusted depending on the antibody concentration and the cell number. This also holds true for secondary antibodies. 2. The procedure described here is suitable for generation of single-cell suspensions suitable for FACS or single-cell sequencing for a broad range of skeletal cells, including LEPR + stromal cells/Cxcl12-abudnant reticular (CAR) cells, osteocalcin+ osteoblasts, articular and other growth plate resident chondrocytes. Additionally, several endothelial populations are also present and small numbers of CTSK+ osteoclast-lineage cells can be isolated. However, this protocol is notably not

Isolation of Skeletal Cells for scRNA-seq

97

suitable for isolation of osteocytes. It is possible that other populations may not be well represented or may require additional optimization for efficient isolation. 3. The endpoint of this procedure will be a single-cell suspension containing large amounts of hematopoietic cells alongside skeletal mesenchymal cells and small amounts of populations, including endothelial cells. For many applications, removal of hematopoietic cells is desirable, and this can be achieved through CD45-based gating strategies when FACS is employed after isolation. Early hematopoietic progenitors, particularly of the erythroid lineage, may express weak CD45, so inclusion of an additional negative selection marker such as glycophorin A is suggested for efficient depletion. In cases where it is not feasible or undesirable to include a FACS step after isolation, magnetic bead-based depletion offers an alternative approach. In some cases, pre-depletion of hematopoietic cells may be employed prior to FACS to increase the frequency of cell types of interest and thereby decrease the instrument time required. However, magnetic bead based depletion strategies must be carefully validated for nonspecific depletion of populations of interest sticking to the depletion column. For this reason, we typically avoid magnetic bead depletion. Alternatively, marrow flushing or centrifugation to remove marrow cells can be employed, but similarly runs the risk of depleting mesenchymal populations of interest or introducing an additional source of variation in population frequencies. A strength of single-cell sequencing studies is that they are relatively robust in the face of contamination with unwanted cell types, so purification need only be robust enough that the populations of interest are adequately represented and to avoid incurring unnecessary expense in sequencing irrelevant cell types. 4. When dealing with vertebral bone, it is important to thoroughly separate any neural tissue from the specimen, as neural support cells can be challenging to distinguish from mesenchymal cells in some forms of analysis due to their sharing many markers with mesenchymal cells. For instance, both Schwann cells and subsets of skeletal mesenchymal cells express CD200 [9]. 5. Microdissection or physical fractionation of the specimen prior to cell isolation can offer a simple but powerful method to later annotate populations identified based on anatomic location. 6. For extremely substantial specimens, such as large samples of human or large animal bone, we have found success by using a chisel to perform the initial physical disassociation of the specimen before progressing to chopping specimens with razor blades. Use eye protection with a chisel or other forceful

98

Shawon Debnath and Matthew B. Greenblatt

method of physical disassociation. Avoid specimen drying during this step by frequent irrigation with medium. 7. The two most important parameters in terms of optimizing cell yield and viability are the physical dissociation of tissue prior to digestion and the conditions of the enzymatic digestion itself. Thus, it is critical that physical dissociation be extremely thorough, ideally almost the point that the specimen takes on the appearance of coarse sand, before proceeding to digestion. For the physical disassociation, we most commonly use chopping with a razor blade performed on a glass plate, such as used for casting SDS-PAGE gels. Mincing with very fine surgical scissors while the specimen is in medium in a 6-well plate is an acceptable alternative. We generally find that mouse femurs until 1 month of age can be thoroughly minced using razor blades. However, efficient mincing of femurs from older mice using razor blades can be difficult to achieve. In that case, the mouse femurs can be crushed using a mortar and pestle. Make sure that the crushed tissue sample is always kept moist. Do not add digestion buffer to the samples while mincing or crushing. Addition of digestion buffer after the tissue is physically dissociated will improve cell recovery. We emphasize that mincing tissue samples is greatly preferred over crushing wherever feasible, as mincing increases cell yield and viability, ultimately helping to preserve rare cell types such as endothelial cells and mesenchymal stem cells. 8. To increase cell recovery, certain parameters may need optimization. This includes (1) the volume of digestion buffer and DNase I solution, (2) the diameter of the falcon tubes, and (3) the conditions and duration of tissue digestion. It is critical to adjust the volume of digestion buffer and DNase I solution according to mouse age. We recommend adding 1 mL of digestion buffer and, subsequently, 1 mL of DNase I solution per mouse femur sample between 0 and 15 days of age. The volume of digestion buffer and DNase I solution needs to be increased for samples from older mice. Please refer to Table 1 for additional information. To ensure optimum digestion, we do not recommend combining samples from multiple mice. Each mouse sample should be kept separate and individually subjected to digestion. Further, the diameter of the falcon tube is important to optimize the tissue digestion process. For example, to digest 8-week-old mouse femurs, 30 mL falcon tubes should be used. Increase in surface area will ensure better tissue digestion, resulting in a better yield of mesenchymal cells. Please refer to Table 1. 9. Additionally, it is imperative to optimize the tissue digestion process. Make sure that the tissue homogenate is not over- or under-digested. We recommend to digest the tissue at 37
C in

Isolation of Skeletal Cells for scRNA-seq

99

an enclosed shaker shaking at 200 rpm. Digesting tissue samples at room temperature instead of using a temperaturestabilized system will lead to poor cell recovery. Shaking the tubes at a modest speed of 200 rpm will increase the enzymatic digestion process by maximizing the contact between minced tissue and the digestion buffer. Speeds should not exceed 225 rpm, as this can lead to enzymatic degradation or negatively impact cell viability. Duration of tissue digestion is another important parameter to consider. Generally, the overall digestion process can range between 30 and 60 min. Digestion of 2-week-old mouse femurs takes about 30 min, whereas 2-month-old mouse femurs will take an hour to digest. However, there can be exceptions. For example, digestion of a bone ossicle/organoid grown in the mouse mammary fat pad only needs 15 min, and 7-day-old mouse femurs can be digested in 15–20 min. We emphasize that there is no fixed time of tissue digestion, and therefore this duration should be optimized for every type of tissue preparation. We recommend monitoring the digestion process at regular intervals, looking to see if the lumps of tissue have started to dissolve in the digestion buffer. As tissue starts to dissolve, the color of the tissue changes from reddish brown to a more whitish hue. The final step of tissue digestion can be confirmed by gently pipetting the homogenate, which can now be easily pipetted up and down and will resuspend in a more homogenous slurry texture. For convenience, please refer to Fig. 1. Also, please refer to the guidelines in Table 1 that shows several examples of optimized tissue digestion conditions. 10. It is challenging to successfully generate a FACS panel containing conjugated antibodies for more than 6 colors. Optimization can be time consuming and requires thorough understanding of the excitation and emission spectra of each fluorophore and the FACS instrument. First, one will need to choose appropriate cell surface markers to identify the population of interest. While designing the FACS panel, the fluorophore conjugates used with each antibody require optimization. The conjugation chemistry varies with different dyes against a specific antibody and vice versa. For example, the BV421 conjugated mouse CD200 antibody do not display the same spectral plot when compared with APC R700 conjugated mouse CD200 antibody in the presence of other colors. Therefore, it is easy to miss an important cell type that is rarely present in the cell population. We therefore, recommend to initially generate single color stains for each antibody using the brightest color wherever possible to confirm the presence or absence of the cells. After initial analysis, the conjugated antibody should be tested in the presence of other conjugated antibodies to check for any discrepancies versus the single

100

Shawon Debnath and Matthew B. Greenblatt

stained sample. Antibodies should also be validated by using proper isotype controls. To set up initial compensation beads should be used. Fluorescence minus one (FMO) controls should be used to set up for additional compensation and to assess background stain for each color. For each antibody, gates should generally be drawn as determined by FMO controls to separate positive and negative populations. 11. Based on size, the cell pellet can be suspended in blocking buffer and, subsequently, primary antibody solution in a volume ranging from 200 μL to 500 μL. For a medium-sized cell pellet (5  106 cells), we recommend a volume of 200 μL. If time is not a constraint, we encourage initially incubating the cells with blocking buffer prior to addition of the primary antibody solution. However, blocking buffer and primary antibody solution can be added simultaneously to speed up the procedure when required. 12. Primary and secondary antibody solutions should be freshly prepared at the time of use. The emission spectra of the fluorophores can change when kept in dilution buffer for more than 10 h. Tandem fluorophores such as PE-Cy7 and BUV737 are the ones most impacted by excessive time in the dilution buffer. All fluorophore conjugated antibodies must be initially diluted in brilliant stain buffer before adding to cells. Conjugated antibodies should never be diluted in FACS buffer. 13. The nozzle size plays an important role to determine the efficiency of sorting and the viability of sorted cells. For example, freshly isolated mesenchymal cells should be sorted with 70 μm nozzle, whereas when cultured, cells may enlarge and need to be sorted with a 100 μm nozzle. Validation experiments must be conducted to optimize sorting conditions. This includes checking cells for viability and function after sorting. References 1. Chan CKF, Gulati GS, Sinha R et al (2018) Identification of the human skeletal stem cell. Cell 175:43–56.e21 2. Mizuhashi K, Nagata M, Matsushita Y et al (2019) Growth plate borderline chondrocytes behave as transient mesenchymal precursor cells. J Bone Miner Res 34(8):1387–1392 3. Chan CKF, Seo EY, Chen JY et al (2015) Identification and specification of the mouse skeletal stem cell. Cell 160:285–298 4. Tikhonova AN, Dolgalev I, Hu H et al (2019) The bone marrow microenvironment at singlecell resolution. Nature 569:222–228 5. Baryawno N, Przybylski D, Kowalczyk MS et al (2019) A cellular taxonomy of the bone marrow stroma in homeostasis and leukemia. Cell 177 (7):1915–1932.e16

6. Debnath S, Yallowitz AR, McCormick J et al (2018) Discovery of a periosteal stem cell mediating intramembranous bone formation. Nature 562:133–139 7. Machado L, Esteves de Lima J, Fabre O, et al (2017) In situ fixation redefines quiescence and early activation of skeletal muscle stem cells. Cell Rep 21(7):1982–1993 8. Greenblatt MB, Ono N, Ayturk UM et al (2019) The unmixing problem: a guide to applying single-cell rna sequencing to bone. J Bone Miner Res 34:1207–1219 9. Chang C-Y, Lee Y-H, Jiang-Shieh Y-F et al (2011) Novel distribution of cluster of differentiation 200 adhesion molecule in glial cells of the peripheral nervous system of rats and its modulation after nerve injury. Neuroscience 183:32–46

Chapter 8 Mapping 5-Hydroxymethylcytosine (5hmC) Modifications in Skeletal Tissues Using High-Throughput Sequencing Fiorella Carla Grandi and Nidhi Bhutani Abstract Cytosine modifications can alter the epigenetic landscape of a cell, affecting the binding of transcription factors, chromatin organizing complexes, and ultimately affecting gene expression and cell fate. 5-Hydroxymethylcytosine (5hmC) modifications are generated by the Ten-eleven-translocation (TET) family of enzymes, TET 1, 2, and 3, through the oxidation of methylated cytosines (5mC). The TET family is capable of further oxidizing 5hmC to 5fC and 5caC, leading to eventual DNA demethylation. However, 5hmC marks can also exist stably in DNA. Stable 5hmC is enriched in the gene bodies of activated genes in multiple tissues, as well as associated with regulatory regions such as enhancers. Alterations to 5hmC patterns have now been found in multiple diseases including osteoarthritis. Here, we describe a method to map 5hmC modifications by next-generation sequencing using a technique based on the selective modification and enrichment of the 5hmC mark. We additionally provide a bioinformatic analysis pipeline to interpret the resulting data. Key words Hydroxymethylcytosine, 5hmC, Pull-down, Next-generation sequencing, Bioinformatics, Gene expression regulation

1

Introduction DNA methylation has long been an epigenetic mark associated with gene silencing. The family of cytosine modifications has now been expanded to include 5hmC, 5fC, and 5caC [1]. Cytosine bases in the CpG context can be modified to 5-methylcytosine (5mC) by the DNMT family of proteins and subsequently oxidized by the TET family of enzymes into 5-hydroxymethylcytosine (5hmC), and two further oxidized productions 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) [1]. Alterations in DNA modifications are associated with cell fate changes during development and tissue regeneration, while changes in these modifications often mark disease states. Multiple skeletal diseases including osteoarthritis, osteoporosis, and rheumatoid arthritis have demonstrated changes in several DNA modifications [2–4]. We have previously shown that

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_8, © Springer Science+Business Media, LLC, part of Springer Nature 2021

101

102

Fiorella Carla Grandi and Nidhi Bhutani

5hmC accumulates during chondrogenesis and is altered during osteoarthritis [5–7]. In addition, 5hmC is increasingly appreciated as a mark of transcriptionally active genes and can be useful in identifying a particular cell type and the epigenetic landscape that defines it [8]. However, the challenge with profiling 5hmC is that it typically represents only 0.05–0.6% of the total DNA in the cell [9] and is highly labile. Below, we detail a protocol for mapping 5hmC by next-generation sequencing. The protocol is based on the selective targeting of the 5hmC mark via treatment with the T4 Phage β-glucosyltransferase that selectively glucosylates the hydroxyl group [10]. The glucose group, which contains an azide, is then labeled using biotin, and this biotinylated-glucose modification can then be selectively pulled-down using commercially available streptavidin beads. The enriched DNA can be used for downstream genome-wide or targeted 5hmC assessment. The protocol consists of four major sections: (1) shearing of the genomic DNA and β-glucosyltransferase treatment for glucosylation of 5hmC moieties, (2) enrichment of DNA fragments, (3) library preparation and sequencing, and (4) data analyses.

2

Materials

2.1 DNA Shearing and Glucosylation Reaction

1. Qubit dsDNA BR Assay Kit (Invitrogen) and Qubit Fluorometer. 2. Covaris S220 Ultrafocused sonicator. 3. Covaris milliTUBE 1 mL AFA Fiber. 4. DNA cleanup columns. 5. PCR Tubes. 6. Active Motif HydroxymethylCollector Kit (see Note 1) or: (a) T4 Phage β-glucosyltransferase. (b) UDP-Azide-Glucose. (c) Biotin solution.

2.2 DNA Fragment Enrichment

1. Active Motif HydroxymethylCollector Kit or: (a) Dynabeads M-280 Streptavidin. (b) Bead washing solution: 10 mM Tris–HCl (pH 7.5), 1 mM EDTA, 2 M NaCl. 2. DNA LoBind Tube 1.5 mL (Eppendorf). 3. TE Buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8).

Mapping 5hmC Modifications using High-Throughput Sequencing

2.3 DNA Library Preparation and Next-Generation Sequencing

3

103

1. NEB Ultra II DNA Library Prep with Sample Purification Beads (NEB #E7103). 2. NEBNext Multiplex Oligos for Illumina.

Methods

3.1 DNA Shearing and Glucosylation Reaction

1. Extract DNA using standard methods for your skeletal tissue of choice. Quantify DNA (see Note 2). 2. Shear DNA: Place 2 μg of DNA in a Covaris tube and shear DNA to 200–300 base pair fragments. Conditions should be optimized for each sample type based on concentration and purity. Ideally, you want 2 μg of DNA in a volume of less than 35.5 μL (see Notes 3–5). 3. Set aside 10% of your fragmented input DNA as your input control for sequencing. This will be used to calculate the background for your assay. Freeze this sample at 20  C until Subheading 3.3. 4. Glycosylation reaction: set up the reactions in a 200 μL PCR tube as in Table 1. Control reactions: A negative control should be set up without the addition of the β-glucosyltransferase. In the skeletal system, ATDC5 cells at day 15 of differentiation can serve as a positive control, as these cells are known to have high levels of 5hmC [5]. 5. Incubate the reaction at 37  C for 1 h in a PCR thermocycler. 6. Spin down the PCR tube to collect all the volume. 7. Add 20 μL of biotin solution (Active Motif) to each reaction. 8. Incubate the reaction for 1 h at 37  C. 9. Purify biotinylated DNA using standard DNA cleanup column. Make sure that you are using a column that is compatible with your input DNA amount.

3.2 DNA Fragment Enrichment

1. Resuspend the streptavidin beads by vortexing and allow to come to room temperature. 2. Set up each enrichment reaction in a 1.5 mL DNA LoBind tube (Eppendorf) a 200 μL according to Table 2. Incubate the reaction for 2–3 h at room temperature with end-to-end rotation (see Note 6). 3. After incubation, spin down the tube to collect the volume at the bottom and place the tube in a magnetic stand. Allow the beads to pellet on the side of the tube. This may take 1–3 min. Remove the supernatant (see Note 7).

104

Fiorella Carla Grandi and Nidhi Bhutani

Table 1 β-glucosyltransferase Reagent

Volume

Buffer

5 μL

10 mM DTT

5 μL

UDP-Azide-glucose

2.5 μL

B-Gluctranserase

2 μL

Sheared DNA

______μL

dH2O

To volume

Total volume

50 μL

Table 2 DNA fragment enrichment Reagents

Volume

Streptavidin beads

25 μL

Binding buffer

25 μL

Purified biotinylation reaction

50 μL

Total volume

100 μL

4. Wash beads gently 5 times with 200 μL binding buffer (Active Motif). For each wash, remove the tube from the stand, pipet up and down gently 2–3 times making sure all beads are fully washed from the sides of the tube, and place the tube back on the magnetic holder. Allow beads to pellet again and repeat wash. 5. After the final wash step, remove the buffer and resuspend the washed beads with 30–50 μL TE buffer. Incubate for 30 min at room temperature with end-to-end rotation. After incubation, remove supernatant. This supernatant contains the 5hmC enriched DNA. 6. Purify the DNA using commercially available DNA cleanup kit for small fragments. 7. Quantify the amount of DNA using Qubit and Bioanalyzer (see Note 8). Store samples at 20  C until you are ready to perform sequencing. 8. Optional: locus-specific qPCR to verify pull-down (see Note 9). (a) Design primers spanning a 90–150 base-pair region on DNA on your target gene of interest.

Mapping 5hmC Modifications using High-Throughput Sequencing

105

(b) Perform a standard qPCR reaction using the Sybr Green system, using 1–5 ng of DNA in each reaction. Delta Ct calculations should be performed using the input as the reference control. (c) Use the negative control to check for the specificity of enrichment. 3.3 DNA Library Preparation and Next-Generation Sequencing

1. Using the quantification of your immunoprecipitated DNA, follow the manufacturer’s instructions for library preparation (see Note 10). If you are planning on pooling several samples into one lane, make sure to use barcodes. We find that it is possible to pool 6–8 samples per Illumina HiSeq 4000 lane. 2. Sequence DNA fragments.

3.4 Data Analysis (See Note 11)

1. The quality of your fastq files can be analyzed using FASTQC [12]. 2. Trim the low quality ends identified using FASTQC using Trimmomatic [13]. After trimming, run the fastq files back through FASTQC to ensure that all low quality reads have been removed. 3. Map reads to the correct species genome using HISAT2 [14]. For mouse and human, we suggest using mm10 and hg19, respectively, as they have the most updated additional epigenetic information from the ENCODE and ROADMAP projects [15–17]. 4. 5hmC peaks can be called using MACS [18], with the input sample as a control for the background to the biotin pulldown. 5. Differentially hydroxymethylated regions (dHMRs) can be called using diffREPS [19]. In order to do this, you must first convert your bam files to bed files using BEDOPS [20] bamtobed tool. 6. Annotation of peaks or dHMRs can be done using BEDOPS closest-feature tool. 7. Motif analysis of the DNA nearby your 5hmC can be useful to determine if specific transcription factors are either being targeted or being used to guide the TET enzymes. This analysis can be done using HOMER [21], which will predict both known and de novo motifs. 8. Visualize your 5hmC reads on gene bodies or other genes of interest using ngsplot [22]. 9. Additional downstream analysis options (see Note 12): (a) Intersect genes with dHMRs or peaks with genes differentially expressed in RNA-sequencing studies.

106

Fiorella Carla Grandi and Nidhi Bhutani

(b) Perform network or pathway analysis of genes with 5hmC peaks or dHMRs using IPA (QIAGEN), Enrichr [23, 24], or STRING [25]. (c) Intersect peaks with ATAC-seq [26] or other ChIP-seq data sets. Peak intersection can be done easily using bed files. We recommend Galaxy [27] as a good online tool for performing these types of analyses quickly. (d) Intersect peaks or dHMRs with enhancers or superenhancers, which can both be defined by a combination of histone modifications [28].

4

Notes 1. We recommend using the standardized Active Motif™ kit, as we find that these reagents give the most reproducible results. It is possible to buy all the component parts separately, but we find this has lower efficiency in the overall enrichment of 5hmC. In this protocol, we present our alterations to the manufacturer’s recommended guidelines, as well as the analysis pipeline we use after sequencing. 2. We recommend using the Qubit or Bioanalyzer assays as they are more accurate than Nanodrop based estimations, especially for DNA that is extracted from tissue samples, which may be more fragmented or contaminated with trace phenol. 3. There are multiple ways to create DNA fragments. We recommend using sonication, as this results in minimal loss of DNA. However, DNA can also be sheared using restriction enzymes. Several platforms exist for DNA sonication, including the Diagenode and Covaris. We have had the best performance with systems that perform sonication in chilled water baths, as this prevents degradation of the fragile 5hmC groups. 4. The input amount of DNA depends on the abundance of 5hmC in your particular skeletal tissue of interest, and the integrity of the sample. In our experience, depending on the degradation of DNA, you may require up to 5 μg of DNA for this assay. In the case of late-stage DMM models of OA, we found that the limited amount of DNA we could extract from the joint was more suitable for targeted analysis of 5hmC by using the Reduced Representation Hydroxymethylation Profiling [11]. 5. To find the ideal shearing conditions, we recommend setting up a reaction with your target DNA amount and doing a timecourse of shearing time. You can check the shearing either by running a gel or by Bioanalyzer. We recommend the Bioanalyzer trace as it will give you a better indication of the fragment

Mapping 5hmC Modifications using High-Throughput Sequencing

107

profile and concentration. A good starting point for the Covaris machine is: duty cycle: 10%, intensity 5, cycles per burst 200, and duration 180 s. 6. The reaction can also be allowed to incubate overnight at 4  C. 7. This supernatant can be analyzed to see what the percent capture is if desired. To do so, the optional qPCR step with the supernatant after DNA extraction as one of the samples. You should see enrichment of your target regions in the pulldown samples compared to the supernatant. 8. Nanodrop will not give an accurate quantification of the fragmented DNA. 9. When performing this enrichment on unknown samples, it can often be useful to check a known gene for 5hmC enrichment to check that enrichment went well. For example, if you are using Day 15 differentiated ATDC5 cells as a positive control, Col2a1 or Acan will be enriched for 5hmC [5]. 10. A variety of different DNA library preparation kits can be used. We recommend NEB’s Ultra II DNA Library Prep with Sample Purification Beads. 11. A variety of different programs can be used to analyze sequencing data. Here we detail the pipeline we use for our sequencing analysis. 12. Make sure that you are selecting only peaks or dHMRS with an adjusted p-value of 0.05 or lower for downstream analysis.

Acknowledgments F.C.G. is supported by the NSF Graduate Research Fellowship. Funds from the Department of Orthopedic Surgery, NIH/NIAMS R03 (R03AR066356) and R01 (R01AR070865) to NB supported this work. References 1. Kohli RM, Zhang Y (2013) TET enzymes, TDG and the dynamics of DNA demethylation. Nature 502:472–479. https://doi.org/ 10.1038/nature12750 2. Klein K, Gay S (2015) Epigenetics in rheumatoid arthritis. Curr Opin Rheumatol 27:76–82. https://doi.org/10.1097/BOR. 0000000000000128 3. Letarouilly J-G, Broux O, Clabaut A (2019) New insights into the epigenetics of osteoporosis. Genomics 111:793–798. https://doi. org/10.1016/j.ygeno.2018.05.001

4. Ramos YFM, Meulenbelt I (2017) The role of epigenetics in osteoarthritis: current perspective. Curr Opin Rheumatol 29:119–129. https://doi.org/10.1097/BOR. 0000000000000355 5. Taylor SE, Li YH, Smeriglio P et al (2015) Stable 5-hydroxymethylcytosine (5hmC) acquisition marks gene activation during chondrogenic differentiation. J Bone Miner Res Off J Am Soc Bone Miner Res 31(3):524–534. https://doi.org/10.1002/jbmr.2711 6. Taylor SEB, Li YH, Wong WH, Bhutani N (2015) Genome-wide mapping of DNA

108

Fiorella Carla Grandi and Nidhi Bhutani

hydroxymethylation in osteoarthritic chondrocytes. Arthritis Rheumatol 67:2129–2140. https://doi.org/10.1002/art.39179 7. Taylor SEB, Smeriglio P, Dhulipala L et al (2014) A global increase in 5-hydroxymethylcytosine levels marks osteoarthritic chondrocytes. Arthritis Rheumatol Hoboken NJ 66:90–100. https://doi.org/10. 1002/art.38200 8. Ecsedi S, Rodrı´guez-Aguilera JR, HernandezVargas H (2018) 5-Hydroxymethylcytosine (5hmC), or how to identify your favorite cell. Epigenomes 2:3. https://doi.org/10.3390/ epigenomes2010003 9. Skvortsova K, Zotenko E, Luu P-L et al (2017) Comprehensive evaluation of genome-wide 5-hydroxymethylcytosine profiling approaches in human DNA. Epigenetics Chromatin 10:16. https://doi.org/10.1186/s13072-017-0123-7 10. Song C-X, Szulwach KE, Fu Y et al (2011) Selective chemical labeling reveals the genomewide distribution of 5-hydroxymethylcytosine. Nat Biotechnol 29:68–72. 101038/nbt.1732 11. Petterson A, Chung TH, Tan D et al (2014) RRHP: a tag-based approach for 5-hydroxymethylcytosine mapping at singlesite resolution. Genome Biol 15:456. https:// doi.org/10.1186/s13059-014-0456-5 12. Babraham Bioinformatics - FastQC A Quality Control tool for High Throughput Sequence Data. https://www.bioinformatics.babraham. ac.uk/projects/fastqc/. Accessed 29 Sep 2019 13. Bolger AM, Lohse M, Usadel B (2014) Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinforma Oxf Engl 30:2114–2120. https://doi.org/10.1093/bio informatics/btu170 14. Kim D, Langmead B, Salzberg SL (2015) HISAT: a fast spliced aligner with low memory requirements. Nat Methods 12:357–360. https://doi.org/10.1038/nmeth.3317 15. The ENCODE Project Consortium (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489:57–74. https://doi.org/10.1038/nature11247 16. Yue F, Cheng Y, Breschi A et al (2014) A comparative encyclopedia of DNA elements in the mouse genome. Nature 515:355–364. https://doi.org/10.1038/nature13992 17. Roadmap Epigenomics Consortium, Kundaje A, Meuleman W et al (2015) Integrative analysis of 111 reference human epigenomes. Nature 518:317–330. https://doi.org/ 10.1038/nature14248 18. Zhang Y, Liu T, Meyer CA et al (2008) Modelbased analysis of ChIP-Seq (MACS). Genome

Biol 9:R137. https://doi.org/10.1186/gb2008-9-9-r137 19. Shen L, Shao N-Y, Liu X et al (2013) diffReps: detecting differential chromatin modification sites from ChIP-seq data with biological replicates. PLoS One 8:e65598. https://doi.org/ 10.1371/journal.pone.0065598 20. Neph S, Kuehn MS, Reynolds AP et al (2012) BEDOPS: high-performance genomic feature operations. Bioinformatics 28:1919–1920. https://doi.org/10.1093/bioinformatics/ bts277 21. Heinz S, Benner C, Spann N et al (2010) Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell 38:576–589. https://doi.org/10. 1016/j.molcel.2010.05.004 22. Shen L, Shao N, Liu X, Nestler E (2014) ngs. plot: Quick mining and visualization of nextgeneration sequencing data by integrating genomic databases. BMC Genomics 15:284. https://doi.org/10.1186/1471-2164-15284 23. Chen EY, Tan CM, Kou Y et al (2013) Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics 14:128. https://doi.org/10.1186/14712105-14-128 24. Kuleshov MV, Jones MR, Rouillard AD et al (2016) Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res 44:W90–W97. https://doi. org/10.1093/nar/gkw377 25. Szklarczyk D, Morris JH, Cook H et al (2017) The STRING database in 2017: qualitycontrolled protein-protein association networks, made broadly accessible. Nucleic Acids Res 45:D362–D368. https://doi.org/10. 1093/nar/gkw937 26. Buenrostro JD, Giresi PG, Zaba LC et al (2013) Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat Methods 10:1213–1218. https://doi.org/10.1038/nmeth.2688 27. Afgan E, Baker D, van den Beek M et al (2016) The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2016 update. Nucleic Acids Res 44:W3–W10. https://doi.org/10.1093/nar/gkw343 28. Hnisz D, Abraham BJ, Lee TI et al (2013) Super-enhancers in the control of cell identity and disease. Cell 155:934–947. https://doi. org/10.1016/j.cell.2013.09.053

Chapter 9 Using FRAP to Quantify Changes in Transcription Factor Dynamics After Cell Stimulation: Cell Culture, FRAP, Data Analysis, and Visualization Kannan Govindaraj and Janine N. Post Abstract Here we show how to measure the mobility of transcription factors using fluorescence recovery after photobleaching (FRAP). Transcription factors are DNA-binding proteins that, upon binding to specific DNA motifs, regulate transcription of their target genes. FRAP is a simple, fast, and cost-effective method, and is a widely used quantitative method to measure the dynamics of fluorescently labeled molecules in solution, membranes, and inside living cells. Dynamics, specified by the immobile fraction, recovery halftime, diffusion constant, and ratio of molecules contributing to different phases of FRAP recovery, can be quantified by FRAP. This can be useful to understand their function in gene regulation. This tutorial is intended to familiarize the reader with the FRAP procedure to quantify transcription factor dynamics using a standard confocal microscope and analysis using MATLAB (MathWorks®). This article will guide the reader through the preconditions of FRAP, and a detailed and step-by-step procedure of preparing cells, bleaching protocol, data analysis in MATLAB, and visualization of the FRAP data. Key words Protein dynamics, Transcription factor activity, Fluorescence recovery, FRAP , SOX9, mGFP, CLSM

1

Introduction Fluorescence recovery after photobleaching (FRAP) is a biophysical technique, developed in the late 1970s by Axelrod et al. [1]. FRAP has been successfully applied to study the mobility of fluorescent molecules in solution. The discovery of green fluorescent protein (GFP [2]) and subsequent advances in imaging technologies further extended the scope of FRAP to study protein and lipid mobility in live cells. FRAP can be used to study the mobility of fluorescently labeled proteins, lipids, and molecules in 2D structures (e.g., the plasma membrane) and in 3D structures (e.g., nuclei and cytoplasm) [3]. However, appropriate FRAP models resembling actual reaction kinetics within the system—2D or 3D—under investigation should be used to interpret the data [4].

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_9, © Springer Science+Business Media, LLC, part of Springer Nature 2021

109

110

Kannan Govindaraj and Janine N. Post

Transcription factors play a key role in the regulation of gene expression, and their binding to DNA precedes its activity. There are numerous theories describing the mechanism of transcription factor activity, reviewed in [5]. Despite many conflicting and consenting theories on their functional mechanism [6], the quantity of transcription factors bound to DNA, affinity, and duration of binding seem to be the major contributing factors in exerting their activity, which could be either transcriptional activation or repression [5, 7]. Thus, quantifying transcription factor dynamics will yield useful information on their transcriptional activity. In this tutorial, we will explain when and how FRAP can be used to measure transcription factor dynamics. We describe the FRAP protocol from preconditions of FRAP, such as required materials and preparatory work, tips and tricks, to data analysis and visualization. We will discuss how to stimulate cells to study transcription factor dynamics in response to external factors [8], as we have done in our studies. To guide the reader through the FRAP procedure, we explain FRAP terminologies, models for fitting FRAP data, data analysis using ImageJ (FIJI), MATLAB (MathWorks®), and data visualization using OriginPro® (OriginLab®) software. Additionally, we added a step-by-step protocol to perform a FRAP experiment using a NIKON confocal microscope as an online supplement. 1.1 Different Methods for Measuring Protein Dynamics

FRAP is a widely used method to study the protein dynamics in live cells due to its simple setup in a laser confocal microscope. Other methods, such as FLAP (fluorescence loss after photoactivation), FLIP (fluorescence loss in photobleaching), FCS (fluorescence correlation spectroscopy), and SMM (single molecule microscopy) are also used to study protein kinetics [3]. These methods have their own advantages and disadvantages, as described in Table 1.

1.2 When to Use FRAP

FRAP can be used to study the signaling mechanism or to map the factors that regulate a cellular process, such as the activity of transcription factors. For this, cells must be transfected with a fluorescent fusion protein, thus the protein of interest is tagged with a fluorescent protein. The choice of transfection method depends on the cell type and the protein of interest. It is recommended to make a stably transfected cell line using, for example, CRISPR/Cas9 [9], if a cell line (which properties are not affected by passage number) is used in the study, or if overexpression of the protein of interest is toxic to the cells. This helps to study the protein of interest at the native expression levels and helps to avoid the transfection step every time before FRAP. If using primary cells or cells which tend to (de)differentiate over a number of passages, transient transfection would be the best option. If the cells are easy to transfect, lipid mediated transfection may be used. Retroviral or other viral-

Protocol for Quantifying Transcriptional Activity by FRAP

111

Table 1 Advantages and disadvantages of various methods used to study protein dynamics Methods Advantages FRAP

Disadvantages

Complex computational modeling is [3, 4, 29] required in case of diffusion coupled recovery l If fast diffusion dominates the FRAP recovery, slow diffusion may not be resolvable

Simple instrumentation setup Many FRAP models reflecting actual processes in the cell are already available l If diffusion and reaction kinetics occur in distinct timescales, protein dynamics can be readily calculated

l

Simple instrumentation setup Procedure is similar to FRAP, a small modification is required for calculations

l

[30, 31]

More precise protein dynamics diffusion rates can be calculated l Diffusion and reaction kinetics can be distinguished even in the diffusion coupled mobility l FCS requires a low level of fluorescent molecules, so it is easier to use on proteins that are expressed at a low concentration

l

Requires complex instrumentation [32–34] Application to study the proteins inside the cell is challenging because of complex diffusion analytical models

Individual fluorescent molecules can be tracked l More precise protein dynamics diffusion rates can be calculated. l Diffusion and reaction kinetics can be distinguished even in the diffusion coupled mobility

l

l l

FLAP/ FLIP

l

FCS

l

SMM

Refs.

l

l

Sensitive to acquisition photobleaching l Reaction kinetics (slow diffusion) dominate fluorescence decay l

Requires ultrasensitive cameras and accurate speckle-tracking programs l Transfection or microinjection should be optimized to obtain very low levels of protein expression l Long exposure time is necessary to get good signal-to-noise ratio, leading to acquisition photobleaching

[3, 35–37]

mediated transfection can be used for hard-to-transfect cells. N.B. check local safety regulations, since in most countries the use of viruses is only allowed in labs with the correct biosafety levels, which is then also true for the location of the microscope (BSL-2 in the USA, MLII in the Netherlands). We do not have a confocal microscope in our MLII lab, so we cannot use viral induction for our FRAP experiments. Since the cells with which we work, human primary chondrocytes (hPCs) and human mesenchymal stem cells (hMSCs), tend to dedifferentiate over a number of passages, we used lipid mediated transient transfection. In these primary cells, we were able to achieve at least a transfection efficiency of 20% (see Subheading 3). For training purposes, we use the immortalized C20/A4 cell line.

112

1.3

Kannan Govindaraj and Janine N. Post

FRAP Principle

In FRAP a small region of interest (ROI, can be circular, square, or rectangular) is photobleached with a high-intensity laser, and the fluorescent recovery is monitored in that ROI in a given timescale (from seconds to minutes). Post-bleach imaging duration and frame-rate is determined based on the mobility of the molecule of interest. The higher the mobility, the lower the imaging time, the higher the frame rate. For example, if the mobility of a transcription is relatively high, with a stable plateau reached within a minute, fluorescence recovery can be recorded for 60 s, at a speed of 4 frames per second (fps). Before photobleaching, a number of pre-bleach images (usually ~10) are recorded for normalization of the fluorescence recovery. This process can be translated into a graph. As an example, we measure the fluorescence recovery of the transcription factor SOX9, the master regulator for cartilage formation [10], see Fig. 1. In this graph, one can observe that the fluorescence in the bleach spot does not reach the level of the original fluorescence. The plateau level is lower than the pre-bleach intensity because some of the FRAP-bleached molecules are immobile within the region of interest that is bleached. Because of their immobility, they do not contribute to the recovery, while at the same time occupying binding sites for incoming unbleached proteins. We therefore name the fraction of molecules that contributes to the recovery the “mobile fraction” and the one that does not contribute the “immobile fraction.”

Fig. 1 Fluorescence recovery curve of SOX9-mGFP. The fluorescence intensity is normalized to 100%. The immobile fraction and recovery half-time are indicated (red dotted lines)

Protocol for Quantifying Transcriptional Activity by FRAP

113

1.4 Mapping Signal Transduction Pathways Regulating Transcription Factor Mobility

For studies of immediate response of transcription factors to extracellular signals, cells can be stimulated with cytokines (before/ during FRAP experiments) in an imaging buffer (see below). For example, we transfected the cells with a fluorescently tagged transcription factor, SOX9-mGFP and treated cells with predicted agonists/antagonists to study their transcriptional regulation [8]. Stimulation time depends on the mechanism of action of cytokine and needs to be determined empirically. If the cytokine exerts fast cellular responses, as we have seen for the SOX9 catabolic factors, such as WNT3a and IL1β, changes were observed within 20 mins after stimulation (Fig. 2, [8]). However, a slightly longer stimulation time (60 min) was required for the SOX9 anabolic factor BMP7 (see Fig. 2).

1.5 Quantitation of Protein Mobility

Two major factors affect the mobility of a molecule in a biological system: diffusion and chemical interactions. The rate of diffusion is determined by the diffusion constant (D) of the molecule, which is dependent on the size of the molecule, the viscosity of the surrounding medium, including physical structures that hinder diffusion, and the temperature. In the cell, proteins interact with other molecules, including other proteins and DNA/RNA. The binding constants of the molecular interactions, binding (kon) and dissociation (koff) with other molecules, will affect mobility of the protein of interest.

Fig. 2 FRAP curves showing SOX9-mGFP mobility changes in C20/A4 cells after treatment with cytokines. Treatment of C20/A4 cells with WNT3a at 10 ng/mL increased SOX9-mGFP mobility at 20 mins (yellow curve) as compared to the control (blue curve). In contrast, BMP7 at 10 ng/mL did not change SOX9-mGFP mobility significantly at 20 min (orange curve) as compared to the control. However, 100 ng/mL of BMP7 with 1 h treatment decreased the SOX9-mGFP mobility significantly (gray curve). This indicates that concentration of cytokines and stimulation time varies among cytokines and should be determined empirically

114

Kannan Govindaraj and Janine N. Post

Protein kinetic data can be calculated from the FRAP measurements. This includes the fraction of immobile molecules (immobile fraction, IF), the recovery half-time (τ½, time it takes for the fluorescence to reach 50% of the final fluorescence), diffusion constant (D), and association (kon) and dissociation (koff) rates [11]. So, using FRAP, one can calculate the diffusion and reaction kinetics of any fluorescently labeled protein/lipid in a cell. If the molecular processes that contribute to the fluorescence recovery occur in distinct timescales, FRAP is a robust method to study the protein kinetics. To calculate the protein kinetic data, different equations/models can be applied: (i) chemical interaction model, (ii) diffusion model, and (iii) reaction-diffusion model. In the chemical interaction model, FRAP curves are fitted with an exponential equation with a single exponent:   F ðt Þ ¼ y0 þ A 1  e t=τ ð1Þ where y0 is the value of the fluorescent intensity at the first postbleach frame, A is the amplitude of mobile population, and τ is the time constant. To calculate the half-time recovery of A, Half  time to recover : t § ¼ ln ð2Þ∗τ

ð2Þ

where t½ is the time-point at which 50% of the fluorescence is recovered, τ is the time constant. Immobile fraction : IF ¼ F I  F E

ð3Þ

where FI is the initial intensity and FE is the end value of the recovered intensity. The effective diffusion constant (D) can be determined from the FRAP curves based on the method from [12], which describes how the effective diffusion constants can be calculated from FRAP curves in combination with its laser bleaching profile. To correct for diffusion during bleaching, the effective radius for the bleaching spot is calculated from the user-defined nominal radius. In combination, a FRAP equation was derived that can be readily fitted to the normalized FRAP data to extract diffusion constants:   K ∗re 2 F ðt Þ ¼ 1  ∗M f þ ð1  M f Þ∗F0 ð4Þ 8∗D∗t þ re 2 þ rn 2 The laser bleaching profile is incorporated in Eq. 4 with K as bleach depth parameter, rn as the user-defined nominal radius, and re as the effective radius. Mf is the mobile fraction, F0 is the postbleach intensity, and D is the effective diffusion constant when considering 3-dimensional diffusion. If the fluorescence recovery is diffusion-coupled (fast and slow diffusion occur in similar timescales) or when many processes occur

Protocol for Quantifying Transcriptional Activity by FRAP

115

in overlapping timescales, complex computational modeling is required for FRAP data analysis. In this case, caution should be applied when calculating diffusion constants and association and dissociation rates from the FRAP data [4]. On the other hand, fluorescence correlation spectroscopy and single molecule microscopy can resolve the different molecular processes occurring at similar timescales and provide precise diffusion constants and association and dissociation rates [3]. For example, in the transcription factor dynamics, diffusion by unbound protein occurs instantly, even before capturing the first post-bleach image. The FRAP curve will be a result of two diffusion processes with distinct timescales, fast (A1) and slow (A2) diffusion. Nonspecific binding of the transcription factor to DNA will be fast moving (A1) and the fluorescence recovery will be slower at the specific binding (A2). The ratio (A1/A2) indicates the amount of contribution of these populations present in the measured cell. In this instant, FRAP curves can be fitted using a two-component fit:     ð5Þ F ðt Þ ¼ y0 þ A 1 1  e t=τ1 þ A 2 1  e t=τ2 where A2 is the amplitude of slow diffusing population, τ1 and τ2 are the time constants of A1 and A2, respectively. 1.6 Analyze FRAP Data

Proper fitting and modeling of the FRAP curve, which reflects the actual biological process, is important to extract useful dynamics information from the FRAP data. Fitting and modeling should consider the number of reactions contributing to the FRAP recovery, the timescales, and whether these reactions occur in similar or distinct timescales. For example, untagged mGFP (OriGene) protein does not bind to any intracellular target and recovers completely and within seconds after photobleaching (see Fig. 3a). FRAP data of mGFP can be easily fit with a single exponential fit (see Fig. 3b, Eq. 1). In contrast, SOX9-mGFP binds to DNA and the FRAP recovery curve is the result of two reactions, namely fast (weakly bound to DNA, nonspecific binding) and slow diffusion (strongly bound to DNA, specific binding). FRAP data of a transcription factor like SOX9 should be fit with two (or more) component fits (see Fig. 3c, Eq. 5). The presence of shoulders in the curve can be an indicator of the number of processes contributing to the FRAP recovery. After bleaching, mGFP quickly recovers and attains a plateau, while SOX9-mGFP slowly recovers and continues to recover (due to exchange at the binding sites) throughout the recovery timescale, with a visible shoulder in the recovery curve (see Fig. 3a, red curve). In our model, we use a two-component fit (see Eq. 5) for transcription factor dynamics, as two processes at distinct timescales

116

Kannan Govindaraj and Janine N. Post

Fig. 3 Fitting FRAP curves. (a) FRAP curves of mGFP (gray) and SOX9-mGFP (red). (b) The mGFP FRAP curve is fit with a one-component fit (blue circles are data-points, red line is the fit) and (c) the SOX9-mGFP curve is fit with a two-component fit (gray and green lines are first- and second-component fit, respectively)

contribute to the FRAP recovery. If more than two processes contribute to FRAP recovery or if these processes occur in similar timescales, this FRAP model cannot be applied. In that case a new model should be built based on initial control experiments for the analysis of FRAP data. 1.7 Explanation of FRAP Parameters

Immobile fraction (IF, Eq. 3) refers to the fraction fluorescent protein bound to its target, e.g., transcription factor bound to DNA. For transcription factors there are two populations, namely fast (A1) and slow (A2) diffusing populations that contribute to the fluorescence recovery (see Fig. 4). The fraction of protein that are weakly bound to DNA constitute the fast diffusion population, whereas the population bound to DNA constitutes the slow diffusing population. The ratio A1/A2 refers to the increase or the decrease of fluorescent proteins in the A1 compared to A2 fraction (i.e., the ratio of the fast diffusing population to the slow diffusing population). If the fast and the slow diffusing populations are equally present inside the nucleus, the value of the A1/A2 ratio will be 1. If the

Protocol for Quantifying Transcriptional Activity by FRAP

117

Fig. 4 Schematic representation of a FRAP recovery curve and explanation of parameters. If fast and slow diffusion occur at two different timescales, the FRAP curve can be split into two phases, as shown. FI: initial intensity, F0: intensity at time point t0 (first post-bleach intensity), FE: end value of the recovered intensity, t½: halftime of recovery, immobile fraction (IF) is the population of SOX9-mGFP bound to DNA. A1 is the amplitude of fast diffusing population of SOX9-mGFP, which is not bound to DNA and contributes to quick recovery. A2 is the amplitude of slow diffusing population of SOX9-mGFP, which interacts on the various binding sites in the DNA

value is less than 1, this indicates that the number of fast diffusing proteins is less than that of the slow diffusing proteins. The diffusion constant (D) is the rate at which a molecule diffuses in a specific area. In a cellular milieu, the term “effective diffusion (Deff)” indicates the recovery that mimics diffusion, but at a rate that is slowed by binding interactions. This is calculated using Eq. 4. The recovery time is the length of time after bleaching, required for the fluorescence recovery to reach a constant value. Recovery half-time (t½, Eq. 3) refers to half of recovery time. Recovery half-time of A1 and A2 indicates half of the time required for the recovery of corresponding phase of the FRAP curve. For detailed information, the reader can refer to [8].

