Ovarian Cancer: Methods and Protocols (Methods in Molecular Biology, 2424) 1071619551, 9781071619551

Citation preview

Methods in Molecular Biology 2424

Pamela K. Kreeger Editor

Ovarian Cancer Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

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

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

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-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.

Ovarian Cancer Methods and Protocols

Edited by

Pamela K. Kreeger Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, USA

Editor Pamela K. Kreeger Department of Biomedical Engineering University of Wisconsin-Madison Madison, WI, USA

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-1955-1 ISBN 978-1-0716-1956-8 (eBook) https://doi.org/10.1007/978-1-0716-1956-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022, Corrected Publication 2022 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover Illustration Caption: See Figure 2 of Chapter 14 for more details. 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 Ovarian cancer is the general term for a variety of malignancies that arise in the fallopian tube and ovary. While subtypes of ovarian cancer differ in frequency and prognosis, high-grade serous is the most common and remains one of the deadliest cancers. This poor outcome results from three major limitations: (1) late-stage diagnosis due to nonspecific symptoms and no definitive screening tests, (2) high rate of recurrence and development of chemoresistant disease, and (3) substantial heterogeneity that has hindered the identification of new therapies. In order to tackle these challenges, researchers need a robust set of methods to understand variation between patients, in vitro models to better study different stages of the disease, and in vivo models to test therapies. We hope this volume can provide such a toolkit and help to standardize studies across different labs to bring the field closer to answers. The volume is organized into four themes: (1) clinical perspectives to orient basic scientists, (2) methods to characterize features of the tumor microenvironment (TME), (3) techniques to isolate cells for in vitro study, and (4) in vitro and in vivo models to study the disease.

Clinical Perspectives While many current ovarian cancer researchers already work closely with gynecological oncologists and pathologists that specialize in ovarian cancer, we recognize that to find solutions to this complex disease we must build a larger network of researchers. To help orient these new researchers, O’Shea provides an overview of the clinical approach to staging and treating ovarian cancer (Chapter 1). In addition, McGregor provides a detailed perspective on the characteristics that pathologists use to provide an ovarian cancer diagnosis (Chapter 2).

Characterizing the TME As one of the initial cancers included in The Cancer Genome Atlas, we have a relatively large amount of information about genomic and proteomic changes associated with ovarian cancer. Unfortunately, these efforts have largely determined that there are relatively few genetic mutations that can be treated with targeted therapies at high prevalence. Therefore, understanding the changes beyond the tumor cell may prove particularly fruitful for identifying methods to slow or stop ovarian cancer progression. To better understand the diversity of cellular phenotypes, Patankar provides a protocol to utilize multiparameter flow cytometry for the immune cells of the ascites (Chapter 3), and Fantl details the adaptation of mass cytometry to ovarian cancer tumors (Chapter 4). In addition, we provide a protocol for isolation of single cells, multicellular aggregates, and the soluble portion of the ascites for utilization in numerous downstream assays (Chapter 5). To construct relevant in vitro models inclusion of the noncellular fraction, or extracellular matrix (ECM), of the tumor

v

vi

Preface

is essential. To determine the organization and relative levels of components of the ECM, Masters details the adaptation of multispectral immunofluorescence in the context of rare tissues such as early lesions (Chapter 6). For tissues that are more abundant, Fogg provides a protocol for quantitative dot blot in combination with immunofluorescence (Chapter 7).

Cell Isolation Of course, to study tumor biology in vitro, one needs to have a way to access the relevant cell types. While there are numerous ovarian cancer cell lines, many of these have been questioned for their relevance to the clinical disease [1]. In addition, there are very few cell lines that are representative of the other cell types in the tumor microenvironment. Therefore, we provide a series of protocols to isolate and culture cell types found in metastatic ovarian tumors. First, Beaumont provides an overview of a method to dissociate tumors, culture tumor cells in a microfluidic trap, and then link phenotypic characteristics (e.g., growth kinetics, surface marker expression) to downstream single cell RNA-seq (Chapter 8). Next, Fuh provides a method to isolate and culture mesothelial cells, the first cell type contacted by tumor cells as they initiate a new metastatic site (Chapter 9). Finally, we have detailed protocols by Eckert to isolate fibroblasts/cancer-associated fibroblasts (Chapter 10) and Mukherjee to isolate adipocytes (Chapter 11). In addition, we refer readers to previously published protocols to culture and differentiate primary macrophages [2] and to culture human fallopian tube epithelial cells, the potential cell of origin for high-grade serous ovarian cancer [3].