2

Materials

2.1 Materials for Cell Culture and Transfection

1. Silicone gasket with 1.5 mm thickness (cat#: 70465-2R2, EMS, USA). 2. Fibreless tissue paper (such as Kimberly-Clark® Professional). 3. 24-Well plates.

118

Kannan Govindaraj and Janine N. Post

4. Microscopic cover glass (12 or 13 mm, Ø). 5. Tweezers to lift the coverslip. 6. Suitable transfection reagent. (a) Lipofectamine LTX with Plus reagent (cat. # 15338030, ThermoFisher scientific) (for C20/A4 cells and human primary chondrocytes). We used 1:1 (w/v) ratio of DNA and Plus reagent and 2:1 (v/v) ratio of Lipofectamine LTX to Plus reagent. (b) Lipofectamine 3000 with P3000 reagent (cat. # L3000015, ThermoFisher scientific) (for hMSCs). We used 1:1.5 (w/v) ratio of DNA and P3000 reagent and 2:1 (v/w) ratio of Lipofectamine 3000 to DNA. 7. Plasmid DNA encoding GFP-tagged protein of interest. 8. Confocal microscope. 9. Cells of interest. 10. Proteins/cytokines for cell stimulation. 11. Microscopic glass slide. 12. Imaging buffer. 2.2 Materials for FRAP

1. 135 mM NaCl.

2.2.1 Imaging Buffer (Tyrode’s Solution, See Note 1)

3. 0.4 mM MgCl2.

2. 10 mM KCl. 4. 1 mM CaCl2. 5. 10 mM HEPES, pH adjusted to 7.2. 6. Imaging buffer with above components was filter sterilized and stored at 20  C. 7. On the day of use, add 20 mM glucose and 0.1% BSA (final concentration after adding to the buffer). Alternatively, medium without phenol red can also be used as imaging media (see Note 2).

2.2.2 Confocal Laser Scanning Microscope

1. Laser scanning confocal microscope (We used a Nikon A1 confocal microscope) with suitable laser lines for excitation of the fluorescent protein. 2. Option to maintain the physiological temperature, CO2 control.

2.3 Materials for FRAP Analysis

1. MATLAB with script (available on request). 2. ImageJ (for drift correction). 3. Alternatively, easyFRAP software can be used.

Protocol for Quantifying Transcriptional Activity by FRAP

2.4 Materials for Data Visualization and Statistical Analysis

3

119

1. Data analysis and graphing software, we used Origin Pro (OriginLab®).

Methods

3.1 Cell Culture and Transfection 3.1.1 Cell Culture

To study the cellular physiology, cells ideally should be maintained in their native state. For example, primary chondrocytes should maintain their chondrogenic potential during the experiments. To prevent dedifferentiation, expand the primary chondrocytes in physioxical (2.5–5% O2) conditions. Use low passage numbers (25) are required. During analysis, use only the average intensity of last 10 pre-bleach images for normalization. 2. Photobleaching: Photobleaching should be instantaneous after the pre-bleach. A small region of interest (ROI) is photobleached with a high laser power. The laser used for the image acquisition can be used for photobleaching as well and the laser power needs to be determined empirically. Duration of bleaching and number of iterations of high-intensity laser pulse should be as low as possible to calculate accurate protein dynamics. Optimal bleaching parameters can be easily determined using a fixed cell (using 4% paraformaldehyde in PBS)

122

Kannan Govindaraj and Janine N. Post

Fig. 6 Optimizing bleaching laser intensity. Bleaching efficiency of SOX9-mGFP in a fixed cell at different percentages and μW (at the objective) of laser power as indicated in the image. Bleaching is inefficient at (a) 10% (6.8 μW) and (b) 25% (17.1 μW) laser power, whereas (c) 50% (34.3 μW) and (d) 100% (69.0 μW) efficiently bleach the SOX9-mGFP and the bleach profile is shown in (e)

expressing the same fluorescent protein (see Fig. 6). The following guidelines can be helpful: (a) Laser power: Usually, 25–100% laser power is used for photobleaching. Higher laser power enables faster bleaching, but is toxic to the cells [21]. It should be optimized to the minimum laser power needed to achieve complete photobleaching of the ROI (see Fig. 6). (b) Size of the ROI: The size of the ROI is usually less than 10% of the total fluorescence distribution area. It also depends on the kinetics of the fluorescent molecule. A size of 1–2 μm (Ø) ROIs with higher magnification objectives (60/100) give a good signal-to-noise ratio (SNR). Smaller ROIs will increase the noise. (c) Shape of the ROI: The ROI can be circular or a rectangle or a square. Again, this depends on the shape/distribution of the fluorescent protein. We use a circular ROI for nuclear proteins, such as transcription factors. However, the calculation formulae may need to be adjusted according to the shape of the ROI [22].

Protocol for Quantifying Transcriptional Activity by FRAP

123

(d) Duration of bleaching: Ideally it should be as short as possible. (e) Number of iterations: It is recommended to achieve complete photobleaching in a single iteration. This is especially important when studying fast kinetics. (f) Scan speed: The higher the scan speed, the faster the bleaching process. However, the bleaching can be inefficient. 3. Post-bleach image acquisition: Acquire sequence of images after photobleaching to monitor the dynamics of fluorescence recovery. The following acquisition guidelines can be used to resolve the dynamic range and good temporal resolution. (a) Acquisition frequency: At least 20 images during the time required for the half of the recovery are needed. (b) Acquisition duration: Needs to be 10 to 50 times longer than the halftime of recovery [1] or until the fluorescence recovery attains a plateau and should be determined empirically. (c) Acquisition photobleaching: To resolve precise protein dynamics, it is necessary to reduce the photobleaching during image acquisition and the guidelines below can be helpful. The general rule is to reduce the photo-toxicity during FRAP, so acquiring sharp and nice images are not the priority. To minimize photobleaching and photo-toxicity during FRAP, the following parameters can be adjusted: 1. Laser power: Laser power should be as low as possible during pre- and post-bleach image acquisition. For example, to image mGFP or mGFP-tagged protein, laser powers of ~0.35–0.5% are ideal on the Nikon confocal microscope. Laser power also depends on the age of the laser. If ~0.35–0.5% laser power does not sufficiently illuminate the fluorophore, increase the laser power. 2. Frame size: Decreasing the frame size will enable faster a scan rate and less light exposure. (For example, a 125  125 or 256  256 frame size can be used instead of 512  512). 3. Frame rate: The higher the frame rate, the lesser the light exposure of the cells. Higher frame rates are recommended especially during the study of rapid kinetics. 4. Averaging: Frame and line averaging should be avoided to achieve faster frame rates and to reduce photo-toxicity and photobleaching. Again, obtaining the best image is not a priority.

124

Kannan Govindaraj and Janine N. Post

5. Pinhole: Opening the pinhole enables to capture more signal and helps to keep the laser power at the minimal level. But, it also enlarges the confocal volume and background signal. However, this will not pose much problem during data analysis as the SNR will be higher for our FRAP settings. 6. Zoom: Use the zoom option to enlarge the imaging area. 7. Photo-stable fluorophores: Use fluorophores which are photo-stable at lower light intensities (such as eGFP/mRFP) to reduce acquisition photobleaching. 3.2.1 Optimal FRAP Parameters

The optimal FRAP parameters for our study of SOX9-mGFP dynamics in the nucleus are given below as an example. Although we used a Nikon A1, the settings should be easily transferable to a confocal microscope of a different manufacturer. 1. Pre-bleach images: 25 images. 2. Bleaching: 1 iteration of high-intensity laser pulse (50%, 488 nm laser). 3. Post-bleach images: 240 images (60 s, imaging time). 4. Objective: 60 (water immersion), 1.4 NA. Alternatively, a silicone oil immersion objective can be used, as long as the refractive index of the objective is close to that of the cell to avoid aberrations. 5. Scan mode: Unidirectional. 6. Frame rate: 4 frames/s(fps, for both pre- and post-bleach imaging). The frame rate can also be set based on the pixel dwell time. 7. Frame size: 256  256 pixels. 8. Averaging: Normal (No averaging!). 9. Pinhole: 1.2 AU (Z-step size: 0.25 μm and Optical sectioning: 0.77 μm). 10. HV (Gain of the detector): ~80–90% (depends on the intensity of the fluorescent protein). 11. Offset (of the detector): 1. 12. Laser power: 0.35% (0.12 μW at the objective). 13. Zoom: 7.09. 14. Pixel size: 0.12 μm. 15. Laser: 488 nm. 16. Bleaching: 50% (34.3 μW, at the objective) laser power, 16 fps (the highest speed possible in the Nikon A1). 17. Export the fluorescent intensity values to a Microsoft Excel document using the “Export” option and save it in a folder. This Excel document will be used for data analysis in MATLAB.

Protocol for Quantifying Transcriptional Activity by FRAP

125

18. Repeat the FRAP experiment for at least 50 cells per condition and collect FRAP image files and Excel documents containing FRAP data. 3.2.2 Performing FRAP Using a Nikon A1 Confocal Laser Scanning Microscope

As a detailed example of a FRAP experiment, we describe performing FRAP on a Nikon A1 confocal microscope. This may help the reader to convert the protocol to their own CLSM. Nikon’s “NIS elements” software provides convenient user interface and FRAP settings. This session will guide the user through FRAP setup in “NIS elements” software. Options and values shown in the snippets below are based on the optimal FRAP parameters we used. In this protocol, we use a nucleus expressing SOX9-mGFP as an intracellular organelle to explain the FRAP settings in the NIS elements software. If your target organelle or region is other than the nucleus, you can replace the term “nucleus” with your target. 1. Turn on the laser, temperature controller, and the microscope system. 2. Fasten a stage inlet suitable to mount microscopic cover glass. 3. Start the “NIS elements” software. 4. Following acquisition and analysis controls need to be opened in the NIS elements software (see Fig. 7). (a) A1plus Compact GUI (View ! Acquisition controls ! A1plus Compact GUI). (b) A1plus Stimulation (View ! Acquisition controls ! A1plus Stimulation). (c) A1plus Scan Area (View ! Acquisition controls ! A1plus Scan Area). (d) ND Stimulation (View ! Acquisition controls ! ND Stimulation). (e) Ti Pad (View ! Acquisition controls ! Ti Pad). (f) Time Measurement (View ! Analysis controls ! Time Measurement). 5. “Ti Pad” contains basic microscopic control options. Set the options as shown below to view the cells in the binocular. (a) Select “60x objective”. (b) Select the light path “E100”. (c) Select “EPI” mode and turn on the shutterns (in the separate physical device). (d) Select “green filter” (for GFP). (e) Zoom option in the “Ti Pad” is the optical zoom. Keep it at 1.00 in the “Ti Pad” and in the microscope as well.

126

Kannan Govindaraj and Janine N. Post

Fig. 7 “A1plus Compact GUI” provides options to control image acquisition. (a) While “Eye port” is selected (red box 2), interlock is automatically activated (red box 3) and the rest of the options are automatically disabled. Samples can be viewed through the binocular. (b) While “Scan” mode is on (red box 1), “Eye port” is not selected (red box 2), “Interlock” is removed (red box 3), “Unidirectional” scan mode is selected (red box 4), “Scan speed” is set to 4 fps, “image size” is set to 256  256 pixels (red box 5), “Averaging” is set to normal (red box 6). “Pinhole” is set to 1.2 AU (red box 7). Needed laser line can be activated by the “laser settings” option (red box 8), “HV” is set to 88, “Offset” is set to 1, “488 nm” laser line is selected and “laser power” is set to 0.35% (red box 9)

6. Set the parameters in the “A1plus Scan Area” window, if necessary. Once the parameters are set, drag and position the scan area (red square) to the middle of the window, and to confirm the position, right click on it (the red square will turn into green). 7. Place a drop of milliQ water on the 60 water objective and place the microscopic glass slide mounted with transfected

Protocol for Quantifying Transcriptional Activity by FRAP

127

cells. The glass slide should be placed in the inverted position, so that the microscopic coverslip is in contact with the objective. 8. Look through the “binocular” and focus the cells using white light. Once the cells are focused, turn on the epi(fluorescent) light (white light can be turned off) and search for the transfected cells. Click the “EPI” button to turn on/off the epi light in the “TiPad” window. Following guidelines can be helpful to select the right cells for FRAP. (a) The transfected cell morphology should be normal. (b) Cells having extra weird fluorescent bodies should be avoided. (c) Avoid cells expressing very low or very high levels of fluorescent proteins. (d) For cells expressing optimal level of fluorescent proteins, “HV” values would be ~85–90% (for above mentioned optimal FRAP parameters). 9. Position a nucleus (or other cellular region of interest) expressing fluorescent proteins in the center of the view area (when looking through the binocular). 10. Change the light path to “L100” and turn on the “PFS” in the “Ti Pad” window. 11. Set the image acquisition parameters in “A1plus Compact GUI” window, as shown in Fig. 10. If “Remove Interlock” button appears in “RED,” click on it, to remove interlock and turn on the “PFS” in the “Ti Pad” window. 12. Make sure that the correct laser is selected (for example, the 488 laser in case of GFP tagged protein); click on “Scan” button (top left corner) in the “A1plus Compact GUI” window. Once the first FRAP measurement is finished after starting NIS elements software, switching between “Eye” and “Scan” mode can be done in one click by “Eye Port” button (see Fig. 7, red box 2). 13. A new window showing the centered nucleus will appear. Focus may need to be slightly adjusted to bring the nucleus into focus. 14. Turn on “pixel saturation indicator” and see if there are any saturated pixels in the nucleus (saturated pixels are shown as red dots by default). If saturated pixels appear, decrease HV values to reduce the number of saturated pixels (a few, >” (red box 3) button and select “Copy All Data” option. Paste it in a separate “.txt” file. (d) Averaged bleach profile and the calculation of effective radius. Effective radius can be calculated from the center axis (red box 4) and using the “μm” scale of the “X”-axis. Effective radius of this bleach profile is 2.2 μm

the light, in practice, the size of the bleached ROI will be always higher than the actual size of the ROI. So, finding the actual bleaching size and effective radius is necessary to calculate precise protein dynamics. To know more about effective radius and nominal radius, refer to [27]. The effective radius of the bleach spot can be calculated as described below. (a) Import the “FRAP image stack” to Fiji. (b) Go to the first post-bleach image (26th image) and draw a straight line across the bleach spot as shown in the Fig. 9a. Center of the straight line should be at the center of the bleach spot. (c) Click on “Plot Profile” option (Analyze ! Plot Profile) to get the fluorescent intensity profile along the line. (d) A new window as shown in Fig. 9 (b or c) will appear. Copy the fluorescent intensity profile data and paste it in a new “.txt” file. (e) Scroll back to the previous image (25th image or the last image of the pre-bleach series). The “straight line” will stay on the image stack. Repeat the steps (iii and iv). (f) MATLAB script recognizes pre-bleach and post-bleach line intensity profiles by the “file name” of the text files and they should be named accordingly. The text file containing post-bleach and pre-bleach fluorescent intensity profile should be suffixed with “ps” and “pe,” respectively.

Protocol for Quantifying Transcriptional Activity by FRAP

133

Remaining characters, especially the file numbers in the file name, should be same for both the files. Refer to the sample datasets in the “Effective Radius” folder. (g) Repeat steps (i – vi) for at least 10 FRAP image stacks and save those intensity profile files in a folder. (Note: the length of the line drawn across the bleach profile at step (ii) should be the same for all FRAP image stacks). (h) There are two MATLAB scripts (“FindRe” and “hdrload”) in the “Effective Radius” folder. Copy the scripts and the intensity files (saved in step vii) to the M ATLAB home directory folder. (i) Run “FindRe” script in the MATLAB; an image with fit with average bleach intensity profile will appear and the effective radius can be calculated from the bleach profile as shown in Fig. 9d. 5. Input the following details in the “FRAPAnalysis” script as shown in Fig. 10. (a) FRAP data files folder location in the “datapath” (line 11). (b) The frame number of first photobleach image (line 20).

Fig. 10 Variables to be entered in the “FRAPAnalysis” Matlab script. FRAP data files folder location is mentioned the “datapath” (red box 1). The frame number of the first post-bleach image and the frame number at which photobleach correction should be started are mentioned in line 20 and 21, respectively (red box 2). The diameter of the bleach ROI is 25 pixels and the pixel size is 0.12 (red box 3). The effective radius is 2.2 as calculated in step 5 (red box 4)

134

Kannan Govindaraj and Janine N. Post

(c) The frame number at which photobleach correction to be started (line 21). (Note: We used mGFP and some of the mGFP proteins underwent triplet state at the start of the FRAP experiment. We acquired 25 pre-bleach images and omitted first 15 images to exclude the triplet state artefacts. So, the photobleach correction starts at 16. If there is no such problem of triplet state artefacts, 10 pre-bleach images are sufficient. In that case, photobleach correction will start from 1.) (d) Diameter of the “Bleaching ROI” and “pixel size” (line 26). (e) Effective radius of the “Bleaching ROI” as calculated in the previous step (line 29). (f) Specify the range of pre-bleach images (line 48): If you have acquired 25 pre-bleach images and want to normalize your data using the last 10 pre-bleach images (i.e., 16–25 frames), then it should be “Dimen ¼ blcor(16:25)”. Adjust the numbers to the range of per-bleach frames needed for data normalization. 6. Once these variables are entered in the MATLAB script, click the “Run” button in MATLAB. MATLAB will automatically extract FRAP data from the excel files and analyze them. 7. Once analysis is complete, a new folder named “Analysis” will be present in the FRAP data files folder. 8. The “Analysis” folder will contain two folders, “SingleFit” and “DoubleFit,” and two Microsoft Excel files, “Consolidated” and “FitResults.” 9. The “DoubleFit” folder will contain all the two-component fit images for individual FRAP measurements. Open the “DoubleFit” folder and check for the proper fitting of the FRAP curves (compare to Fig. 3). If there are any improperly fit images, note the file name of the FRAP experiment and exclude it from further FRAP analysis. 10. The Microsoft Excel document named, “Consolidated” will contain processed FRAP intensity values (corrected for acquisition photobleach and background signals) for individual FRAP measurements. Average those values to get the final FRAP recovery curve. 11. Another Microsoft Excel document named “FitResults” will contain the dynamics data, such as immobile fraction, half-time to recover, diffusion constant values for both single and double component fits. 12. Continue FRAP analysis in a Microsoft Excel sheet containing “Two component fit” values for FRAP studies involve two components for fluorescence recovery, such as transcription

Protocol for Quantifying Transcriptional Activity by FRAP

135

Table 2 Abbreviations in the “FitResults” Microsoft Excel sheet are explained (scripts available upon request) Abbreviation Data type y0

Fluorescence intensity value at the “Bleaching ROI” of the first photobleach image

A1

Amplitude of the fast diffusing population

A2

Amplitude of the slow diffusing population

I0

Fluorescence intensity value of the first image of the FRAP image stack at the “Bleaching ROI”

tau1

Time constant of fast diffusing population (A1)

tau2

Time constant of slow diffusing population (A2)

thalf1

Half-time to recover for fast diffusing population calculated from “tau1”

thalf2

Half-time to recover for slow diffusing population calculated from “tau2”

IF

Immobile fraction

D11

Effective diffusion constant of fast diffusing population (A1)

D21

Effective diffusion constant of slow diffusing population (A2)

Ratio

A1/A2 ratio

factors. Data name abbreviation in the “FitResults” Microsoft Excel sheet is described in Table 2. 13. Some of the data in the “FitResults” Microsoft Excel sheet are for reference purpose only. Useful dynamics data are thalf1, thalf2, IF, diffusion constants, and ratio of A1/A2. 14. Calculate mean and standard deviation for the useful dynamics data in the Microsoft Excel sheet itself. These data can be used for tabular visualization of the FRAP data. 3.4 Data Visualization and Statistics 3.4.1 Data Visualization

3.4.2 Statistics

FRAP measures a variety of dynamics data at the single cell level resolution. Proper graphical representation of the FRAP data would effectively communicate underlying biological information, such as different aspects of protein dynamics and presence of different cell populations, etc. We use Origin® (93E) software (OriginLab, Northampton, Massachusetts, USA) for making boxplots, interval plots, scatterplots, and statistical analysis. Fluorescence intensity values are plotted against the function of time in the X-axis and shown in the scatterplot (see Fig. 11a). Interval plots or boxplot with individual data points is a good choice to visualize protein dynamics data (see Fig. 11b, c). Our data does not show a normal distribution because of the cellular heterogeneity. We therefore used a Mann-Whitney U-test to calculate statistical significance.

136

Kannan Govindaraj and Janine N. Post

Fig. 11 Visualizing protein dynamics data as measured by FRAP. (a) Scatter plot showing fluorescence recovery curves. (b) Boxplot and (c) interval plot showing immobile fraction data

4

Notes 1. Maintaining physiological pH and osmolarity during imaging is important to keep the cells alive and normal. We used Tyrode’s solution [28] buffered with HEPES during imaging. 2. In case of using media without phenol red as imaging media, either CO2 supply or HEPES buffering is required during imaging. Avoid incubating cells (in well plates) in the imaging buffer in the CO2 incubator. 3. Wash the coverslips with “90% EtOH + 10% glacial acetic acid” solution and rinse with 70% EtOH. Autoclave and keep sterile until use (see Subheading 2.2.1 day 1). 4. We used a silicone gasket with 1.5 mm thickness (cat#: 704652R2, EMS, USA). We did cut the small circular well and made the circle slightly bigger to allow for a larger imaging area (see Subheading 2.2.1, item 1). 5. Coverslips in the 24-well plates can be easily lifted using a 10 μl tip and a tweezer. Practice this mounting step before doing the procedure with cells. This step needs prior practice! Gently press the coverslip with a fibreless tissue paper (such as Kimberly-Clark® Professional, Subheading 2.2.1, item 3). 6. Check if the XY drift occurs with every FRAP measurement; if it occurs frequently, check whether the glass side with coverslip is properly mounted in the stage inlet. Also, if there is a XY drift, mark it in the corresponding Microsoft Excel file containing FRAP data, so that during data validation, corresponding FRAP image stack can be readily picked for image registration (see Subheading 3.2.2). 7. If the “PFS” is turned on, usually the Z drift will not occur (see Subheading 3.2.2).

Protocol for Quantifying Transcriptional Activity by FRAP

137

8. You can create shortcuts for “Template Matching” and “Plot Z-axis profile” options in the Fiji (Plugins ! Shortcuts ! Add Shortcuts ! Select preferred shortcut key and command). This will help to save time during data validation (see Subheading 3.2, step 5).

Acknowledgements We thank Anne K. Kenworthy for kindly providing the MATLAB script to calculate the effective diffusion. Their MATLAB script is partly incorporated in our script. We thank Samantha L. Schwartz and Diane S. Lidke for their FRAP script and useful discussion. We thank Irene Siemerink-Konings for reading this article for the clarity check and suggestions for improvement. References 1. Axelrod D, Koppel DE, Schlessinger J, Elson E, Webb AW (1976) Mobility measurement by analysis of fluorescence photobleaching recovery kinetics. Biophys J 16:1055–1069. https://doi.org/10.1016/ S0006-3495(76)85755-4 2. Tsien RY (1998) The green fluorescent protein. Annu Rev Biochem 67:509–544. https://doi.org/10.1146/annurev.biochem. 67.1.509 3. Fritzsche M, Charras G (2015) Dissecting protein reaction dynamics in living cells by fluorescence recovery after photobleaching. Nat Protoc 10(5):660–680. https://doi.org/10. 1038/nprot.2015.042 4. Sprague BL, McNally JG (2005) FRAP analysis of binding: proper and fitting. Trends Cell Biol 15(2):84–91. https://doi.org/10.1016/j.tcb. 2004.12.001 5. Pan Y, Tsai C-J, Ma B, Nussinov R (2010) Mechanisms of transcription factor selectivity. Trends Genet 26(2):75–83. https://doi.org/ 10.1016/j.tig.2009.12.003 6. Spivakov M (2014) Spurious transcription factor binding: non-functional or genetically redundant? BioEssays 36(8):798–806. https://doi.org/10.1002/bies.201400036 7. Orphanides G, Reinberg D (2002) A unified theory of gene expression. Cell 108 (4):439–451. https://doi.org/10.1016/ S0092-8674(02)00655-4 8. Govindaraj K, Hendriks J, Lidke DS, Karperien M, Post JN (2019) Changes in fluorescence recovery after photobleaching (FRAP) as an indicator of SOX9 transcription

factor activity. Biochimica et Biophysica Acta Gene Regul Mech 1862(1):107–117. https:// doi.org/10.1016/j.bbagrm.2018.11.001 9. Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA, Charpentier E (2012) A programmable dual-RNA–guided DNA endonuclease in adaptive bacterial immunity. Science 337 (6096):816. https://doi.org/10.1126/sci ence.1225829 10. Bi W, Deng JM, Zhang Z, Behringer RR, de Crombrugghe B (1999) Sox9 is required for cartilage formation. Nat Genet 22:85. https:// doi.org/10.1038/8792 11. Rayan G, Guet J-E, Taulier N, Pincet F, Urbach W (2010) Recent applications of fluorescence recovery after photobleaching (FRAP) to membrane bio-macromolecules. Sensors (Basel, Switzerland) 10 (6):5927–5948. https://doi.org/10.3390/ s100605927 12. Blaney Davidson EN, Vitters EL, van Lent PL, van de Loo FA, van den Berg WB, van der Kraan PM (2007) Elevated extracellular matrix production and degradation upon bone morphogenetic protein-2 (BMP-2) stimulation point toward a role for BMP-2 in cartilage repair and remodeling. Arthritis Res Ther 9 (5):R102. https://doi.org/10.1186/ar2305 13. Shaner NC, Patterson GH, Davidson MW (2007) Advances in fluorescent protein technology. J Cell Sci 120(24):4247. https://doi. org/10.1242/jcs.005801 14. Nagy P, Vereb G, Post JN, Friedl€ander E, Szo¨l˝si J (2005) Novel single cell fluorescence loo approaches in the investigation of signaling at the cellular level. In: Damjanovich S

138

Kannan Govindaraj and Janine N. Post

(ed) Biophysical aspects of transmembrane signaling. Springer, Berlin, Heidelberg, pp 33–70. https://doi.org/10.1007/3-540-26511-2_2 15. Zacharias DA, Violin JD, Newton AC, Tsien RY (2002) Partitioning of lipid-modified monomeric gfps into membrane microdomains of live cells. Science 296(5569):913. https:// doi.org/10.1126/science.1068539 16. Shaner NC, Steinbach PA, Tsien RY (2005) A guide to choosing fluorescent proteins. Nat Methods 2:905. https://doi.org/10.1038/ nmeth819 17. Cardarelli F, Tosti L, Serresi M, Beltram F, Bizzarri R (2012) Fluorescent recovery after photobleaching (FRAP) analysis of nuclear export rates identifies intrinsic features of nucleocytoplasmic transport. J Biol Chem 287 (8):5554–5561. https://doi.org/10.1074/ jbc.M111.304899 18. Wu B, Piatkevich KD, Lionnet T, Singer RH, Verkhusha VV (2011) Modern fluorescent proteins and imaging technologies to study gene expression, nuclear localization, and dynamics. Curr Opin Cell Biol 23(3):310–317. https:// doi.org/10.1016/j.ceb.2010.12.004 19. Nissim-Rafinia M, Meshorer E (2011) Photobleaching assays (FRAP & FLIP) to measure chromatin protein dynamics in living embryonic stem cells. J Vis Exp 52:2696. https:// doi.org/10.3791/2696 20. Goldring MB, Birkhead JR, Suen LF, Yamin R, Mizuno S, Glowacki J, Arbiser JL, Apperley JF (1994) Interleukin-1 beta-modulated gene expression in immortalized human chondrocytes. J Clin Invest 94(6):2307–2316. https://doi.org/10.1172/JCI117595 21. Post JN, Lidke KA, Rieger B, Arndt-Jovin DJ (2005) One- and two-photon photoactivation of a paGFP-fusion protein in live Drosophila embryos. FEBS Lett 579(2):325–330. https://doi.org/10.1016/j.febslet.2004.11. 092 22. Siggia ED, Lippincott-Schwartz J, Bekiranov S (2000) Diffusion in inhomogeneous media: theory and simulations applied to whole cell photobleach recovery. Biophys J 79 (4):1761–1770. https://doi.org/10.1016/ S0006-3495(00)76428-9 23. Tseng Q (2011) Template matching and slice alignment - ImageJ plugins. https:// sitesgooglecom/site/qingzongtseng/tem plate-matching-ij-plugin. Accessed Jan 2020 24. Tseng Q, Wang I, Duchemin-Pelletier E, Azioune A, Carpi N, Gao J, Filhol O, Piel M, The´ry M, Balland M (2011) A new micropatterning method of soft substrates reveals that different tumorigenic signals can promote or

reduce cell contraction levels. Lab Chip 11 (13):2231–2240. https://doi.org/10.1039/ C0LC00641F 25. Koulouras G, Panagopoulos A, Rapsomaniki MA, Giakoumakis NN, Taraviras S, Lygerou Z (2018) EasyFRAP-web: a web-based tool for the analysis of fluorescence recovery after photobleaching data. Nucleic Acids Res 46 (W1):W467–W472. https://doi.org/10. 1093/nar/gky508 26. Blumenthal D, Goldstien L, Edidin M, Gheber LA (2015) Universal approach to FRAP analysis of arbitrary bleaching patterns. Sci Rep 5:11655. https://doi.org/10.1038/ srep11655 27. Day CA, Kraft LJ, Kang M, Kenworthy AK (2012) Analysis of protein and lipid dynamics using confocal fluorescence recovery after photobleaching (FRAP). Curr Prot Cytom C HAPTER:Unit2.19-Unit12.19. https://doi. org/10.1002/0471142956.cy0219s62 28. Lidke DS, Nagy P, Heintzmann R, Arndt-Jovin DJ, Post JN, Grecco HE, Jares-Erijman EA, Jovin TM (2004) Quantum dot ligands provide new insights into erbB/HER receptor–mediated signal transduction. Nat Biotechnol 22:198. https://doi.org/10.1038/nbt929 29. Mueller F, Mazza D, Stasevich TJ, McNally JG (2010) FRAP and kinetic modeling in the analysis of nuclear protein dynamics: what do we really know? Curr Opin Cell Biol 22 (3):403–411. https://doi.org/10.1016/j. ceb.2010.03.002 30. Beaudouin J, Mora-Bermu´dez F, Klee T, Daigle N, Ellenberg J (2006) Dissecting the contribution of diffusion and interactions to the mobility of nuclear proteins. Biophys J 90 (6):1878–1879. https://doi.org/10.1529/ biophysj.105.071241 31. Fritzsche M, Lewalle A, Duke T, Kruse K, Charras G (2013) Analysis of turnover dynamics of the submembranous actin cortex. Mol Biol Cell 24(6):757–767. https://doi.org/10. 1091/mbc.E12-06-0485 32. Oh D, Zidovska A, Xu Y, Needleman Daniel J (2011) Development of time-integrated multipoint moment analysis for spatially resolved fluctuation spectroscopy with high time resolution. Biophys J 101(6):1546–1554. https:// doi.org/10.1016/j.bpj.2011.08.013 33. Kim SA, Heinze KG, Schwille P (2007) Fluorescence correlation spectroscopy in living cells. Nat Methods 4:963. https://doi.org/ 10.1038/nmeth1104 34. Herrick-Davis K, Grinde E, Cowan A, Mazurkiewicz JE (2013) Fluorescence correlation spectroscopy analysis of serotonin, adrenergic,

Protocol for Quantifying Transcriptional Activity by FRAP muscarinic, and dopamine receptor dimerization: the oligomer number puzzle. Mol Pharmacol 84(4):630–642. https://doi.org/10. 1124/mol.113.087072 35. Mendoza MC, Besson S, Danuser G (2012) Quantitative fluorescent speckle microscopy (QFSM) to measure actin dynamics. Curr Protoc Cytom Chapter 2:Unit2.18. https://doi. org/10.1002/0471142956.cy0218s62 36. Watanabe N (2010) Inside view of cell locomotion through single-molecule: fast F/G-actin

139

cycle and G-actin regulation of polymer restoration. Proc Jpn Acad Ser B Phys Biol Sci 86 (1):62–83. https://doi.org/10.2183/pjab. 86.62 37. Watanabe N, Mitchison TJ (2002) Singlemolecule speckle analysis of actin filament turnover in lamellipodia. Science 295 (5557):1083–1086. https://doi.org/10. 1126/science.1067470

Chapter 10 Quantitative Molecular Models for Biological Processes: Modeling of Signal Transduction Networks with ANIMO Sakshi Khurana, Janet Huisman, Stefano Schivo, and Janine N. Post Abstract Computational modeling of biological networks is increasing in popularity due to the increased demand for understanding biological processes. This understanding requires integration of a variety of experimental data that allows understanding of complex mechanisms regulating cell and tissue function. However, the mathematical complexity of many modeling tools have thusfar prevented broad adaptation and effective use by molecular biologists. In this chapter, we show by example how one can start building a model in ANIMO and how to adapt the model to experimental data. We show how this model can be used for simulating network activities, testing hypotheses, and how to improve the model using wet-lab data. Key words Modeling tool, Computational model, Signal transduction network, Signaling pathways, Signaling cross talk, WNT, NFkB, TGF beta

1

Introduction

1.1 The Need of Modeling for Biological Networks

Over the past decades, advancement in high-throughput technologies have led to a rapid accumulation of genomic and proteomic data. It has provided a great wealth of information by studying the gene regulation and gene-protein interactions in health, as well as in pathological states. However, not all data are utilized in an efficient manner, and changes in cellular signaling are often incompletely explored. This necessitates a shift from the central dogma of “hypothesis to experiment to models” toward “big data to models to hypothesis to new experiment to additional data and more inclusive models” [1]. One effective way of utilizing information obtained from genomic and proteomic data is by building computational models. These models can be an option for storing large and complex data as well as for hypothesis generation. Models can be used as a tool to combine data from existing literature. For example, data obtained on a protein studied in context of one disease can be used to build a model. This model can then be used to predict the mechanism of this protein in the

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_10, © Springer Science+Business Media, LLC, part of Springer Nature 2021

141

142

Sakshi Khurana et al.

context of another disease by adding tissue-specific proteins and genes. Later, these predictions can be confirmed by wet lab experiments, thereby validating the model predictions and obtaining important information about the molecular mechanism involved in the latter tissue. If similar experiments were performed by wet lab experiments only, then one would have to test many concentrations and time points for each molecule of interest, leading to longer time investment to understand the mechanism. Recently, our group presented a model that was built to understand the signaling cross talk between the wingless related integration site (WNT) and Interleukin 1 beta (IL1β) signaling pathways in osteoarthritis. The model was based on a study conducted on cancer cells, showing that IL1β induced nitric oxide (NO) production, which upregulated WNT signaling [2]. Our model predicted that these two pathways are linked in chondrocytes via induced nitric oxide synthase (iNOS), which was confirmed by wet lab experiments in human chondrocytes [3]. We later used this knowledge to incorporate the IL1β into a larger network of 7 signal transduction pathways that regulate cartilage formation and homeostasis. In that large network, which we called ECHO (for Executable CHOndrocyte), we show that healthy cartilage is partially protected against the effects of IL1β exposure due to the presence of the WNT antagonists DKK1 and FRZB [4]. This way computational models can be used to find novel cross talk between pathways in a relatively short time span. In addition to storage of data or as predictive tool, models can be used as an alternative for the testing of drugs in animals. Preliminary attempts are already made to limit the animal testing by developing software that can predict the outcome of various drug screening assays. In this direction, researchers have developed software to test drug compounds and found 87% accuracy with the animal testing data [5]. Environmental protection agencies have announced that any funding to mammalian testing will be stopped by 2035 [6, 7], hence necessitating a step toward more software developments as an alternative for animal testing. 1.2

Building a Model

Computational models can graphically represent biological networks based on information obtained from literature, in combination with a formalization of the network in terms of activity. In contrast to small networks, for large and complex networks it is impossible to add all the components of the network. For complex networks, it would be ideal to include at least the most important molecules. However, it is not always easy to figure out which molecules play key roles in the functioning of the network. Many biological events are, in effect, changes in activity, comparable to switches / dimmers. For example, changes in concentration, posttranslational modifications (i.e., phosphorylation, methylation) or localization of a protein, or changes in gene expression are causal

Modeling with ANIMO

143

factors for downstream effects. Therefore, the state or concentration of molecules can be described in terms of activity. For many molecules the actual biological activity is not measured, especially not in the context of the cell. Therefore, the more active a molecule, the more it will affect regulation of downstream factors. To gain insight into important events led by the molecular interactions, the following guidelines should be considered: 1. Choose to measure a molecule that shows differential downstream effects for the modeling network activity, i.e., that can be used as an output for validating and / or optimization of the network model. 2. Include overlapping treatments to normalize experimental data between days or assay batches. 3. Include a positive control for each of the measurements to get an indication of the potential maximum activity of the molecules in the network. This allows scaling of network activities between 0 and 100% to construct a nondimensional model. This way, one can omit the need for precise intracellular concentrations. 4. Include a negative control (t ¼ 0) for background activity levels. 5. To study the dynamics of network in an efficient way, it is important to include multiple time-points. Single time-point measurements provide poor insight into the dynamic behavior of a system. Consider the following factors to decide how many time-points should be measured. Ideally, measurements are taken from time 0 until the system reaches a steady state. When peak dynamics are expected, 3 time points are the absolute minimum: one before the peak, one as close as possible to the peak, and one after the peak. Time range for a biological network depends on the direct effects or indirect effects of upstream molecules. In case of signal transduction, direct or primary effects take 240/480 min whereas indirect effects might take 24–48 h. High order indirect effects might even take weeks or months [8]. These higher order effects may be, for example, effects regulating gene expression of genes that encode growth factors and cytokines, whose expression will be new input for the network. 1.3 Precision of Models

Generally, the aim of models is to replicate the behavior of biological systems based on the known properties of individual components [9]. The precision of a model, i.e., the detail in which the biological network is represented in terms of number of molecules and interactions, depends on the availability of experimental data, the need for detailed information, and available computational tools. Models suffer less from experimental issues

144

Sakshi Khurana et al.

like noise and resolution, thus providing more precise insights at the cellular and molecular level. Specifically, models based on ordinary differential equations, which are more parameter intensive, represent biological interactions more precisely [10]. Whereas modeling tools, such as models based on logic descriptions (Boolean, Fuzzy) as well as the modeling tool we will use in this chapter, ANIMO, suffer more with precision errors as they are based on single parameter k-values [11]. 1.4 Comparison of some Modeling Formalisms and Tools

Various computational tools are available for building models that define the dynamics of biological networks. In this part, we describe the formalism that is underlying some of the most-used modeling tools. Tools based on formalisms described by Boolean and fuzzy logic models are often used. Tools that belong to this category, such as CytoCopteR [12], GINsim [13], and BooleSim [14], are based mainly on discrete transitions and can be used to model interactions among a large variety of proteins. These tools make it easier to perform model building and model validation, and can be used for predictions. On the other side of the tool-spectrum, ordinary differential equations (ODEs) make a purely continuous model and can efficiently represent biochemical reactions for large and complex networks where mass-action approximations are appropriate [10, 11, 15]. Examples of tools using ODEs are Odefy [16], COPASI [17], and CellDesigner [18]. An actively growing field for computing graphical representations of biological networks through literature or high-throughput data has been gaining interest in the last few years [10]. In this field, biological networks are typically represented with the help of graphs, where each component, e.g., gene or protein, is represented by nodes and potential interactions by edges. Edges can typically represent a wide range of stimulatory or inhibitory interaction modes from direct physical binding to correlated gene expression or phosphorylation by kinase, etc. For the purpose of making computational modeling of dynamic biological networks available to researchers without extensive mathematical training, we previously developed ANIMO: Analysis of Networks with Interactive Modeling [8, 11, 19]. ANIMO is a tool that helps a biological researcher to build and analyze a protein interaction network, providing a visual representation of the activity profile over time. In order to formalize and correctly analyze the models, ANIMO networks are based on a Timed Automata model, that can be described as piecewise linear approximation of ODE models. More information about ANIMO and the comparison with other tools, as well as example signal transduction models, can be found in Schivo et al. [11]. In this chapter, we want to give a working example of how to build a computational model. In this model one can add data and test hypotheses. Wet-lab experiments can be used to validate and improve the model, as also described in this chapter.

Modeling with ANIMO

2

145

Materials

2.1 Computational Materials

To run ANIMO, a desktop or laptop computer is needed with the following software installed: 1. Java (java.com). 2. UPPAAL (www.uppaal.org) (see Note 1). 3. Cytoscape (www.cytoscape.org) (see Note 2). An extensive user manual for installation and for creating a network can be found here: http://fmt.cs.utwente.nl/tools/ animo/ The software to run Java-based programs is provided for free by Oracle. More information is available on the java.com website. Cytoscape is an open source project released under the terms of the GNU Lesser General Public License. UPPAAL is developed by a collaboration by the universities of Uppsala (Sweden) and Aalborg (Denmark), and is free for noncommercial applications in academia only. All software works under Windows, Mac, and the most common GNU/Linux OS (see Note 3). Models used in this chapter, as well as raw data files, can be found online via: https://www.utwente.nl/en/tnw/dbe/research/additionalmaterials/

2.2 Wet-Lab Materials

1. 96-Well plates (BD biosciences, with white walls and a clear bottom such that they are suitable for cell culture and allow for luciferase activity measurements). 2. Cells that are being studied (C2C12, mouse myoblasts, ATCC ® CRL-1772™). 3. Cell culture media (DMEM), Thermo Fisher. 4. Fetal bovine serum (FBS), Thermo Fisher. 5. Penicillin-Streptomycin (Pen/Strep), Thermo Fisher. 6. Phosphate-buffered saline (PBS), Thermo Fisher. 7. Trypsin, Thermo Fisher. 8. Lipofectamine™ 2000 Transfection Reagent, Thermo Fisher Scientific. 9. Reporter DNA (16xTOPFlash (TCF Reporter Plasmid 21-170, Sigma Aldrich). 10. Renilla DNA (pRL-CMV, Promega,for normalization of transfection efficiency). 11. Opti-MEM reduced serum medium (ThermoFisher). 12. Lipofectamine LTX (ThermoFisher).

146

Sakshi Khurana et al.

13. Stimulation factors TGF-β3 (R&D systems, 8420-B3-005), BIO (Sigma, B1686-5MG), and IL-1β (BioLegend, 575102). 14. Dual-Glo® kit (Promega, E2940). 15. Varioskan™ LUX (ThermoFisher).

multimode

microplate

reader

16. Nuclease-free 1.5 mL Eppendorf tubes.

3

Methods

3.1 Preliminary Model and Hypotheses

1. Determine the pathways that need to be studied. In this example, the cross talk between the canonical WNT, NFκB, and TGF-β pathways will be determined. These pathways play an important role in various cellular processes, such as cell survival, differentiation, and growth, but their cross talk is not yet fully understood. The most important molecules and interactions in these pathways are shown in Fig. 1 [20–22]. 2. In this example, the model building was initiated by placing nodes in ANIMO for the compounds that can be used to stimulate each of these pathways (see Fig. 2a). The shape of the nodes corresponds to the type of molecule they represent (cytokine, receptor, transcription factor, gene, etc.) while the

Fig. 1 Cross talk between the canonical WNT, NFκB, and TGF-β pathways as found in literature [20–22]. These pathways play an important role in various cellular processes, such as cell survival, differentiation, and growth

Modeling with ANIMO

147

color indicates the activity of each node. A red color represents a low activity and a green node suggests high activity of the molecule. Literature research is necessary to determine the most important ligands for the network as seen in Fig. 1. In general, one would start modeling the canonical pathways, unless it is already described that noncanonical pathways are important for cross talk between pathways. WNT is the general ligand to indicate all WNT ligands that stimulate the canonical β-catenin pathway, e.g., WNT3A, WNT5b, and WNT10a [23]. BIO [24], 6-bromoindirubin-30 -oxime, is a potent, reversible, and ATP-competitive GSK-3α/β inhibitor, and is used to stimulate the Wnt pathway (see Note 4). Interleukin 1 beta (IL-1β) activates the NFκB pathway [25] and TGF-β3 stimulates the TGF-β pathway [26] (see Note 5). 3. Next, a literature search will help to determine which downstream transcription factors and genes are up- or downregulated when each of these compounds is added to the cell. This will describe each general pathway (Wnt, NFκB, and TGF-β) without its cross talk. For example, it was found that for the Wnt pathway, the transcription cofactor β-catenin binds to the transcription factors TCF-LEF [27] to initiate transcription of metalloproteinases (MMPs 2, 7 and 9 [28]) and that TGF-β3 binds to a receptor of the TGF-β superfamily (Activin receptorlike kinases 1–7, ALK [29]), which promotes the formation of the Smad2/3/4 complex, which in turn increases transcription of the genes with a SMAD Binding Element (SBE) in the promotor [30–32]. Examples of these genes are TGF-β-induced protein (TGFBIP), transmembrane prostate androgen induced RNA (TMEPA), and matrix metalloprotease 2 (MMP2) [33]. 4. Add the transcription factors and genes to the model generated with the ANIMO software (see Fig. 2b). Add edges to describe the direction of the interaction, e.g., activation or inhibition, and include dynamics as found in literature (see Notes 6–8) [8]. 5. More extensive literature research is then used to investigate the cross talk between each of the pathways. Additional nodes can be added if necessary and nodes that overcomplicate the model, for example, because they only interact with one other protein in the same pathway, can be removed if this does not interfere with the dynamics of the downstream nodes. Interactions can be added in the ANIMO model as edges (see Fig. 2c). In our example, it was found that NFκB inhibits β-catenin [34, 35] and that Smad7 inhibits NFκB, β-catenin and the TGF-β receptors 1 (TGFβR1, ALK5) and 2 (TGFβR2) [36–39].

148

Sakshi Khurana et al.

Fig. 2 Flow of generating an ANIMO model to simulate pathway interactions. (a) Start with compounds that can be used to stimulate each of these pathways. (b) Add transcription factors and genes that are up- or downregulated when each of these compounds is added to the cell. (c) Add interactions that describe the cross talk between the studied pathways

6. After completing the model with all nodes and edges, a simulation can be run to generate hypotheses (see Fig. 3). In our example, we are interested in the influence of the stimuli on the activation of TCF/LEF and subsequent target gene expression. Depending on the chosen scenario and reaction constants, different hypotheses will be generated (see Note 9). The rate constants can, at this stage, be manipulated until the results roughly correspond to what is described in literature. However, it is recommended to do the fine-tuning of the model after the wet-lab experiments.

Modeling with ANIMO

149

Fig. 3 Hypotheses of the activation of the TCF/LEF luciferase reporter with different stimulation factors generated using the ANIMO model as shown in Fig. 2c

7. In Fig. 3, the hypotheses of the activation of a TCF/LEF luciferase reporter are shown with different stimulation factors. It was chosen to focus on TCF/LEF since this transcription factor is found downstream of the canonical WNT pathway and its expression is dependent on the cross talk between all three modeled pathways, as is shown in Fig. 2c. The activity of TCF/LEF is depicted with colors, where red indicates a low activity and green suggests high activity of the transcription factor. As can be seen in Fig. 3, the model predicts that TGF-β3 by itself will not activate TCF/LEF, while WNT will strongly activate TCF/LEF. IL-1β will only partially activate TCF/LEF due to its direct activation of TCF/LEF while at the same time inhibiting the β-catenin production. A combination of TGF-β3 and WNT is predicted to fully activate TCF/LEF, indicating that TGF-β3 does not inhibit WNT in activating TCF/LEF. Stimulating with both TGF-β3 and IL-1β is not sufficient to activate TCF/LEF and adding both WNT and IL-1β only partially activates TCF/LEF. Lastly, it is expected that a stimulation with all three stimuli (TGF-β3, WNT, and IL-1β) will lead to activation of TCF/LEF. 3.2 Wet-Lab Experiments

Model predictions can be validated either by looking at existing literature that was not used for model-building, or by performing wet-lab experiments. We chose the latter, since we want to validate both the model topology, i.e., the cross talk between the pathways of interest, and the model parameters. 1. Many different techniques can be used to determine the up- or downregulation of gene transcription and protein translation. To be able to get an indication of the TCF/LEF expression and validate the hypotheses as shown in Fig. 3, a reporter assay can be used. This assay measures the activity of a specific gene promoter by transfection of a vector containing the promotor of our gene of interest upstream of a reporter gene, such as

150

Sakshi Khurana et al.

Table 1 TCF/LEF-luciferase expression in C2C12 mouse myoblasts for different stimulation factors (10 ng/mL TGF-β3, 1 μM BIO, and 10 ng/mL IL-1β) at different time points (17 h, 22 h, 41 h, and 48 h after stimulation). N ¼ 4 17 hours

22 hours

41 hours

48 hours

Condition

Mean

Std

Mean

Std

Mean

Std

Mean

Std

1

Control

1.000

1.186

1.000

0.956

1.000

0.952

1.000

0.922

2

TGF-β3

1.620

1.584

1.538

1.455

0.846

0.760

1.309

1.032

3

BIO

3.320

3.367

0.529

0.445

0.380

0.360

0.557

0.515

4

IL-1β

1.122

1.174

0.875

0.749

0.532

0.556

0.909

0.768

5

TGF-β3 + BIO

5.007

4.623

1.108

0.789

0.865

1.027

1.345

1.136

6

TGF-β3 + IL-1β

1.898

2.337

1.322

0.995

1.154

1.079

1.749

1.456

7

BIO + IL-1β

4.766

5.161

0.871

0.711

0.414

0.393

0.871

0.943

Fig. 4 Luciferase expression under control of the TCF/LEF promoter in C2C12 mouse myoblasts for different stimulation factors (TGF-β3, BIO, and IL-1β) at different time points (17 h, 22 h, 41 h, and 48 h after stimulation). N ¼ 4

luciferase, green fluorescent protein, or β-galactosidase. In our example, the expression of TCF-LEF is studied by a reporter assay using bioluminescent luciferase (see Note 10). 2. For the reporter assay, 96-well plates are seeded with C2C12 mouse myoblasts at a density of 20,000 cells per well in a

Modeling with ANIMO

151

volume of 100 μL cell culture medium (DMEM supplemented with 1% v/v FBS). C2C12 cells are a cell line that can differentiate into myoblasts and myocytes (muscle cells) (see Note 11). For each time point a separate plate was used and the outer wells are filled with sterile PBS to reduce evaporation of medium. 3. For the transfection, lipofectamine (Lipofectamine™ LTX Transfection Reagent, Thermo Fisher Scientific) is used with some modifications, see Note 4. 4. For each well, 0.1 μg of reporter DNA (16xTOPflash and 0.005 μg of renilla DNA (pRL-CMV) is combined with 0.1 μL PLUS reagent and diluted into 5 μL of Opti-MEM in a 1.5 mL tube. In a separate tube, for each well, 0.305 μL of Lipofectamine is diluted into 5 μL of Opti-MEM. 5. The diluted DNA solution is transferred to the diluted Lipofectamine solution, mixed well by pipetting, and incubated for 5 min at room temperature to form DNA-Lipofectamine complexes. 6. 10 μL of DNA-Lipofectamine complexes is added to each well. The cells are then incubated overnight at 37  C and 5% CO2. Is this without adding DMEM+FBS after 3–6 h? 7. The next day, medium is replaced by fresh culture medium (DMEM +1 v/v% FBS (see Note 12) + 1 v/v% Pen/Strep) containing the stimulation factors (10 ng/mL TGF-β3, 1 μM BIO, 10 ng/mL IL-1β or a combination of these, 75 μL per well and in quintuplicate) and incubated at 37  C and 5% CO2 (see Note 13). 8. At four different time points (17, 22, 41, and 48 h after stimulation), the reporter assay is performed. At each time point, Dual-Glo® Luciferase Reagent (75 μL per well) is added to the culture medium and incubated for 10 min. 9. After incubation, TCF/LEF-luciferase luminescence is measured using a VarioSkan LUX. Emission firefly luciferase measured at 550–570 nm. 10. Next, Dual-Glo® Stop & Glo® Reagent (75 μL per well) is added to the culture medium and incubated for 10 min. 11. After incubation, renilla-luciferase luminescence is measured at 480 nm using the Varioskan LUX (ThermoFisher). 12. Repeat steps 7–11 for each different time point. 13. Analyze the obtained data (see Note 14). Use the renillaluciferase data to normalize for differences in transfection efficiency. Next, average the values for identical conditions and calculate the standard deviation (see Table 1 and Fig. 4).

152

Sakshi Khurana et al.

14. In Fig. 4, the normalized luciferase expression under the control of the TCF/LEF promoter in C2C12 mouse myoblasts can be seen. It is clearly visible that for all conditions, the expression is highest at the first time point (17 h after stimulation) and decreases at subsequent time points. It is expected that this observed peak-behavior is due to the fairly short halflife of the luciferase reporter. If BIO was used to stimulate the cells, either by itself or in combination with other stimuli, an increase in expression was observed compared to the control. For TGF-β3 and IL-1β, no significant increase was seen in the luciferase expression. It is unknown whether the expression 17 h after stimulation is maximum or if it had been even higher at earlier time points. Additional wet-lab experiments at earlier time points will have to be conducted to obtain this data. 3.3 Validation and Adjustment of Model

1. Identify the differences between the wet-lab results and the hypotheses generated with the preliminary model. These can be differences in expression (e.g., the wet-lab results show an increase in expression for a specific gene while this gene is downregulated in the preliminary model) or in time scale (see Note 15). In our example, the expression levels correspond fairly well with the hypotheses, except that peak behavior is seen in the wet-lab results (see Fig. 4, expression is downregulated after 17 h for all conditions), which was not yet included in the preliminary model. 2. Add missing nodes and/or connections to the model. For example, adding a specific node might explain certain behavior as observed in the wet-lab experiments or a specific inhibition dynamic can solve problems in the simulated expression. This completely depends on the specific model and might not always be necessary. An additional literature search might help identifying missing nodes or connections. 3. Incorporate peak behavior in the model if this is not already done. In our example, peak behavior was observed in the experimental results, but was not yet incorporated in the model. This peak behavior can be explained by the fact that luciferase fluorescent intensity is measured, which has a fairly short half-life. By adding auto-inactivation/ inhibitory loops to each node, this peak behavior could be added to the model to better fit the experimental data, as described before [8]. In Fig. 5, the optimized model is shown for our example, which is based on Fig. 2c and improved using wet-lab data. 4. Optimize the rate parameters k to fit the wet-lab data. This can be done manually by changing each and every rate parameter until the simulations match the data from the wet-lab experiment. However, for larger models this can get very time consuming and complicated. To make this process a little easier,

Modeling with ANIMO

153

Fig. 5 Optimized ANIMO model based on the model as shown in Fig. 2c. Wet-lab results are used to improve the model. Auto-inactivation/inhibitory loops were added to each node to incorporate peak behavior in the model

the function “optimize k-value(s)” in ANIMO can be used (see Note 16). For our example, results after optimization can be found in Fig. 6. It is also important to remember that the network topology (i.e., the nodes currently included in the model and their interactions) should have priority on parameter values: if a network model fits a data set only with a very specific parameter choice, it is unlikely that the current topology of the model is a useful representation of the biological process. This is because ANIMO interactions are a relatively rougher abstraction of biochemistry than the model network’s topology is. 5. Use more data to optimize the model even further. This data can be obtained by performing more wet-lab experiments (for example, by measuring more time points or using another technique such as Western Blotting or qPCR to obtain data on other nodes in the model) or by using data from other research groups or papers. 6. In Fig. 6, a comparison can be seen between the wet-lab experimental data (in blue) and the optimized ANIMO model (in red). The general behavior of the model corresponds with the wet-lab data for all tested combinations of stimuli. However, to be able to improve the model even further, more wet-lab data will be necessary. Furthermore, these results only focus on the cross talk of TGF-β3 and IL-1β on the WNT signaling pathway. To investigate the other cross talks that are

154

Sakshi Khurana et al.

Fig. 6 Results of optimization of rate parameters k to obtain a model behavior (red) that corresponds with the obtained wet-lab data (blue) for various stimuli

incorporated in the model, such as the effect on WNT and IL-1β on the TGF-β pathway, a new wet-lab experiment will need to be setup where, for example, an SBE-luciferase reporter is used instead of a TCF/LEF-luciferase reporter. With the current model, it is already possible to generate hypotheses for this experiment. The cross talk between WNT and the TGF-β pathway is especially interesting, since TGF-β induces cartilage differentiation while WNT induces bone formation. By better understanding the cross talk between these pathways, as can be modeled with ANIMO, it is possible to better control and direct the differentiation into either bone or cartilage, which is of great importance in the field of tissue engineering.

Modeling with ANIMO

4

155

Notes 1. For UPPAAL installation: Unzip the downloaded file to a known location: UPPAAL will be installed there. Complete U PPAAL installation. For macOS: – Open the UPPAAL installation location in Finder. – Drop the UPPAAL.App icon in your Applications folder. – Copy the verifyta executable file to a known location. The installation of UPPAAL is complete. For windowsOS / LinuxOS: – Open a console, type \cd PATH_TO_THE_UPPAAL_DI RECTORY” and press Return; PATH_TO_THE_UPPAA L_DIRECTORY is the path to the directory where you installed UPPAAL. – Type \java -jar uppaal.jar “and press Return. – The license for UPPAAL will be automatically acquired, and the main window of UPPAAL user interface will appear: you may now close that window. 2. Installing ANIMO: ANIMO is free only for academic use. For commercial licenses, please contact us at [email protected]. – Run Cytoscape. – Click the menu command Apps - > App Manager: the App Manager window will open. – Click on ANIMO in the list of available apps and then on the Install button. ANIMO will be automatically downloaded and installed. You can close the App Manager window. – ANIMO will automatically try to locate the verifyta executable included in UPPAAL, which is needed to verify Timed Automata models. If you know the location where you installed UPPAAL, you can just stop the process (press on the small X on the right), and indicate the location of the verifyta executable. You can find it where it was copied in the bin (bin-Linux, bin-Win32,. .. depending on your operating system) directory inside the UPPAAL installation directory, where it was unzipped. ANIMO is correctly installed and ready to be used. 3. In case of problems accessing a web site with Microsoft Internet Explorer, we advise to try with a different web browser (such as Mozilla Firefox or Google Chrome) or to update Internet Explorer.

156

Sakshi Khurana et al.

4. BIO is used as a stimulus for the activation of the WNT pathway. It acts by inhibiting glycogen synthase kinase 3β (GSK3β) such that GSK3β cannot phosphorylate β-catenin. Unphosphorylated β-catenin is not degraded and will therefore accumulate in the cell, where it can interact with TCF/LEF transcription factors to initiate transcription of target genes [40]. 5. For each node that is added in ANIMO, the following parameters need to be selected: a name to describe the compound, a type (e.g., kinase, receptor, transcription factor, gene), the maximum activity (usually kept at 100) and the initial activity (usually 0 except for the stimulating compounds, which are either 100 (added to the cell) or 0 (not added). The option “plotted” needs to be selected to be able to run a simulation for that node. 6. Try to make the model as simple as possible in this stage. Only include the most important transcription factor(s) and focus on one or two genes that are of interest to study. 7. For each edge that is added in ANIMO, the following parameters need to be selected: the effect of the edge (activation or inhibition), a scenario for the reaction kinetics and a reaction constant k. The description field can be used to include literature that supports each specific interaction. 8. In this example, the following k-values were chosen for the initial model (see Fig. 2c)

From node

To node

kInfluence Scenario value

TGF-β3

Type 1/2 receptor

Activation 1

0.004

Type 1/2 receptor

Smad2/3

Activation 1

0.004

Smad2/3

Smad2/3/4

Activation 1

0.004

Smad2/3/4

SBE transcription factor

Activation 1

0.002

SBE transcription factor

SBE luciferase

Activation 1

0.001

WNT

β-catenin

Activation 1

0.004

β-catenin

TCF/LEF Activation 1 transcription factor

0.002

TCF/LEF TCF/LEF luciferase transcription factor

Activation 1

0.001

TCF/LEF luciferase

Activation 1

0.001

CCND1 gene

(continued)

Modeling with ANIMO

157

From node

To node

kInfluence Scenario value

IL-1β

NFκB

Activation 1

0.004

NFκB

NFκBia gene

Activation 1

0.001

Smad2/3

Smad7

Activation 1

0.004

Smad2/3/4

β-catenin

Activation 1

0.004

Smad7

NFκB

Inhibition 1

0.004

Smad7

β-catenin

Inhibition 1

0.004

Smad7

Type 1/2 receptor

Inhibition 2

0.004

β-catenin

Type 1/2 receptor

Inhibition 1

0.004

NFκB

TCF/LEF Activation 1 transcription factor

0.002

NFκB

Smad7

Activation 1

0.004

NFκB

CCND1 gene

Activation 1

0.001

NFκB

β-catenin

Inhibition 1

0.004

9. The choice for the scenario and reaction constant k have a large effect on the results of the simulation. For the initial model, most edges are set to scenario 1 (unless described differently in literature). For the reaction constant it generally holds true that complex steps such as transcription of a gene are slow and thus have a low reaction constant k, while other processes such as the binding of a stimulation factor to a receptor or a transcription factor to a gene are generally fast and have a high reaction constant k. For each edge, an initial approximation of the reaction constant k needs to be made, if possible based on literature, which can be optimized using results from wet-lab experiments [8]. 10. It is important to note that measuring a reporter is not identical to measuring the protein directly. In a reporter assay, the measured signal is a result of a transcription factor (in this case TCF/LEF) binding to a promotor region in a gene, resulting in transcription of a protein bound to a reporter such as luciferase. The luciferase then produces a fluorescent signal after addition of a substrate, of which the intensity can then be measured. Since this process includes many additional steps that sometimes have a limited life-time, the measured response can differ from the result that would be obtained when directly measuring the transcription factor TCF/LEF. 11. C2C12 mouse myoblasts is an immortalized mouse myoblast cell line that undergoes rapid proliferation and can be differentiated into myoblasts and osteoblasts. They are a popular cell

158

Sakshi Khurana et al.

line for studying biochemical pathways. They can be easily transfected and can be stimulated with a variety of cytokines and growth factors. 12. 1% Serum is used to starve the cells of signals and reduce background signal of pathway activity, so that the effect of the addition of stimuli is clearly visible. 13. Each well plate should contain the same conditions, such that they can be compared at different time points. Each plate should also contain 5 replicates of each condition to be able to calculate the average signal and thereby reduce errors. For the stimuli used in this example concentrations of 10 ng/mL TGF-β3, 1 μM BIO, and 10 ng/mL IL-1β were added to the cells. 14. TCF/LEF-luciferase luminescence and Renilla-luciferase luminescence data was obtained for each well at four different time points. Per well, the TCF/LEF-luciferase data is normalized by dividing the TCF/LEF-luciferase luminescence data by the Renilla-luciferase luminescence data. Next, the mean luminescence and standard deviation is calculated for each condition at each time point. A subsequent normalization is done by dividing the mean value for each condition by the mean value of the control, in which no stimuli were added to the cells. Finally, the data is plotted for each condition at each time point, as can be seen in Fig. 4. 15. This is mainly a problem when peak behavior is observed, where a certain protein or gene is initially upregulated after addition of a stimulus, but is later inhibited by another protein that is formed further down the pathway. 16. Before the function “optimize k-value(s)” in ANIMO can be used, it is essential to already have obtained a graph shape that is similar to the graph shape found in the wet-lab experiments. The function “optimize k-value(s)” is only able to change the k-values, meaning that the graph can be stretched horizontally and vertically. However, the shape of the graph will never change, e.g., if no peak is visible in the model before optimizing the k-values, a peak will never appear after optimization and nodes and/or connections need to be changed, added or removed to the model before optimization. In this example, peak behavior was added to the model by adding an additional inhibitory node to each node that was already present. The node that represents a specific protein or gene activates this inhibitory node with a specific k-value, while the inhibitory node inhibits the same protein or gene with a slightly higher k-value. In this way, an inhibitory loop is included in the model and peak behavior can be included to obtain a graph shape that is similar to the one found in the wet-lab experiments.

Modeling with ANIMO

159

References 1. Liang Y, Kelemen A (2017) Dynamic modeling and network approaches for omics time course data: overview of computational approaches and applications. Brief Bioinform 19(5):1051–1068. https://doi.org/10.1093/ bib/bbx036 2. Du Q, Zhang X, Liu Q, Zhang X, Bartels CE, Geller DA (2013) Nitric oxide production upregulates Wnt/β-catenin signaling by inhibiting Dickkopf-1. Cancer Res 73 (21):6526–6537. https://doi.org/10.1158/ 0008-5472.CAN-13-1620 3. Zhong L, Schivo S, Huang X, Leijten J, Karperien M, Post JN (2017) Nitric oxide mediates crosstalk between interleukin 1β and WNT signaling in primary human chondrocytes by reducing DKK1 and FRZB expression. Int J Mol Sci 18(11):2491. https://doi.org/ 10.3390/ijms18112491 4. Schivo S, Khurana S, Govindaraj K, Scholma J, Kerkhofs J, Zhong L, Huang X, van de Pol J, Langerak R, van Wijnen AJ, Geris L, Karperien M, Post JN (2019) ECHO, the executable CHOndrocyte: a computational model to study articular chondrocytes in health and disease. Cell Signal 68:109471. https://doi. org/10.1016/j.cellsig.2019.109471 5. Luechtefeld T, Marsh D, Rowlands C, Hartung T (2018) Machine learning of toxicological big data enables read-across structure activity relationships (RASAR) outperforming animal test reproducibility. Toxicol Sci 165 (1):198–212. https://doi.org/10.1093/ toxsci/kfy152 6. Grimm D (2019) U.S. EPA to eliminate all mammal testing by 2035. https://www. sciencemag.org/news/2019/09/us-epa-elimi nate-all-mammal-testing-2035 7. Grimm D (2019) EPA plan to end animal testing splits scientists. Science 365(6459):1231. https://doi.org/10.1126/science.365.6459. 1231 8. Scholma J, Schivo S, Urquidi RA, Pol JVD, Karperien M, Post JN (2014) Biological networks 101 : Computational modeling for molecular biologists. Gene 533(1):379–384. https://doi.org/10.1016/j.gene.2013.10. 010 9. Brodland GW (2015) How computational models can help unlock biological systems. In: Seminars in cell & developmental biology. Elsevier, pp 62–73. https://doi.org/10. 1016/j.semcdb.2015.07.001 10. Morris MK, Saez-Rodriguez J, Sorger PK, Lauffenburger DA (2010) Logic-based models

for the analysis of cell signaling networks. Biochemistry 49(15):3216–3224. https://doi. org/10.1021/bi902202q 11. Schivo S, Scholma J, van der Vet PE, Karperien M, Post JN, van de Pol J, Langerak R (2016) Modelling with ANIMO: between fuzzy logic and differential equations. BMC Syst Biol 10(1):56. https://doi.org/10. 1186/s12918-016-0286-z 12. Terfve C, Cokelaer T, Henriques D, MacNamara A, Goncalves E, Morris MK, Mv I, Lauffenburger DA, Saez-Rodriguez J (2012) CellNOptR: a flexible toolkit to train protein signaling networks to data using multiple logic formalisms. BMC Syst Biol 6(1):133. https://doi.org/10.1186/1752-0509-6-133 13. Chaouiya C, Remy E, Mosse´ B, Thieffry D (2004) Qualitative analysis of regulatory graphs: a computational tool based on a discrete formal framework. In: Benvenuti L, De Santis A, Farina L (eds) Positive systems. Lecture Notes in Control and Information Science, vol 294. Springer, Berlin. https://doi. org/10.1007/978-3-540-44928-7_17 14. Bock M, Scharp T, Talnikar C, Klipp E (2013) BooleSim: an interactive Boolean network simulator. Bioinformatics 30(1):131–132. https://doi.org/10.1093/bioinformatics/ btt568 15. Schivo S, Scholma J, Karperien M, Post JN, Van De Pol J, Langerak R (2014) Setting parameters for biological modelsWith ANIMO. In: Electronic Proceedings in Theoretical Computer Science, EPTCS, pp 35–47. doi:https:// doi.org/10.4204/EPTCS.145.5 16. Krumsiek J, Po¨lsterl S, Wittmann DM, Theis FJ (2010) Odefy - from discrete to continuous models. BMC Bioinformatics 11(1):233. https://doi.org/10.1186/1471-2105-11233 17. Mendes P, Hoops S, Sahle S, Gauges R, Dada J, Kummer U (2009) Computational modeling of biochemical networks using COPASI. In: Maly IV (ed) Methods in molecular biology (methods and protocols)-systems biology, vol 500. Vol systems biology. Humana Press, New York, NY. https://doi.org/10.1007/ 978-1-59745-525-1_2 18. Matsuoka Y, Funahashi A, Ghosh S, Kitano H (2014) Modeling and simulation using celldesigner. In: Miyamoto-Sato E, Ohashi H, Sasaki H, Nishikawa J-I, Yanagawa H (eds) Methods in molecular biology (methods and protocols)- Transcription factor regulatory networks., vol 1164. Humana Press,

160

Sakshi Khurana et al.

New York, NY. https://doi.org/10.1007/ 978-1-4939-0805-9_11 19. Schivo S, Scholma J, Wanders B, Camacho RAU, van der Vet PE, Karperien M, Langerak R, van de Pol J, Post JN (2014) Modeling biological pathway dynamics with timed automata. Ieee J Biomed Health 18 (3):832–839. https://doi.org/10.1109/Jbhi. 2013.2292880 20. Eisenmann DM (2005) Wnt signaling. https://doi.org/10.1895/wormbook.1.7.1. Accessed 19 Dec 2019 21. Murphy AM, Wong AL, Bezuhly M (2015) Modulation of angiotensin II signaling in the prevention of fibrosis. Fibrogenesis Tissue Repair 8:7–7. https://doi.org/10.1186/ s13069-015-0023-z 22. Morlon A, Munnich A, Smahi A (2005) TAB2, TRAF6 and TAK1 are involved in NF-κB activation induced by the TNF-receptor, Edar and its adaptator Edaradd. Hum Mol Genet 14 (23):3751–3757. https://doi.org/10.1093/ hmg/ddi405 23. MacDonald BT, Tamai K, He X (2009) Wnt/ beta-catenin signaling: components, mechanisms, and diseases. Dev Cell 17(1):9–26. https://doi.org/10.1016/j.devcel.2009.06. 016 24. Tran FH, Zheng JJ (2017) Modulating the wnt signaling pathway with small molecules. Protein Sci 26(4):650–661. https://doi.org/10. 1002/pro.3122 25. Verstrepen L, Bekaert T, Chau TL, Tavernier J, Chariot A, Beyaert R (2008) TLR-4, IL-1R and TNF-R signaling to NF-κB: variations on a common theme. Cell Mol Life Sci 65 (19):2964–2978. https://doi.org/10.1007/ s00018-008-8064-8 26. Massague´ J (1998) TGF-β signal transduction. Annu Rev Biochem 67(1):753–791. https:// doi.org/10.1146/annurev.biochem.67.1.753 27. Cadigan KM, Waterman ML (2012) TCF/LEFs and Wnt signaling in the nucleus. Cold Spring Harb Perspect Biol 4(11): a007906. https://doi.org/10.1101/ cshperspect.a007906 28. Wu B, Crampton SP, Hughes CCW (2007) Wnt signaling induces matrix metalloproteinase expression and regulates T cell transmigration. Immunity 26(2):227–239. https://doi. org/10.1016/j.immuni.2006.12.007 29. Heather G, Chun P (2006) Activin receptorlike kinases: structure, function and clinical implications. Endocr Metab Immune Disord Drug Targets 6(1):45–58. https://doi.org/ 10.2174/187153006776056585

30. Frederick JP, Liberati NT, Waddell DS, Shi Y, Wang X-F (2004) Transforming growth factor β-mediated transcriptional repression of c-myc is dependent on direct binding of Smad3 to a novel repressive Smad binding element. Mol Cell Biol 24(6):2546. https://doi.org/10. 1128/MCB.24.6.2546-2559.2004 31. Higuera GA, Hendriks JA, van Dalum J, Wu L, Schotel R, Moreira-Teixeira L, van den Doel M, Leijten JC, Riesle J, Karperien M, van Blitterswijk CA, Moroni L (2013) In vivo screening of extracellular matrix components produced under multiple experimental conditions implanted in one animal. Integr Biol 5 (6):889–898. https://doi.org/10.1039/ c3ib40023a 32. Jonk LJC, Itoh S, Heldin C-H, ten Dijke P, Kruijer W (1998) Identification and functional characterization of a Smad binding element (SBE) in the JunB promoter that acts as a transforming growth factor-β, activin, and bone morphogenetic protein-inducible enhancer. J Biol Chem 273 (33):21145–21152. https://doi.org/10. 1074/jbc.273.33.21145 33. Ranganathan P, Agrawal A, Bhushan R, Chavalmane AK, Kalathur RKR, Takahashi T, Kondaiah P (2007) Expression profiling of genes regulated by TGF-beta: differential regulation in normal and tumour cells. BMC Genomics 8:98–98. https://doi.org/10.1186/14712164-8-98 34. Hyun Hwa C, Hye Joon J, Ji Sun S, Yong Chan B, Jin Sup J (2008) Crossregulation of β-catenin/Tcf pathway by NF-κB is mediated by lzts2 in human adipose tissue-derived mesenchymal stem cells. Biochim Biophys Acta 1783(3):419–428. https://doi.org/10.1016/ j.bbamcr.2007.08.005 35. Du Q, Geller DA (2010) Cross-regulation between Wnt and NF-κB signaling pathways. For Immunopathol Dis Therap 1(3):155–181. https://doi.org/10.1615/For umImmunDisTher.v1.i3 36. Wang J, Zhao J, Chu ESH, Mok MTS, Go MYY, Man K, Heuchel R, Lan HY, Chang Z, Sung JJY, Yu J (2013) Inhibitory role of Smad7 in hepatocarcinogenesis in mice and in vitro. J Pathol 230(4):441–452. https://doi.org/10. 1002/path.4206 37. Guo X, Wang X-F (2009) Signaling cross-talk between TGF-beta/BMP and other pathways. Cell Res 19(1):71–88. https://doi.org/10. 1038/cr.2008.302 38. Huang N, Li W, Wang X, Qi S (2018) MicroRNA-17-5p aggravates lipopolysaccharideinduced injury in nasal epithelial cells by

Modeling with ANIMO targeting Smad7. BMC Cell Biol 19(1):1–1. https://doi.org/10.1186/s12860-018-01525 39. Hayashi H, Abdollah S, Qiu Y, Cai J, Xu Y-Y, Grinnell BW, Richardson MA, Topper JN, Gimbrone MA, Wrana JL, Falb D (1997) The MAD-related protein Smad7 associates with the TGFβ receptor and functions as an antagonist of TGFβ signaling. Cell 89(7):1165–1173.

161

https://doi.org/10.1016/S0092-8674(00) 80303-7 40. Huang X, Zhong L, Hendriks J, Post JN, Karperien M (2018) The effects of the WNT-signaling modulators BIO and PKF118-310 on the chondrogenic differentiation of human mesenchymal stem cells. Int J Mol Sci 19(2):561. https://doi.org/10.3390/ ijms19020561

Part II In Vivo Models of Skeletal Tissue Injury, Degeneration, and Repair

Chapter 11 Generation and Characterization of Mouse Models for Skeletal Disease Gabrielle E. Foxa, Ye Liu, Lisa M. Turner, Alexander G. Robling, Tao Yang, and Bart O. Williams Abstract Our laboratories have used genetically engineered mouse models (GEMMs) to assess genetic contributions to skeletal diseases such as osteoporosis and osteoarthritis. Studies on the genetic contributions to OA are often done by assessing how GEMMs respond to surgical methods that induce symptoms modeling OA. Here, we will describe protocols outlining the induction of experimental OA in mice as well as detailed descriptions of methods for analyzing skeletal phenotypes using micro-computerized tomography and skeletal histomorphometry. Key words microCT, Histomorphometry, Osteoarthritis

1

Introduction The analysis of skeletal phenotypes of genetically modified mice has contributed to advancements of our knowledge about skeletal development and disease and has provided the foundation for the development of several FDA-approved treatments for bone-related diseases. Our laboratories have used genetically engineered mouse models (GEMMs) to assess genetic contributions to skeletal diseases such as osteoporosis and osteoarthritis. Studies on the genetic contributions to OA are often done by assessing how GEMMs respond to surgical methods that induce symptoms modeling OA. The generation of GEMMs is applicable to all tissue and disease types and discussion of methodologies for their generation is available in many other contexts [1, 2]. Here, we focus on several methods specific for generating and analyzing skeletal phenotypes in mice. We first describe the generation of an OA-like model via surgical destabilization of the medial meniscus. We will then describe the evaluation of skeletal phenotypes via microcomputerized tomography (microCT) and histomorphometry.

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_11, © Springer Science+Business Media, LLC, part of Springer Nature 2021

165

166

2

Gabrielle E. Foxa et al.

Materials

2.1 Surgical Destabilization of the Medial Meniscus (DMM)

1. Experimental mice of the same strain divided into DMM surgery group (n
5) and sham surgery control group (n
5). Age of mice should be decided according to the goal of experiments, but generally mice at or older than 3 months of age are optimal as their skeletal maturity is reached. All mice should be housed in a gnotobiotic facility, according to the National Research Council’s Guide and Institutional Animal Care and Use Committee-approved protocols for the care and use of experimental animals (see Note 1). 2. Isoflurane machine: VetEquip Tabletop laboratory animal anesthesia system (VetEquip # 901820). 3. Isoflurane (Piramal # NDC 66794–017-25). 4. Surgical microscope (Leica # M651). 5. Veterinary operating Mediheat V500).

table

heated

mat

(Peco

#

6. 9 mm Wound forceps and clips (Reflex # RS-9260, RS-9262). 7. Wound clip remover (Reflex # RS-9263). 8. Hair removal lotion. 9. Alcohol antiseptic wipes. 10. Micro-iris scissors. 11. Micro-surgical knife (#3 handle and #10/ #11 blade). 12. Sharppoint 15 5 mm blade micro-surgical knife. 13. Epinephrine (1:1000 dilution). 14. Amoxicillin (20 mg/kg). 2.2 Osteoarthritis (OA) Histology

1. Histological tissue cassette case. 2. Surgical scissors. 3. Tweezer. 4. Deionized water. 5. Ethanol 70%, 80%, 90%, 95%, 100%. 6. Neutral buffered formalin (NBF) 10%. 7. Formic acid 5%. 8. Xylene. 9. Paraffin (Paraplast X-tra). 10. Oven. 11. Stock solution A: 10 g hematoxylin, 500 mL 80% ethanol. 12. Stock solution B: 20 g ferric chloride, 475 mL distilled water, 5 mL hydrochloric acid (36.5–38%).

Generation and Characterization of Mouse Models for Skeletal Disease

167

13. Modified Weigert’s Iron Hematoxylin: equal volume stock solutions A and B (see Note 2). 14. Acid-alcohol 1.0%: 5 mL hydrochloric acid (36.5–38%), 495.0 mL 70% ethanol. 15. Acetic acid 1.0%: 5.0 mL glacial acetic acid, 495 mL distilled water. 16. Fast Green 0.08%: 0.20 g Fast Green, 250.0 mL distilled water. 17. Safranin O 1%: 2.5 g Safranin O, 250 mL distilled water. 2.3 Microcomputed Tomography

All standards, equipment, and software for scanning and analysis are produced by Bruker microCT, Kontich, Belgium and Micro Photonics Inc., Allentown, PA. 1. Neutral buffered formalin (NBF) 10%. 2. Deionized (DI) water. 3. Ethanol (EtOH) 70%. 4. Gauze cut to length of the sample. 5. Plastic conical. 6. Bruker Skyscan 1172. 7. Skyscan 1172 (Scanning software) Version 1.5.26.0. 8. Software NRecon Version 1.7.4.6. 9. GPUReconServer Version 1.7.4.2. 10. Data Viewer Version 1.5.6.3 software.

2.4 Bone Mineral Density Calibration

1. Bone mineral density standards 0.25 and 0.75.

2.5 Trabecular Analysis of Microcomputed Tomography Images

1. CT Analyzer Version 1.18.8.0 software.

2.6 Cortical Analysis of Microcomputed Tomography Images

1. CT Analyzer Version 1.18.8.0 software.

2.7

1. CTvol Version 2.3.2.0 software.

Bone Modeling

2.8 Fluorochrome Labeling of Bone

2. CT Analyzer Version 1.18.8.0 software.

1. Calcein Green 1.0% stock solution: In a 50 mL conical tube, dissolve 400 mg NaHCO3 in 20 mL of saline (0.9% NaCl). Slowly add 200 mg calcein (Sigma-Aldrich # C0875). Calcein will foam if added quickly. pH to 7.4 using 1 N NaOH or 1 N HCl and filter sterilize into evacuated 20 mL glass vial. Wrap in foil and store at 4  C (see Note 3).

168

Gabrielle E. Foxa et al.

2. Calcein Green 0.10% working solution: Dilute the Calcein Green 1.0% stock solution in saline. Pipette 2.0 mL of 1% calcein stock and 18.0 mL of saline. Wrap in foil and store at 4  C (see Note 4). 3. Alizarin Complexone 2.5% stock solution: In a 50 mL conical tube, dissolve 400 mg NaHCO3 in 20 mL of saline (0.9% NaCl). Slowly add 500 mg alizarin complexone (Sigma-Aldrich # A3882). Alizarin will foam if added quickly. pH to 7.4 using 1 N NaOH or 1 N HCl and filter sterilize into evacuated 20 mL glass vile. Wrap in foil and store at 4  C. 4. Alizarin Complexone 0.20% working solution: Dilute 2.5% stock solution with saline to 0.2%. Pipette 0.8 mL of 2.5% alizarin stock and 9.2 mL of saline. 5. Demeclocycline HCL 0.75% working solution: In a sterile 200 mL bottle, dissolve 1 g demeclocycline hydrochloride (Sigma-Aldrich # D6140) in 133 mL of sterile saline (0.75% solution). Place on a rocker plate and rock gently for 20 min. Aliquot into light-proof 1.5 mL tubes and store at 20  C. 2.9

Fixation

1. DI water. 2. Ethanol 70%. 3. Neutral Buffered Formalin (NBF) 10%.

2.10

Infiltration

1. Dremel hand tool. 2. Automated tissue processor. 3. Ethanol 70%, 95%, 100%. 4. Fisherbrand Histocassette Tissue Cassette. 5. Infiltration solution: 85% Methyl methacrylate (MMA) and 15% dibutyl phthalate (DBP). 6. Methyl methacrylate (MMA) 100%. 7. NBF 10%. 8. Rapid Infiltration Processor (RIP). 9. Xylene. 10. Embedding molds. 11. Glass culture tubes.

2.11 Embedding and Cross-Sectioning

1. 600-grit Sandpaper. 2. Allen wrench. 3. Binder clips. 4. DI water. 5. Embedding solution: 85% MMA, 15% DBP, and 0.5% Perdadox 16.

Generation and Characterization of Mouse Models for Skeletal Disease

169

6. Eukitt Mounting Medium Electron Microscopy Sciences # 15320. 7. Extech Labcut 150. 8. 3”.006” ½ Diamond Wafering Blade. 9. Glass culture tubes with rubber stoppers. 10. Glass slides. 11. Lab Armor Bead Bath. 12. Micrometer. 13. Paper towel. 14. Sonicator. 15. Small weight. 2.12 MMA Embedding and Coronal Sectioning

1. Automatic rotary microtome. 2. Bead bath. 3. Clamps. 4. Coating solution: 85 mL DI water, 1 g gelatin powder, and 15 mL glycerol incubate the solution at 42  C in a water bath until dissolved or visible particles are no longer visible (see Note 5). 5. Cutting solution: 10 mL 70% ethanol, 90 mL distilled water, and 2 drops of soap (see Note 6). 6. Embedding solution: 85% MMA, 15% DBP, and 0.5% Perdadox 16. 7. Ethanol 70%, 75%. 8. Forceps. 9. Fine tip paint brush. 10. Glass slides. 11. Kimwipes. 12. MetaServ 250 Single Grinder/Polisher. 13. Paperclips. 14. Paper towels. 15. Petri dish. 16. Plastic film. 17. Rubber roller. 18. Silicone stopper. 19. Oven. 20. Prefabricated mold for embedding. 21. Acetone. 22. Cellusolve. 23. DI water.

170

Gabrielle E. Foxa et al.

2.13 Golder’s Trichrome Stain

1. Wiegert’s Iron Hematoxylin: Prepare fresh by combining 50 mL Wiegert’s Solution A and 50 mL Wiegert’s Solution B. 2. Ponceau Fuchsin solution A: 1 g Ponceau, 100 mL DI water. 3. Ponceau Fuchsin solution B: 1 g Acid fuchsin, 100 mL DI water. 4. Ponceau Fuchsin stock solution: 30 mL Ponceau fuchsin solution A, 10 mL ponceau fuchsin solution B. 5. Acetic acid 1%. 6. Ponceau Fuchsin working solution: Dilute 10 mL ponceau fuchsin stock solution in 150 mL 0.2% acetic acid. 7. Phosphotungstic Acid (PTA) Orange G: Combine and filter 2 g Orange G, 4 g phosphotungstic acid, and 100 mL DI water. 8. Light Green: combine and filter 0.2 g Light Green, 0.2 mL acetic acid, and 100 mL DI water. 9. DI water. 10. Ventana Symphony tissue processor.

2.14 Paraffin Embedding and Sectioning

1. DI water. 2. Embedding center (Leica EG1150H). 3. Ethylenediaminetetraacetic acid (EDTA) 10%: pH 7.4. 4. Ethanol 70%, 80%, 95%, 100%. 5. Fisherbrand Histocassette Tissue Cassette. 6. Forceps. 7. Glass slides. 8. Kimwipes. 9. NBF 10%. 10. Oven. 11. Paraffin/tissue infiltration media. 12. Steel microtome knife. 13. Xylene. 14. Tissue processor. 15. Water bath.

2.15

TRAP Stain

1. Sigma-Aldrich product 387-A: Acid Phosphatase, Leukocyte (TRAP) kit. 2. Aqueous mounting media. 3. Glass slides. 4. Glass coverslips.

Generation and Characterization of Mouse Models for Skeletal Disease

2.16

Slide Imaging

171

1. ET-DAPI/FITC/Texas Red filter set (Chroma product # 69002). 2. Leica Aperio AT2 High Volume, Digital Whole Slide Scanning. 3. Lumen Dynamics X-Cite Series 120 Q. 4. QImaging QIClick Digital CCD Camera. 5. Zeiss Axio Imager.A2. 6. BIOQUANT OSTEO 19.2.6. software.

2.17 Histomorphometry

3 3.1

1. BIOQUANT OSTEO 19.2.6 software.

Methods DMM Surgery

The methods described below that we utilize in the laboratory have been developed and optimized based on several previous studies [3–6]. Operate the anesthesia machine according to the VetEquip user guide and operating manual (http://www.vetequip.com/ pdfs/LAAS%20 Manual.pdf). DMM surgery should be performed on both knees of each individual mice of the DMM group, and sham surgery will be performed to another group of mice for controls. Surgery should be carried out with standard sterile techniques, i.e., preoperative instrument sterilization and disinfection, and cleansing of incision area. 1. Anesthetize the mice with 4% isoflurane and 1 L/min oxygen in the VetEquip tabletop laboratory animal anesthesia chamber. 2. Once the mice are anesthetized, transfer them to a mask and maintain with 3% isoflurane and 1 L/min oxygen. 3. Place the mice on a veterinary operating table heated mat to maintain body temperature during surgery. 4. To bare the incision area, apply hair removal lotion on the medial side of knee for 3 min. Then wipe with alcohol antiseptic wipes. 5. In order to expose the medial side of stifle joint capsule, use a micro-iris scissor or a micro-surgical knife to open a 3 mm longitudinal incision through the skin and subcutaneous tissue over the distal patella to proximal tibial plateau. 6. Expose medial meniscotibial ligament (MMTL), which anchors medial meniscus (MM). During the process, avoid destabilizing patellar tendon, medial collateral ligament, and articular cartilage.

172

Gabrielle E. Foxa et al.

7. To prevent bleeding that may block the vision for performing transection, one drop of 0.1% epinephrine can be applied at the incision site (optional). 8. Under a surgical microscope can assist accurate transect the MMTL with a micro-surgical knife 15 5 mm blade) to release the MM. Locating ligament first, this can assist the identification of region.

transection, (sharppoint the patellar the incision

9. The sham group mice will be operated through the same approach without MMTL transection. 10. Close the incision together with a 9 mm wound forceps and clips. 11. To prevent joint infection, Amoxicillin (20 mg/kg) can be injected after surgery (optional). 12. Remove the clips with a wound clip remover after wound has healed. 3.2 Osteoarthritis (OA) Histological Grading

At 4 to 6 weeks (or at specific time point according to the experimental design) post DMM surgery, OA phenotypes can be graded or assessed by histological analysis, gene expression assay, microCT, pain behaviors, or gait analysis. Here we focus on the histological grading. 1. Collect knees from mice and arrange in a histological tissue cassette case (see Note 7). 2. To fix the tissue, place the tissue cassettes with mouse knees in 10% NBF for 48 h at 4  C. Rinse in DI water 3 for 10 min each time and store in 70% ethanol at 4  C. 3. To decalcify the samples, place the fixed tissues with cassettes in a beaker containing at least 15 volumes of 5% formic acid on a swirling/rocking platform for 7 days (see Note 8). 4. Rinse the tissues in DI water 4 for 10 min each time. 5. Dehydrate the tissues in 70%, 95%, and 100% ethanol for multiple rounds 1–2 h each (see Note 9). 6. Clear tissues in xylene for 1 h 2. 7. Infiltrate tissues with paraffin (Paraplast X-tra) at 58  C for 1 h 2. 8. Embed tissues into wax blocks at 62  C. 9. Rehydrate tissues for histology using xylene, absolute alcohol, 95% ethanol, 80% ethanol, then rinse in distilled water (see Note 10). 10. Cut 5 mm sections of the tissues embedded in the blocks (see Note 11).

Generation and Characterization of Mouse Models for Skeletal Disease

173

11. Bake the slides at 60  C overnight and cool at room temperature for 20–30 min. 12. Deparaffinize and hydrate in 70% ethanol and air dry. 13. Incubate the slides with Weigert’s Iron Hematoxylin for 30 min. 14. Wash the slides gently in running tap water for 10 min. 15. Differentiate by single dip in 1% acid-alcohol. 16. Rinse the slides in DI water 3. 17. Incubate the slides with 0.08% Fast Green for 5 min. 18. Rinse the slides in 1.0% acetic acid for 10 s (blot, no rinse). 19. Incubate the slides in 1% Safranin O for 30 min (blot, no rinse). 20. Dehydrate the sections with 1 change of 95% ethanol and two changes of 100% ethanol by 1 dip each and air dry. 21. Clear the slides in xylene for 1 min, 2 and coverslip. 22. Score histologically to evaluate OA progression in a blinded manner following the semiquantitative scoring system established by Sonya S. Glasson et al. The OA phenotype of all four quadrants of joint, including the medial femoral condyle, medial tibial plateau, lateral femoral condyle, and lateral tibial plateau should be scored (see Note 12). 3.3 Imaging by Microcomputed Tomography

These methods have been adapted and optimized for use in our laboratories from several sources [7–9]. Methodology is specific to femurs from mice at 3–6 months of age. All microCT methods were performed under the recommended guidelines of the SkyScan software Standard Operating Procedures by Bruker MicroCT (Kontich, Belgium). 1. Fix limbs in 10% NBF for 48 h. Rinse limbs in deionized (DI) water and store in 70% ethanol. 2. Prepare the limbs by wrapping them in gauze saturated in 70% ethanol. Load the limbs into a conical submerged in 70% ethanol while preventing air bubbles from entering. Samples should not be overlapping and should be snug in the container. 3. Turn the Skyscan 1172 on by turning key to “START” position. The key will automatically snap back into the “ON” position. 4. Open Skyscan 1172 software and start the X-ray tube by pushing the X-ray icon. A pop-up window will open with a progress bar as the tube ages. Once the X-ray is on, an indicator window will open. 5. Press the “grab image” icon to obtain a continuous preview from the X-ray camera and establish scan settings (see Note 13).

174

Gabrielle E. Foxa et al.

6. Perform a flat field correction (see Note 14). 7. Open the chamber on the scanner by pressing the “specimen door” icon, fasten the sample to a stage, and secure the stage on the rotation axis by turning the brass ring to tighten it. 8. Close the chamber by pressing the “specimen door” icon, turn the X-ray back on, and restore the settings established in step 5. 9. Set an oversize scan (see Note 15). 10. Select the “Start scanning” button in the Scout Scan window to open the settings dialog window and complete scan settings (see Note 16). 11. Click “OK” to start the scan. 12. After scanning is complete, open NRecon and load a dataset by choosing “Open Dataset” in the “Actions” tab and choosing one of the images from a scan. Choosing one image will open the entire dataset associated with the sample. 13. Perform a thermal correction (see Note 17). 14. Choose a dynamic range based on the histogram of the sample (see Note 18). 15. Correct beam-hardening (see Note 19). 16. Correct misaligned segments (see Note 20). 17. Reduce ring artifacts (see Note 21). 18. Smooth the images (see Note 22). 19. Choose a region of interest (ROI) to save (see Note 23). 20. Choose a volume of interest (VOI) to save (see Note 24). 21. Reconstruct and save the images (see Note 25). 22. After the images have been reconstructed, open Data Viewer and load a reconstructed dataset from a “Rec” folder of a sample. 23. After the sample is open, click the “load for 3D viewing” icon. 24. Use the three views to straighten the sample. Orient all samples in a project the same way. 25. Under the “View” tab, choose “Single Volume of Interest (VOI),” and fit the sample within each of the boxes to crop the dataset and define a volume of interest (VOI). 26. Save the volume of interest (see Note 26). 3.4 Bone Mineral Density Calibration

1. Scan, reconstruct, and orient the bone mineral density standards following the steps under “Imaging by Microcomputed Tomography 3.3.” 2. Open the CTan software and load one of the density standards by choosing the “open image or dataset” icon.

Generation and Characterization of Mouse Models for Skeletal Disease

175

3. Define the upper bound of the density standard by scrolling to the top and choosing a slice that includes the entire density standard. Right click the slice in the dataset window and choose “Set the Top of Selection.” Repeat for the lower bound but choose “Set the Bottom of the Selection.” 4. Define the region of interest (ROI) to include only the density standard (see Note 27). 5. Save the ROI (see Note 28). 6. Open the new ROI file for the density standard by selecting the “open image or dataset icon” and choosing one of the images in the ROI folder. 7. Apply the ROI to the open file (see Note 29). 8. Identify and record the attenuation coefficient (see Note 30). 9. Repeat steps 2–8 for the other density standard. 10. Under “File,” choose “Preferences,” and under the “Histogram” tab, choose “Calibrate.” 11. Enter the known min and max BMD density values of the density standards (Ex: 0.25 and 0.75) followed by the corresponding attenuation coefficients recorded in step 8. 12. Click “OK” to calibrate the density range. 3.5 Trabecular Analysis of Microcomputed Tomography Images

1. In the CTan software, select the “open image or dataset” icon. Choose a dataset within a newly created VOI folder. 2. Identify the reference point for the ROI, use a 0.25 mm offset, and create an ROI 2.5 mm in height (see Note 31). 3. Create a region of interest to include trabecular bone only, by manually drawing an ROI (see Note 32) or use an automated program and save the regions. If manually drawing ROIs, continue to step 6 after this step. To make an automated program to create an ROI, continue to step 4. 4. Begin the automated trabecular ROI process by, saving the ROI defined in step 2 and determining the thresholding scale (see Note 33). 5. Open the ROI saved in step 4 by choosing the “open an existing image or dataset” icon and choosing an image within the ROI folder for the sample. Build a task list to create, analyze, and save an automated trabecular ROI (see Note 34). After building the task list, skip to step 9. 6. Open a sample’s manual ROI by choosing the “open an existing image or dataset” icon and choosing an image within the trabecular ROI folder for the sample. 7. Apply the manual ROI to the open file and determine the thresholding scale (see Note 35).

176

Gabrielle E. Foxa et al.

8. Choose the “custom processing preview” icon and select the “Internal” tab to build a task list to analyze the trabecular bone (see Note 36). 9. Run the task list on multiple samples using the batch manager (see Note 37). 3.6 Cortical Analysis of Microcomputed Tomography Images

1. In the CTan software, select the “open image or dataset” icon. 2. Identify the reference point for the ROI and set the height to 0.6 mm (see Note 38). 3. Create a region of interest to include the entire area of the bone (see Note 39). 4. Save the cortical ROI (see Note 40). 5. Open a sample ROI by choosing the “open an existing image or dataset” icon and choosing an image within the cortical ROI folder for the sample. 6. Apply the ROI to the open file and determine the thresholding scale (see Note 41). 7. Build a 2D analysis task list using a combination of functions until the cortical bone that remains in the “binary selection preview” closely matches the raw images (see Note 42). 8. Build a histogram Task List by constructing a task list the same as the “2D analysis Task List” except for the 2D analysis and 3D model functions. Add ROI shrink-wrap, reload, histogram functions to the end of the list (see Note 43). 9. Run the task list on multiple samples using the batch manager. Repeat for each task list (see Note 44).

3.7 Modeling Skeletal Phenotypes

Several modifications can be made when generating visual representations of scanned samples. The guidelines below are for a basic, gray-scale model. 1. Using CTvol, open a 3D file by selecting the “open 3D object” icon and choosing the p3g file within the ROI folder of the desired sample. 2. Use the wheel on the computer mouse to zoom in and out on the sample. 3. Select “Options” and choose “Objects” to open up the “Objects” window. Adjust color, emission, diffusion, reflection, and opacity as desired within this window. 4. Select the “move camera” or “move objects” icon and use your mouse to adjust the sample to the desired position and angle. 5. Save the Image by selecting “Save Image” from the “File” dropdown menu.

Generation and Characterization of Mouse Models for Skeletal Disease

3.8 Mouse Fluorochrome Injections

177

Labeling schedules will vary depending on anticipated phenotype and project design. The following schedule is for mice 6 months of age with fluorochrome labeled surfaces at four different time points. 1. At 4 weeks of age, perform intraperitoneal (IP) injections on mice with 0.10% working calcein solution. Inject 0.2 mL per 20 g of body weight for a 10 mg/kg dose (see Note 45). 2. At 10 weeks of age, perform IP injections on mice with 0.20% working alizarin complexone solution. Inject 0.2 mL per 20 g of body weight for a 20 mg/kg dose (see Note 45). 3. At 24 weeks of age, perform IP injections on mice with 0.75% working demeclocycline hydrochloride solution. Inject 0.250 mL per 20 g mouse 150 IP/100 subcutaneous for a 75 mg/kg dose (see Note 45). 4. At 25 weeks of age, perform IP injections on mice with 0.10% working calcein solution. Inject 0.2 mL per 20 g of body weight for a 10 mg/kg dose. 5. Sacrifice animals at 26 weeks of age.

3.9

Infiltration

1. Fix the femurs in 10% NBF for 48 h, followed by storage in 70% ethanol. 2. Remove excess tissue from femurs and move to individual histocassette tissue cassettes. 3. Using the Dremel hand tool, notch each bone ¼ to ½ of the way through in the midshaft region. Return each bone to its cassette. 4. Dehydrate and clear the samples for 2–4 h in 70% ethanol, 95% ethanol, 100% ethanol, and xylene (see Note 46). 5. Transfer the specimens to xylene until infiltration. 6. Within the RIP unit infiltrate the samples under 15 psi using 100% MMA for 24 h (infiltration 1), the infiltration solution for 24 h (infiltration 2), followed by the infiltration solution for at least 2 days (infiltration 3). 7. Transfer bone tissue to individual, prefabricated embedding molds for coronal sectioning or glass culture tubes for crosssectioning.

3.10 Femoral Embedding and Cross-Sectioning

When processing tissues that are light sensitive due to fluorochrome labeling, keep in the dark in between processing steps. 1. Add a few drops of glass culture tube embedding solution into a labeled glass culture tube and insert the femur, ensuring the head of the femur is closest to the open end of the tube (see Note 47).

178

Gabrielle E. Foxa et al.

2. Cap the glass tube with a rubber stopper and transfer the beaker into the bead bath at 25  C for optimal polymerization. 3. Check the glass tubes once daily until the polymerization process is complete (see Note 48). 4. Remove the polymerized rod containing the femur from the glass culture tube by gently shattering the surrounding glass. 5. Load the resulting plastic rod into the specimen holder of the Extech Labcut 150. Secure the rod into the specimen holder by tightening down the adjustment bolts with an Allen wrench. 6. Line up the plastic rod with the diamond wafering blade to the midshaft of the femur and move the specimen arm so it sits at the back of the Labcut 150. 7. Place either the unattached reed switch magnet or the canopy in place so that the reed switch is covered. 8. Turn the instrument on, set the speed to 166.2 rpm, select the “rev” button to start the blade. 9. Turn on the lubricant pump and slowly pull the specimen arm forward. 10. Sit the rod atop the saw blade and begin to section the bone at the midshaft or slightly more distal (see Note 49). 11. Save the distal end for MMA embedding and coronal crosssectioning or discard it. 12. Adjust the micrometer on the Extech Labcut 150 with the diamond wafering blade to the desired thickness for the first wafer. Ensure the rod is advancing over the blade (see Note 50). 13. Section desired number of wafers. 14. Remove any plastic dust particles from the wafers by dipping each in a sonicator filled with deionized water for 3–5 s. Dab off any excess water. 15. Place a nickel size pool of Eukitt mounting media onto the back of a labeled glass slide and mount the wafers in the pool. 16. Cover the glass slide with 2–3 layers of 1–2 ply paper towel cut into slide size strips. 17. Place a blank glass slide on top of the paper towel strips and fasten both slides together with 2–3 binder clips and allow the slide to air dry overnight. 18. Remove the binder clips, blank slide, and paper towel strips from the surface of the slide with mounted wafers to be sanded. 19. Using a table micrometer, zero out the slide thickness by placing a portion of the glass slide that does not contain a wafer between the clamps. Tighten the clamp on the slide and clear the counter.

Generation and Characterization of Mouse Models for Skeletal Disease

179

20. Open the micrometer clamp and move the slide so one of the mounted wafers is measurable. Close the clamp down onto the wafer and obtain a starting measurement of the wafer. Repeat this step for all wafers. 21. Using 600 grit sandpaper, gently sand the surface of the wafers until each measures 40–50 μm. Remeasure each wafer using the micrometer throughout the sanding process in order to determine when the desired thickness has been obtained. 22. Once all wafers have been sanded to the desired thickness, wipe the sanding remnants from the surface of the slide with a damp paper towel. 23. Add a thin layer of Eukitt mounting media to the surface of the wafers and add a glass coverslip. 24. Add 2–3 small weights to the top of the coverslip to ensure any air bubbles are removed and allow the slides to air dry overnight. 3.11 MMA Embedding and Coronal Sectioning of Femurs

When processing tissues that are light sensitive due to fluorochrome labeling, keep in the dark in between processing steps. 1. Add a thin layer of prefabricated mold embedding solution to each prefabricated mold and allow the reagent to become sticky/tacky. 2. Orient the tissue for coronal sectioning and fill remainder of the mold with prefabricated mold embedding solution. 3. Add a paper label containing the specimen ID, and a paper clip bent to a 45 –90 angle to each mold. 4. Cap the top of the mold with a silicone stopper and transfer each mold to the bead bath at 25  C, ensuring the containers are covered. 5. Remove molds from the bead bath daily and check until polymerization is complete (see Note 51). 6. Remove fully polymerized molds to grind and polish using the MetaServ 250 Single Grinder/Polisher (see Note 52). 7. Apply a thin layer of the coating solution to a glass slide. 8. Place the embedded tissue mold in the microtome chuck and face into the mold until the desired tissue is exposed. 9. Brush the exposed face of the block with a fine tip paint brush, soaked in the cutting solution, and initiate the automatic microtome to begin slicing a section from the block at 5 μm. 10. Cut a small rectangle from a Kimwipe, and soak in 70% ethanol. 11. Apply the Kimwipe to the surface of the mold and initiate cutting on the microtome at a very slow speed.

180

Gabrielle E. Foxa et al.

12. As the Kimwipe and plastic section are slowly cut, grasp them together with a pair of forceps. Continue to carefully pull the section off the block as it is cut away by the microtome. 13. Transfer the section and wipe into a petri dish of 75% ethanol. Carefully stretch the section if wrinkles are visible using two pairs of forceps. 14. Once flat, place a piece of plastic film, just larger than the section, on top of the section to keep it free of wrinkles. 15. Transfer the section and plastic film combination to the coated slide. 16. Cover the slide with a thin sheet of plastic and lightly squeeze out any excess ethanol using a gentle wiping motion with a Kimwipe. 17. Cover the section with paper towel and roll flat with rubber roller. 18. Each day sectioning is completed, stack the slides together with a blank slide on the top and bottom of the stack and place in a clamp. Transfer each stack to an oven set at 60  C, overnight. 19. Remove slides from the oven and inspect for damage. 20. Begin deplasticizing slides by soaking them in cellusolve for 1 h. 21. Place slides in acetone for 2–5 min. 22. Rinse slides in DI water for 2–5 min. 23. Continue to “Goldner’s Trichrome Staining 3.12” step 1 for staining or step 10 to only coverslip for fluorochrome imaging. 3.12 Goldner’s Trichrome Staining

1. After deplasticizing samples, place slides in Wiegerts Hematoxylin for 20 min. 2. Rinse slides in running Tap Water for 10 min. 3. Place slides in working solution of Ponceau Acid Fuchsin for 10 min. 4. Place slides in 1% acetic acid for 30 s. 5. Place slides in filtered PTA-Orange G for 4 min. 6. Place slides in 1% acetic acid for 30 s. 7. Place slides in Light Green solution for 10 min. 8. Place slides in 1% acetic acid for 1 min. 9. Rinse slides in 2 changes of distilled water, for 1 min each. 10. Coverslip using the Ventana Symphony by using the proprietary “Coverslip Only” program.

Generation and Characterization of Mouse Models for Skeletal Disease

3.13 Paraffin Embedding and Sectioning of Femurs

181

1. Fix the femurs in 10% NBF for 48 h, followed by storage in 70% ethanol. 2. Remove excess tissue from femurs and move to individual histocassette tissue cassettes. 3. Decalcify samples in 10% EDTA pH 7.4, stirring for 10–14 days at 4  C (see Note 53). 4. Dehydrate, clear, and infiltrate the samples in cassettes within a tissue processor using 70% ethanol for 60 min, 80% ethanol for 60 min, 95% ethanol for 60 min 2, 100% ethanol for 60 min 3, xylene for 30 min 2, paraffin for 30 min, and paraffin for 45 min. 5. Remove the cassettes from the tissue processor and move them to the embedding center. 6. Set the heated areas of the embedding center between 60  C and 70  C. 7. Remove the tissue from the cassette and place it in an appropriately sized heated mold. Orient all the samples the same way depending on the type of cut desired (sagittal or coronal). 8. Hold the tissue specimen down with forceps while partially filling the mold with molten paraffin. Secure the tissue by quickly cooling the base of the mold (see Note 54). 9. Place the cassette bottom atop the mold and fill with paraffin. 10. Cool the blocks in a cooling area to set the paraffin until it has solidified. 11. Remove the blocks from the mold by pulling upward on the cassette. 12. Rough cut/face each block to expose the desired surface area of the tissue(s). Precool paraffin blocks with the tissue side down in a tray of ice water to facilitate sectioning. 13. Using a steel microtome knife or disposable blade, cut the appropriate sections (see Note 55). 14. For histological sections, place the sections on the pre-labeled glass slides with the IDs that correspond to each block being sectioned. Mount sections by floating a paraffin ribbon into a water bath set between 36  C and 42  C. 15. Dry paraffin sections in a lab oven set to 60  C for 10–20 min. 16. Remove the sections from the oven and allow the slides to cool at room temperature. 17. Deparaffinize slides by soaking in fresh xylene for 3 min 3, hand-dip slides into 100% ethanol 10, repeat in fresh 100% ethanol, hand-dip slides in 95% ethanol 10, repeat in fresh 95% ethanol, and rinse in DI water. 18. Store slides in DI water until TRAP staining.

182

3.14

Gabrielle E. Foxa et al.

TRAP Staining

1. Use the Sigma-Aldrich product 387-A: Acid Phosphatase, Leukocyte (TRAP) kit and protocol to stain deparaffinized samples. 2. Coverslip the slides using an aqueous mounting media and glass coverslips.

3.15

Imaging Slides

1. Image the trichrome stained samples at 20 magnification using a brightfield microscope (Leica Aperio AT2—High Volume, Digital Whole Slide Scanning). 2. Image fluorochrome labeled samples at 20 magnification (Zeiss Axio Imager.A2, Lumen Dynamics X-Cite Series 120 Q, QImaging QIClick Digital CCD Camera) using objective filter ET-DAPI/FITC/Texas Red (Chroma product # 69002). Use BIOQUANT OSTEO 19.2.6 software and follow the BIOQUANT OSTEO 19.2.6 “Image Region and Image Menu Introduction.”

3.16 Histomorphometry

The analysis steps are specific to BIOQUANT OSTEO 19.2.6 Manual. 1. To quantify trabecular bone parameters, open BIOQUANT OSTEO 19.2.6 software and create a new dataset following “The Data Region: Introduction.” 2. Measure trabecular parameters (Goldner’s Trichrome stained, Fluorochrome labeled, TRAP stained slides) by following the “Rodent Trabecular Bone Protocol.” 3. To quantify cortical bone parameters, start by creating new dataset following “The Data Region: Introduction.” 4. Measure cortical bone parameters by following the “Basic Cortical Bone Protocol.”

4

Notes 1. Male mice are more susceptible to developing DMM-induced osteoarthritis [10] thus considered suitable materials for such experiments, but the selection of mouse genders should accommodate the goals of specific studies. 2. Mix equal volume of stock solutions A and B to make the Modified Weigert’s Iron Hematoxylin working solution right before use. The working solution has a shelf life of about 1 week when stored in dark. 3. Wrap in foil to prevent exposure as the solution is light sensitive. 4. Stock solution is diluted to the 10 mg/kg recommended dose for injecting mice.

Generation and Characterization of Mouse Models for Skeletal Disease

183

5. Once prepared, the coating solution is stable at room temperature for 1 week. 6. Once prepared, the solution is stable at room temperature until expiration date of the ethanol lot. 7. Sacrifice the mice in a CO2 chamber. Dissect both hind limbs from the hip of the mice using surgical scissors and tweezers. Remove about 1/3 of femur from the proximal end and 1/3 of the tibia from the distal end using surgical scissors. Straighten each knee and place in a labeled histological tissue cassette with the patella facing up. This will help the following embedding and sectioning in the frontal orientation. 8. Another common method for decalcification is incubation in 14% EDTA solution for 10 days. EDTA-decalcified tissues generally have better result for immunohistochemistry than formic-acid-decalcified ones. However, the Safranin O stain shows better color intensity on the formic-acid-decalcified sections. 9. Dehydrate the tissues in the following order: 70% ethanol for 1 h, 95% ethanol for 1 h, absolute ethanol for 1 h 2, absolute ethanol for 1.5 h, and absolute ethanol for 2 h. 10. Deparaffinize sections, 2 changes of xylene, 3 min each, then rehydrate in 2 changes of absolute alcohol, 2 min each. Incubate in 95% ethanol for 2 min followed by 80% ethanol for 2 min. Wash sections briefly in distilled water. 11. According to the recommendation of Osteoarthritis Research Society International (OARSI) [11], the knee joints should be paraffin-embedded in frontal orientation and sectioned through the entire joint at 80 mm interval. Each section should be 5 mm. 12. The mouse OA knee damage categories followed by scores in parentheses include normal (0), loss of Safranin O with no structural changes (0.5), small fibrillations with no loss of cartilage (1), vertical clefts down to the layer immediately below the superficial layer and some loss of surface lamina (2), vertical clefts/erosion to the calcified cartilage extending to 75% of the articular surface (6). Use mean value of the scores from the 4 quadrants of an individual knee joint for statistical analyses. 13. To establish settings, double-click the indicator window to open the X-ray energy settings. Set the voltage to 60 kV and

184

Gabrielle E. Foxa et al.

current to 167 μA. The “always maximize power” box should be checked, which will automatically adjust the power to 10 W and thus adjust voltage or current accordingly. Set the pixel resolution to 2000  1200 and the X-ray camera image pixel size to 6 μm. Apply a 0.5 mm aluminum filter. 14. Open “Preferences” in the “Options” tab. Un-check the two flat-field correction boxes and click “OK.” Select “Acquisition Modes” in the “Options” tab. Locate the window with the arrow which corresponds to the filter, camera position, and pixel size. Set the exposure to 360 and adjust the voltage and current windows if needed. Check the option, “Acquire bright + dark for current mode.” Grab image to obtain a live preview and right click to view the intensity profile. The average intensity should be ~60%. Note: If the average intensity is not ~60%, allow the X-ray to stabilize and adjust the exposure until it reaches ~60%. Open “Preferences” in the “Options” tab. Check the two flat-field correction boxes and click “OK.” The average intensity should be 88% (2%). 15. Select “Set Oversized Scan” in the “Actions” tab. Click “Scout Scan” and allow the scan to run to the bottom of the stage. Enter the sample name in the “Prefix” window. Select the “Top” button and click the top of the sample, followed by the “Bottom” button and click the bottom of the sample. Repeat this process for each sample in the scout scan preview window. 16. Designate a file location by choosing “Browse” and select a location. Set rotation step (deg.) to 0.4, averaging (frames) to 5, random movement to 10. Un-check the following: use 360-degree rotation, open chamber after scanning, camera offset and partial width. Check “X-Ray off after scanning.” 17. Select “X/Y alignment with a reference scan” under the “Actions” tab. Move the adjustable box to cover the sample. Adjust the “Max. x-shift from start point (+)” and “Max. y-shift from start point (+)” to 20. Change the “Matching criteria and method” to “Least-Square.” Change the “Projection pairs to match” to “All.” Click “Match,” and after the application is complete, click “Accept.” 18. In the “Start” tab, select the area in the sample to preview. For a mouse femur, there will be multiple segments. Select “Preview.” In the “Output” tab, change the graph to log view. To toggle between graphs, double-click the graph. Click the tab underneath the histogram and set the minimal value to 0. Adjust the maximal value, indicated by the red line, by dragging the line and placing it just after the peaks from the dataset end. Use this dynamic range for the rest of the samples in the project.

Generation and Characterization of Mouse Models for Skeletal Disease

185

19. In the “Output” tab, right click and drag over a nonporous part of a sample. This will open an absorbance vs. distance graph. This works best when using cortical bone. Record the shape of the graph (cupping indicates beam-hardening, arching indicates over-correction of beam-hardening, and flat indicates little to no beam hardening). Repeat two or three times in different proximal and distal regions of the sample. Click on the “Settings” tab and perform the necessary corrections for each graph type (increase the beam-hardening correction percentage for cupping, decrease the beam-hardening correction percentage for arching, and no further correction needed for flat graphs). Adjust beam-hardening settings until the graphs become flat. The same beam-hardening settings are used for the rest of the samples in a project. For a mouse femur, this number typically ranges from 20% to 40%. 20. In the “Start” tab, select an area in the sample and click the “Fine Tuning” button and select “Post-alignment.” Set number of trials to 5 and set parameter steps to 0.5 or 1. Click “Start” and refer to the “Output” tab. Toggle through the choices using the arrow buttons and choose an option that minimizes changes in pores, edges, and dense particles. If the fine tuning does not work in the range selected, manually change the misalignment compensation in the “Settings” tab. Repeat fine tuning for each part of the sample (there should be 3 or more segments for a mouse femur). 21. In the “Settings” tab, check “Ring Artifacts Reduction,” and set to 10. In the “Start” tab, select an area in the sample, click the “Fine Tuning” button, and select “Ring Artifacts Reduction.” Set number of trials to 5 and set parameter steps to 5. Click “Start” and refer to the “Output” tab. Toggle through the choices using the arrow buttons and choose an option that best reduces ring artifacts. This will only be done once per sample. 22. This application should only be used if there is noise in the background of the scan, which can be viewed in the “Output” tab. To turn on smoothing, click the “Settings” tab and check the “Smoothing” box. Adjust the levels until the noise in the background is canceled. To see the effect of smoothing, choose the “Start” tab, select an area in the sample and click “Preview.” Use the same smoothing settings for the rest of the samples in a project. 23. In the “Start” tab, select an area in the sample to view and click “Preview.” In the “Output” tab, check “Use ROI,” and choose a shape. Move and adjust the size of the ROI in the preview window so that the sample is inside the ROI. Expand the ROI

186

Gabrielle E. Foxa et al.

in different segments of the femur to confirm the whole sample is within the ROI (from proximal to distal epiphysis). 24. Choose the “Start” tab to see an overview of the scan. Move the red lines at the top and bottom of the scan to include the area of interest. 25. There are two ways to reconstruct and save the images. Option 1: Choose the “Start” tab and click the “Start” button. Option 2: Choose the “Start” tab and click the “Add to batch button.” Continue to load more datasets to the batch by choosing the “Add to batch button” after reconstructing each sample. When all datasets are complete, click “Start Batch” in batch manager. Folders with reconstructed files will appear in the scan folders. 26. Under the “Actions” tab, choose “Save” and “Select a VOI to Save.” Establish settings in the “Options for saving a VOI” window under “Saving options.” In the Resize X/Y option, choose 1. In the View to save option, choose Transaxial (X-Y). Create a new VOI folder within the reconstruction folder. Click “OK” to save the VOI. 27. Select the “regions of interest preview” icon. Select the “interpolated region of interest” drop-down menu. Select “Round.” Adjust the circle so that it is slightly smaller than the density standard and place it in the center of the density standard. 28. Under “File,” choose “Save from ROI.” Create a new ROI folder and un-check “Create folder named VOI.” Check “Save ROI-file” and click “OK” to save the ROI file. 29. Select the “regions of interest preview” icon followed by the “load regions of interest” icon. Open the “.roi” file. 30. Select the “binary selection preview” icon. Select the “From dataset” tab in the “Histogram” window. Select the “Attenuation coefficient” tab. Scroll to the bottom and record the mean (total). 31. Scroll through the images to locate where the growth plate separates. In the dataset window right click the slice chosen and click “Selection Reference.” Double-click the black bar in the dataset window to open the “Selection” window. Under the “Analytic” tab set offset 0.25 mm and height to 2.5 mm. 32. To manually draw an ROI, choose the “regions of interest preview” icon. Scroll to the lower boundary of the ROI and draw a region that includes trabecular bone but excludes the growth plate and cortical bone. Scroll toward the upper boundary of the ROI and continue drawing skipping several images in between. The ROIs drawn will interpolate automatically. To save, choose “Save from ROI” under the “File” tab. Create a new folder for the trabecular ROI and un-check

Generation and Characterization of Mouse Models for Skeletal Disease

187

“Create Folder named VOI” and check “Save ROI-file.” Click “OK” to save the ROI. 33. Save the ROI by choosing “Save from ROI” under the “File” tab. Create a new folder for the trabecular ROI. Uncheck “Create Folder named VOI” and check “Save ROI-file.” Click “OK” to save the ROI. To determine the thresholding scale, select the “binary selection preview” icon. Select the “From dataset” tab in the histogram window. Choose the “automatic thresholding” icon in the histogram window. Record the upper and lower threshold values. For a project, the thresholds for all samples should be recorded and then averaged. 34. To build a task list to create a trabecular ROI and run analysis, select the “custom processing preview” icon and select the “Internal” tab to build a task list. Click + to add a function to a task list. Do not run the task list yet. Add the functions in the following order with the described settings: thresholding (Global—enter the values determined in step 4), despeckle (type: sweep, 3D space, remove: all except largest object, and apply to: image), ROI shrink-wrap (mode: shrink-wrap, 2D space, check “stretch over holes,” diameter: 32 pixels—adjust as needed), bitwise operations (image ¼ image XOR region of interest), morphological operations (type: opening, 3D space, kernel: round, radius: 2, and apply to: image), morphological operations (type: closing, 3D space, kernel: round, radius: 2, and apply to: image), morphological operations (type: opening, 3D space, kernel: round, radius: 4, and apply to: image), morphological operations (type: closing, 3D space, kernel: round, radius: 4, and apply to: image), morphological operations (type: opening, 3D space, kernel: round, radius: 6, and apply to: image), morphological operations (type: closing, 3D space, kernel: round, radius: 6, and apply to: image), morphological operations (type: opening, 3D space, kernel: round, radius: 8, and apply to: image), morphological operations (type: closing, 3D space, kernel: round, radius: 8, and apply to: image), morphological operations (type: opening, 3D space, kernel: round, radius: 9, and apply to: image), morphological operations (type: closing, 3D space, kernel: round, radius: 9, and apply to: image), morphological operations (type: opening, 3D space, kernel: round, radius: 10, and apply to: image), morphological operations (type: closing, 3D space, kernel: round, radius: 10, and apply to: image), morphological operations (type: erosion, 2D space, kernel: round, radius: 3, apply to: image), bitwise operations (region of interest ¼ COPY Image), reload (apply to: image), despeckle (type: sweep, 2D space, remove: all except largest object, apply to: region of interest), despeckle (type: remove pores, 2D space,

188

Gabrielle E. Foxa et al.

detected: by image borders, apply to: region of interest), save bitmaps (apply to ROI, file format: BMP, check “custom subfolder” and name the folder, check “convert to monochrome” 1 bit), histogram (unit: bone mineral density, choose 3D space, check “Inside VOI,” choose a file name and location), thresholding (global thresholding, use the values previously recorded for thresholding), despeckle (remove white speckles, choose 3D space, volume: less than 6 voxels, apply to image), despeckle (remove black speckles, choose 3D space, volume: less than 6 voxels, apply to image), 3D analysis (check “basic values,” “additional values,” “trabecular thickness” and “trabecular separation,” under the “save results as” region, select a format to save the data output, and name the output of the dataset), and 3D model (apply to: image inside ROI, type of file: *.p3g, algorithm: Double-Time Cubes, check “Custom filename” and name the file). 35. To apply the ROI to the open file, select the “regions of interest preview” icon. Select the “load regions of interest” icon and open the “.roi” file. To determine the thresholding scale, select the “binary selection preview” icon. Choose the “From dataset” tab in the histogram window and select the “automatic thresholding” icon in the histogram window. Record the upper and lower threshold values. For a project, the thresholds for all samples should be recorded and then averaged. 36. To add functions to a task list, click + on a highlighted function. Add the following functions to the task list in the following order with the described settings: histogram (unit: bone mineral density, choose 3D space, check “Inside VOI,” choose a file name and location), thresholding (global thresholding, use the values previously recorded for thresholding), despeckle (remove white speckles, choose 3D space, volume: less than 6 voxels, apply to image), despeckle (remove black speckles, choose 3D space, volume: less than 6 voxels, apply to image), 3D analysis (check “basic values,” “additional values,” “trabecular thickness” and “trabecular separation,” under the “save results as” region, select a format to save the data output, and name the output of the dataset), and 3D model (apply to: image inside ROI, type of file: *.p3g, algorithm: DoubleTime Cubes, check “Custom filename” and name the file). 37. Select the “BATMAN” icon. Choose the “Add” button and load a trabecular ROI dataset (from step 3 if manual method was used and step 4 if automated method was used). Under the “Add” arrow button, choose “Load ROI” and select the “. roi” file. Add all samples within a project and click “Start” to run the task list on all samples.

Generation and Characterization of Mouse Models for Skeletal Disease

189

38. To define a reference point, first select the “regions of interest preview” icon. Define the length of the bone by using the Z-position in the “Dataset” window and calculate 45% of the femur’s length. Move the distance calculated from the distal end of the femur. Select the slice in the “Dataset” window. To make the position the selection reference, right-click and choose “Selection Reference.” Double-click the black bar in the dataset window to open the “Selection” window. Under the “Analytic” tab set offset 0 mm and height to 0.6 mm. 39. At this location in the diaphysis, mostly cortical bone remains. To create an ROI that contains the cortical bone, start by choosing the “regions of interest preview” icon. Scroll to the lower boundary of the ROI. Click the “interpolated region of interest” dropdown and choose “Elliptic.” Include the entire area of the bone within the ellipse. Scroll toward the upper boundary of the ROI and continue moving the ellipse to cover the entire tissue, skipping several images in between. The ellipses will interpolate automatically. 40. To save the cortical region of interest (ROI) choose “Save from ROI” under “File.” Create a new cortical ROI folder. Uncheck “Create folder named VOI” and check “Save ROI-file.” Click “OK” to save the ROI. 41. To apply an ROI, select the “regions of interest preview” icon. Select the “load regions of interest” icon and open the “.roi” file. To determine a thresholding scale, start by selecting the “binary selection preview” icon. Select the “From dataset” tab in the histogram window. Choose the “automatic thresholding” icon in the histogram window. Record the upper and lower threshold values. For a project, the thresholds for all samples should be recorded and then averaged. 42. Choose the “custom processing preview” icon and select the “Internal” tab to build a task list. Add the following functions to the task list in the order with the described settings: thresholding (global threshold, use values recorded from step 6), despeckle (remove white speckles, choose 3D space, volume: less than 6 voxels, apply to image), despeckle (remove black speckles, choose 3D space, volume: less than 6 voxels, apply to image), ROI shrink-wrap (mode: shrink-wrap, 2D space, check “stretch over holes,” choose appropriate number of pixels for phenotype for diameter, morphological operations and bitwise operations (use a combination as needed to rid the sample of any trabecular bone so that only cortical bone remains), despeckle (use in between morphological operations as necessary to rid the region of trabecular bone, type: sweep, 2D or 3D space depending on sample, remove: all except largest object, apply to: image), 2D analysis (under the “add to report” region

190

Gabrielle E. Foxa et al.

choose “all results,” check “Append Summary results to file,” and choose a file name and location), and 3D model (apply to: image inside ROI, type of file: *.p3g, algorithm: double-time cubes, check “custom filename” and name the file). 43. Add the following functions to the end of the task list in order with the described settings: ROI shrink-wrap (mode: shrinkwrap, 3D space to ensure that only the bone will be measured when calculating BMD, check “stretch over holes,” diameter: choose appropriate number of voxels for phenotype), reload (apply to: image), and histogram (unit: bone mineral density, 3D space, check “Inside VOI” and “Append Summary results to file,” and choose a file name and location). 44. To use Batch Manager, select the “BATMAN” icon. Choose the “Add” button and load a cortical ROI dataset. Under the “Add” arrow button, choose “Load ROI” and select the “.roi” file. Load all samples within a project and click “Start” to run the task list on all samples. 45. To minimize complications with IP injections, dosage can be split between IP and subcutaneous injections. 46. In a tissue processor, in order, clear the samples for 2–4 h each, in the following solutions: 70% ethanol, 95% ethanol 2, 100% ethanol 3, and xylene 2. 47. Add a paper label containing the specimen ID to the end of the tube. Fill the remainder of the tube with embedding solution until the tube is approximately ¾ full. 48. Monitor the volume of embedding solution for evaporation and add additional embedding solution to ensure the samples remain submerged, when necessary. Polymerization is complete when the embedding solution has turned to a hardened plastic, the tube is smooth, and no longer tacky. 49. When the saw blade cuts completely through the rod, the saw blade will automatically stop as it sits on the specimen arm. 50. Example: a 15 mm setting on the micrometer will yield a 35 μm wafer. 51. Polymerization is complete when the embedding solution has turned to a hardened plastic, the mold is smooth, and no longer tacky. 52. Expose the surface of the polymerized block using the grinder by holding the back of the block so that the face is in contact with the abrasive paper. While the block is still wet, apply pressure as needed and rotate the paper to polish away the surface of the mold. 53. Replace with fresh EDTA after 7 days. Samples should be flexible before processing.

Generation and Characterization of Mouse Models for Skeletal Disease

191

54. To prevent cross-contamination, wipe the forceps with a clean Kimwipe prior to placing them back into the forceps warmers. 55. For histological sections, 5 μm is the standard thickness.

Acknowledgements Work in our laboratories has been or is supported by the Van Andel Institute and the following grants from NIH: BOW (AR068668), AGR (AR053237), and TY (AG061086). References 1. Youlten SE, Baldock PA (2019) Using mouse genetics to understand human skeletal disease. Bone 126:27–36 2. Williams BO, Warman ML (2017) CRISPR/ CAS9 technologies. J Bone Miner Res 32 (5):883–888 3. Kamekura S et al (2005) Osteoarthritis development in novel experimental mouse models induced by knee joint instability. Osteoarthr Cartil 13(7):632–641 4. Glasson SS, Blanchet TJ, Morris EA (2007) The surgical destabilization of the medial meniscus (DMM) model of osteoarthritis in the 129/SvEv mouse. Osteoarthr Cartil 15 (9):1061–1069 5. Welch ID et al (2009) The retinoic acid binding protein CRABP2 is increased in murine models of degenerative joint disease. Arthritis Res Ther 11(1):R14 6. Fang H et al (2018) Early changes of articular cartilage and subchondral bone in the DMM mouse model of osteoarthritis. Sci Rep 8 (1):2855

7. Bouxsein ML et al (2010) Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. J Bone Miner Res 25(7):1468–1486 8. Parfitt AM et al (1987) Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry nomenclature committee. J Bone Miner Res 2(6):595–610 9. Kedlaya R et al (2013) Sclerostin inhibition reverses skeletal fragility in an Lrp5-deficient mouse model of OPPG syndrome. Sci Transl Med 5(211):211ra158 10. Ma HL et al (2007) Osteoarthritis severity is sex dependent in a surgical mouse model. Osteoarthr Cartil 15(6):695–700 11. Glasson SS et al (2010) The OARSI histopathology initiative - recommendations for histological assessments of osteoarthritis in the mouse. Osteoarthr Cartil 18(Suppl 3): S17–S23

Chapter 12 Drill Hole Models to Investigate Bone Repair Zhijun Li and Jill A. Helms Abstract Our understanding of the mechanisms underlying fracture healing is rapidly developing and is contributing to new therapeutic strategies to enhance repair. To gain new insights, animal models must also evolve. From initially imprecise, uncontrolled bone defects we now have precise injury models that still capture all of the stages and phases of bone repair yet do so in a highly reproducible manner. The simple mono-cortical defect model allows assessment of bone repair through a cartilage intermediate, e.g., endochondral ossification, as well as direct bone repair, e.g., intramembranous healing. Cellular contributions of the periosteum can be distinguished from contributions originating in the bone marrow. In this chapter, we focus on the advantages of this bone repair model, as well as its limitations. Key words Animal model, Bone regeneration, Fracture healing, Orthopedics

1

Introduction Bone fractures are one of the most common traumatic injuries and their etiologies typically involve falls and/or accidents [1]. Fractures can also occur because of underlying disease processes, e.g., cancer and osteoporosis that weaken the bone structure and increase the likelihood of fracture. Fracture severity is typically classified according to whether it is closed or open (compound), simple or comminuted, and infected or not [2, 3]. The more severe the fracture, the longer is the healing time. Despite the known importance of these variables, few of the conditions are replicated in animal models. Standardized animal models of bone repair typically simplify fracture severity, in hopes of producing reproducible injuries. For example, one commonly model using blunt trauma created by ophthalmic forceps or three-point bending equipment that is delivered to the tibia or femur and which creates a closed compound fracture [4, 5]. These fractures can then be allowed to heal with or without stabilization. In stabilized scenarios, an intramedullary “nail” (e.g., an insect pin) is typically inserted into the femur to

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_12, © Springer Science+Business Media, LLC, part of Springer Nature 2021

193

194

Zhijun Li and Jill A. Helms

align the bone ends. This method of stabilization is relatively easy, but it also makes it impossible to ascertain how cells in the bone marrow cavity contribute to the repair process [6, 7]. External, e.g., Ilizarov, fixators can be adapted to mouse models [8, 9] but tend to be technique-sensitive and time-consuming. To circumvent these issues, we and others have used a drill hole to model the stages of fracture healing [10–12]. The method is simple, highly reproducible, and requires only a few surgical tools. Producing a mono-cortical “drill hole” in the tibia does not require the placement of an intramedullary nail for stabilization; the stabilization is provided by the second, intact cortex. Consequently, the contribution of cells from the marrow cavity and surrounding soft tissues to the healing process is not impinged upon by the presence of internal or external devices [13, 14]. The holes can be of different diameters but in our hands, a smaller diameter injury is preferable because the animal experiences minimal discomfort, and in the tibia, the healing process is identical whether the injury is large or small. The animal has several anatomical sites that are amenable for drilling. In the appendicular skeleton, several sites have been used including the proximal third of the tibia, the femoral neck, and the metatarsus [15, 16]. The mandible, maxilla, and calvaria can also be used [10, 17, 18]. In all appendicular locations, the surgery consists of producing a complete osteotomy in one cortex, while leaving the far cortex intact; the resulting gap between the bone ends constitutes the fracture (see Fig. 1). The diameter of the osteotomy should be approximately equal to the thickness of the cortex (i.e., in mice, ~0.4 mm, and in rats, ~0.8 mm) in order to provide sufficient stabilization of the cut bone ends. If the diameter of the osteotomy is increased, then the risk of spontaneous fracture of the far cortex is also increased. When this happens, the fracture is considered unstable and healing ensues via a large, amorphous cartilage callus [19]. In rodents, the fracture healing process is complete within ~21 days. The initial inflammatory phase, characterized by the accumulation of granulation tissues, takes place between postsurgery days 1 and 3. The soft (cartilage) callus stage ranges from post-surgery days 3 to 10; the hard callus phase is between postsurgery days 10 and 14, and thereafter remodeling takes over until the original bone contour is achieved. One major advantage of this drill hole model is that while the endosteum heals via intramembranous ossification (see Fig. 1), the injured periosteum heals via endochondral ossification (see Fig. 2). This spatial separation between endochondral and intramembranous healing within the same injury site is especially useful when trying to distinguish a disruption in hypertrophic cartilage formation from an overall disruption in bone repair [20].

Drill Hole Models to Investigate Bone Repair

195

Fig. 1 (a) General morphological view of fibular, (b) sagittal view, (c) coronal view of 2D/3D reconstruction of intact fibular, red arrow points the surgical site. (d) Safranin O stained images show the monocortical defect in

196

Zhijun Li and Jill A. Helms

Fig. 2 (a) Pentachrome stained images show the intact bone and healing site on post-injury (PID), (b) day 3, (e) 7, and (f) 28. (c) The PCNA expressing levels and (d) Collagen type II shows by immunostaining on PID3. Abbreviations: gp growth plate, cb cortical bone, bm bone marrow

A second major advantage of the drill hole model is that the defect can be filled with a biomaterial, a biologic, a scaffold, and/or cells, and their contribution to the healing process can be readily assessed. For example, we have placed collagen sponges soaked in recombinant BMP2 into drill holes (see Fig. 3), thus eliciting in rodents the same robust, endochondral bone repair from the periosteum [21] that is reported in patients [22]. Moreover, the bone that forms is heterotopic [23], also as reported in patient populations [24]. Drill hole defects can also receive osteogenic cells, whose contribution to the healing process can be monitored (see Fig. 3). For example, osteogenic cells that co-expressed luciferase, the enzyme that cleaves luciferin to produce photons of light, were placed into a tibial defect, then in vivo live imaging was used to monitor the cells as a function of time (see Fig. 3). Histology and immunostaining with a luciferase antibody confirmed that the loss of signal was not due to a loss of cells; rather, the diminishment in signal was because the osteogenic cells secreted a mineralized matrix, the ability of the photons of light to pass through the denser tissue resulted in a loss of signal (see Fig. 3).

ä Fig. 1 (continued) the fibular. (e) Schematics shows gap between the bone ends constitutes the fracture. Pentachrome stained images show healing site on post-injury (f) day 2 and (i) 21. Safranin O stained image indicates heals via intramembranous ossification on post-injury (g) day 6; intramembranous healing show by (h) TRAP stained image at the same time point. Abbreviations: gp growth plate, cb cortical bone, bm bone marrow

Drill Hole Models to Investigate Bone Repair

197

Fig. 3 (a) Schematics and (b, c) Aniline blue stained images show the drill hole defects filled with a scaffold with BMP2 and contribution to the healing process can be readily assessed. (d) Drill hole defects can receive osteogenic cells that co-expressed luciferase, the enzyme that cleaves luciferin to produce photons of light, were placed into a tibial defect, (e) then in vivo live imaging was used to monitor the cells as a function of time. (f) Drill hole defects can be with growth factors(L-Wnt3a), pentachrome stained images show healing site with (g) L-Wnt3a and with (h)L-PBS on PID 7. (i) The model can be adapted to reflect an osteonecrotic bone lesion, (j) Aniline blue stained image shows the healing site on PID1. (k) DAPI staining show osteocytes at the cut bone ends at the same time point. (l) The adapted drill holes model can be filled with a biomaterial, a biologic, a scaffold, and/or cells, and their contribution to the healing process can be readily assessed (m, n). Abbreviations: gp growth plate, cb cortical bone, bm bone marrow

198

Zhijun Li and Jill A. Helms

The model can be adapted to reflect an osteonecrotic bone lesion (see Fig. 3). Although many animal models of osteonecrosis exist [25], most are challenging to perform [26–28] and none fully replicate the complex etiology of osteonecrosis in patients [29– 31]. Cryoablation can be performed immediately after creating the drill hole, by placing the top of a metal drill bit in contact with a piece of dry ice, then positioning the drill tip on the cut bone ends [32]. The loss of DAPI staining in osteocytes at the cut bone ends can be used to verify the extent of cryoablation [32]. Using cryoablation, a simple drill hole injury can be converted into a nonhealing osteonecrotic bone defect, which mimics at least some of the features of osteonecrosis in patients. The drill hole model has other advantages. Drilling sites are often adjacent to anatomical landmarks such as on the tibial crest or mid-shaft of the diaphysis of the femur, which allow for easy identification in tissue block during sectioning. Drill holes can also be produced in the calvaria, either at the midline overlying the sagittal sinus (see Fig. 4) or in the center of the frontal or parietal bones (see Fig. 4). In the case of calvarial defects, care should be taken to avoid any damage to the underlying dura mater (see Fig. 4). The absence of a large marrow space means that even very small diameter, e.g., 0.3 mm, calvarial defects are slow to heal. Larger calvarial defects are often referred to as “critical size” defects, although the absolute accuracy of that nomenclature has been disputed [33] . There are limitations to drill hole models. First, the holes created are small, and consequently surgeries typically require the use of a microscope. Second, in the absence of nearby landmarks, it may be difficult to identify the injury site once it has healed. The healing time is also short, and at least in the appendicular skeleton, the contributions of the marrow to the healing process are substantial. As a consequence, the potential benefit of allogeneic cells or biomaterials may not be appreciated because of the robust endogenous healing response. In these cases, it may prove beneficial to slow natural healing by cryoablation [32] or by genetic [20] or biochemical [34] means.

2

Materials

2.1 Instrument Preparation

As a survival surgery, sterile instruments must be used for each animal. Therefore, all surgical instruments are placed in an envelope and autoclaved prior to use. When performing multiple animal surgeries at one time, instruments may be sterilized between animals using a glass bead sterilizer (30 s stay at 240–270  C). Take care to ensure that instruments are no longer hot before touching any tissues. Gloves should be changed between animals, and personal protective equipment (PPE) including a head cover, face mask, and shoe covers should always be worn.

Drill Hole Models to Investigate Bone Repair

199

Fig. 4 (a) Skeletal staining show calvarial anatomy. (b) Drill holes can also be produced at the midline overlying the sagittal sinus or in the center of the frontal or parietal bones. (c, d, e) 3D and 2D reconstruction of calvarial defects model. (f) Schematics and (g) Aniline blue stained images show the case of calvarial defects. (h) Calvarial defects can filled with a biomaterial, a biologic, a scaffold, and/or cells, (i) aniline blue stained image and (k) BrdU immunoassayed image show their contribution to the healing process. Abbreviations: P parietal bone, F frontal bone, IP intraparietal bone, N nasal bone, MN mandible Table 1 Anesthetic agents Agent name

Dosage (mg/kg)

Route

Ketamine hydrochloride

70–100 mg/kg

Intraperitoneal (IP)

Xylazine

5–10 mg/kg

Intraperitoneal (IP)

Isoflurane

2–5% gas with oxygen

Inhalation (INH)

2.2 Anesthetic Agents

See Table 1.

2.3

See Table 2.

Analgesic Agents

2.4 Anti-Infective Agents

See Table 3.

200

Zhijun Li and Jill A. Helms

Table 2 Analgesic agents

Agent name

Dosage (mg/kg)

Route

Buprenorphine 0.1 mg/kg Carprofen

Duration and frequency of administration

Subcutaneous Every 6–12 h for 3 days post-op, then as needed. (SQ)

5–10 mg/kg Subcutaneous Every 24 h for 3 days, then as needed (SQ)

Buprenorphine 0.5–1.0 mg/ Subcutaneous Buprenorphine SR will be given before the start of SR kg (SQ) surgery, and will last up to 72 h post-op.

Table 3 Anti-infective agents

Agent name

Dosage (mg/kg)

Cefazolin

25 mg/kg

Subcutaneous The first dose intraoperatively and second dose 2 h (SQ) postoperatively

Enrofloxacin

10 mg/kg

Subcutaneous Only one dose intraoperatively, continue for a total of (SQ) 2 days

Route

Duration and frequency of administration

Buprenorphine 0.5–1.0 mg/ Subcutaneous Buprenorphine SR will be given before the start of SR kg (SQ) surgery, and will last up to 72 h post-op.

3

Methods

3.1 Animal Preparation

1. All animals are weighed prior to surgery and pre-surgical weights are recorded on the surgical record. 2. Ophthalmic lubricant is placed in each eye to avoid corneal drying during surgery. 3. Anesthetize the animal as per established regimens, such as indicated in Table 1. 4. Remove hair with an electric clipper, then clean loose hairs with ethanol pad. 5. At this stage, analgesics including Buprenorphine or Buprenorphine SR are given. 6. Next, the shaved skin is aseptically prepared and draped according to APLAC Guidelines for Rodent Survival Surgery. Sterilization of the area is performed using cotton-tipped applicators, alternating between a Betadine solution and 70% ethanol for a total of 3 times. 7. The animal is gently placed on top of a clean, absorbent surface positioned on a validated heat source (e.g., circulating water

Drill Hole Models to Investigate Bone Repair

201

blanket, slide warmer, instant heat device) to prevent hypothermia during surgery. 8. Before making an incision, the depth of anesthesia is confirmed via a toe-pinch; the withdrawal reflex should be abolished before proceeding. 3.2

Anesthetic

3.2.1 Isoflurane Anesthesia

1. Place the animal in the induction chamber. 2. Adjust the oxygen flowmeter to 0.8–1.5 L/min. 3. Adjust the isoflurane vaporizer to 3–5%. 4. Use the mask connected to the Bain circuit. 5. Adjust the flowmeter to 0.4–0.8 L/min. 6. Adjust the isoflurane vaporizer to 2–2.5%.

3.2.2 Injectable Anesthesia

1. Injectable anesthetic volume depends on sex, age, strain, body weight, and the body condition of the animal. 2. Prepare the cocktail solution the day before surgery. Immediately before use, make sure that the solution is thoroughly mixed. 3. In mice, injectable anesthetics are best administered via IP and IV routes. 4. Duration of anesthesia is approximately 20–30 min.

3.3 Long Bone Cortical Hole Drilling

1. After adequate anesthesia, an 8–10 mm skin incision is made in anterior surface of the tibia or on the lateral surface of the femur. 2. A full-thickness flap is raised at the incision sites. 3. The overlying muscle is gently dissected off the tibia or femur. 4. The site is gently irrigated with sterile saline. Immediately thereafter, a small hole is created using a dental drill running at 1000–2000 rpm. The drill must be new and sharp; otherwise too much pressure must be exerted to advance the drill and the risk of touching the far cortex is increased. 5. The hole created in the anterior tibial plateau or in the mid-shaft of femur should have an outer diameter of approximately 0.8–1.0 mm. 6. The drill hole should penetrate through a single tibial/femur cortex only. No irrigation is required after creating the drill hole. Bleeding usually stops spontaneously. 7. After creating the skeletal injury, soft tissues are closed with absorbable suture. 8. The skin incision is closed with nonabsorbable sutures. 9. The nonabsorbable sutures are removed 7–10 days later.

202

Zhijun Li and Jill A. Helms

10. Antibiotics are typically not required, although any signs of prolonged inflammation, e.g., longer than 24 h, may require the use of antibiotics; check with veterinary staff to ascertain whether this is necessary (see Note 1). 3.4 Calvaria Cortical Bone Hole Drilling

1. Once adequate anesthesia has been verified, a 3 mm skin incision is made in the mid-line at the caudal portion of the skull (see Note 2). 2. The periosteum is gently elevated and reflected. 3. A drill hole is created using of a micro-dissecting trephine and sterile saline irrigation. 4. Take extreme care to leave the underlying dura mater undisturbed. 5. The skin incision is closed with nonabsorbable suture. 6. The nonabsorbable sutures are removed 7–10 days later (see Note 3).

3.5 Parameters Monitored After Surgery

1. Animals are transferred into a clean cage which is partially over a 37  C heating pad. This allows animals, once ambulatory, to choose whether to stay on the warm side or on the non-heated side. 2. Separate awake animals from recovering animals. 3. Animals are allowed to ambulate freely but are closely monitored for any evidence of pain as indicated by licking, biting, crouching, vocalizations, teeth chattering, stiff walking, failure to eat, self-imposed isolation/hiding, self-mutilation, rapid breathing, open mouthed breathing, squinting, muscle rigidity, lack of muscle tone, twitching/trembling, or abnormal posture.

3.6 Early Euthanasia Criteria

1. The most likely complication of skeletal surgery is that the injury site becomes infected. During the initial recovery period (0–3 days post-surgery) closely evaluate each animal for signs of inflammation/infection, such as redness and swelling, or signs of exudate at the surgical site. 2. If signs of infection are found, consult with veterinarians about the use of antibiotics. If signs of infection persist, the animal should be euthanized. 3. The post-surgical weight of each mouse is monitored daily. If any animal exhibits loss of weight greater than 10%, they are typically removed from the study and euthanized.

Drill Hole Models to Investigate Bone Repair

4

203

Notes 1. A surgeon’s technique is a major influence on the outcome of any animal surgery. Proper sterile technique and tissue handling both decrease postoperative complications, increase survival rates, and ensure reproducible results. Whenever there is a doubt about an aspect of surgical care, consult a veterinarian. 2. Handle all tissues gently, avoiding any unnecessary tissue damage. Remember, healing duration depends on the overall health of the surgical site. 3. The size of the hole is of utmost importance when developing any bone drilling model on any sites. Consider the smallest diameter hole possible, to lessen complicates and avoid unnecessary pain during the recovery process.

Acknowledgements We thank previous Helms lab members including Drs. Philipp Leucht, Benjamin Salmon, Bo Liu, Wei Jing, and Sylvain Mouraret for their invaluable contributions over the years to our cumulative understanding of bone healing. References 1. Bell A, Templeman D, Weinlein JC (2016) Nonunion of the femur and tibia: an update. Orthop Clin North Am 47(2):365–375 2. Papakostidis C, Kanakaris NK, Pretel J, Faour O, Morell DJ, Giannoudis PV (2011) Prevalence of complications of open tibial shaft fractures stratified as per the GustiloAnderson classification. Injury 42 (12):1408–1415 3. Ktistakis I, Giannoudi M, Giannoudis PV (2014) Infection rates after open tibial fractures: are they decreasing? Injury 45 (7):1025–1027 4. Cheng L, Ye F, Yang R, Lu X, Shi Y, Li L et al (2010) Osteoinduction of hydroxyapatite/ beta-tricalcium phosphate bioceramics in mice with a fractured fibula. Acta Biomater 6 (4):1569–1574 5. Kayal RA, Siqueira M, Alblowi J, McLean J, Krothapalli N, Faibish D et al (2010) TNF-alpha mediates diabetes-enhanced chondrocyte apoptosis during fracture healing and stimulates chondrocyte apoptosis through FOXO1. J Bone Miner Res 25(7):1604–1615

6. O’Neill KR, Stutz CM, Mignemi NA, Burns MC, Murry MR, Nyman JS et al (2012) Microcomputed tomography assessment of the progression of fracture healing in mice. Bone 50 (6):1357–1367 7. Einhorn TA (1995) Enhancement of fracturehealing. J Bone Joint Surg Am 77(6):940–956 8. Tay BK, Le AX, Gould SE, Helms JA (1998) Histochemical and molecular analyses of distraction osteogenesis in a mouse model. J Orthop Res 16(5):636–642 9. Choi P, Ogilvie C, Thompson Z, Miclau T, Helms JA (2004) Cellular and molecular characterization of a murine non-union model. J Orthop Res 22(5):1100–1107 10. Leucht P, Jiang J, Cheng D, Liu B, Dhamdhere G, Fang MY et al (2013) Wnt3a reestablishes osteogenic capacity to bone grafts from aged animals. J Bone Joint Surg Am 95 (14):1278–1288 11. Tanaka K, Tanaka S, Sakai A, Ninomiya T, Arai Y, Nakamura T (2010) Deficiency of vitamin A delays bone healing process in

204

Zhijun Li and Jill A. Helms

association with reduced BMP2 expression after drill-hole injury in mice. Bone 47 (6):1006–1012 12. Behr B, Leucht P, Longaker MT, Quarto N (2010) Fgf-9 is required for angiogenesis and osteogenesis in long bone repair. Proc Natl Acad Sci U S A 107(26):11853–11858 13. Jawad MU, Fritton KE, Ma T, Ren PG, Goodman SB, Ke HZ et al (2013) Effects of sclerostin antibody on healing of a non-critical size femoral bone defect. J Orthop Res 31 (1):155–163 14. Yang F, Wang J, Hou J, Guo H, Liu C (2013) Bone regeneration using cell-mediated responsive degradable PEG-based scaffolds incorporating with rhBMP-2. Biomaterials 34 (5):1514–1528 15. Christou C, Oliver RA, Pelletier MH, Walsh WR (2014) Ovine model for critical-size tibial segmental defects. Comp Med 64(5):377–385 16. Berner A, Reichert JC, Woodruff MA, Saifzadeh S, Morris AJ, Epari DR et al (2013) Autologous vs. allogenic mesenchymal progenitor cells for the reconstruction of critical sized segmental tibial bone defects in aged sheep. Acta Biomater 9(8):7874–7884 17. Yuan X, Pei X, Zhao Y, Li Z, Chen CH, Tulu US et al (2018) Biomechanics of immediate postextraction implant osseointegration. J Dent Res 97(9):987–994. https://doi.org/ 10.1177/0022034518765757 18. Chen T, Li J, Cordova LA, Liu B, Mouraret S, Sun Q et al (2018) A WNT protein therapeutic improves the bone-forming capacity of autografts from aged animals. Sci Rep 8(1):119 19. Colnot C, Romero DM, Huang S, Helms JA (2005) Mechanisms of action of demineralized bone matrix in the repair of cortical bone defects. Clin Orthop Relat Res 435:69–78 20. Leucht P, Kim JB, Currey JA, Brunski J, Helms JA (2007) FAK-mediated mechanotransduction in skeletal regeneration. PLoS One 2(4): e390 21. Minear S, Leucht P, Jiang J, Liu B, Zeng A, Fuerer C et al (2010) Wnt proteins promote bone regeneration. Sci Transl Med 2 (29):29ra30 22. Neagu TP, Tiglis M, Cocolos I, Jecan CR (2016) The relationship between periosteum and fracture healing. Romanian J Morphol Embryol 57(4):1215–1220

23. Minear S, Leucht P, Miller S, Helms JA (2010) rBMP represses Wnt signaling and influences skeletal progenitor cell fate specification during bone repair. J Bone Miner Res 25 (6):1196–1207 24. Lorenzo J, Horowitz M, Choi Y (2008) Osteoimmunology: interactions of the bone and immune system. Endocr Rev 29 (4):403–440 25. Fan M, Peng J, Wang A, Zhang L, Liu B, Ren Z et al (2011) Emu model of full-range femoral head osteonecrosis induced focally by an alternating freezing and heating insult. J Int Med Res 39(1):187–198 26. Allen MR (2009) Bisphosphonates and osteonecrosis of the jaw: moving from the bedside to the bench. Cells Tissues Organs 189 (1–4):289–294 27. Allen MR, Kubek DJ, Burr DB, Ruggiero SL, Chu TM (2011) Compromised osseous healing of dental extraction sites in zoledronic acidtreated dogs. Osteoporos Int 22(2):693–702 28. Velez R, Soldado F, Hernandez A, Barber I, Aguirre M (2011) A new preclinical femoral head osteonecrosis model in sheep. Arch Orthop Trauma Surg 131(1):5–9 29. Zalavras CG, Lieberman JR (2014) Osteonecrosis of the femoral head: evaluation and treatment. J Am Acad Orthop Surg 22(7):455–464 30. Hernigou P, Flouzat-Lachaniette CH, Delambre J, Poignard A, Allain J, Chevallier N et al (2015) Osteonecrosis repair with bone marrow cell therapies: state of the clinical art. Bone 70:102–109 31. Hamadeh IS, Ngwa BA, Gong Y (2015) Drug induced osteonecrosis of the jaw. Cancer Treat Rev 41(5):455–464 32. Salmon B, Liu B, Shen E, Chen T, Li J, Gillette M et al (2017) WNT-activated bone grafts repair osteonecrotic lesions in aged animals. Sci Rep 7(1):14254 33. Cooper GM, Mooney MP, Gosain AK, Campbell PG, Losee JE, Huard J (2010) Testing the critical size in calvarial bone defects: revisiting the concept of a critical-size defect. Plast Reconstr Surg 125(6):1685–1692 34. Leucht P, Lam K, Kim JB, Mackanos MA, Simanovskii DM, Longaker MT et al (2007) Accelerated bone repair after plasma laser corticotomies. Ann Surg 246(1):140–150

Chapter 13 Generation and Experimental Outcomes of Closed Femoral Fracture in Mice Joseph L. Roberts, Christopher W. Kinter, and Hicham Drissi Abstract Fracture healing requires the integration of many cell types, growth factors, and cytokines that cannot be adequately studied using in vitro and in silico models. This has prompted the development of highly informative in vivo animal models to understand the complexities of fracture repair. Here, we describe a modified procedure for mice, first developed for rats by Bonnarens and Einhorn, that does not require a skin incision or suturing. This procedure involves boring a hole through the skin and articular surface of the femoral condyle with a 25-gauge needle, fixation with a K-wire, and creation of a transverse mid-diaphyseal fracture using a three-point bending fracture device. Fracture healing can be assessed using a variety of techniques, including microcomputed tomography, torsion testing, histological and histomorphometric analyses, and assessment of gene expression. There are many orthopedic trauma applications of this murine femoral fracture model ranging from assessment of safety and efficacy of novel therapeutics to the influence of specific genes on bone repair. Key words Fracture, Murine, Bone repair, Healing, Closed fracture, micro-CT, Torsion testing

1

Introduction The first report of the standard closed fracture in rats was described by Bonnarens and Einhorn in 1984 [1]. It has since been adapted to other model systems such as mice, which have become one of the most widely accepted translational models to study fracture phenotypes [2–4]. The availability of numerous transgenic mouse strains coupled with the simplicity, speed, and reproducibility of this procedure have proven useful for studying both genetic and therapeutic influences on healing [5]. Studies using this model have led to significant advancements in our understanding of the mechanisms by which fractures heal under both pathological and treatment conditions. After administration of anesthetics, this method involves the creation of a hole in the femur medullary canal by perforating the skin and articular surface of the femoral condyle with a 25-gauge

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_13, © Springer Science+Business Media, LLC, part of Springer Nature 2021

205

206

Joseph L. Roberts et al.

needle. The K-wire is then inserted into the medullary canal using a retrograde approach and covered by the skin. Finally, the operated limb is held perpendicular to the line of fracture, beneath a blunt guillotine device, wherein a transverse fracture is created by threepoint bending. Administration of preoperative analgesics helps to manage pain. Fractured femurs are typically harvested at multiple timepoints to evaluate the progression of healing. Fracture healing and quality of newly formed bone can be quantitatively assessed using microcomputed tomography (micro-CT) imaging and torsion testing, respectively. Radiographic scoring is another useful qualitative measure of fracture healing. Typically, specimens are fixed in formalin as needed and processed for histology. Histological and histomorphometric assessments allow for the assessment of fracture callus matrix composition through quantification of the distribution of cartilage, bone, and fibrous tissue. Histological sections can be further utilized for immunohistochemical analyses to detect cells and tissues that express growth factors, cytokines, and other proteins of interest during fracture healing. RT-qPCR from RNA isolated from whole callus or from histological sections is also a powerful tool to gain valuable insight into the relationship between different experimental models and treatments and expression of genes during fracture healing.

2

Materials

2.1 Closed Fracture Model

1. Personnel: It is optimal to have at least two researchers involved in this surgery. The first researcher (assistant) is needed to surgically prepare the mice, manipulate non-sterile objects, and position the mouse. The second researcher (surgeon) performs the surgeries under sterile conditions. 2. Anesthetics and analgesics: 2% isoflurane/oxygen mixture delivered via nose-cone inhalation and buprenorphine SR (1 mg/kg administered via subcutaneous injection). 3. Standard sterile surgical equipment: hemostat, chlorohexidine, 25-gauge needles, wire cutter, an appropriate length of 0.01500 diameter K-wire (average length per adult mouse femur: 15.0–15.5 mm; 316 LVM stainless steel, Small Parts), sterile drapes. All surgical equipment should be opened and maintained in a sterile fashion. 4. Preoperative preparation: Hair clippers, chlorohexidine, 70% isopropyl alcohol, gauze, Puralube Vet ointment, scale, heating pad. 5. Blunt 3-point bending guillotine device. 6. Sterile gloves, masks, bouffant cap, sterile gown. 7. X-ray device.

Generation of Murine Femoral Fracture

2.2 Microcomputed Tomography (Micro-CT)

207

1. A micro-CT machine. 2. Portable hard drive or storage device to extract and store data. 3. Scanning medium such as 70% ethanol or 1 phosphatebuffered saline. 4. Parafilm. 5. A testing protocol that indicates area of the scan/duration, tube voltage (kVP), current x time product (mAs), and special resolution (μm). 6. Foam or gauze. 7. Forceps.

2.3

Torsional Testing

1. Polymethylmethacrylate (PMMA) or dental cement (e.g., Stoelting Cat# 51458). 2. Forceps. 3. 1 cm2 Aluminum hollow cubes. 4. Cold saline solution. 5. Gauze. 6. Camera with video capabilities. 7. Caliper for measurements. 8. Personal protective equipment: gloves, laboratory coat, safety glasses. 9. Waterproof/chemical-resistant laboratory marker. 10. Torsion mechanical testing apparatus/computer with a load cell that is sensitive enough to detect sample differences in the order of 0.5–5 Nmm with appropriate tools to manipulate the machine.

2.4 Radiographic Scoring

1. Personnel: at least two trained individuals. 2. Anteroposterior and lateral radiographs. 3. Scoring criteria and assessment sheets.

2.5

Histology

1. Paraffin embedding: 10% Neutral buffered formalin, an orbital shaker, a chelator such as 14% EDTA solution (pH 7.1–7.2) for decalcification of experimental specimens, 70%, 95%, and 100% ethanol, xylene, paraffin wax, fume hood, razor blades, metal base molds for embedding, embedding cassettes, chemical resistant marker or pencil, and an embedding center to make paraffin blocks. 2. Paraffin sectioning: Paraffin blocks to be sectioned, a microtome capable of cutting sections as thin as 5 μm, a hot water bath (set at 45  C), positively charged slides, and a rack for drying slides.

208

Joseph L. Roberts et al.

3. Staining materials: Containers filled with 70%, 95%, and 100% ethanol, xylene, distilled water, slide staining rack, oven capable of reaching 56  C, cover slips, and appropriate mounting solution. 4. Immunohistochemistry materials: 1 Tris–buffered saline, 0.1% Tween-20 (TBST), quenching solution (3% hydrogen peroxide), blocking buffer (5% normal goat serum in TBST), a water bath capable of reaching 55  C, microscope with camera, an appropriate antigen retrieval solution (e.g., sodium citrate), humidified chamber, hydrophobic pen, and appropriate diluted primary and conjugated secondary antibodies (e.g., HRP-conjugated, fluorochrome-conjugated). 5. Histomorphometric quantification: Appropriate analysis software and microscope with camera. 2.6 RNA Isolation and RT-qPCR

1. Cryogenic tissue homogenizer or a mortar and pestle. 2. Liquid nitrogen. 3. Temperature-controlled centrifuge. 4. TRIzol. 5. Chloroform. 6. 100% Isopropanol. 7. 75% Ethanol. 8. RNase-free water. 9. Heating block capable of reaching 55–65  C. 10. Nuclease-free microcentrifuge tubes, pipettes, and filter pipette tips (1000 μL, 200 μL, 10 μL). 11. Surgical scissors and forceps.

2.6.1 RNA Isolation from Histological Sections

1. Freshly cut histological sections. 2. RNase-away. 3. RNase-free 10 μL tips. 4. Inverted light microscope with a 4 objective. 5. Xylene or other appropriate deparaffinization solution. 6. Kit capable of extracting RNA from formalin-fixed paraffinembedded tissue (e.g., Qiagen RNeasy FFPE or RecoverAll total nucleic acid isolation kit for FFPE). 7. Nuclease-free microcentrifuge tubes.

2.6.2 Reverse Transcription and RT-qPCR

1. UV spectrophotometer or Nano-drop machine. 2. cDNA synthesis kit. 3. Thermal cycler. 4. Real-time PCR machine.

Generation of Murine Femoral Fracture

209

5. Nuclease-free water. 6. Forward and reverse primers for genes of interest and housekeeping genes. 7. PCR plates or strips compatible with machine. 8. Centrifuge with plate holder attachments. 9. PCR tubes.

3

Methods

3.1 Closed Femoral Fracture Model

Surgeries should be performed in a clean surgical suite or procedure room on a surface disinfected with 70% ethanol or Quatricide (see Note 1). 1. Record body weight of mouse prior to fracture (see Notes 2 and 3). 2. Anesthetize the mouse by placing it in a container with air flow consisting of 2% isoflurane/oxygen mixture. 3. Once unresponsive to stimuli (e.g., pedal reflex), remove the anesthetized mouse and place on a heating pad, coat eyes with Puralube vet ointment, insert nose into nose cone with continuous flow of 2% isoflurane/oxygen mixture, and administer Burprenorphine SR (1 mg/kg subcutaneously) (see Note 4). 4. Shave the entire surgical leg and adjacent skin well with clippers, wipe shaven limb with gauze soaked in chlorohexidine, then gauze soaked in 70% isopropyl alcohol. 5. Transfer the mouse to the surgical table covered in sterile drapes, place head inside a nose cone to deliver continuous 2% isoflurane/oxygen mixture in a supine position. 6. Cut drapes to expose a hole only large enough to slip the surgically prepared limb through. 7. The surgeon holds the operated limb with the middle finger positioned underneath the knee and the index finger resting on the quadriceps muscle. The thumb is held against the tibia for balance. 8. Using their other hand, the surgeon perforates the skin and articular surface of the femoral condyle with a 25-gauge needle, rotating (without applying pressure) gently in a single direction until the medullary cavity is punctured (see Notes 5 and 6) (see Fig. 1a). 9. The surgeon picks up a K-wire of an appropriate length (average 15–15.5 mm) using a hemostat.

210

Joseph L. Roberts et al.

Fig. 1 Creation of mid-diaphyseal femoral fracture. (a) Perforate the skin and articular surface of the femur using a 25G needle to puncture the medullary canal. (b) Using a hemostat, a K-wire is inserted into the medullary canal until it reaches the proximal end of the femur. (c) Excess K-wire is cut leaving 1–2 mm exposed. (d) The exposed K-wire is bent to form a 90 angle using a hemostat, then (e) covered under the skin. (f) The surgeon positions the pinned femur perpendicular to the blunt guillotine and pressure is applied to create a mid-diaphyseal fracture

10. The needle is removed from the marrow cavity and the K-wire is inserted into the medullary canal using a retrograde approach until it reaches the proximal end of the femur (see Note 5). Correct depth of insertion is felt with resistance upon meeting cortical bone of the proximal femur (see Fig. 1b). 11. Excess K-wire is cut with wire cutters, leaving approximately 3–4 mm of K-wire that is bent to form a 90 angle, then covered under the skin without suturing (see Fig. 1c–e). 12. The surgical assistant holds the mouse in a prone position, while the surgeon positions the pinned femur perpendicular to the blunt guillotine (see Fig. 1f). 13. With their free hand, the surgical assistant raises the lower stage which applies pressure via three-point bending until an audible break is noted. 14. Take an X-ray to confirm location, type, and severity of the fracture. 15. The animal is placed in a clean cage that has food pellets located at the bottom in addition to the normal food and water.

Generation of Murine Femoral Fracture

211

16. Monitor mice daily for the first 3 days after fracture to ensure that they bear weight on their operated limbs and are eating and drinking. 17. Take radiographs at intervals (e.g., 7, 14, 21 days postfracture) to ensure that the K-wire is maintaining fixation of the fracture and track healing. 3.2

Micro-CT

Micro-CT is used to examine fracture healing at a later timepoints, mainly 14 days post-fracture. Fractured femurs can be scanned at earlier timepoints (i.e., day 7 and 10 post-fracture) to acquire 3D images, but analyses cannot be accurately completed due to the small amount of mineralized tissue within the callus at earlier timepoints. 1. Fix dissected femurs in a sufficient volume of 10% neutral buffered formalin for 1 week at 4  C prior to scanning. 2. Wash fixed bones in 1 PBS for 30 min at room temperature on an orbital shaker. Dispose of PBS and replace with fresh 1 PBS. 3. Turn on the machine and verify the pre-programmed testing conditions. 4. Carefully remove the K-wire pin from the medullary canal and discard in a designated sharps container. 5. Wrap bones in foam or gauze and insert into a scanning tube filled with appropriate scanning medium (e.g., 70% ethanol or 1 PBS). Cover top of tube with parafilm to prevent evaporation during scanning. 6. Dock the attachment and close machine. 7. Repeat steps 4 and 5 if machine is equipped with a carousel that allows batch scanning. 8. Acquire a scout-view, select the areas of interest and analysis program. The machine computes the estimated completion time of the scan (see Note 7). 9. Start the scan and wait for the indicated time until the next sample(s) may be processed. 10. If machine is not equipped with a carousel, repeat steps 2–6 until all the samples are scanned. 11. Reconstruct scan for further analysis. 12. Images are then contoured (and a global threshold is applied if applicable) to segment out existing cortical bone and analyzed (see Fig. 2).

3.3

Torsion Testing

Care must be taken to ensure the safety of both the individual performing the test and those in the surrounding area when biomechanical testing is performed. The ideal timepoint for torsion

212

Joseph L. Roberts et al.

Fig. 2 Microcomputed tomography (micro-CT) of fracture callus. (a) Manual contouring of the day 14 fracture callus to exclude existing cortical bone. (b) Example of micro-CT reconstructions of a murine fracture callus at day 14 post-fracture

testing is at the discretion of the researchers. Torsion testing is typically most informative of quality of regenerated bone in the later stages of fracture healing (21 days post-fracture). 1. At time of bone harvest, carefully remove muscle without disturbing the callus leaving the K-wire inside the medullary cavity. Wrap bone in saline-soaked gauze and store at 20  C until testing. 2. When ready to perform testing, defrost the bone in a lukewarm water bath until thawed. 3. Turn on the torsion mechanical testing apparatus with appropriate load cell attached. 4. Slowly and carefully pull the surgical K-wire from the fractured femora. Rewrap the bones in saline-soaked gauze and place on ice (see Fig. 3a). 5. Measure the total length, maximum outer diameter within the diaphyseal region, and minimum outer diameter within the diaphyseal region of the femur with a caliper and record. 6. Take pictures of all bones before testing and label as “before failure” pictures. 7. Position the femoral condyles inside a 1 cm2 square mold. Pour the PMMA within the mold to cement each extremity (see Note 8), wrap bone in cold saline soaked gauze (see Note 9), and allow it to solidify. Rotate the femur and place the femoral head/greater trochanter inside another 1cm2 square mold. Fill with PMMA and allow to solidify (see Note 10). Measure and record gauge length for all specimens. Take pictures to confirm sample alignment (see Fig. 3b–g). 8. Start and configure the software for the calibration. The testing outputs/data analysis should be set (if applicable).

Generation of Murine Femoral Fracture

213

Fig. 3 Potting and torsion mechanical testing of fractured femora. (a) The K-wire is carefully removed from the fractured limb using forceps. (b) 1 cm2 aluminum cubes are positioned and (c) PMMA powder is added to the cube until full (d). (e) PMMA liquid is added using a pipette. (f) The femur is carefully positioned inside of PMMA using forceps and allowed to solidify. (g) The opposite end of the femur is then potted. The fully potted femur is placed in a torsion machine (h) and torsion is applied until failure (i). (j) Typical data obtained from torsion testing

9. Place the first specimen in the testing apparatus and tighten one end into place (see Fig. 3h). Remove the gauze around the bone. 10. Being careful not to damage the specimen, slowly tighten the second end until locked. After the position is set, zero the torque and run the test while recording a video. Record inner and outer maximum diaphyseal diameters for all samples after testing. Take pictures to confirm site of failure (see Fig. 3i). 11. Repeat this procedure for all samples. 12. Turn off all machinery. 13. Extract the data and analyze the torque curves to verify testing accuracy (see Fig. 3j). 3.4 Radiographic Scoring

1. Randomly shuffle or code radiographs of fractures to avoid any experimental group associations (see Note 11). 2. The observers evaluate anteroposterior and lateral radiographs in a blinded fashion using a validated radiographic scoring system, subdivided into the following categories: (a) periosteal and endosteal reaction, (b) callus opacity, and (c) cortical remodeling and bridging (see Note 12) (see Fig. 4).

214

Joseph L. Roberts et al.

Category

0

1

2

3

Periosteal and Endosteal Reaction

No reaction

Mild reaction

Moderate reaction

Marked reaction

Callus Opacity

No evidence of mineralization

Heterogeneous with minimal mineralization and cortices well demarcated

Heterogeneous with moderate mineralization and partially confluent with the cortices

Confluent with the cortices, uniform

All cortical edges seen but illdefined

Minimal cortical union (3 cortical edges visible) without reformation of the medullary canal

Partial cortical union (1-2 cortical edges visible) with visible medullary canal

Cortical Remodeling and Bridging

No apparent remodeling

4

Complete cortical union (no cortical edges visible) with well demarcated medullary canal

Fracture 1

Fracture 2

Fracture 3

Periosteal and Endosteal Reaction

3

3

3

Callus Opacity

1

2

3

Cortical Remodeling and Bridging

1

2

4

Fig. 4 Radiographic scoring criteria and representative scores of three independent fractured femora

(a) Periosteal and endosteal reaction is determined using a four-point scoring system: 0 ¼ no reaction, 1 ¼ mild reaction, 2 ¼ moderate reaction, and 3 ¼ marked reaction. (b) Callus opacity is determined using a four-point scoring system: 0 ¼ no evidence of mineralization, 1 ¼ heterogeneous with minimal mineralization and cortices well demarcated, 2 ¼ heterogeneous with moderate mineralization and partially confluent with the cortices, and 3 ¼ confluent with the cortices, uniform. (c) Cortical remodeling and bridging are determined using a five-point scoring system: 0 ¼ no apparent remodeling, 1 ¼ all cortical edges seen but ill defined, 2 ¼ minimal cortical union (three cortical edges visible) without reformation of the medullary canal, 3 ¼ partial cortical union

Generation of Murine Femoral Fracture

215

(1–2 cortical edges visible) with visible medullary canal, and 4 ¼ complete cortical union (no cortical edges visible) with well-demarcated medullary canal. 3. Collect evaluation sheets, average scores for each reviewer, and perform appropriate statistical analyses. 3.5

Histology

3.5.1 Paraffin Embedding

The methodology outlined below describes general procedures for histological fixation, sample preparation, sectioning, and histomorphometric quantification. Due to the wide variety of staining procedures, the experimenter is advised to consult relevant literature specific to the staining procedures needed to visualize desired tissue components (see Note 13). 1. After euthanasia, collect operative limbs and carefully remove excessive soft tissue and muscle. Some soft tissue should be left to maintain the structural integrity of the periosteum and the fractured bone. 2. Fix bones in 10% neutral buffered formalin (20:1 fixative to tissue ratio) at 4  C for 1 week. 3. Rinse fixed bones in PBS for 30 min on an orbital shaker. 4. Place bones in labeled embedding cassettes and submerge in a beaker containing a sufficient volume of a decalcification agent (e.g., 14% EDTA) (see Note 14). Place submerged bones on an orbital shaker at room temperature. Change decalcification solution every 2–3 days, for 2–3 weeks. Decalcification is complete when bone is pliable and can be confirmed by X-ray. 5. Rinse decalcified bones twice by submerging in 1 PBS for 30 min with agitation. 6. To dehydrate bones, place cassettes in four separate changes of 70% ethanol and 95% ethanol for 20 min each. Followed by one solution of 100% ethanol for 30 min, then overnight in 100% ethanol with gentle shaking. The next day submerge the bones in xylene for 10 min each for a total of 4 changes. 7. The bones are then placed in heated xylene:paraffin (1:1 ratio; 56  C) for 3 h, then allowed to solidify at room temperature overnight. The following morning the xylene:paraffin containing the bones is melted, then the bones are placed in three separate heated liquid paraffin (paraffin 1, 2, 3) solutions for 2 h each at 56  C. 8. Remove bones from cassettes and carefully retrieve the intramedullary pin from each fractured femur. 9. Turn-on embedding station to heat paraffin tank to 56  C and to cool the cold plate. 10. Embed bones, with the femoral head pointing up, in hot paraffin wax positioned in the center of a metallic mold on an embedding station and allow the blocks to cool.

216

Joseph L. Roberts et al.

11. Remove any air bubbles that form in the center of the wax mold. 12. Allow samples to cool on the embedding station for 30 min to 1 h before removing them from the metal molds and storing at 4  C or  20  C. 3.5.2 Paraffin Sectioning

1. Label all positively charged slides with specimen identification (e.g., animal ID, genotype) and section number with a chemical-resistant marker or pencil. 2. Turn on the hot water bath at least 30 min before sectioning. 3. Place a sharp microtome blade in the microtome. 4. Remove paraffin blocks from the refrigerator or  20  C freezer and place on ice with paraffin side facing down toward the ice. 5. Carefully position each block in the microtome and coarsely section (15–20 μm) until the tissue is exposed. 6. Place the block on ice for 10–20 min. 7. Section the block at a setting of 5–7 μm until the mid-sagittal plane is reached (see Note 15). 8. Carefully place sections on surface of the water in the water bath to stretch the sections. 9. Allow paraffin sections to stand for 1–2 min in the water bath, before carefully placing onto slides. 10. Repeat this procedure until the sample is fully sectioned or until desired number of slides have been collected. 11. Place slides on a slide drying rack and allow slides to dry overnight at room temperature.

3.5.3 Histological/IHC Staining

1. Bake slides in an oven set at 56  C for 15–30 min or until paraffin is melted. 2. Rehydrate slides (generally 10 min in xylene, 5 min in xylene, 1 min in 100% ethanol, 1 min in 100% ethanol, 1 min in 95% ethanol, 1 min in 95% ethanol, 5 min in deionized water). 3. Conduct histological staining/IHC procedure (differs depending on experiment: consult online literature) on rehydrated slides (see Notes 16–18). 4. Dehydrate slides (generally 1 min in 95% ethanol, 1 min in 100% ethanol, 1 min in 100% ethanol, 1 min in xylene, 1 min in xylene) (see Note 19). 5. Add appropriate mounting medium, carefully apply cover slip and remove any bubbles by gently applying pressure to the cover-slipped slide (see Note 20). 6. Allow slides to dry at room temperature overnight.

Generation of Murine Femoral Fracture 3.5.4 Histomorphometric Quantification

217

1. Visualize slides under a microscope using an appropriate magnification. 2. Use software such as ImageJ, ImagePro, and Osteomeasure for image analyses. Different analyses may require different software (consult chosen software documentation for additional details). Mineralized tissue, fibrous tissue, cartilage matrix can be quantified, and normalized to the total callus area. The total number of cells and cells expressing certain factors can be counted and their total normalized to the area of analyses (see Note 21). 3. Analyze data using appropriate statistical approaches.

Gene Expression

Special care must be taken to ensure that solutions and work area are not contaminated with RNases/DNases, which can compromise the validity of the experiment. It is recommended to spray and wipe all laboratory benches, pipettes, and pipette tip boxes with RNase Away prior to starting the experiment.

3.6.1 RNA Isolation from Whole Fracture Callus

1. Isolate fracture calluses following euthanasia by removing all surrounding soft tissue and muscle, removing the pin, and surrounding bone from the intact fracture callus. Wrap the callus in labeled aluminum foil and flash freeze in liquid nitrogen. Store bones at 80  C until ready for further processing.

3.6

2. Homogenize calluses into a fine powder using either a tissue homogenizer or a mortar and pestle, maintained at 80  C. 3. Place homogenized samples in a nuclease-free microcentrifuge tube and add an appropriate volume of TRIzol. Incubate for 5 min at room temperature. 4. Add 0.2 mL of chloroform per 1 mL TRIzol reagent to the samples and shake tubes vigorously by hand for 15 s. 5. Incubate samples at room temperature for 3 min and subsequently centrifuge them at 12,000  g for 15 min at 4  C. 6. Carefully isolate the upper clear aqueous phase and pipette it into a new tube, taking care not to disturb the white interphase or lower pink aqueous phase. 7. Add 0.5 mL of 100% isopropanol per 1 mL TRIzol used, mix by pipetting, and incubate at room temperature for 10 min. Centrifuge the mixture at 12,000  g for 10 min at 4  C. 8. Remove the supernatant from the tube, leaving only the RNA pellet, which is washed in 1 mL 75% ethanol per 1 mL TRIzol used. 9. Briefly vortex the mixture and centrifuge at 7500  g for 5 min at 4  C.

218

Joseph L. Roberts et al.

10. Carefully aspirate the ethanol wash using a pipette and air-dry the pellet for 10–15 min. It is important to prevent complete drying of the pellet. 11. Depending on the size of the RNA pellet, resuspend with appropriate volume of RNase-free water (approximately 10–30 μL). 12. Incubate the resuspended pellet in a heating block set at 65  C for 5 min (or 55  C for 10 min). 13. Proceed with reverse transcription or store isolated RNA at 80  C for downstream applications. 3.6.2 RNA Isolation from Histological Sections

1. Cut fresh sections (7–10 μm) from paraffin-embedded fractured femurs. 2. Carefully place sections on surface of the water in the water bath to stretch the sections. 3. Allow paraffin ribbon sections to stand for 1–2 min in the bath, before carefully placing onto slides. 4. Repeat this procedure until a total of 5–7 sections have been collected. 5. Allow sections to dry on a drying rack that has been cleaned with RNase-Away. 6. Spray microscope stage with RNase-Away. 7. Visualize fracture callus under microscope and using a sterile 10 μL pipette tip carefully scrape the callus (see Fig. 5). The scraped tissue will adhere to the pipette tip via electrostatic forces. 8. Place scraped tissue into a nuclease-free microcentrifuge tube. 9. Repeat until all slides have been scraped.

Fig. 5 Histological sections of a day 14 fracture callus before and after scraping the callus (red outline)

Generation of Murine Femoral Fracture

219

10. Deparaffinize the collected scraped tissue using an appropriate deparaffinization reagent (i.e., xylene) according to the RNA isolation from FFPE kit manufacturer’s instructions. 11. Isolate RNA using an RNA isolation from FFPE kit according to manufacturer’s instructions. 12. Proceed with reverse transcription or store isolated RNA at 80  C for downstream applications. 3.6.3 Reverse Transcription

1. Remove RNA from 80  C and thaw on ice. 2. Determine RNA concentration and purity using a Nanodrop or UV spectrometer (see Note 22). Repeat for each sample. 3. Reverse transcribe the same quantity of RNA for each sample (typically 1 μg) according to the cDNA kit. cDNA obtained from reverse transcription can be diluted and stored at 20  C.

3.6.4 Quantification of Gene Expression

1. Prepare a master mix for each gene of interest and housekeeping gene, containing forward primer, reverse primer, and SYBR green mix (if SYBR green is used). Vortex tubes to mix, and centrifuge briefly (see Note 23). 2. First add an appropriate volume of master mix to each well of the PCR plate. Then add the cDNA obtained from each callus to the corresponding wells of the PCR plate. There should be at least two, but ideally three replicates per cDNA and gene of interest. It is also recommended to include a no-template control (NTC) which only contains master mix and no cDNA. 3. Seal plate with plastic cover tightly and centrifuge briefly. 4. Place plate into the RT-qPCR machine. Add labels to the software and run PCR using the optimized conditions for the sample of interest (see Note 24). 5. Record CT values and enter them into an excel spreadsheet for data analysis. 6. Perform data analysis for obtained CT values, relating expression of genes of interest to housekeeping genes (see Note 25).

4

Notes 1. All metallic equipment and the surgical tray should be cleaned with enzymatic detergent prior to surgery. All metallic equipment, gauze, and drapes should be wrapped in surgical drapes and autoclaved prior to surgery. Needles should be opened in a sterile fashion over the opened sterilized surgical tray. 2. It is important to monitor body weight during healing and to determine the dose of analgesic to administer.

220

Joseph L. Roberts et al.

3. It is recommended to use mice that are older than 10 weeks of age for fracture healing experiments. 4. Analgesics can also be administered postoperatively, but any drug administration needs to be preapproved in the animal protocol. Buprenorphine-HCL (Buprenex®) is an alternative analgesic to buprenorphine SR but requires daily intramuscular injections during the initial 3-day postoperative monitoring period. NSAIDs should not be used as routine analgesics because they can negatively influence fracture healing [6]. 5. Before performing this procedure, it is suggested that the surgeon practice inserting the K-wire. 6. An alternative to the method described here involves creation of a medial parapatellar incision to expose the femoral condyles prior to boring the hole in the medullary cavity. This would require suturing and administration of antibiotics. 7. It is recommended to scan ex vivo samples with the following parameters: 6 μm resolution, 70 kVp tube voltage, and tube current 114 mA, with 200 ms integration time. Scan parameters can change depending on the particular experiment. 8. There are several alternatives to potting specimens with PMMA. Common examples include low-temperature Field’s metal and Bondo®. Fields metal is melted using a double burner and poured into the metal molds, which then solidifies in approximately 5 min. Bondo® (polyester resin sold in kits) forms moldable putty when mixed with a hardener, which subsequently sets and hardens into a solid geometry. 9. PMMA is hardened through an exothermic reaction wherein a liquid methylmethycrylate monomer (should not exceed three times the volume of the potted specimen region) is added to a polymer powder. Bones should be kept in chilled saline-soaked gauze when potting to preserve the biomechanical properties of the bone. 10. The time for polymerization of PMMA is variable and depends on the ratio of PMMA powder to liquid. 11. Researchers should ensure that group associations remain coded throughout the evaluation process and that they do not introduce any external bias. 12. It is best to have either orthopedic surgeons or experienced orthopedic researchers who are familiar with the scoring scale participate as observers for radiographic evaluation. 13. Processing of bones including decalcification can influence staining and other downstream applications. When possible, all bones belonging to a single experiment should be processed at the same time to minimize variability within experiments.

Generation of Murine Femoral Fracture

221

14. Many decalcification agents are available, and the most appropriate one is dependent on the desired downstream application. 10%–14% EDTA is a slower decalcifier but works well for most histological stains used to assess fracture healing. Aciddecalcifiers work quicker but should not be used if there is an interest in completing enzymatic staining like TRAP. 15. Occasionally the embedded bones contain residual mineral. If this is observed, the block should be surface decalcified by removing the block from the microtome and placing it face down in an appropriate decalcification solution (e.g., 14% EDTA) for 30–60 min. This will only remove mineral near the surface, so additional surface decalcification may be required as the paraffin block is further sectioned. 16. Different staining procedures require different materials and dyes. To determine if additional materials are needed and to familiarize the researcher with each specific staining procedure, it is recommended to consult online literature before beginning an experiment. Tissue processing can also affect tissue processing. The staining time for each stain must be optimized for each experiment. 17. The tissue of interest will determine the type of histological staining used. Hematoxylin and Eosin is useful for gross morphology and identifying callus bone, while Safranin O/Fast Green and Alcian Blue/Hematoxylin staining is useful for identifying cartilage. 18. IHC primary antibody dilutions should be verified in the researcher’s lab with positive and negative controls prior to any IHC staining. 19. Dehydration of stained slides should not be completed for certain stains such as TRAP because it can weaken staining intensity. Slides with these stains should be coverslipped using an aqueous mounting medium and sealed using fingernail polish by painting around the edge of the coverslip. 20. If the applied pressure does not remove air bubbles from under the coverslip, soak the slide in xylene until the coverslip falls off. Add mounting medium to the exposed tissue and apply a new coverslip. 21. There are different techniques for quantifying the number of cells after IHC staining. Some studies set a baseline level of staining intensity and count positive cells above that level. Other studies count all cells with moderate staining and above. 22. Ideally RNA should have a 260/280 ratio of 1.8–2, with a 260/280 ratio of 2 indicating pure RNA. An abnormal 260/280 ratio suggests contamination with ethanol, phenol,

222

Joseph L. Roberts et al.

guanidine, or other reagents used to isolate RNA. Very low concentrations of RNA (3 times the thickness of the adjacent cartilage)

Table 3 Assessment of osteophyte maturity in mouse knee joints Osteophyte maturity

Features

0

None

1

Predominantly cartilage

2

Mix of cartilage and bone

3

Predominantly bone

Note 14). Depending upon the selected antibody or the tissue processing (fixation, decalcification method, etc.), the optimal antigen retrieval method, the primary and secondary antibody concentration, or the antibody blocking solution will vary, and therefore each IHC protocol should be carefully optimized. Below is provided an example of an IHC protocol using the C1,2C antibody (IBEX #50–1035) on formalin-fixed, paraffin-embedded knee sections in mice post-DMM. 3.4.1 Immunoperoxidase Staining

1. Deparaffinize sections in xylene, 2 times for 10 min each. 2. Rehydrate gradually through a series of graded ethanol concentrations: 100% (2 times, 3 min each); 95% (2 min); 85% (1 min); 70% (1 min); 50% (1 min). Finally, wash in 1
PBS (2 times, 5 min each). 3. Perform antigen retrieval treatment with 2 mg/mL hyaluronidase for 30 min at 37  C in a humid chamber (see Note 15). 4. Wash in 1
PBS buffer (3
each for 5 min). 5. Incubate for 15 min at RT in 3% hydrogen peroxide in dH2O to quench endogenous peroxidase activity. 6. Wash in 1
PBS buffer (2
each of 5 min). 7. Incubate in 1.5% normal goat blocking solution in a humid chamber for at least 1 h at room temperature (see Note 16).

DMM Mouse Model of Post-traumatic Osteoarthritis

245

Fig. 7 Representative photomicrograph images of Safranin O-stained coronal sections taken from the knee (wild type mouse at 8 weeks post-DMM surgery) with yellow outlines indicating osteophytes. The composition of the osteophyte is dependent on its maturity: Early osteophytes are composed mainly of cartilage (a), and then transition to a combination of both cartilage and bone (b) before they develop into mature osteophytes that consist mainly of bone (c). The size and maturity of the osteophyte often correlates with the severity of OA found in the joint post surgery, with small osteophytes consisting mainly of cartilage observed in joints with a low histological score (d), and mature osteophytes consisting mainly of bone observed in highly damaged joints (e). Osteophyte formation is modest in wild type mice subjected to DMM surgery, but may be enhanced by certain gene modifications. Graphs can be constructed based on scoring of multiple knee joints using Tables 2 and 3

8. Remove blocking solution from slides and incubate with primary antibody C1,2C at 1:200 dilution or the corresponding isotype-matched negative control IgG (see Note 17) in the normal goat blocking buffer with 1.5% Tween overnight at 4  C in a humid chamber. 9. Wash in 1
PBS buffer (2
each of 5 min). Optional: Add 0.5% Tween to the 1
PBS buffer from this step if the background is expected to be high. 10. Incubate with biotin-conjugated secondary antibody for 1 h at room temperature in a humid chamber. 11. Wash in 1
PBS buffer (2
each of 5 min). 12. Incubate with avidin biotin enzyme reagent for 30 min at room chamber in a humid chamber. 13. Wash in 1
PBS buffer (2
each of 5 min).

246

Kirsty L. Culley et al.

14. Incubate with DAB chromogen or Vector NovaRed until desired stain intensity develops. Comparative slides (e.g., wild type versus knockout and positive and negative controls) should be monitored to determine the proper development time. 15. Wash sections in dH2O for 2 min to stop the reaction. 16. The provided C1,2C IHC protocol does not include a counterstaining step. If required, counterstaining methods include: (a) Fast green 0.2% solution: stain for 2 min, followed by quick washing in 95% EtOH. (b) Hematoxylin (filtered): Stain for 30 s, followed by immediate washing with several changes of dH2O, then with 95% EtOH. 17. Rinse quickly in 100% EtOH, and then perform a second wash in 100% EtOH for 2 min. 18. Incubate in Xylene (2
for 5 min each). 19. Mount slides with 2 drops of Vectamount medium, cover with a glass coverslip, and observe by light microscopy. 3.4.2 Immunofluorescence Staining Protocol

The immunofluorescence (IF) protocol is commonly used when there is a need to detect multiple cellular targets by simultaneous labeling; a mix of primary antibodies is followed by a combination of secondary antibodies conjugated to diverse fluorochromes emitting light at different wavelengths. The following protocol is intended as a general guide for immunofluorescence on paraffinembedded sections: 1. Follow steps 1–4 of the IHC protocol (see Subheading 3.4.1). 2. Incubate with normal blocking serum for 1 h at room temperature in a humid chamber (see Note 16). 3. Incubate with optimized primary antibody concentration in a humid chamber overnight at 4  C (see Note 17). If using a primary antibody conjugated with a fluorochrome, omit step 5. 4. Wash in 1
PBS buffer with 0.5% Tween 20 (2
each of 5 min). 5. Incubate with fluorochrome-conjugated secondary antibody for 2 h at room temperature in a humid chamber (make sure from this step onward the samples are shielded from light). Secondary antibodies are conjugated to a wide range of fluorochromes to suit the users’ needs (e.g., IgG-FITC, IgG–TR, IgG-CY3, IgG-Cy5, Alexa Fluor or DyLight, Chromeo, SureLight, etc.). As cartilage autofluoresces in the green spectrum, it is recommended to use fluorochromes that do not fall within this spectrum.

DMM Mouse Model of Post-traumatic Osteoarthritis

247

6. Wash in 1
PBS buffer (2
each of 5 min) in a slide container covered with aluminum foil. 7. Mount slides with an antifade mounting medium (e.g., ProLong Gold antifade reagent with DAPI) and follow drying instructions of the manufacturer. 8. Visualize staining with a fluorescence microscope. 3.5 Cartilage Microdissection for RNA and DNA Extraction for Gene Expression and DNA Methylation Analyses

For nucleic acid isolation, cartilage is dissected from the femoral heads and tibial plateaus of the knee and homogenized in TRIzol (Invitrogen). Below (see Subheadings 3.6 and 3.7) we describe methods to isolate total RNA using mirVana™ miRNA isolation kits (Ambion), for work with mRNA and miRNA, or the RNeasy® Mini (Qiagen) kits, following the manufacturer’s instructions with additional modifications. Following these methods, an average of 50–80 ng of high-quality total RNA is obtained from the articular cartilage isolated from one knee joint. This should serve as a guide for the number of mice required to achieve a sufficient amount of RNA for gene expression analyses. For RNA isolation, it is critical to place each leg in RNAlater (steps 1–5) as early as possible to obtain good RNA integrity; thus, one person completes steps 1–7 and a second person completes steps 8–12. Person 1 1. Sacrifice mouse and immediately remove hind legs. 2. Use dissection scissors to quickly remove as much soft tissue from the legs as possible. 3. Place the legs in separate Petri dishes and cover with ice-cold 1
PBS. 4. While working on one leg, keep the other leg in 1
PBS on ice, complete steps 5–7 on the other leg. 5. Dissect remaining muscle and tendon with a scalpel and size 15 blade under the microscope. Remove dissected soft tissues from the Petri dish, as these tissues may be sources of RNases. 6. Transfer cleaned leg to a fresh Petri dish and wash with ice-cold 1
PBS. 7. Transfer the washed leg into a fresh small Petri dish and cover with RNAlater. Keep the leg on ice in RNAlater until ready to complete steps 8–13. Person 2: Keep legs in RNAlater at all times during the following steps in the Petri dish placed under the dissection microscope, use a scalpel with size 11 blade to separate the tibia and femur and expose the articular surfaces.

248

Kirsty L. Culley et al.

8. Remove any remaining soft tissue or tendon surrounding the articular surfaces with a scalpel or small dissection scissors. Be careful not to damage the cartilage. 9. Place the bones in a fresh Petri dish and cover with RNA later, as small amounts of soft tissue will make dissected cartilage difficult to decipher. 10. While the leg is bathed in RNAlater use a scalpel with size 11 blade to carve the cartilage from the articular surfaces of the tibia and femur. 11. Place the harvested cartilage in RNase-free Eppendorf tube containing RNAlater (enough to completely cover the cartilage sample). Samples can be pooled in order to obtain a higher RNA yield. 12. Either: (a) Proceed immediately with the RNA isolation using the mirVana kit and the modified protocol below; or. (b) Place the cartilage in RNAlater overnight at 4  C. Remove excess RNA later in the morning, snap freeze, and store at 80  C until ready to isolate RNA. RNAlater is not required for DNA isolation from microdissected articular cartilage, which is performed as described above using clean ice-cold 1
PBS in all steps (1–10). After microdissection (step 11), place articular cartilage in nuclease-free Eppendorf tube and, step 12(a) proceed immediately with DNA isolation as detailed in Subheading 3.8; or step 12(b) snap-freeze tissues in liquid Nitrogen and store (liquid Nitrogen) until use. High-quality DNA suitable for epigenomic analyses is obtained using Gentra® Puregene® DNA isolation kits (Qiagen), following the manufacturer’s instructions with additional modifications (described in Subheading 3.8). Following this method, an average of 200–450 ng of high-quality DNA is obtained from microdissected articular cartilage from one knee joint. This should serve as a guide for the number of mice required to achieve a sufficient amount of DNA for analyses. 3.6 RNA Isolation from Cartilage Using a Modified mirVana miRNA Isolation Kit (Ambion) Protocol

The following procedure is performed according to the manufacturer’s instructions (Ambion) (see Note 18), except that additional phenol:chloroform steps have been introduced to help improve the 260/280 values of the RNA isolated. To obtain good RNA integrity (RIN) values, all reagents, pipet tips, and Eppendorf tubes must be certified nuclease-free. In addition, isolation should be completed on a clean work space treated with RNAse Zap (Ambion). 1. Place the Eppendorf tube containing cartilage sample on ice and add 500 μL of TRIzol to the tube. Homogenize the cartilage sample in TRIzol using a 1.5-mL RNAse-free pestle

DMM Mouse Model of Post-traumatic Osteoarthritis

249

and handheld homogenizer, for approximately 10 min (or until confident that the cartilage is fully homogenized). Each knee can be homogenized separately, with the option of pooling multiple samples onto the mirVana filter column at Step 18. 2. Add 200 μL TRIzol (to total volume of 700 μL) and vortex to mix. 3. Transfer TRIzol with disrupted tissue to QIAshredder to further homogenize and disrupt cells: centrifuge for 2 min at 16,000
g. 4. Transfer sample to a fresh nuclease-free Eppendorf tube, taking care not to take or disrupt the pellet of matrix that will be visible. 5. Add 300 μL TRIzol and vortex to mix (total volume: 1 mL). 6. Add 200 μL chloroform:IAA per 1 mL TRIzol and vortex for 15 s. 7. Incubate samples on ice for 5 min. 8. Centrifuge at 13,000
g for 10 min at 4  C. 9. Remove the aqueous phase, taking care not to disrupt the interface, and transfer to new RNase-free tube (make note of volume). 10. Add 1 volume of phenol:chloroform:IAA to the aqueous phase and vortex for 15 s. 11. Incubate on ice for 10 min. 12. Centrifuge at 16,000
g for 15 min at 4  C. 13. Remove aqueous phase and transfer to new RNase-free tube (make note of volume). 14. Add 1 volume of chloroform:IAA to the aqueous phase, vortex for 15 s to mix. 15. Incubate on ice for 5 min. 16. Centrifuge at 10,000
g, 10 min, 4  C. 17. Remove aqueous phase and transfer to new RNase-free tube (make note of volume). 18. Add 1.25
aqueous volume of nuclease-free 100% ethanol (room temperature) to the aqueous phase. From this step onward, reagents from the mirVana miRNA isolation kit will be used. These steps will isolate total RNA, including miRNA, within the same fraction. However, the kit also provides the option for isolating miRNA and mRNA in different fractions (refer to the user’s manual included with the kit for further information).

250

Kirsty L. Culley et al.

1. For each sample, place a mirVana filter cartridge into a collection tube. Add the lysate/ethanol mix to the filter, 700 μL at a time, and centrifuge (~15 s, 10,000
g, room temperature). Discard flow-through. (For sample volumes larger than 700 μL repeat in successive applications to the same filter). 2. Apply 700 μL of miRNA wash solution 1 (mirVana Kit) to column and centrifuge (~5 to 10 s, 10,000
g, room temperature). Discard flow-through. 3. Apply 500 μL of miRNA wash solution 2/3 (mirVana Kit) to column and centrifuge (~5–10 s, 10,000
g, RT). Discard flow-through. 4. Repeat step 19. 5. Spin column to remove residual fluid (1 min, 10,000
g, room temperature). 6. Transfer filter cartridge to fresh collection tube. 7. Add 100–50 μL nuclease-free water (room temperature) to the filter and leave for 1 min. 8. Centrifuge to elute (~30 s, 16,000
g) (see Note 19). 9. Apply the eluted 50 μL to the filter cartridge and spin again (~30 s, 16,000
g) (see Note 20). 10. To a 50 μL-volume of RNA sample, add 0.1 volume (5 μL) of DNase buffer and 1 μL DNase (all reagents provided with the DNA-free Kit from Ambion). 11. Incubate at 37  C for 25 min. 12. Add 0.1 volume (5.5 μL) of inactivating reagent, mix well, and incubate at room temperature for 2 min (mixing occasionally, at least 3 times). 13. Centrifuge at 10,000
g, 1.5 min, 4  C. 14. Remove the supernatant and place in a nuclease-free 1.5mL tube. 15. Quantify RNA using a NanoDrop spectrophotometer. 16. RNA integrity value can be assessed at this point using a Bioanalyzer (Agilent) through a service usually provided by the institutional genomics core. 3.6.1 Ethanol Precipitation

This precipitation step can be completed to concentrate the RNA if the concentration obtained above is not enough for required analyses. 1. On ice, add 94 μL of ice-cold water to each 50 μL of RNA sample (volumes should now be about 144 μL). 2. Add 16 μL of cold 3 M NaOAc, pH 5.0, buffer to each sample to make give a concentration of around 0.3 M. Mix by pipetting up and down. Final volume will be 160 μL.

DMM Mouse Model of Post-traumatic Osteoarthritis

251

3. To each sample add 2 μL of 20 μg/μL UltraPure Glycogen. Mix by pipetting up and down (see Note 21). 4. Add 2.5 parts of ice-cold 100% ethanol to each sample— pipette up and down to mix. 5. Place the samples at 80  C for at least 1 h (can be left overnight). 6. Centrifuge at maximum speed, 10 min at 4  C. 7. The RNA pellet should now be visible in the Eppendorf tube. Decant ethanol. 8. Add 750 μL of ice-cold 70% ethanol to the pellet, gently vortex, and centrifuge at 10,000
g for 5 min. 9. The RNA pellet should now be visible. Decant ethanol. Centrifuge samples quickly (approximately 5 s) to collect any remaining ethanol in the bottom of the tube and remove it with a filtered, nuclease-free tip. 10. Resuspend RNA in 15 μL of sterile nuclease-free water. 11. Re-assess RNA concentration spectrophotometer.

using

a

NanoDrop

12. Assess RIN value using a Bioanalyzer (Agilent). 3.7 RNA Isolation from Cartilage Using a Modified RNeasy Mini RNA Isolation Kit (Qiagen) Protocol

This modified RNA isolation protocol combines the TRIzol® RNA extraction method with the Qiagen Rneasy® mini kit protocol. The protocol modifications are intended to improve RNA purity (A260/280), while preserving RNA integrity (RIN > 7). All reagents, pipet tips, and Eppendorf tubes must be certified nuclease-free. In addition, isolation should be completed on a clean workspace treated with RNAse Zap (Ambion). 1. Label 2 mL microcentrifuge tubes compatible with TissueLyser and dispense 1 stainless steel bead per tube. 2. Place microdissected cartilage (see Note 22) in the 2 mL microcentrifuge tube along with 700 μL of Trizol and homogenize the cartilage at 300 Hz for 2 min using the TissueLyser II. If needed, repeat the step again once. See Note 23 for subsequent homogenization. (a) Alternatively, if no TissueLyser is available, homogenize the cartilage using 1.5 mL RNAse-free pestle and handheld homogenizer, for approximately 5–10 min on ice, as described in Subheading 3.6. 3. Transfer the homogenized sample in 700 μL of TRIzol to a QIAshredder to further homogenize and disrupt cells: centrifuge for 2 min at 16,000
g.

252

Kirsty L. Culley et al.

4. Transfer the flow-through to a fresh nuclease-free Eppendorf tube, taking care not to disrupt or take the pellet of matrix that will be visible. 5. Add 300 μL TRIzol and vortex to mix (total volume: 1 mL). 6. Add 200 μL chloroform:IAA per 1 mL TRIzol and vortex for 15 s. 7. Incubate samples on ice for 5 min. 8. Centrifuge at 13,000
g for 10 min at 4  C. 9. Remove the aqueous phase, taking care not to disrupt the interface, and transfer to a new nuclease-free tube (make note of volume). 10. Add 1 volume of phenol:chloroform:IAA to the aqueous phase and vortex for 15 s. 11. Incubate on ice for 10 min. 12. Centrifuge at 16,000
g for 15 min at 4  C. 13. Remove aqueous phase and transfer to new nuclease-free tube (make note of volume). 14. Add 1.25
aqueous volume of 100% molecular-grade ethanol (room temperature) to the aqueous phase (for example, add 625 ul of 100% ethanol to 500 ul of aqueous phase). 15. Mix thoroughly by inverting the tube and let it stand at room temperature for 2 min. From this step onward, reagents from the RNAeasy RNA isolation kit will be used and procedure will be followed as per manufacture’s instruction (refer to the user’s manual included with the kit for further information). 1. Add the lysate/ethanol mix to the Qiagen mini column filter, 700 μL at a time, and centrifuge (~15 s, 10,000
g, room temperature). Discard flow-through. (For sample volumes larger than 700 μL repeat in successive applications to the same filter). 2. Wash the column with 350 μL of RW1 wash buffer, by centrifuging for ~15 s at 10,000
g, room temperature. 3. DNase treatment: add 80 μL of DNase mixture directly on the column (without touching the filter or the walls of the column), incubate at room temperature for 15 min. (a) To make DNase mixture: add 70 μL of RDD buffer and 10 μL of DNase. 4. Add 350 μL of RW1 wash buffer to wash the column, and centrifuge for ~15 s at 10,000
g, room temperature. Discard the flow-through.

DMM Mouse Model of Post-traumatic Osteoarthritis

253

5. Add 500 μL of RPE buffer and centrifuge for ~15 s at 10,000
g, room temperature. Discard the flow-through. 6. Repeat washing with 500 μL RPE buffer, centrifuge for ~2 min at 10,000
g, room temperature. Discard the collection tube. 7. Transfer the column to new collection tube and centrifuge for 1 min at 10,000
g, to remove any residual RPE buffer from the column. Discard the collection tube. 8. Transfer the column to fresh nuclease-free 1.5 mL Eppendorf tube. Add 30 μL of RNase-free water to the filter (without touching the filter), let the column stand for 1 min at room temperature. Centrifuge for 1.5 min at 10,000
g, room temperature to elute the RNA. (a) Optional: to concentrate and/or fully elute RNA, add back the same 30 μL of eluted nuclease-free water to the filter and repeat the elution step 7 (see Note 20). 9. Quantify RNA using a NanoDrop spectrophotometer. 10. RNA integrity value can be assessed at this point using a Bioanalyzer. 3.8 DNA Isolation from Cartilage Using a Modified Gentra Puregene DNA Isolation Kit (Qiagen) Protocol

The Gentra® Puregene® DNA isolation protocol has been modified to increase DNA yield and purity. All reagents, pipet tips, and Eppendorf tubes must be certified nuclease-free, and all isolation steps should be completed in a clean workspace. 1. Place freshly isolated or frozen knee cartilage in 1.5 mL nuclease-free Eppendorf tube (see Note 22 for comment regarding required amount of starting material). 2. Homogenize the cartilage using 1.5 mL pestle and handheld motorized homogenizer. 3. Add 300 μL of cell lysis solution to the grounded cartilage, and heat the sample at 65  C for 30 min. 4. Add 1.5 μL of proteinase K to the sample tube, shake, and incubate at 55  C for 1 h. If cartilage is not completely dissolved, continue heating at 55  C for up to 3 hrs. Cartilage must be completely digested before moving to the next steps. 5. Add 1.5 μL of RNase A solution, mix thoroughly by inverting the tube, and incubate at 37  C for 45 min. 6. Quickly transfer the sample to ice and allow to cool for 1 min. 7. Add 100 μL of protein precipitation solution and vortex vigorously for 20 s to mix, incubate on ice for 5 min. 8. Centrifuge sample tube for 3 min at 13000
g. A visible whitecolored precipitated protein pellet is formed at the bottom of the tube.

254

Kirsty L. Culley et al.

9. If pellet is loose, repeat the incubation for 5 min on ice and repeat step 8. 10. Transfer the supernatant to a fresh nuclease-free 1.5 mL Eppendorf tube. 11. Add 300 μL of ice-cold molecular-grade isopropanol to the supernatant. 12. Add 0.5 μL of glycogen (stock concentration: 20 μg/μL) to the isopropanol and supernatant mixture, and invert 50 times. 13. Incubate the tube at 20  C for 1.5 h. Centrifuge for 10 min at 13000
g. 14. A transparent DNA pellet will be visible at this stage. Remove the supernatant carefully and discard without disturbing the DNA pellet. 15. Add 300 μL of 70% ethanol and invert tube several times to wash DNA. Centrifuge at 13000
g for 1 min. 16. Carefully discard the supernatant without disturbing the DNA pellet (whitish, at this stage). 17. Repeat steps 17 and 18 twice, to ensure a thorough washing of the DNA pellet. 18. Carefully discard the supernatant and remove remaining excess ethanol by completely (and carefully) draining the tube by inverting on Kimwipes, taking care that the pellet remains in the tube. Avoid overdrying the DNA pellet as DNA will be difficult to resuspend. 19. Add 50 μL of DNA hybridization buffer to the tube with the dry DNA pellet, incubate at 65  C for 1 h to dissolve DNA. 20. If needed, incubate at room temperature (15–25  C) overnight with gentle shaking. 21. Quantify DNA using a NanoDrop spectrophotometer, and perform required quality control steps to ensure that the quality is sufficient for downstream analyses.

4

Notes 1. To induce Cre-recombinase-mediated deletion of the floxed gene only in adult chondrocytes, mice receive three intraperitoneal injections of either tamoxifen (knockout) or vehicle alone (wild-type control) at 2-day intervals of 2.5 mg per 10 g of body weight of mouse. However, the concentration and dose should be chosen carefully depending on the age of the animal, and therefore the current literature should be reviewed. The final injection is scheduled 1 week prior to surgery, which is usually performed at 12 weeks of age,

DMM Mouse Model of Post-traumatic Osteoarthritis

255

allowing the mice time for complete gene ablation and recovery from tamoxifen treatment before anesthesia. To observe the effects of gene modification on OA development, comparisons should be done in tamoxifen- versus vehicle-treated littermates. 2. Different doxycycline concentrations have been used successfully, with no reported adverse effect [35, 36], but with differences in the time required for transgene activation to occur due to the varying amount of time required for doxycycline to clear from the mouse system. The concentrations achieved by the doses administered are not sufficient to inhibit collagenase activity. Adequate controls have to be used for comparison after the DMM surgery [36]. 3. All procedures must be approved by the Institutional Animal Care and Use Committee (IACUC). Before live animals are used, all personnel must obtain CLAS orientation and training, including demonstration to veterinarians the surgical procedure on cadaveric mice. 4. See Glasson et al. [37] for excellent photographs and schematics to guide the surgery. 5. It is recommended that one surgeon perform all surgeries involving each genetically modified mouse strain to reduce variability. Comparisons between wild type and knockout or transgenic strains are completed using littermates. Non-operated or sham controls need to be checked to ensure that there is no nonspecific effect of the genetic modification. 6. The fat pad may also be left in place and merely transected to allow access and visualization of the MMTL. 7. Significant postoperative pain and debility are not anticipated. Mice are monitored postoperatively by the Veterinary Staff, and if there is evidence of pain, suffering, or illness, analgesia and/or other treatment will be administered at their discretion. Particular attention will be paid to ensure that animals are ambulating normally after the procedure. For further information on analgesic dugs, please see reviews [39, 42]. 8. Since only male mice are used in the DMM model because of the protection by estrogen in females [37], it is necessary to establish the following procedures to avoid aggressive behavior. Mice are housed together prior to surgery (completed as soon as possible post weaning if males are not from the same litter), and the same mice are housed together post-surgery in a cage containing some of their original dirty bedding. Administration of analgesics is continued according to the protocol. Surgical mice are monitored for the first few hours after surgery and immediately the next morning. If a dominant male is noticed, it is immediately separated from the group. The

256

Kirsty L. Culley et al.

remaining mice in the cage are monitored daily for fighting and any new emerging dominant male is removed. A minimum of 3 mice must remain housed together post surgery to encourage the level of activity required to promote OA initiation, development, and progression. 9. Immediately after surgery, place one mouse tunnel per cage to provide enrichment and promote the level of activity required to promote OA initiation, development, and progression. 10. For histology, mice may be sacrificed by CO2 inhalation. For RNA extraction, if a CO2 tank is not immediately available, cervical dislocation may be required to avoid rigor mortis before tissues can be dissected from the joints and placed in extraction buffer. 11. Alternatively to test if the samples are decalcified, a needle can be passed through the bone of the sample. If no resistance is felt, the sample can be considered decalcified. 12. During sectioning check carefully every fifth slides to ensure the knee is correctly orientated in the paraffin. The lateral femoral condyle will usually appear in the first sections. Refer to Figs. 4 and 5 to gain a good concept of the proper knee orientation. If after 15 to 20 collected slides all four quadrants are not identified, discard the knee. 13. Previous studies by Glasson et al. [37] found that the mean maximum histological scores at 4 weeks postoperatively for the unoperated, sham surgery, DMM, and ACLT groups, respectively, were (S.E.M.) 1.1  0.1, 1.0  0.3, 3.7  1.5, and 4.3  0.4; and at 8 weeks the mean maximum scores were 1.2  0.3,1.2  0.2, 4.1  0.3, and 5.0  0.4. 14. The conditions for immunohistochemistry and immunofluorescence (e.g., antigen retrieval method, primary and secondary antibody concentration) may vary and require optimization. The provided protocols are guidelines and have been optimized for applying the specified antibodies to mouse knee joints utilizing the reagents indicated. 15. Hyaluronidase treatment is just one of many available antigen retrieval methods. In general, the retrieval method depends upon the antibody selected and the tissue processing, and therefore requires optimization for the antibody used for detection of specific antigen epitopes in a given tissue. It is advisable to perform optimization steps comparing conditions without antigen retrieval with one or two antigen retrieval methods. Retrieval methods include: (a) Heat retrieval in a sodium citrate buffer pH 6.0 (20 min at 95  C). (b) Treatment with hyaluronidase (2 mg/mL 30 min at 37  C).

DMM Mouse Model of Post-traumatic Osteoarthritis

257

(c) Treatment with pepsin (5 mg/mL in 0.02% HCL, for 45 min at 37  C). (d) Treatment with 0.05% saponin solution for 30 min RT. 16. Blocking solutions are used to reduce background and diminish nonspecific staining. Many blocking methods exist and the correct method should be used for your chosen antibody. If normal serum is used for blocking, the correct serum should be chosen to avoid interaction with primary and secondary antibodies, and the tissue being stained. Ideally the serum chosen should be derived from the same species in which the secondary antibody is raised, or from an unrelated species. Increasing the incubation time with blocking serum can further reduce background staining. In addition, adding non-ionic detergents (such as Tween) can reduce nonspecific hydrophobic interactions and help permeabilize the tissue to reach intracellular epitopes. 17. Optimal antibody concentrations should be determined empirically, by titration in the blocking buffer. Always incubate slides with positive (tissue known to express your protein of interest control or commercially available positive controls) and negative controls (isotype non-immune immunoglobulin control at the same concentration as the primary antibody, or a tissue that does not express the antigen). 18. The mirVana miRNA isolation kit (Ambion) can be purchased with or without phenol. This protocol is optimized using the mirVana kit without phenol, with the addition of TRIzol (Invitrogen). 19. Some downstream sequencing protocols require a minimum amount of RNA at a certain concentration (ng) per μL. The 50 μL elution volume is a smaller volume than that recommended by the manufacturer, but can result in more concentrated RNA, which can sometimes prevent the need to complete ethanol precipitation in order to concentrate the sample. 20. Re-apply the eluted sample back onto the filter to increase both yield and concentration of RNA. 21. Glycogen can interfere with some downstream sequencing methods and should be investigated before use as a carrier. 22. Pooled microdissected cartilage from 3 knee joints yielded optimal results (high purity and integrity, and required concentration) for RNA-Seq (and RTqPCR validation) and Reduced Representation of Bisulfite Sequencing analyses in cartilage samples retrieved at different time points after DMM surgeries. 23. If some fragments or cartilage remain intact after the TissueLyser step, use 1.5-mL RNAse-free pestle and handheld homogenizer as described in Subheading 3.6.

258

Kirsty L. Culley et al.

Acknowledgements Research related to this topic was supported by National Institutes of Health grants R21-AG049980, R01-AG022021, and RC4-AR060546. Kirsty L. Culley, Purva Singh, Mary B. Goldring, and Miguel Otero contributed equally to this work. References 1. Poulet B, Ulici V, Stone TC, Pead M, Gburcik V, Constantinou E, Palmer DB, Beier F, Timmons JA, Pitsillides AA (2012) Time-series transcriptional profiling yields new perspectives on susceptibility to murine osteoarthritis. Arthritis Rheum 64 (10):3256–3266. https://doi.org/10.1002/ art.34572 2. Poulet B, Hamilton RW, Shefelbine S, Pitsillides AA (2011) Characterizing a novel and adjustable noninvasive murine joint loading model. Arthritis Rheum 63(1):137–147. https://doi.org/10.1002/art.27765 3. Ko FC, Dragomir C, Plumb DA, Goldring SR, Wright TM, Goldring MB, van der Meulen MC (2013) In vivo cyclic compression causes cartilage degeneration and subchondral bone changes in mouse tibiae. Arthritis Rheum 65 (6):1569–1578. https://doi.org/10.1002/ art.37906 4. Sato T, Konomi K, Yamasaki S, Aratani S, Tsuchimochi K, Yokouchi M, Masuko-HongoK, Yagishita N, Nakamura H, Komiya S, Beppu M, Aoki H, Nishioka K, Nakajima T (2006) Comparative analysis of gene expression profiles in intact and damaged regions of human osteoarthritic cartilage. Arthritis Rheum 54(3):808–817 5. Aigner T, Fundel K, Saas J, Gebhard PM, Haag J, Weiss T, Zien A, Obermayr F, Zimmer R, Bartnik E (2006) Large-scale gene expression profiling reveals major pathogenetic pathways of cartilage degeneration in osteoarthritis. Arthritis Rheum 54(11):3533–3544 6. Glasson SS (2007) In vivo osteoarthritis target validation utilizing genetically-modified mice. Curr Drug Targets 8(2):367–376 7. Little CB, Fosang AJ (2010) Is cartilage matrix breakdown an appropriate therapeutic target in osteoarthritis--insights from studies of aggrecan and collagen proteolysis? Curr Drug Targets 11(5):561–575. https://doi.org/10. 2174/138945010791011956 8. Bernardo BC, Belluoccio D, Rowley L, Little CB, Hansen U, Bateman JF (2011) Cartilage intermediate layer protein 2 (CILP-2) is expressed in articular and meniscal cartilage

and down-regulated in experimental osteoarthritis. J Biol Chem 286(43):37758–37767. https://doi.org/10.1074/jbc.M111.248039 9. Yasuhara R, Ohta Y, Yuasa T, Kondo N, Hoang T, Addya S, Fortina P, Pacifici M, Iwamoto M, Enomoto-Iwamoto M (2011) Roles of beta-catenin signaling in phenotypic expression and proliferation of articular cartilage superficial zone cells. Lab Investig 91 (12):1739–1752. https://doi.org/10.1038/ labinvest.2011.144 10. Lodewyckx L, Cailotto F, Thysen S, Luyten FP, Lories RJ (2012) Tight regulation of winglesstype signaling in the articular cartilage - subchondral bone biomechanical unit: transcriptomics in Frzb-knockout mice. Arthritis Res Ther 14(1):R16. https://doi.org/10.1186/ ar3695 11. Loeser RF, Olex AL, McNulty MA, Carlson CS, Callahan MF, Ferguson CM, Chou J, Leng X, Fetrow JS (2012) Microarray analysis reveals age-related differences in gene expression during the development of osteoarthritis in mice. Arthritis Rheum 64(3):705–717. https://doi.org/10.1002/art.33388 12. Nuka S, Zhou W, Henry SP, Gendron CM, Schultz JB, Shinomura T, Johnson J, Wang Y, Keene DR, Ramirez-Solis R, Behringer RR, Young MF, Hook M (2010) Phenotypic characterization of epiphycan-deficient and epiphycan/biglycan double-deficient mice. Osteoarthritis Cartilage 18(1):88–96. https://doi.org/10.1016/j.joca.2009.11.006 13. Glasson SS, Askew R, Sheppard B, Carito B, Blanchet T, Ma HL, Flannery CR, Peluso D, Kanki K, Yang Z, Majumdar MK, Morris EA (2005) Deletion of active ADAMTS5 prevents cartilage degradation in a murine model of osteoarthritis. Nature 434(7033):644–648 14. Stanton H, Rogerson FM, East CJ, Golub SB, Lawlor KE, Meeker CT, Little CB, Last K, Farmer PJ, Campbell IK, Fourie AM, Fosang AJ (2005) ADAMTS5 is the major aggrecanase in mouse cartilage in vivo and in vitro. Nature 434(7033):648–652 15. Little CB, Barai A, Burkhardt D, Smith SM, Fosang AJ, Werb Z, Shah M, Thompson EW

DMM Mouse Model of Post-traumatic Osteoarthritis (2009) Matrix metalloproteinase 13-deficient mice are resistant to osteoarthritic cartilage erosion but not chondrocyte hypertrophy or osteophyte development. Arthritis Rheum 60 (12):3723–3733 16. Wondimu EB, Culley KL, Quinn J, Chang J, Dragomir CL, Plumb DA, Goldring MB, Otero M (2018) Elf3 contributes to cartilage degradation in vivo in a surgical model of posttraumatic osteoarthritis. Sci Rep 8(1):6438. https://doi.org/10.1038/s41598-01824695-3 17. Culley KL, Lessard SG, Green JD, Quinn J, Chang J, Khilnani T, Wondimu EB, Dragomir CL, Marcu KB, Goldring MB, Otero M (2019) Inducible knockout of CHUK/IKKalpha in adult chondrocytes reduces progression of cartilage degradation in a surgical model of osteoarthritis. Sci Rep 9(1):8905. https://doi.org/ 10.1038/s41598-019-45334-5 18. Ismail HM, Miotla-Zarebska J, Troeberg L, Tang X, Stott B, Yamamoto K, Nagase H, Fosang AJ, Vincent TL, Saklatvala J (2016) Brief report: JNK-2 controls aggrecan degradation in murine articular cartilage and the development of experimental osteoarthritis. Arthritis Rheumatol 68(5):1165–1171. https://doi.org/10.1002/art.39547 19. Rowe MA, Harper LR, McNulty MA, Lau AG, Carlson CS, Leng L, Bucala RJ, Miller RA, Loeser RF (2017) Reduced osteoarthritis severity in aged mice with deletion of macrophage migration inhibitory factor. Arthritis Rheumatol 69(2):352–361. https://doi.org/ 10.1002/art.39844 20. Miotla Zarebska J, Chanalaris A, Driscoll C, Burleigh A, Miller RE, Malfait AM, Stott B, Vincent TL (2017) CCL2 and CCR2 regulate pain-related behaviour and early gene expression in post-traumatic murine osteoarthritis but contribute little to chondropathy. Osteoarthritis Cartilage 25(3):406–412. https://doi. org/10.1016/j.joca.2016.10.008 21. Hwang SM, Feigenson M, Begun DL, Shull LC, Culley KL, Otero M, Goldring MB, Ta LE, Kakar S, Bradley EW, Westendorf JJ (2018) Phlpp inhibitors block pain and cartilage degradation associated with osteoarthritis. J Orthop Res 36(5):1487–1497. https://doi. org/10.1002/jor.23781 22. Holyoak DT, Chlebek C, Kim MJ, Wright TM, Otero M, van der Meulen MCH (2019) Low-level cyclic tibial compression attenuates early osteoarthritis progression after joint injury in mice. Osteoarthritis Cartilage 27 (10):1526–1536. https://doi.org/10.1016/j. joca.2019.06.005

259

23. Echtermeyer F, Bertrand J, Dreier R, Meinecke I, Neugebauer K, Fuerst M, Lee YJ, Song YW, Herzog C, Theilmeier G, Pap T (2009) Syndecan-4 regulates ADAMTS-5 activation and cartilage breakdown in osteoarthritis. Nat Med 15(9):1072–1076. http://www. nature.com/nm/journal/v15/n9/suppinfo/ nm.1998_S1.html 24. Lin AC, Seeto BL, Bartoszko JM, Khoury MA, Whetstone H, Ho L, Hsu C, Ali AS, Alman BA (2009) Modulating hedgehog signaling can attenuate the severity of osteoarthritis. Nat Med 15(12):1421–1425 25. Sampson ER, Hilton MJ, Tian Y, Chen D, Schwarz EM, Mooney RA, Bukata SV, O’Keefe RJ, Awad H, Puzas JE, Rosier RN, Zuscik MJ (2011) Teriparatide as a chondroregenerative therapy for injury-induced osteoarthritis. Sci Transl Med 3(101):101ra193. https://doi. org/10.1126/scitranslmed.3002214 26. Chockalingam PS, Sun W, Rivera-Bermudez MA, Zeng W, Dufield DR, Larsson S, Lohmander LS, Flannery CR, Glasson SS, Georgiadis KE, Morris EA (2011) Elevated aggrecanase activity in a rat model of joint injury is attenuated by an aggrecanase specific inhibitor. Osteoarthritis Cartilage 19(3):315–323. https:// doi.org/10.1016/j.joca.2010.12.004 27. Johnson K, Zhu S, Tremblay MS, Payette JN, Wang J, Bouchez LC, Meeusen S, Althage A, Cho CY, Wu X, Schultz PG (2012) A stem cellbased approach to cartilage repair. Science 336 (6082):717–721. https://doi.org/10.1126/ science.1215157 28. Rai MF, Hashimoto S, Johnson EE, Janiszak KL, Fitzgerald J, Heber-Katz E, Cheverud JM, Sandell LJ (2012) Heritability of articular cartilage regeneration and its association with ear-wound healing. Arthritis Rheum 64:2300. https://doi.org/10.1002/art.34396 29. Hashimoto S, Rai MF, Janiszak KL, Cheverud JM, Sandell LJ (2012) Cartilage and bone changes during development of post-traumatic osteoarthritis in selected LGXSM recombinant inbred mice. Osteoarthritis Cartilage 20 (6):562–571. https://doi.org/10.1016/j. joca.2012.01.022 30. Nakamura E, Nguyen MT, Mackem S (2006) Kinetics of tamoxifen-regulated Cre activity in mice using a cartilage-specific CreER(T) to assay temporal activity windows along the proximodistal limb skeleton. Dev Dyn 235 (9):2603–2612. https://doi.org/10.1002/ dvdy.20892 31. Dao DY, Jonason JH, Zhang Y, Hsu W, Chen D, Hilton MJ, O’Keefe RJ (2012)

260

Kirsty L. Culley et al.

Cartilage-specific beta-catenin signaling regulates chondrocyte maturation, generation of ossification centers, and perichondrial bone formation during skeletal development. J Bone Min Res 27(8):1680–1694. https://doi. org/10.1002/jbmr.1639 32. Henry SP, Jang CW, Deng JM, Zhang Z, Behringer RR, de Crombrugghe B (2009) Generation of aggrecan-CreERT2 knockin mice for inducible Cre activity in adult cartilage. Genesis 47(12):805–814. https://doi.org/10.1002/ dvg.20564 33. Henry SP, Liang S, Akdemir KC, de Crombrugghe B (2012) The postnatal role of Sox9 in cartilage. J Bone Min Res 27 (12):2511–2525. https://doi.org/10.1002/ jbmr.1696 34. Gossen M, Bujard H (1992) Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc Natl Acad Sci U S A 89(12):5547–5551 35. Grover J, Roughley PJ (2006) Generation of a transgenic mouse in which Cre recombinase is expressed under control of the type II collagen promoter and doxycycline administration. Matrix Biol 25(3):158–165. https://doi.org/ 10.1016/j.matbio.2005.11.003 36. Xu L, Polur I, Servais JM, Hsieh S, Lee PL, Goldring MB, Li Y (2011) Intact pericellular matrix of articular cartilage is required for unactivated discoidin domain receptor 2 in the mouse model. Am J Pathol 179

(3):1338–1346. https://doi.org/10.1016/j. ajpath.2011.05.023 37. Glasson SS, Blanchet TJ, Morris EA (2007) The surgical destabilization of the medial meniscus (DMM) model of osteoarthritis in the 129/SvEv mouse. Osteoarthr Cartil 15 (9):1061–1069 38. Ma HL, Blanchet TJ, Peluso D, Hopkins B, Morris EA, Glasson SS (2007) Osteoarthritis severity is sex dependent in a surgical mouse model. Osteoarthritis Cartilage 15 (6):695–700. https://doi.org/10.1016/j. joca.2006.11.005 39. Flecknell PA (1996) Laboratory animal anesthesia, 2nd edn. Academic Press, London 40. Glasson SS, Chambers MG, Van Den Berg WB, Little CB (2010) The OARSI histopathology initiative - recommendations for histological assessments of osteoarthritis in the mouse. Osteoarthritis Cartilage 18(Suppl 3): S17–S23. https://doi.org/10.1016/j.joca. 2010.05.025 41. Loeser RF, Olex AL, McNulty MA, Carlson CS, Callahan M, Ferguson C, Fetrow JS (2013) Disease progression and phasic changes in gene expression in a mouse model of osteoarthritis. PLoS One 8(1):e54633. https://doi. org/10.1371/journal.pone.0054633 42. Jenkins WL (1987) Pharmacologic aspects of analgesic drugs in animals: an overview. J Am Vet Med Assoc 191(10):1231–1240

Chapter 15 Immunostaining of Skeletal Tissues Crystal Idleburg, Madelyn R. Lorenz, Elizabeth N. DeLassus, Erica L. Scheller, and Deborah J. Veis Abstract Immunostaining is the process of identifying proteins in tissue sections by incubating the sample with antibodies specific to the protein of interest, then visualizing the bound antibody using a chromogen (immunohistochemistry or IHC) or fluorescence (immunofluorescence or IF). Unlike in situ hybridization, which identifies gene transcripts in cells, immunostaining identifies the products themselves and provides information about their localization within cells (nuclear, cytoplasmic, or membrane) or extracellular matrix. This can be particularly important in the context of bone and cartilage because they contain many cell types as well as matrix components, each with distinct protein expression patterns. As the number of antibodies continues to grow, this technique has become vital for research laboratories studying the skeleton. Here, we describe a detailed protocol for antibody-based in situ analysis of bone and associated tissues, addressing specific issues associated with staining of hard and matrix-rich tissues. Key words Immunostaining, Immunohistochemistry, Immunofluorescence, Bone, Cartilage, Decalcification, Fixation, Antibodies, Antigen retrieval

1

Introduction In both clinical and research studies, histology-based methods are critical for describing phenotypes in patients and in experimental organisms. There are 3–4 main steps for immunostaining: 1. Incubation with antigen-specific primary antibody. 2. Incubation with an enzyme-, biotin-, or fluorophoreconjugated secondary antibody. 3. Detection of secondary antibody via an enzymatic reaction that produces a colored precipitate (IHC). 4. Imaging using standard light (IHC) and/or confocal fluorescence microscopy (IF).

Crystal Idleburg and Madelyn R. Lorenz contributed equally to this work. Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_15, © Springer Science+Business Media, LLC, part of Springer Nature 2021

261

262

Crystal Idleburg et al.

However, the tissue collection and processing steps that come prior to immunostaining are crucial and can greatly affect the image quality. This is particularly true for bone and cartilage where it is necessary to decalcify tissue while maintaining matrix components such as proteoglycans. Therefore careful attention must be paid to each step from tissue harvest and fixation to decalcification and antigen retrieval [1]. Mistakes and overprocessing at any of these steps can damage antigenic epitopes, tissue morphology, or adhesion of tissue to slides, making it difficult to assess morphology and obtain good staining. 1.1

Fixation

Fixation is the process of treating tissue with solutions that preserve gross morphology as well as molecular structures within the tissue and should be started as soon as possible after harvest [2, 3]. Penetration of fixative is determined by the size and nature of the tissue of interest. Soft tissues and small pieces of tissue will fix faster than larger or harder tissues. The standard fixative for paraffin embedding is 10% neutral buffered formalin (NBF), while the most typical for frozen sections is 4% paraformaldehyde (PFA). However, when applied properly, either fixative can be used for each embedding method. As with most fixatives, these solutions preserve tissue by cross-linking the proteins. Therefore, tissues fixed in 10% NBF and 4% PFA usually require antigen retrieval before incubation with primary antibody. Because of the cross-linking action, it is important to avoid over-fixation as this can lead to excessive cross-linking, which may mask the antigens, or dehydration, which may produce an undulation artifact during sectioning [2]. Fixation in neutral buffered zinc formalin, which prevents excessive cross-linking, should also be considered for frozen sections, as it often eliminates the need for antigen retrieval steps [4]. Although unfixed frozen sections are useful in some situations such as reporter cell lines or mice that express fluorescent proteins, fixation is almost always required when performing antibody-based detection, and results are generally better when tissue is fixed up front, rather than dipping sections into fixative. To ensure proper preservation when working with bone or cartilage, it is necessary to clean away any unwanted soft tissue such as skin and muscle. This allows for fixative penetration in a timely manner and avoids under- or over-fixation. It also makes it easier to orient the bone during embedding. Additionally, undesired autofluorescence of blood cells can be removed by perfusing the animal with phosphate-buffered saline (PBS) and 10% NBF or 4% PFA to rinse and fix the vasculature, followed by a post-fix of immersion in 10% NBF or 4% PFA [5].

1.2

Decalcification

Before embedding bone in paraffin or OCT, it is essential to soften the tissue by lowering the calcium content (i.e., decalcification). The duration of decalcification and degree of calcium ion removal

Immunostaining of Skeletal Tissues

263

are influenced by the solution used. Most commercial solutions are acids, either mineral or organic, and soften bones quickly, but they can easily damage the tissue and are generally not compatible with immunostaining. Another less common method is the use of an ion-exchange resin to exchange ammonium ions for calcium ions. While this method yields the best morphology, especially in bone marrow specimens, it is also the most expensive option [2]. The most useful decalcifying method for immunostaining is treatment with 14% ethylene diamine tetraacetic acid (EDTA) [2, 3, 6]. This gentler chelating agent may decalcify hard tissues more slowly but is less likely to damage tissue or affect antigenicity. Even with EDTA, it is important to monitor and optimize the decalcification duration. Failure to do so can lead to poor morphology and weak antibody-based staining. Lastly, if decalcification is not an option for your desired antigen or preparation, specialized tape transfer methods can be used to capture calcified sections of frozen or resinembedded tissues [7]. 1.3

Antigen Retrieval

Due to the cross-linking action of most fixatives, it is often necessary to unmask antigens before staining [8]. The choice of retrieval method will vary according to the antigens and antibodies used. There are several methods of antigen retrieval but they fall into two main categories, enzyme digestion and heat treatment. Each retrieval method presents its own challenges and needs optimization for different specimen types. Enzyme digestion requires precision in pH and duration of treatment because different tissues will digest at different rates. The challenge in heat retrieval is in treating the tissue long enough to ensure antigen retrieval without causing it to lift off from the slide, which is a common problem when working with cartilage and bone. For frozen sections, if retrieval is necessary, enzymatic methods should be of primary consideration due to the fragility of hydrated, aqueous-treated tissues. Antigenicity in frozen sections can also be improved by use of a buffer solution containing detergent (such as Tween 0.05–0.1%).

1.4

Data Analysis

To accurately interpret staining, it is important to know the standard morphology and staining patterns in the tissue of interest. Textbooks on histology, pathology, and developmental biology can be a good resource for identifying the cells and structures. To interpret the staining itself, the first priority is in determining whether the signal (whether a chromogen or fluorescence) is specific or represents nonspecific background. Having both negative and positive controls is crucial in making this determination. Negative control slides can be generated in two ways: no primary antibody or isotype- and species-matched immunoglobulin or serum (which contains immunoglobulin) instead of primary antibody step [8, 9]. While the no primary negative controls are usually acceptable, the isotype control is the gold standard because it is possible to

264

Crystal Idleburg et al.

see interactions between nonimmune immunoglobulins from one species and target tissues from another. When using fluorophores, the “no primary” control indicates native tissue autofluorescence. Another good option for negative control comparisons is the use of knockout tissue or cell lines with all the same reagents as test slides, including the primary antibody. Ideally, control slides should have no staining at all. However, some background is often unavoidable and usually related to specific structures/cells, and thus the negative control slides must be directly compared to the test slides in similar areas to demonstrate specificity. Other important factors to consider are which tissues, cells, or organelles are stained and whether the staining pattern makes physiological sense based on known molecular pathways. If knockout/wild-type pairs of the exact site and experimental conditions are available, staining in the wild-type samples can be interpreted as positive. When such pairs are not in hand, positive control slides from other sites or conditions with previously established antigen expression are very useful here, although different tissue types may have quite different staining patterns as well as nonspecific background. Complementary techniques such as in situ RNA hybridization, which identifies gene expression in specific cells, laser capture microdissection with RNA analysis, or tissue fractionation with protein or RNA analysis can also be used to verify findings.

2

Materials

2.1 IHC on Paraffin Sections

1. Phosphate-buffered saline (PBS). 2. Citrate buffer, pH 6: Make 0.1 M stock solutions of citric acid and trisodium citrate. To 450 mL dd water add 9 mL of citric acid stock solution and 41 mL of sodium citrate stock solution. The pH of this final solution should be about 6.0  0.1. 3. 10% Neutral buffered formalin (NBF). 4. 4% Paraformaldehyde (PFA) can be purchased as a 16% stock. Making your own from powder is hazardous, and respiratory precautions must be taken. 5. Peroxidase block (3% H2O2 in methanol): 25 mL 30% H2O2 to a final of 250 mL in 100% methanol chilled at 20  C. 6. Positively charged slides such as Fisherbrand Superfrost Plus (Thermo). 7. Vectastain ABC Kit (Vector Labs). 8. DAB Chromogen (Biocare). 9. 14% Free acid EDTA, pH 7.2–7.4 (EDTA Decalcification Buffer): Mix 140 g EDTA free acid with 700 mL distilled water. While stirring, slowly add 30 mL ammonia hydroxide

Immunostaining of Skeletal Tissues

265

at 30-min intervals (for a total of 90 mL ammonia). Check the pH. EDTA will not dissolve until pH is close to 7.2 If not up to 7.2, add the remaining 10 mL ammonia hydroxide dropwise to get to pH 7.2, while constantly stirring. Add ~300 mL distilled water for final volume of 1 L. It is critical that this solution is made properly. If you overshoot the pH, do not attempt to correct with HCl—just start over, as the excess chloride ions will prevent proper EDTA chelation of ionized calcium [2]. 10. Methanol. 11. Graded ethanols (30%, 50%, 70%): It is least expensive to dilute from a purchased 70% stock, but 95% can also be used. These concentrations do not have to be very exact. Adding about 50 or 70 mL of 70% ethanol and diluting up to 100 mL in ddH2O is sufficient. 12. Xylene. 13. Coplin jars. 14. Humidity chamber: Any container with a lid can be lined with damp paper towels to make one, and several companies sell them. 15. Mounting medium, xylene compatible such as Richard-Allan Scientific™ Mounting Medium (Thermo). 16. Proteinase K in 10 mM Tris–HCl, pH 7.4–8.0. Make a 10 mg/ mL Proteinase K (Roche) stock solution in ddH2O. 1 M Tris– HCl pH 8.0 is commercially available. Dilute 1 M Tris–HCl stock solution at 1:100 dilution (10 μL/mL) to make a 10 mM Tris–HCl diluent solution. To 1 mL of this diluent solution, add 1 μL of the Proteinase K stock solution to get a final concentration of 10 μg/μL of Proteinase K. 17. 10% Blocking serum in PBS (see Note 1). 18. Primary antibody. 19. Secondary antibody (enzyme or biotin-conjugated). 20. Coverslips. 21. Mounting media, such as Permount (Fisher) or Cytoseal (Richard-Allan). 2.2 IF on Frozen Sections

1. Tris-NaCl-Tween Buffer (TNT): Combine 50 mL 1 M Tris– HCL pH 7.4, 25 mL 3 M NaCl, and 250 μL Tween-20. Bring the solution to 500 mL total using dH2O and mix. Filter the solution at .2 μm for best results. PBS with 0.1% Tween 20 (PBST) may also be used in place of TNT. 2. 10% Neutral buffered formalin (NBF), buffered aqueous zinc formalin (ZF), or 4% paraformaldehyde (PFA). PFA can be purchased as a 16% stock. Making your own PFA from powder is hazardous, and respiratory precautions must be taken.

266

Crystal Idleburg et al.

3. 10% Blocking serum in TNT (see Note 1). 4. Primary antibody. 5. Fluorophore-conjugated secondary antibody. 6. DAPI counterstain. 7. Positively charged slides such as Fisherbrand Superfrost Plus (Thermo). 8. Coverslips. 9. PAP pen, such as SuperPAP with 4 mm tip (Biotium). 10. Aspirator vacuum pump. 11. Disposable freeze-safe cassettes for embedding. 12. 14% Free acid EDTA, pH 7.2 (EDTA Decalcification Buffer): See Subheading 2.1 for more details. 13. Humidity chamber. 14. Glycerol-based aqueous mounting medium, Fluoromount-G Mounting Medium (Affymetrix).

such

as

15. Clear nail polish.

3

Methods

3.1 IHC on Paraffin Sections 3.1.1 Tissue Preparation

1. If perfusing for rodent studies, an approved dose of anesthetic should be administered as determined by your local animal safety committee. When the animal is fully sedated, it can be perfused with an intracardial injection of 1 M PBS, followed by 10% NBF or 4% PFA (typically 10 mL to 25 mL each). Constant flow rate can be applied with a peristaltic syringe pump or other method to improve the quality of the perfusion fixation [5]. 2. Immediately after dissection, fix bones in 10% neutral buffered formalin or 4% paraformaldehyde for 24–72 h at room temperature or 4  C. Fixative volume should be 15–20 times tissue volume. To ensure complete penetration, tissue should be agitated on a shaker during fixation (see Note 2). 3. Rinse tissue in PBS or distilled deionized water (ddH2O) 6 times, 15 min each. 4. Decalcify in 14% free acid EDTA, pH 7.2–7.4, with rocking, changing solution daily (on weekdays only is OK). The number of days required for decalcification of mouse bones is as follows (see Note 3): (a) Embryo > E17.5: 1–2 days. (b) Postnatal days (P) 1–4: 3 days. (c) P5-P10: 4–5 days.

Immunostaining of Skeletal Tissues

267

(d) P10-P21: 7–10 days. (e) Adults: 10–14 days. 5. Rinse tissue in PBS or ddH2O 6 times, 15 min each, to stop the decalcification process (see Note 4). 6. Dehydrate the tissue through a series of ethanol solutions, with rocking. (a) 30% ethanol for 15 min. (b) 50% ethanol for 15 min. (c) 70% ethanol for 15 min. 7. Place tissue in the tissue processor for dehydration, clearing, and infiltration before embedding. Typically, a 4-h processing program works well for most machines with mouse long bones (see Note 5). 8. When processing is complete, the bones are embedded in paraffin. It is important to determine the plane of interest before this point in order to orient the tissue properly. 9. Tissue sections are then cut at 5 μm using a microtome, floated onto a 45  C water bath and placed on color frost slides (see Note 6). 10. Slides are dried at room temperature overnight before storage (see Note 7). 3.1.2 Antibody-Based Staining

1. Heat slides in a 55  C oven for 1 h (see Note 8). 2. Incubate slides in Xylene 3  5 min. 3. Incubate slides in 100% ethanol 3  3 min. 4. Incubate slides in 95% ethanol 2  3 min. 5. Incubate slides in 70% ethanol 1  3 min. 6. Incubate slides in 50% ethanol 1  3 min. 7. Incubate slides in 30% ethanol 1  3 min. 8. Rinse in ddH2O 3  5 min. 9. DO NOT LET YOUR SLIDES DRY OUT AFTER THIS POINT. Incubate slides in Peroxidase block for 10 min (see Note 9). 10. Rinse slides once in ddH2O (2–3 dips) and then wash in PBS 3  5 min. 11. Start the antigen retrieval process by incubating slides in citrate buffer in a covered Coplin jar at 55  C overnight (see Note 10). 12. Rinse slides in PBS 3  5 min. 13. Block endogenous biotin with avidin/biotin block according to manufacturer instructions. We find that this is important in skeletal tissues.

268

Crystal Idleburg et al.

14. Incubate in 10% serum diluted in PBS for 60 min at room temperature in humidified chamber (see Note 1). 15. Drain off serum solution. Do not rinse slides or let the slides dry completely. 16. Incubate in primary antibody diluted in PBS with 1.5% serum (see Note 11) overnight at 4  C or for 1 h at room temperature in humidified chamber (see Note 12). 17. Rinse slides 3  5 min in PBS. 18. Incubate sections in biotinylated secondary antibody for 30 min, following data sheet from manufacturer for dilutions (see Note 13). 19. Rinse slides 3  5 min in PBS. 20. Prepare ABC solution as per manufacturer’s instructions and incubate with slides for 30 min. 21. Prepare DAB substrate. Also prepare an extra Coplin jar with water to stop the substrate reaction (see Note 14). 22. For developing the slides you will need a light microscope. Lay out all your reagents—the substrate solution, several Coplin jars with water (see Note 15), and slides should all be easy to reach. Some reactions take as little as 30 s before developing background so there is little margin for error at this point. 23. Place your positive (+ve) and negative ( ve) control slides on the microscope stage and add a drop of substrate solution to each. The ideal time interval will give you the most intense staining in your +ve control without giving any staining in the ve. The maximum staining time should be 5 min or less. Place the slides in the extra Coplin jars of ddH2O to stop the reaction. Do not return them to jars with undeveloped slides since this will start the substrate reaction prematurely. 24. Develop the rest of the slides one at a time at the optimal time determined in step 23 Once a slide has developed put it in the extra Coplin Jar with water. 25. Rinse slides well in ddH2O. 26. Counter stain the slides in Harris hematoxylin for 30 s to 1 min. 27. Wash in running tap distilled water for 10 min. 28. Dehydrate and clear through 2 changes of 95% ethanol, 3 changes of 100% ethanol, and then xylene. 29. Add coverslip with mounting medium. In a fume hood, place a drop or thin line of mounting medium on the edge of the coverslip on the benchtop and touch it with the edge of an inverted slide at a 45o angle, and gently bring it down onto coverslip. Avoid bubbles under the coverslip (see Note 16).

Immunostaining of Skeletal Tissues

269

Wipe the excess xylene and mounting medium with gauze or Kimwipe prior to viewing under microscope. Allow slides to dry and xylene to evaporate in a fume hood. 3.2 IF on Frozen Sections 3.2.1 Tissue Preparation

1. Perfusion, fixation, decalcification, and washing are the same as Subheading 3.1.1, steps 1–5 above. 2. Incubate tissue in 30% sucrose solution at 4  C for 3–5 days. This helps to prevent formation of damaging ice crystals during freezing/embedding. Tissue should sink to the bottom of the container when ready for embedding. 3. Embed sucrose-permeated tissue in OCT Compound (TissueTek), paying close attention to orientation (see Note 17). Store frozen blocks at –80  C until sectioning. 4. Tissue sections are then cut at the desired thickness (generally from 10–100 μm) using a cryostat and placed on color frost slides (see Note 18). 5. Store frozen slides at –80  C for long-term storage.

3.2.2 Antibody-Based Staining

NOTE: All steps should be performed in a humidified chamber. 1. Remove slides from freezer and allow to thaw in a humidifying chamber for 10 min. 2. Draw a PAP pen barrier around tissue prior to rinsing for maximum adherence of hydrophobic barrier to slide (see Note 19). 3. Pipette enough PBS to create surface tension over the tissue inside the hydrophobic barrier (generally 200 μL or more) onto the slide to wash OCT from the tissue for 5 min. Then aspirate the liquid. 4. Pipette PBST buffer to rinse for 5 min. Then aspirate the liquid. 5. Incubate in 10% serum diluted in TNT for 60 min at room temperature in humidified chamber (see Note 1). Then aspirate the liquid. 6. Incubate in primary antibody diluted in TNT (see Notes 11 and 12). Close lid and wrap in plastic wrap, taking care not to disturb the slides. Incubate slides at 4  C. For incubation time suggestions, see Note 20. 7. After incubation, aspirate the antibody and rinse with PBST 3  10 min. 8. Incubate in secondary antibody diluted in TNT (see Notes 11 and 12). Close lid and wrap in plastic wrap, taking care not to disturb the slides. Incubate slides at 4  C. For incubation time and temperature suggestions, see Note 21. 9. After incubation, aspirate the antibody and rinse with PBST 3  10 min.

270

Crystal Idleburg et al.

10. If the mounting media used contains DAPI, skip to step 12. If not, add DAPI counterstain cocktail to slide and incubate for 5–7 min. Aspirate the liquid immediately to avoid overstaining. For DAPI concentration suggestions, see Note 22. 11. Aspirate the liquid immediately to avoid overstaining. Rinse with PBST 3  5 min. 12. Place 2–3 drops or a thin line of glycerol-based aqueous mounting media along the edge of a coverslip. Carefully place the slide (tissue down) on top, beginning on the edge with the mounting media. This helps to prevent bubbles. 13. Seal slide edges with clear nail polish. This prevents the slides from drying out and media from escaping, as aqueous media does not fully set. 14. Keep slides at 4  C. Room temperature slides may dry out, mold, or lose fluorescence more quickly (see Note 23).

4

Notes 1. Blocking and primary antibody incubation serum should be from the same species as the secondary antibody (i.e., if the secondary antibody is goat anti-rabbit then goat serum should be used to block slides). 2. For example, 1–2 mouse femurs and/or tibiae should be placed in a 15 mL tube with at least 10 mL fixative. Most tissues will be properly fixed in 24 h, but large bones, such as from rabbits, might require longer fixation. However, antigenicity can be reduced with longer fixation so optimization of fix time may be needed. 3. The first time you do this, include a test bone of the same type as you will analyze in the decalcification and use this one to bend and compress. A fully decalcified bone should bend easily and not break. 4. Total rinse time should be about 2 h. 5. First step in processor should be 70% ethanol. Whole bone specimens from larger animals may require 6–8 h processing times. 6. Slides can be checked using light, darkfield, or phase microscopy at this point for proper plane of section. 7. Do not skip this step and go straight to heating slides at 55  C. The tissue is likely to fall of the slides during staining if you do this. 8. Alternatively, slides may be baked overnight at 55  C. 9. Start with 10 min and extend if background is high on negative controls. The 30% hydrogen peroxide can also be diluted in

Immunostaining of Skeletal Tissues

271

PBS instead of cold methanol. In addition, there are commercial peroxidase block alternatives such as Biocare 1, which may be preferable because they have shorter incubation times. 10. Slides can also be heated to 95  C in citrate buffer for 10 min, followed by cooling in hot buffer for 15 min. However, overnight citrate buffer treatment is preferable to high heat because it preserves tissue morphology better. If the high heat method is used, make sure the buffer does not come to a full boil as this will cause the tissue to fall off the slide. Other alternative retrieval methods include enzymatic digestion at 37  C with proteinase K (10 μg/mL in 10 mM Tris–HCl pH 7.4–8.0 for 20 min) or hyaluronidase (1% in PBS for 30 min). Avoid using reagents generated in donkeys (serum or antibodies) when using hyaluronidase since it increases background staining. In addition, some antigens do not require retrieval. This is a step that must be optimized for each antigen. 11. Data from the antibody manufacturer are good sources for determining the primary and secondary antibody concentrations to use. However, users may have to run serial dilution experiments to determine the optimal concentration for specific tissue/antibody combinations. 12. Depending on the size of your tissue, you will need 50–200 μL of antibody solution. Thick frozen sections may require up to 300 μL. Cut a piece of parafilm the same size as the slide and float this on top to retain the antibody over the tissue, taking care not to have any bubbles. Do not use a glass or plastic coverslip. You do not need to use the parafilm for shorter incubation, but make sure the tissue is covered with the solution. Do not allow the tissue dry or you will get very high background. 13. Alternatively, incubate in secondary antibody conjugated to horseradish peroxidase, diluted with 1.5% serum in PBS for 30 min at room temperature. If you do this, then skip step 20 and go straight to step 21. 14. We usually use DAB solution from Biocare, although other chromagens are available. DAB generates a brown precipitate and is very carcinogenic. Take care to use gloves and follow manufacturer’s instructions. 15. Deionized water is sufficient in most cases, but milliQ or ddH2O is fine to use if available. 16. You can use gentle pressure to push small bubbles to the edge of the coverslip. If the bubbles are very large, you probably did not use enough mounting medium. Put the whole slide with the coverslip back into xylene to float off the coverslip and start over. If you try to pry off the coverslip, you are likely to damage the tissue.

272

Crystal Idleburg et al.

17. Embedding frozen tissue is best accomplished using a cryoplate for even freezing of the sample. 18. Section thickness depends on structure or antigen of interest. Thicker sections can be used for larger structures, such as vasculature or nerves. Thinner sections may be used for molecular or cellular antigens of interest. Of note, stronger detergents may be needed to facilitate antibody penetration into thick sections. Section thickness is also limited by decreasing optical transparency and subsequent image quality. 19. The PAP pen may need to be vortexed prior to use. Take care to only press the pen tip gently on the slide, as the liquid may rush and pool. If greater adherence to the slide is needed, the PAP pen outline may be re-enforced by spreading with a cotton tip applicator. This barrier allows for use of smaller aliquots of antibody and also prevents drying out of tissue over long periods of time. 20. Thin sections (10–20 μm) may be incubated in primary antibody overnight. Thick sections (20 μm+) should be incubated for 48-h to allow maximum penetration and equilibration. 21. Secondary antibody incubation times may vary. Thin sections (10–20 μm) may be incubated at room temperature for 1–3 h. Thick sections (20 μm+) should be incubated for 24 h at 4  C. 22. DAPI concentration may range from 0.1 μg/mL to 10 μg/ mL. For best results, troubleshoot concentrations when a new batch is received for your desired application. For example, high concentrations will result in densely stained, more uniform nuclei, while low concentrations will reveal differences between chromocenters and chromatin compaction if imaging of nuclear architecture is needed. 23. The quality of the fluorescence will begin to degrade after the first few months of storage. Use of mounting media containing DAPI may require imaging within the first few days of staining. For best results, specimens should be imaged as soon as possible to ensure the most accurate representation of the antigens. References 1. Bord S (2003) Protein localization in wax-embedded and frozen sections of bone using immunohistochemistry. Methods Mol Med 80:237–247 2. Freida CL, Hladik C. Histotechnology.; A self instructional text , 2009 3. Sheehan DC, Hrapchak BB (1980) Theory and Practice of Histotechnology. 2nd ed. Bancroft JD, Stevens J, editors. St. Louis, MO, The CV Mosby Company

4. L’Hoste RJ Jr, Tourres MA (1995) Using Zinc Formalin as a Routine Fixative in the Histology Laboratory. Lab Med 26(3):210–214 5. Gage GJ, Kipke DR, Shain W (2012) Whole animal perfusion fixation for rodents. J Vis Exp 65 6. Hao Z, Kalscheur VL, Muir P (2002) Decalcification of bone for histochemistry and immunohistochemistry procedures. J Histotechnol 25 (1):33–37

Immunostaining of Skeletal Tissues 7. Serowoky MA, Patel DD, Hsieh JW, Mariani FV (2018) The use of commercially available adhesive tapes to preserve cartilage and bone tissue integrity during cryosectioning. BioTechniques 65(4):191–196 8. Battifora H (1999) Quality assurance issues in immunohistochemistry. J Histotechnol 22 (3):169–175

273

9. Watkins S (2003) Immunohistochemistry. In: Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA et al (eds) Current protocols in molecular biology. John Wiley & Sons Inc., Hoboken, NJ. p. Supplement 7 Unit 14.6

Chaptert 16 Mapping Regional Cortical Bone Responses to Local Changes in Loading and Systemic Stimuli Sara H. Windahl, Peter J. Delisser, and Gabriel L. Galea Abstract Quantification of cortical bone mass and architecture using μCT is commonplace in osteoporosis and osteoarthritis research. Different groups often report substantially divergent mouse cortical bone responses to nominally comparable interventions. In the case of studies assessing bones’ responses to externally applied loading, these differences are commonly associated with methodological differences in the loading regime. This chapter describes a widely published, standardized method of in vivo mouse tibia axial loading to produce lamellar bone formation. Despite uniform application of axial loading, changes in bone mass are highly site-specific within individual bones. For example, the mouse proximal tibia rapidly accrues new bone following axial loading, but this osteogenic response tapers to produce undetectable differences distally. Consequently, the bone sites selected for comparisons substantially influence the magnitude of differences observed. Application of the freely available Site Specificity software allows site-specific responses to be identified by rapidly quantifying cortical bone mass at each 1% site along the bone’s length. This highcontent screening tool has been informatively applied to study the local effects of changes in loading as well as systemic interventions including hormonal treatment and aging. Automated multisite analyses of cortical mass is increasingly identifying site-specific effects of “systemic” interventions such as global gene deletions. Biological mechanisms underlying this apparent regionalization of cortical responses are largely unknown but may start to be elucidated by increasingly widespread application of Site Specificity methods. Key words Mechanical loading, Micro-computed tomography, Site specificity

1

Introduction Mechanical loading is the primary functional determinant of bone mass and architecture. Increases in loading-engendered mechanical strain locally trigger new bone formation, whereas reduced loading during disuse favors resorption [1]. The resulting mechanoadaptive feedback loop, called the Mechanostat, locally adapts bone mass to its load-bearing function [2]. Mechanisms underlying the Mechanostat have largely been elucidated through the

Sara H. Windahl and Gabriel L. Galea contributed equally to this work. Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_16, © Springer Science+Business Media, LLC, part of Springer Nature 2021

275

276

Sara H. Windahl et al.

application of in vivo loading models to the bones of various vertebrates [3]. Axial loading of the mouse tibia is now the most commonly applied in vivo loading model. Within the loaded bone, osteocytes and osteoblasts rapidly respond through wide-ranging transcriptomic changes [4], increased osteoblast and osteocyte metabolic activity [1, 5], and predictable downregulation of the anti-osteogenic Wnt antagonist sclerostin [6, 7]. However, the resulting loading-induced changes in cortical bone mass reported by different groups vary markedly. Some of these differences can be attributed to methodological changes such as insertion of rest periods, different waveforms, or studying mice of different ages [3]. Here we describe a widely published, standardized method of in vivo mouse tibia axial loading to produce lamellar bone formation [8–11]. Prior to performing in vivo loading it is essential to measure the strength of cadaveric bones in order to calculate peak loads which need to be applied to generate comparable strains between experimental groups. In wild-type mice, the change in cortical bone mass following loading is linearly related to peak strain when habitual loading is eliminated [12, 13]. The methods required for this strain-gauging procedure have been extensively described elsewhere [14, 15]. The standard in vivo loading protocol described here is intended to generate lamellar bone formation in 16-week-old C57BL/6 mice of both genders. An additional confounding variable is the site of the bone in which functional adaptation is assessed. While seemingly paradoxical, applying mechanical loading axially through the mouse tibia is now known to cause bone gain in a site-specific manner [16, 17]. The proximal tibia shows a much greater increase in cortical bone mass than the distal tibia following 2 weeks of tri-weekly axial loading [13]. Although the biological mechanisms underlying this site-specificity are unknown, engineering-inspired hypotheses such as the effect of greater proximal curvature have been proposed to explain it [18]. What is even more perplexing is that “systemic” interventions such as deletion of osteogenic or antiosteogenic genes also regionally alter cortical bone mass. One of the first studies to describe genetically determined site-specificity investigated the effect of ERα deletion in osteoclasts on femoral bone mass [19]. Measurement of femoral bone mineral density in 20 equal segments along the length of the femur using dual X-ray absorptiometry identified genotype-related differences proximally, but not distally. The time and resources required to scan an entire mouse bone initially limited the analysis of multiple sites along each bone using micro-computed tomography (μCT). As μCT scanning became faster and more accessible, various groups reported changes in bone mass in response to loading as well as systemic interventions at multiple sites along the length of the bone [17, 20, 21]. We further extended this multisite analysis approach, initially to provide a global picture of compartmentalized cortical bone

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . .

277

phenotypes caused by Prkcα deletion [22]. This global gene deletion produced very regional in-growth of trabecular-like bone into the endosteal cavity at specific sites of long bones. Problems faced early-on in applying site-specific analyses of bone mass included requirement for extensive computing power to analyze each μCT slice through an adult mouse bone (typically ~2000 slices at 10 μm voxel sizes), needing to exclude the fibula from studies focused on the tibia, and difficulties in aligning each bone at a meaningful landmark. To address these we developed a simple Site Specificity workflow, which was validated and reported as a sensitive, rapid screening tool [23]. Site Specificity follows a very simple workflow with open source MATLAB-executed code. Similar workflows have also been described by others [24]. Application of Site Specificity allows quantitative discrimination between regional and global changes in cortical mass in response to changes in loading, aging, genetic mutations, and hormonal treatment [10, 23, 25]. In this chapter we detail the methods involved in axially loading the mouse tibia, imaging the tibia using μCT analysis, and screening for site-specific versus global changes in cortical bone mass.

2

Materials

2.1 Strain Gauging and In Vivo Axial Tibial Loading

1. Strain gauge, 120 Ω: Micro Measurements (Vishay Precision group, NC, USA), catalog number EA-06-031DE-120. 2. PCT-3 M gauge installation tape, Micro Measurements. 3. M-Flux, Micro Measurements. 4. Single conductor wire e.g.134 AWP, Micro Measurements. 5. M-Bond 200 adhesive and Catalyst, Micro Measurements. 6. Tin lead soldering wirer 60/40 0.8 mm. 7. Soldering station. 8. Multimeter, e.g., Digital multimeter, auto-range, VWR (Lutterworth, UK), catalog number 620–1920. 9. Dissection microscope with extra light. 10. Glass plate. 11. Cotton swabs. 12. Scalpels and forceps. 13. Isopropanol or 70% ethanol. 14. Polyurethane spray (can be purchased from local hardware stores).

278

Sara H. Windahl et al.

2.2 Ex Vivo μCT Imaging

1. There are several μCT instruments. Methods and settings described in Subheading 3.2 Ex vivo μCT imaging below is optimized for the use of an 1172 model, Bruker microCT (Aartselaar, Belgium). 2. BMD calibration phantoms, 0.25 and 0.75 g/cm3, Bruker microCT (Aartselaar, Belgium).

2.3 Site Specificity and Statistical Analysis

3

1. MATLAB® 2012 or later. 2. FIJI (NIH, ImageJ) with BoneJ installed. 3. IBM SPSS Statistics.

Methods

3.1 Strain Gauging and In Vivo Axial Tibial Loading

3.1.1 Preparation of the GAUGES

Pre-gauging is performed to ensure that matched osteogenic strains are engendered by the loads applied to each experimental group. Pre-gauging should be performed before the loading experiment on a subset of animals from the same group to be used in the loading experiment. When multiple groups are to be compared, the load:strain relationship should be calculated for each group (see Note 1). Protocols related to anesthetizing and working with live mice are beyond the scope of this chapter. 1. Place the gauge on the glass board (shining side up, Fig. 1). 2. Secure one of the long edges of the gauge to the board with adhesive masking tape. 3. Cut three sides of the gauge as shown in Fig. 1. 4. Lightly scratch the terminals with the blade to improve solder adherence. 5. Apply a drop of M-Flux to each terminal. 6. Apply solder to each terminal.

Fig. 1 Sites for trimming the strain gauge. The gauge is attached to a glass plate at the bottom of the image. The gauge is then cut with a scalpel in the indicated order 1–4

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . .

279

Fig. 2 Photomicrograph of strain gauge with the attached wires before the final cut. Solder points (arrows) with wire attached. The side cuts have been made and the top cut is all that remains to be made. The top edge of the gauge is immobilized by masking tape. The scale bar indicates 0.5 mm

7. Prepare two lengths of wires to be soldered onto the strain gauge: (a) Trim wire lengths to connect the gauge from the mouse’s leg to the strain gauge unit. (b) Remove the insulation from the end of the wires. (c) Dip the ends in M-Flux. (d) Apply solder to the ends. 8. Solder the wires to the terminals of the gauge (see Fig. 2). 9. Test the soldered gauge with a multimeter to ensure the resistance remains within the tolerance of the gauge (120
0.2 Ω is perfect; 118–121 Ω should work, below 120 is better than above) and then tape the wires together. 10. Cut the gauge, loosening it from the tape (line 4, Fig. 1). 11. Spray the gauge with polyurethane. 12. Test the gauge again with the multimeter. 3.1.2 Attach the Strain Gauge to the Tibia of a Recently Euthanized Mouse to Pre-gauge (See Fig. 3)

1. Euthanize the mouse using an ethically approved method which the experimenter is competent and confident performing. 2. Locally dissect the area of the right tibia where the strain gauge should be applied. 3. Wipe the bone with a cotton swab dipped in EtOH or isopropanol to degrease the bone and facilitate attachment of the gauge.

3.1.3 Attachment of the Gauge to the Mouse Tibia

1. Place a drop of catalyst and a drop of glue on a piece of foil. 2. Dip the gauge in catalyst.

280

Sara H. Windahl et al.

Fig. 3 Attachment site of the gauge. The gauge is attached so that the center of the gauge is located approximately at 37% of the bone’s length proximally. The arrows indicate where the top of the gauge will be located

3. Immediately dip the gauge in glue and touch a clean bit of the foil surface to remove excess glue. Too much glue will disable the function of the gauge. 4. Hold the gauge steady on the bone with forceps to attach it to the bone. See Fig. 3 for the correct site. The center of the gauge should be where the tibia is the widest. The upper part of the gauge should be positioned where there is an indention at the medial side. 5. Test the gauge with a multimeter. 6. Secure with polyurethane (on gauge and bone if stored cold in PBS before use). 3.1.4 Pre-gauging Protocol

Place the Limb to Be Pre-gauged in Your Loading Device and Subject it to Multiple, Graded Loads, Recording the Strain Gauge Resistance at Each Load A typical pre-gauging waveform is described below: 1. Ramp

Load

0.5 N/s to 0.5 N (pre-load)

2. Dwell

Load

7s

3. Ramp

Load

460 N/s to 6 N (peak compressive load)

4. Dwell

Load

5s

5. Ramp

Load

460 N/s to 0.5 N (continued)

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . . 6. Dwell

Load

281

2s

7. Repeat 3–6 two times, recording the resistance value achieved each time (in total three recordings per load). 8. Repeat 3–7 with increasing peak loads in step 3. 9. Ramp

Displace

1.9 mm/s to 0.8 mm (unload)

Test each load at least three times on each bone (see table above). The resulting load versus gauge resistance plot can then be used to generate load:strain graphs by using the gauge factor to convert resistance values into strain values. 3.1.5 Axially Load the Tibia of Anesthetized Mice

3.2 Ex Vivo μCT Imaging 3.2.1 Dissect the Left (Control) and Right (Exogenously Loaded) Tibia

Once you know what loads need to be applied experimentally, place an anesthetized mouse in the loading device so that the knee is fixed in the upper cup, and the foot is fixed in the lower cup. Make sure that the foot and knee are aligned so that the joint is loaded axially (see Note 2). The sequence below describes a standard rest-inserted loading program applying 40 cycles at 11 N compressive load. 1. Ramp

Load

0.5 N/s to 0.5 N (pre-load)

2. Dwell

Load

2s

3. Ramp

Load

393 N/s to 11 N (peak load, Note 3)

4. Dwell

Load

0.05 s

5. Ramp

Load

393 N/s to 0.5 N

6. Dwell

Load

10 s (Rest period)

7. Repeat

3–6

39 repeats (repetitions)

8. Ramp

Displace

1.9 mm/s to 1.9 mm (Unload)

1. Euthanize the mouse using an ethically approved method which the experimenter is competent and confident performing. 2. Using sharp-tipped scissors, make a skin incision over the femur. The skin can then be manually pulled distally over the calcaneus (see Note 3). 3. Proximally, flex the knee (stifle) and insert the sharp tip of a 10A scalpel laterally to medially through the popliteus to sever muscle and tendon attachments. Similarly, cut through the straight patellar ligament and collateral ligaments on either side of the knee. This should destabilize the joint, allowing you to insert the scalpel into the joint to cut the collateral ligaments and disarticulate the femur from the tibia (see Note 4).

282

Sara H. Windahl et al.

4. Distally, flex the ankle (hock) and cut through the intertarsal joints. It is not necessary to disarticulate the talus from the tibia and doing so can occasionally damage the medial malleolus and artifactually shorten the tibia, affecting bone measurement. 3.2.2 Fix and Dehydrate the Tibia 3.2.3 μCT Scan Each Bone

1. Fix the bone and muscle in chilled 4% PFA for 2 days. 2. Dehydrate the tibia and muscle in 70% ethanol (see Note 5). Methods and settings described here may need to be adapted to the specific μCT scanner available following the manufacturers’ instructions. 1. Prepare the bone by rolling the bone in non-PVC cling film and place in the container (i.e., a straw). Attach the straw to the μCT stage stand before screwing it into the scanner. If the bone can be scanned dry, make sure the bone has dried out for ~10 min before the scan. Evaporation during the scan can alter the density values. If the bone needs to remain moist, then it can be scanned in 70% ethanol within a cut-down 1 mL syringe using syringe stoppers above and below the bone to keep the bone stable and avoid evaporation. 2. Select the parameters that best suit your bone. For mouse bones, we use a 0.5 mm aluminum filter to reduce beam hardening. Beam hardening is what happens when the bone preferentially absorbs low energy photons and results in a dark looking surface. The filter will decrease the low energy output and improve the image. The whole tibiae and surrounding muscles are typically imaged with a voxel size of 4.8 μm (110 μm3). The applied X-ray voltage is 49–50 kV, current of 200 mA, with 0.5 mm aluminum filtration. Scans are obtained over 180 degrees with a 0.6-degree rotation step. The images can be reconstructed and binarized with global thresholding (values: 1.000–1.160) using the NRecon Bruker software. 3. If the entire bone does not fit on one scanning run, an “oversize scan” may be needed to scan larger bones, which can then be merged digitally in post-processing to generate a single bone series. 4. Once scanning is complete, replace bone in 70% ethanol solution and proceed with processing for additional tests (see Note 6).

3.2.4 Reconstruction

Reconstruct the RAW 2D Radiographic Images Generated from the CT Scanner Following the Manufacturer’s Instructions This should provide reconstructed *.bmp files representing cross-sections through the bone, which can be used in Site Specificity analysis (see Note 7).

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . .

3.3 Site Specificity and Statistical Analysis

283

1. Place the reconstructed cross-section μCT images for each bone in a single folder, henceforth the “Inpath.” Ensure there are no blank (“spacer”) slices in the folder as bone length is calculated from the number of slices in this folder. 2. Create a local copy of the SiteSpecificityV2.m script. This can be downloaded from https://www.researchgate.net/publica tion/295858230_SiteSpecificityV2 and is available with the original manuscript [23]. 3. Open MATLAB®, direct it to the directory containing SiteSpecificityV2.m and open the script for editing. 4. Modify the Inpath line (line 13) to the folder containing the bone to be analyzed (see Note 4). For example, change this line to read:inpath ¼ ‘C:\Site specificity\Example\’ 5. Save the script file and run it. The script will create an “Output” folder within the Inpath. It will count the number of images of the folder and identify the images corresponding to each 1% site along the bone’s length (see Note 8). The corresponding 100 images each undergo the following processing (see Fig. 2a, b): (a) Locally adaptive binarization to identify the bone and exclude soft tissue. (b) Identification of the largest continuous object (the tibia) and exclusion of smaller objects (the fibula and trabecular bone). (c) Quantification of the binarized area. This is exported as “Bone Area” and corresponds to Cortical Area (Ct.Ar) in conventional μCT analysis. The output report will display this as pixels2 and should be converted to mm2 using the voxel size known from the μCT reconstruction settings. (d) Quantification of empty space within the binarized area (see Note 9). This is reported as “Marrow Area” (i.e., Ma. Ar) in pixels2. (e) The perimeter of the binarized area is also provided. If required, Bone Area and Marrow Area can be summer to calculate Total Tissue Area (Tt.Ar). 6. The Output folder will now contain 100 images saved as .gif files corresponding to each 1% site along the bone’s length, .txt result files with the cross-sectional measurements at each site, and a single .csv file amalgamating the results (see Note 10). The .gif files retain the same pixel dimensions as the original image but are of much smaller size such that they can be easily processed in additional software if needed.

284

Sara H. Windahl et al.

7. To quantify parameters beyond those directly calculated by Site Specificity, drag and drop the Output folder into Fiji. This will open the subsampled cross-section images as a stack. Additional parameters, such as moments of inertia and crosssectional thickness, can now be calculated using the BoneJ plugin [26]. 8. Iteratively repeat the process until all bones have been analyzed. The resulting data can now be graphed as in Fig. 4c to confirm successful analysis. If using Site Specificity as a screening tool with a small number of bones, you can stop at this point. The following steps describe statistical comparison between treatment groups using the mixed model procedure in SPSS.

Fig. 4 Execution and application of the Site Specificity workflow. (a) Schematic representation of Site Specificity analysis applied to the tibia of a 3-week-old mouse. The region of the bone analyzed by Site Specificity are shown in the binarized reconstruction (right). Parts of the fibula (red highlight) not connected to the tibia are identified and excluded. Images corresponding to each 1% site along the bone’s length are identified. Proximal (including the growth plate, green) and distal sites are excluded from the analysis. (b) Cross-section through the proximal tibia and fibula (red) at the 20% site from the proximal end. The binarized tibia images are saved and can be exported into other analysis software such as BoneJ. Cortical area (Ct.Ar), Marrow area (Ma.Ar), and external perimeter are provided as part of the in-built workflow. (c) Quantification of Ct.Ar and Ma.Ar in the tibiae of five 3-week-old mice showing the expected pattern of cortical bone differences in typical inter-mouse variability

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . .

285

(a) Mixed models are used to test the effect of random effects (e.g., cage number), fixed effects (e.g., loading), and fixed covariates (e.g., % site along the bone’s length). Strengths and limitations of mixed models are described elsewhere (e.g., [27]). (b) Set up a datasheet with the following headings: Animal ID, Site (% bone length), and Treatment (e.g., loading, drug, genotype). Enter all data in a single list under each of these headings. Set the appropriate variable types in the SPSS “Variable view” tab. (c) In SPSS, select the linear mixed models option. Specify the Animal ID as the “subject,” Site as the “repeated measure” (see Note 11), and click continue. Set up your model as needed on the next screen (see the SPSS help function on your version of the software if you are not familiar with how to do this). (d) An important outcome of Site Specificity analysis is quantification of whether the response to treatment is significantly different between different sites. This is shown by a significant Site by Treatment interaction (indicates the response to treatment is significantly different between different sites). (e) To request a direct comparison between different sites with a Bonferroni post hoc comparison, click “Paste” to obtain the model syntax and add the following text at the end of the syntax before the full stop: /EMMEANS ¼ TABLES(Treatment*Site) COMPARE (Treatment) ADJ(Bonferroni) (f) Post hoc comparisons are generated in a table under the subheading “Pairwise Comparisons” (see Notes 12 and 13). (g) The same syntax can now be rerun for other bone parameters in the same experiment.

4

Notes 1. Be careful when applying a treatment/genetic modification which alters bone structure before starting a loading experiment. If the same load is applied to bones with different mass and architecture, different levels of osteogenic strain will be engendered. 2. If the tibia is not placed directly vertical between the upper and lower cups, the knee will be forced forward or back, which is likely to rupture the cruciate ligaments. If this happens in an anesthetized mouse, cull the mouse before it wakes up.

286

Sara H. Windahl et al.

3. Most materials testing devices will avoid exceeding the specified load and, at rapid loading rates with short dwell times, often fail to achieve the desired peak load. You may need to set a slightly higher programmed peak load in order to achieve the desired force (e.g., set a peak load of 11.2 N to achieve a load of 11 N). 4. It is usually easy to evert and unroll mouse skin off the tibia up to the calcaneus, but harder to fully remove it. There is no need to pull it further. If the tibia is to be scanned and analyzed, ensure the femur is grasped instead to avoid fracturing the fibula. The fibula is itself a loading-responsive bone which can be analyzed separately. 5. Disarticulating the tibia and femur can be tricky in some mice. It is essential to obtain the entire tibia to direct the analysis at each percent of the bone’s length. If the experimenter finds this difficult at first, cut through the distal femur instead. Handling tissues with wet gloves on wet tissue paper sometimes helps and a dissection microscope can be used if needed. 6. The same bone can be used for histology and histomorphometry end points as well as μCT scanning. For histology, the bone should be fixed (e.g., in paraformaldehyde) prior to dehydration. It is often helpful to trim away some of the overlying muscles to aid fixative penetration but be careful not to break the fibula. 7. SiteSpecifictyV2.m is intended to analyze .bmp image files. If a different file type is used, this can be changed in line 26 (“d ¼ dir([inpath, ‘*.bmp’]);”). 8. The level of subsampling is set to 100 images, with each image corresponding to a 1% site along the bone’s length by default. This subsampling can be changed in line 31 (“num_samples ¼ 100;”). 9. Large pores and blood vessels which breach the cortex in a single slice effectively cause the marrow area to become continuous with the background, producing a small or zero Ma.Ar measurement which should be excluded from subsequent analyses. Straight trans-cortical breaches in a single slice are rare in the cortex of young adult mice and the Ct.Ar measurement primarily used to screen bone mass is unaffected. Workarounds are therefore not directly implemented in Site Specificity, but if these breaches are problematic (e.g., genetic model with very low bone mass), the subsampled binarized images saved in the Output folder can easily be reanalyzed in conventional μCT analysis software (e.g., CtAN, Bruker) implementing a shrinkwrap function. The small number and size of subsampled images means these can be batch-processed very quickly.

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . .

287

10. If the extremes of the bone have very faint signal which is of soft-tissue equivalence, such that they binarize to produce a black image, that slice will be excluded from the analysis. The total number of output images and values will therefore be less than 100. Additionally, the proximal and distal extremes of the bone contain trabecular bone which cannot be analyzed using Site Specificity. We therefore exclude the proximal and distal 10–15% of the tibia from our analyses. This may not be necessary when analyzing less trabecular bone such as the mouse ulna. 11. The structure of the model used will vary depending on the endpoint. When testing whether an intervention enhances or diminishes the response to loading, it may be more appropriate to calculate the percentage difference between loaded and control limbs for each mouse and use this loading response as the dependent variable. This is especially true if the intervention itself changes baseline bone mass (see Strain Gauging methods above). 12. Displaying regions of significant differences between treatment groups across multiple sites can be challenging. We and others have adopted various methods to show this in previous publications. 13. Calculating the statistical power to identify differences as statistically significant using a complex mixed model approach is challenging. It is possible to simulate changes in bone mass by resizing the Output images in Fiji by a known percentage. This provides an indication of the proportion of sites at which that percentage change would be detected as statistically significant. In our initial validation of this analysis method [23], a 10% change in Ct.Ar was detected as significant at 98.8% of sites analyzed. References 1. Hillam RA, Skerry TM (1995) Inhibition of bone resorption and stimulation of formation by mechanical loading of the modeling rat ulna in vivo. J Bone Miner Res 10:683–689 2. Skerry TM (2006) One mechanostat or many? Modifications of the site-specific response of bone to mechanical loading by nature and nurture. J Musculoskelet Neuronal Interact 6:122–127 3. Meakin LB, Price JS, Lanyon LE (2014b) The Contribution of Experimental in vivo Models to Understanding the Mechanisms of Adaptation to Mechanical Loading in Bone. Front Endocrinol (Lausanne) 5:154

4. Galea GL, Meakin LB, Harris MA, Delisser PJ, Lanyon LE, Harris SE, Price JS (2017) Old age and the associated impairment of bones’ adaptation to loading are associated with transcriptomic changes in cellular metabolism, cellmatrix interactions and the cell cycle. Gene 599:36–52 5. Skerry TM, Bitensky L, Chayen J, Lanyon LE (1989) Early strain-related changes in enzyme activity in osteocytes following bone loading in vivo. J Bone Miner Res 4:783–788 6. Moustafa A, Sugiyama T, Prasad J, Zaman G, Gross TS, Lanyon LE, Price JS (2012) Mechanical loading-related changes in

288

Sara H. Windahl et al.

osteocyte sclerostin expression in mice are more closely associated with the subsequent osteogenic response than the peak strains engendered. Osteoporos Int 23:1225–1234 7. Robling AG, Niziolek PJ, Baldridge LA, Condon KW, Allen MR, Alam I, Mantila SM, Gluhak-Heinrich J, Bellido TM, Harris SE, Turner CH (2008) Mechanical stimulation of bone in vivo reduces osteocyte expression of Sost/sclerostin. J Biol Chem 283:5866–5875 8. Bergstrom I, Isaksson H, Koskela A, Tuukkanen J, Ohlsson C, Andersson G, Windahl SH (2018) Prednisolone treatment reduces the osteogenic effects of loading in mice. Bone 112:10–18 9. Lionikaite V, Henning P, Drevinge C, Shah FA, Palmquist A, Wikstrom P, Windahl SH, Lerner UH (2019) Vitamin A decreases the anabolic bone response to mechanical loading by suppressing bone formation. FASEB J 33:5237–5247 10. Meakin LB, Todd H, Delisser PJ, Galea GL, Moustafa A, Lanyon LE, Windahl SH, Price JS (2017) Parathyroid hormone’s enhancement of bones’ osteogenic response to loading is affected by ageing in a dose- and timedependent manner. Bone 98:59–67 11. Svensson J, Windahl SH, Saxon L, Sjogren K, Koskela A, Tuukkanen J, Ohlsson C (2016) Liver-derived IGF-I regulates cortical bone mass but is dispensable for the osteogenic response to mechanical loading in female mice. Am J Physiol Endocrinol Metab 311: E138–E144 12. Meakin LB, Galea GL, Sugiyama T, Lanyon LE, Price JS (2014a) Age-related impairment of bones’ adaptive response to loading in mice is associated with sex-related deficiencies in osteoblasts but no change in osteocytes. J Bone Miner Res 29:1859–1871 13. Sugiyama T, Meakin LB, Browne WJ, Galea GL, Price JS, Lanyon LE (2012) Bones’ adaptive response to mechanical loading is essentially linear between the low strains associated with disuse and the high strains associated with the lamellar/woven bone transition. J Bone Miner Res 27:1784–1793 14. Javaheri B, Bravenboer N, Bakker AD, van der Veen A, de Souza RL, Saxon L, Pitsillides AA (2019a) In vivo models of mechanical loading. Methods Mol Biol 1914:369–390 15. Melville KM, Robling AG, van der Meulen MC (2015) In vivo axial loading of the mouse tibia. Methods Mol Biol 1226:99–115 16. Carrieroa A, Pereirab AF, Wilson AJ, Castagno S, Javaheri B, Pitsillides AA, Marenzana M, Shefelbine SJ (2018) Spatial

relationship between bone formation and mechanical stimulus within cortical bone: Combining 3D fluorochrome mapping and poroelastic finite element modelling. Bone Rep 8:72–80 17. Sugiyama T, Saxon LK, Zaman G, Moustafa A, Sunters A, Price JS, Lanyon LE (2008) Mechanical loading enhances the anabolic effects of intermittent parathyroid hormone (1-34) on trabecular and cortical bone in mice. Bone 43:238–248 18. Dodge T, Wanis M, Ayoub R, Zhao L, Watts NB, Bhattacharya A, Akkus O, Robling A, Yokota H (2012) Mechanical loading, damping, and load-driven bone formation in mouse tibiae. Bone 51:810–818 19. Nakamura T, Imai Y, Matsumoto T, Sato S, Takeuchi K, Igarashi K, Harada Y, Azuma Y, Krust A, Yamamoto Y, Nishina H, Takeda S, Takayanagi H, Metzger D, Kanno J, Takaoka K, Martin TJ, Chambon P, Kato S (2007) Estrogen prevents bone loss via estrogen receptor alpha and induction of Fas ligand in osteoclasts. Cell 130:811–823 20. de Souza RL, Pitsillides AA, Lanyon LE, Skerry TM, Chenu C (2005) Sympathetic nervous system does not mediate the load-induced cortical new bone formation. J Bone Miner Res 20:2159–2168 21. Javaheri B, Carriero A, Wood M, de Souza R, Lee PD, Shefelbine S, Pitsillides AA (2018) Transient peak-strain matching partially recovers the age-impaired mechanoadaptive cortical bone response. Sci Rep 8:6636 22. Galea GL, Meakin LB, Williams CM, HulinCurtis SL, Lanyon LE, Poole AW, Price JS (2014) Protein kinase Calpha (PKCalpha) regulates bone architecture and osteoblast activity. J Biol Chem 289:25509–25522 23. Galea GL, Hannuna S, Meakin LB, Delisser PJ, Lanyon LE, Price JS (2015) Quantification of alterations in cortical bone geometry using site specificity software in mouse models of aging and the responses to ovariectomy and altered loading. Front Endocrinol (Lausanne) 6:52 24. Javaheri B, Herbert E, Hopkinson M, Al-Jazzar A, Pitsillides AA (2019b) Sost haploinsufficiency provokes peracute lethal cardiac tamponade without rescuing the osteopenia in a mouse model of excess glucocorticoids. Am J Pathol 189:753–761 25. Orriss IR, Lanham S, Savery D, Greene NDE, Stanier P, Oreffo R, Copp AJ, Galea GL (2018) Spina bifida-predisposing heterozygous mutations in planar cell polarity genes and Zic2 reduce bone mass in young mice. Sci Rep 8:3325

Mapping Regional Cortical Bone Responses to Local Changes in Loading. . . 26. Doube M, Klosowski MM, Arganda-Carreras I, Cordelieres FP, Dougherty RP, Jackson JS, Schmid B, Hutchinson JR, Shefelbine SJ (2010) BoneJ: free and extensible bone image analysis in ImageJ. Bone 47:1076–1079

289

27. Yang J, Zaitlen NA, Goddard ME, Visscher PM, Price AL (2014) Advantages and pitfalls in the application of mixed-model association methods. Nat Genet 46:100–106

Chapter 17 Pain and Activity Measurements David H. H. Molstad and Elizabeth W. Bradley Abstract Musculoskeletal pain contributes significantly to chronic pain experienced by adults and to health care use. This chapter details several methods to evaluate pain and physical activity in mice that can be applied to preclinical orthopedic models. These methods include the von Frey filament assay that measures mechanical allodynia, open-field activity assays for evaluation of ambulation, and incapacitance measurements to determine static weight bearing. Key words Functional assays, Mechanical allodynia, von Frey, Incapacitance, Evoked pain, Open-field activity assays, Static weight bearing

1

Introduction One of the main reasons patients visit physicians for a musculoskeletal problem is not the physical damage of tissue degeneration, but the associated pain [1]. Moreover, musculoskeletal pain is poorly controlled and contributes significantly to opioid use [2, 3]. Musculoskeletal pain can be either nociceptive (e.g., fracture, torn ligament) or neuropathic (e.g., osteoarthritis, low back pain). Nociceptive pain arises when there is potential for tissue damage or as a result of tissue damage; this type of pain is acute (e.g., pain lasting under 6 months) and has a protective function [4]. In contrast, neuropathic pain, otherwise known as chronic pain, is initiated by the central nervous system [4]. Whereas acute pain serves a protective function, chronic pain does not and instead can lead to disability [5]. Functional measurements of pain and locomotion in preclinical models in the research setting may therefore be helpful to assess analgesic benefits of potential therapeutics. In addition, inclusion of pain assessment in a research design may also shed light to mechanistic action. Measurements of allodynia and hyperalgesia in preclinical models include evoked mechanical-sensitivity, static and dynamic weight bearing, gait and spontaneous pain measurements

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7_17, © Springer Science+Business Media, LLC, part of Springer Nature 2021

291

292

David H. H. Molstad and Elizabeth W. Bradley

[6]. In this chapter we will detail methods to perform von Frey filament assays, open-field activity measurements, and static weight bearing (i.e., incapacitance) determinations. The methods described below are for use in mouse models, but could be adapted for use in other rodents. In combination, these functional assessments of pain are useful tools for evaluation of pain in murine models of musculoskeletal conditions. These musculoskeletal models include pain associated with cartilage degradation due to aging, as well as chemically induced or surgically induced models of joint destruction [7– 9]. Bone pain arising as a result of induced fracture, bone defects, or tumor metastasis to bone can also be evaluated using these functional assessments [10, 11]. The von Frey filament assay quantifies evoked mechanical allodynia or hyperalgesia in murine models (see Fig. 1). The theory behind the von Frey assay is that in a pain state otherwise innocuous stimuli, such as a light touch, elicit a pain response [12]. In this method, animals are placed in a suspended enclosure over a mesh grid and monofilament fibers with defined bending forces are applied to the plantar surface of the hind limb. If the force applied from the filament is nocifensive, the animal withdraws its paw. Underlying pain influences the threshold force needed to elicit a nocifensive response and causes a shift in the pain response curve (see Fig. 1); thus, the von Frey assay measures mechanical allodynia by determining the force required to produce nocifensive responses.

Fig. 1 Pain sensitization enhances responses to mechanical stimuli. Centralization of the pain response during chronic pain shifts the pain response curve from baseline pain (gray curve) to sensitized pain (blue curve). Allodynia (black shaded region) results when stimuli that are normally not perceived as painful become nocifensive. In contrast, hyperalgesia (red shaded region) characterizes an increased response to painful stimuli

Pain and Activity Measurements

293

The testing period consists of two phases. The first is testing before experimental intervention (e.g., injury) to establish von Frey percent withdrawal responses for each animal. The second phase is testing following intervention. The von Frey monofilament test can be repeated on each animal over the duration of the study to determine how interventions affect pain associated with musculoskeletal injury. For example, one would expect paw withdrawal percentage to increase following destabilization of the medial meniscus surgery [8]. In a fracture-healing model, paw withdrawal percentage would decrease along with bone healing. There are a number of factors that can influence the reproducibility of data obtained from the von Frey filament assays [13]. The advantages of the von Frey filament assays are that it is an elicited response and that handling-induced stress is eliminated because the test subject is unrestrained within the apparatus. The limitations of the approach are the long acclimation periods and labor-intensive testing. It is also subject to reader bias, and subject responses may vary depending on the reader (e.g., male or female reader). There are three different methods for performing the von Frey filament assay: the up-down method, the ascending method, and the percent response method [5]. In this chapter, we will describe the percent response method, as this approach has a lower chance of a learned response and is less labor intensive. In this method, animals are first acclimated to the testing enclosure and test operator. This is followed by baseline testing (e.g., before injury) and testing following experimentation. Monofilaments of increasing strength are then tested for a defined number of times and the percent withdrawal for each filament is determined. Open-field activity arenas are used to track the movement of research animals in three-dimensional space. These arenas monitor distance traveled, active ambulating time, resting time, rearing time and rearing events, average speed and stereotypic movement time (e.g., small movements such as grooming). These measurements are made in discrete increments, which can be summed to reflect total activity over a period of time (e.g., 10 increments of 2 min for a total of 20 min). A two-dimensional grid of infrared photocell pairs is used to define the x-y axes within the arena. A second set of infrared photocells placed 2.5 cm above the arena floor monitor movement in the z-axis due to rearing. Simultaneous interruption of the infrared beams places the animal within this threedimensional grid of the arena. A single arena, or set of arenas run in parallel, can be used to track changes in activity over time during an experiment. Time of day will impact animal activity, so it is best to use a consistent time of day for all experimental recordings. Similarly, the environment the activity arenas are placed in can also impact animal activity. The arena should be free of other animals and sudden and/or loud noises.

294

David H. H. Molstad and Elizabeth W. Bradley

The incapacitance assay measures static weight bearing of each hind limb of an animal. The animal is within an animal holder at an incline such that most of the animal’s weight is placed on the hind limbs. A force plate below each hind limb then measures spontaneous postural changes reflected by the relative amount of weight placed on each limb. It can therefore be used to determine how an injury to one limb affects how much weight the animal will place on the affected and contralateral limb [14]. The advantage of this method is the spontaneous measurement of postural changes. The disadvantage of this method is that it relies on the animal voluntarily assuming the correct position within the enclosure. This method can also only be used in models of unilateral injury.

2

Materials

2.1 von Frey Filament Assays

1. von Frey monofilament 0.008–300 g.

fibers,

Semmes-Weinstein,

2. Ugo Basile Grid Platform for Plantar Stimulation (Stoelting, Wood Dale, IL, Product #57816). 3. Multiple Configuration Animal Enclosure (Stoelting, Wood Dale, IL, Product #57823 or Bioseb, Pinellas Park, FL, BIO-PVF). 2.2 Open-Field Activity Measurements

1. Activity Animal Meter: Opto-Verimex 5 Auto-Track, Columbus Instruments, Columbus, OH, or an equivalent system. 2. PC Station with appropriate hardware.

2.3 Static Weight Bearing Incapacitance Assays

1. Incapacitance meter (Bioseb, Pinellas Park, FL, BIO-SWBTOUCH-M) or equivalent instrument. (a) Animal holder/restrainer (rat or mouse). (b) Control unit. (c) Platform with sensors. (d) Footswitch to start/stop experiment hands free. 2. PC Station with appropriate hardware. 3. Lab tape.

Pain and Activity Measurements

3

295

Method

3.1 von Frey Filament Assays 3.1.1 Acclimation Period

Two weeks prior to baseline readings: 1. Introduce animals to experimental chambers several times each week for 1–2 h. 2. On the day of the assay, allow the mice to become acclimated to their surroundings for about 15–30 min, or until the mice are no longer exploring.

3.1.2 On Day of Testing

1. Assemble the animal enclosure/grid platform (see Fig. 2). Ensure that each enclosure is fully accessible through the underlying mesh of the grid platform. Note: Opaque material should be used to separate mice so that they cannot see each other during acclimation or testing periods (see Fig. 2). 2. Place animals in enclosure chambers. Let animals acclimate to enclosure until they are no longer actively exploring. If animals are properly acclimated to the enclosure chambers, this should take under 15 min (see Note 1).

Fig. 2 Equipment for the von Frey Monofilament assays. (a) Assembly of the von Frey equipment includes multiple animal enclosures placed above a suspended grid platform. Mice are placed within chambers so that the plantar surface can be accessed through the mesh grid. An opaque wall separates chambers so that animals cannot see each other during acclimation and evaluation periods. (b) A mouse placed within the animal enclosure chamber demonstrating access of the hind limb plantar surface through the mesh grid. (c) Monofilaments ranging in fiber bowing weight used to perform the von Frey assay. (d) The von Frey monofilament is applied until the fiber bows to achieve a defined force

296

David H. H. Molstad and Elizabeth W. Bradley

3. Testing should not be performed while the animal is rearing, grooming, or actively exploring the animal enclosure chamber (see Note 1). 4. Start with a low monofilament fiber bowing weight (e.g., 0.16 g for mice). 5. Apply the fiber perpendicularly to the center of the plantar surface of the hind paw of the animal’s affected limb (e.g., right or left) five times. Record the numbers of paw withdrawals out of five tests. 6. Repeat the monofilament stimulation for the affected limb of each animal within the animal enclosure. 7. Test the contralateral paw of each animal (e.g., five monofilament stimuli). 8. Retest each hind limb (affected followed by contralateral) two times for a total of 15 potential paw withdrawal responses per paw. 9. Repeat steps 5–8 using increasing fiber bowing weights (e.g., 0.4 g and 0.6 g for mice). 10. Determine the average percent paw withdrawal by dividing the number for withdrawals for each hind limb of each animal by 15 (total number of tests), then converting to percentage. Report values for a fiber bowing weight that elicits 50% paw withdrawal in the control baseline condition. 3.2 Open-Field Activity Measurements

1. Transport mice to the area where activity arenas are located. Allow mice 15 min in cages to recover from transport before starting analyses. 2. Open the Opto-track software. Open a new experiment, select the location to save the file, and name the file. 3. Select arenas to be used in each run by checking the appropriate checkbox within the software. Adjust the total time for the analysis to 20 min or desired alternate time. 4. Verify that the software is localizing objects in the threedimensional space by tracing along the arena surface. Place one animal in each arena. Allow the animal to acclimate within the arena for 2 min (see Note 2). 5. Hit the “Start All” button within the Opto-track software. Ensure the software is tracking the animal accurately. Allow time to elapse. 6. When the experiment is complete, return animals to cages. Select “Analyze Experiment,” and then export all bin files. This will save a .csv file to the location you selected that can be viewed with Excel software.

Pain and Activity Measurements

297

7. Open the .csv file in Excel. To obtain the total distance and time measurements, sum all measured intervals. The speed of the animal can be obtained through the average of all speed values obtained during the experiment. 8. Determine changes between experimental groups and or over time. 3.3 Static Weight Bearing Incapacitance Assays

1. Transport mice to the area where the incapacitance meter is located. Allow mice 15 min in cages to recover from transport before starting analyses. 2. Place the animal within the animal enclosure (see Fig. 3). Taping down the animal’s tail to the base of the enclosure will help the animal assume correct placement. 3. Ensure that the animal has both upper limbs placed equally on the inclined portion of the animal enclosure and that both hind feet are located on the force plate (see Fig. 3 and Note 3). 4. Use the footswitch to take a 5-s measurement. 5. Repeat for a total of three consecutive 5-s measurements. Use the average of these three measurements. 6. Determine the percent weight bearing on the affected limb using the following equation: % weight on ipsilateral limb ¼ (weight placed on ipsilateral limb/summed weight placed on both limbs)  100% 7. Determine changes between experimental groups and/or over time.

Fig. 3 Incapacitance measurement equipment. The incapacitance assay measures static weight bearing on each hind limb of a mouse. (a) Side view of a mouse placed in the incapacitance chamber. (b) Forward view of a mouse within the incapacitance chamber. (c) Rear view of a mouse placed in an incapacitance chamber. Note the foot on the left is not contacting the force plate within the chamber. Lab tape is also used to secure the animal’s tail to prevent the animal from turning around within the chamber

298

4

David H. H. Molstad and Elizabeth W. Bradley

Notes 1. During the acclimation period of the von Frey assay, the reader should be present. Acclimation of the experimental animals to the enclosures will ensure that habituation occurs on test days. Fiber bowing weights can be adjusted if a response is either not elicited or is saturated (e.g., paw withdrawal is either 0% or 100%). During testing, be careful not to brush or scratch the plantar surface with the monofilament. Contact with the outer margins of the plantar surface should also be avoided. A positive response is evident if the animal actively withdraws its paw from the stimulus. 2. The activity assay software should display the correct location (lower left hand is 0,0 and the upper right is 30,30 within the Cartesian plane, see (Fig. 4) and register that the z-plane is broken. If software is not registering location correctly, ensure that the arena walls or other items do not break infrared beams. 3. For static weight bearing assays, the animal must assume the correct positioning within the enclosure before an accurate reading can be taken. This includes the animal facing forwards with both forelimbs placed on the incline and both hind feet placed on the force plates. Likewise, it is difficult to determine if an animal will not place its foot on the force plate due to pain or due to non-compliance.

Fig. 4 Open-field activity assays. (a) An open-field activity assay arena. (b) Each activity arena tracks the three-dimensional movement of a mouse within the arena. Cartesian coordinates for each corner of the arena are shown and can be used to verify correct tracking of an animal by the infrared grid

Pain and Activity Measurements

299

Acknowledgements This work was made possible by training and research grants from the National Institutes of Health (AR065397 and AR072634) and the University of Minnesota Stem Cell Institute and Board of Regents. These contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. References 1. Hawker GA (2017) The assessment of musculoskeletal pain. Clin Exp Rheumatol 35(Suppl 107):8–12 2. Carnide N, Hogg-Johnson S, Cote P, Irvin E, Van Eerd D, Koehoorn M, Furlan AD (2017) Early prescription opioid use for musculoskeletal disorders and work outcomes: a systematic review of the literature. Clin J Pain 33:647–658 3. Moshfegh J, George SZ, Sun E (2018) Risk and risk factors for chronic opioid use among opioid-naive patients with newly diagnosed musculoskeletal pain in the neck, shoulder, knee, or low back. Ann Intern Med 170 (7):504–505 4. Nicholson B (2006) Differential diagnosis: nociceptive and neuropathic pain. Am J Manag Care 12:S256–S262 5. Deuis JR, Dvorakova LS, Vetter I (2017) Methods used to evaluate pain behaviors in rodents. Front Mol Neurosci 10:284 6. Miller RE, Malfait AM (2017) Osteoarthritis pain: what are we learning from animal models? Best Pract Res Clin Rheumatol 31:676–687 7. Bradley EW, Carpio LR, McGee-Lawrence ME, Castillejo Becerra C, Amanatullah DF, Ta LE, Otero M, Goldring MB, Kakar S, Westendorf JJ (2016) Phlpp1 facilitates posttraumatic osteoarthritis and is induced by inflammation and promoter demethylation in human osteoarthritis. Osteoarthr Cartil 24 (6):1021–1028

8. Hwang SM, Feigenson M, Begun DL, Shull LC, Culley KL, Otero M, Goldring MB, Ta LE, Kakar S, Bradley EW, Westendorf JJ (2017) Phlpp inhibitors block pain and cartilage degradation associated with osteoarthritis. J Orthop Res 36(5):1487–1497 9. Xu M, Bradley EW, Weivoda MM, Hwang SM, Pirtskhalava T, Decklever T, Curran GL, Ogrodnik M, Jurk D, Johnson KO, Lowe V, Tchkonia T, Westendorf JJ, Kirkland JL (2017) Transplanted senescent cells induce an osteoarthritis-like condition in mice. J Gerontol Ser A Biol Sci Med Sci 72:780–785 10. Majuta LA, Guedon JG, Mitchell SAT, Kuskowski MA, Mantyh PW (2017) Mice with cancer-induced bone pain show a marked decline in day/night activity. Pain Rep 2:e614 11. Minville V, Laffosse JM, Fourcade O, Girolami JP, Tack I (2008) Mouse model of fracture pain. Anesthesiology 108:467–472 12. Lolignier S, Eijkelkamp N, Wood JN (2015) Mechanical allodynia. Pflugers Archiv 467:133–139 13. Mogil JS (2017) Laboratory environmental factors and pain behavior: the relevance of unknown unknowns to reproducibility and translation. Lab Anim 46:136–141 14. Pomonis JD, Boulet JM, Gottshall SL, Phillips S, Sellers R, Bunton T, Walker K (2005) Development and pharmacological characterization of a rat model of osteoarthritis pain. Pain 114:339–346

INDEX A

E

Annulus fibrosus (AF)...............................................41–51 Antibodies .............................21, 91–93, 95, 96, 99, 100, 196, 208, 221, 225, 229, 230, 236, 242, 244, 245, 256, 257, 261–266, 268, 270–272 Antigen retrieval................ 208, 244, 256, 262, 263, 267

Endosteal mesenchymal progenitors ..........30, 33, 35–38 Enzymatic digestions .............29–38, 42, 90, 98, 99, 271

B Bone marrow...........................15, 18, 19, 25, 29–38, 55, 57, 72, 194, 263 Bone repair ........................................................... 193–203 Bones ...................................3–11, 15–26, 29, 30, 32, 33, 35, 37, 38, 55, 57, 89, 90, 96, 97, 99, 154, 167, 174–178, 182, 185, 186, 188–190, 193, 194, 196, 198, 201–203, 206, 210, 212–214, 217, 220, 221, 224, 234, 242, 248, 256, 262, 263, 266, 267, 270, 275–287, 292, 293

C Cartilage .................................... 21, 41, 53–69, 112, 142, 154, 171, 183, 194, 206, 217, 221, 223–225, 230–232, 240, 242, 245, 247–254, 257, 262, 263, 292 Cell isolation.......................................... 45–47, 51, 90, 97 Chondrocytes ................................ 36, 41, 42, 48, 53–69, 96, 111, 118, 119, 142, 254 Chondrogenesis ..................................15, 18, 20, 21, 102 Closed fracture ..................................................... 205, 206 Co-culture ................................53, 54, 58–61, 63, 65, 68 Collagenase................................ 4–11, 31, 37, 43, 46, 47, 50, 56, 57, 91, 232, 255 Colony forming unit-fibroblast (CFU-F).............. 30, 33, 37, 38 Cryopreservation............................................................. 72 Cultures .............................. 5, 6, 8, 9, 11, 16, 19–21, 23, 25, 26, 29–38, 42–44, 46, 48–50, 54, 58–61, 63, 67, 68, 72, 74–78, 80–82, 86, 87, 109–137, 145, 151, 168, 169, 177, 178, 224, 230

D Decalcification .............................21, 183, 207, 214, 220, 221, 228–229, 237, 244, 262–264, 266–268, 270

F Fixation ............................... 50, 168, 211, 214, 228–229, 236, 237, 244, 262, 266, 268, 270 Fluorescence activated cell sorting (FACS) ............89–91, 93–97, 99, 100 Fluorescence recovery ....... 112, 114–117, 121, 123, 134 Fluorescence recovery after photobleaching (FRAP)...................................................... 109–137 Fractures ............................ 193, 194, 205–222, 291, 292

G Gene expression regulation ....................... 4, 50, 65, 110, 142, 143, 172, 217–219, 222, 226, 230–233, 247, 264

H Healing .............................193, 194, 196, 198, 203, 205, 206, 210–212, 219–221, 225, 293 Histology ........................37, 38, 50, 166, 172, 196, 206, 207, 214–217, 236, 256, 263, 286 Histomorphometry .............................165, 171, 182, 286 Hydroxymethylcytosine ....................................... 101–107

I Immunofluorescence (IF).......................... 114, 116, 135, 229, 230, 236, 239, 245, 247, 256, 261, 265, 268–270 Immunohistochemistry (IHC).................. 183, 208, 216, 221, 229, 230, 236, 239, 242, 244–246, 256, 261, 264, 266–268 Immunostaining........................... 21, 196, 237, 261–263 Incapacitance ............................................... 292, 294, 297 Induced pluripotent stem cells (iPSCs) ..................71–73, 78, 80–85 Intervertebral disc (IVD) .........................................41–51 Isolation ........................ 3–11, 18, 29–38, 42, 43, 51, 57, 89, 90, 96, 97, 202, 208–209, 217–219, 230–232, 247–254, 257

Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221, https://doi.org/10.1007/978-1-0716-0989-7, © Springer Science+Business Media, LLC, part of Springer Nature 2021

301

OSTEOPOROSIS

302 Index

AND

OSTEOARTHRITIS

M

P

Matrix deposition......................................................53, 54 Mechanical allodynia..................................................... 292 Mechanical loading ..................................... 225, 275, 276 Mesenchymal stem cells .......................29, 54, 72, 74, 78, 80–82, 96, 98, 111, 119 Mesenchymal stromal cells (MSCs) ................. 15–27, 29, 53–69, 78, 80 Micro-computed tomography (micro-CT/μCT) ......206, 207, 210, 276 Mouse embryonic fibroblasts (MEFs) ...... 71, 73, 75–78, 80–82, 86

Pain ..........172, 202, 203, 206, 225, 255, 291–293, 298 Protein dynamics.................................110, 121, 123, 132

N Next-generation sequencing ............................... 102–104 Nucleus pulposus (NP).............................................41–51

O Open field activity assays............................................... 298 Osteoarthritis ............................101, 102, 142, 165, 166, 172, 173, 182, 183, 223–258, 291 Osteoblasts ................................4, 5, 7, 8, 10, 11, 30, 36, 89, 96, 157, 276 Osteocytes ...........................................3–12, 97, 198, 276

R Regenerative medicine ................................................ v, 54 Reprogramming ..............................71–73, 78, 80–82, 85 RNA and DNA extraction ..................230–232, 247, 248

S Single cell sequencing ......................................... v, 89–100 Site specificity ...................................... 276–278, 282–287 Static weight bearing .................................. 292, 294, 298 Surgical models .................................................... 223–257

T Tissue engineering ................................................. 54, 154 Torsional testing............................................................ 207 Transcription factor activity.......................................... 110 Trophic effects...........................................................54, 55

V Von Frey .......................................................292–296, 298