In Vitro and In Vivo Models Identifying new targets and testing therapies require model systems for the many different stages of tumor development. The contribution from Russo provides an in vitro system to study the effects of different perturbations on ciliary beating in the fallopian tube epithelium (Chapter 12), while the protocol from Dean enables the study of metastasis from the fallopian tube to the ovary (Chapter 13). The next stage of metastasis to distal organs such as the omentum can be modeled using the methods detailed by Hilliard (Chapter 14). Finally, Mehta describes how to isolate cancer stem cells and generate cocultures with other cell types in the tumor microenvironment (Chapter 15). While in vitro models enable the study of isolated stages in disease progression, therapeutic assessment will require in vivo models. To fill this need, protocols are provided for the intraperitoneal xenograft/syngeneic models by Pradeep (Chapter 16), patient-derived xenografts with the potential to add patient-matched tumor-infiltrating lymphocytes by Powell (Chapter 17), and the orthotopic model in combination with surgery/chemotherapy that mimics clinical treatment of human disease by McCormack (Chapter 18). Finally, Miller provides a new protocol to combine in vivo models with high-content imaging of cellular signaling (Chapter 19).

Preface

vii

Concluding Thoughts We hope that these contributions improve the ability of researchers to study ovarian cancer and find new therapeutic approaches. We would note that while these protocols were optimized in the context of ovarian cancer, many of them will be broadly useful to the study of other cancers and diseases. I am grateful to John Walker and the editorial staff at Methods in Molecular Biology for providing the opportunity to develop this volume and to the authors and reviewers for meeting deadlines despite the challenges of the COVID-19 pandemic. Madison, WI, USA

Pamela K. Kreeger

References 1. Domcke S, Sinha R, Levine DA, Sander C, Schultz N (2013) Evaluating cell lines as tumour models by comparison of genomic profiles. Nat Commun 4:2126. https://doi. org/10.1038/ncomms3126 2. Classen A, Lloberas J, Celada A (2009) Macrophage activation: classical versus alternative. Methods Mol Biol 531:29–43. https://doi.org/10.1007/978-1-59745-396-7_3 3. Fotheringham S, Levanon K, Drapkin R (2011) Ex vivo culture of primary human fallopian tube epithelial cells. J Vis Exp (51). https://doi.org/10.3791/2728

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

PART I

CLINICAL PERSPECTIVES

1 Clinical Staging of Ovarian Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrea S. O’Shea 2 Pathologic Classification of Ovarian Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stephanie M. McGregor

PART II

v xi

3 11

TUMOR MICROENVIRONMENT

3 Multiparameter Flow Cytometry for Detailed Characterization of Peritoneal Immune Cells from Patients with Ovarian Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Jessica Vazquez, Dagna Sheerar, Aleksandar K. Stanic, and Manish S. Patankar 4 Mass Cytometry for the Characterization of Individual Cell Types in Ovarian Solid Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Veronica D. Gonzalez, Ying-Wen Huang, and Wendy J. Fantl 5 Processing and Analysis of Ascites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Hannah M. Micek, Molly J. Carroll, Lisa Barroilhet, and Pamela K. Kreeger 6 Multispectral Staining and Analysis of Extracellular Matrix . . . . . . . . . . . . . . . . . . . 105 Carine M. Renner, Mike R. Visetsouk, Pamela K. Kreeger, and Kristyn S. Masters 7 Quantitative Analysis of the Extracellular Matrix by Immunoblot . . . . . . . . . . . . . 121 Kaitlin C. Fogg

PART III

CELL ISOLATION

8 Multiparameter Single-Cell Characterization of Ovarian Intratumor Heterogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kristin G. Beaumont, Christina Andreou, Ethan Ellis, and Robert Sebra 9 Culturing Primary Human Mesothelial Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mary Mullen, Hollie Noia, and Katherine Fuh 10 Isolation of Normal and Cancer-Associated Fibroblasts . . . . . . . . . . . . . . . . . . . . . . Katarzyna Zawieracz and Mark A. Eckert 11 Isolation of Primary Normal and Cancer-Associated Adipocytes from the Omentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abir Mukherjee

ix

135

147 155

167

x

Contents

PART IV

MODEL SYSTEMS

12

Isolation of Fallopian Tube Epithelium for Assessment of Cilia Beating Frequency (CBF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Angela Russo and Joanna E. Burdette 13 Ex Vivo Ovarian Culture to Model the Initial Metastasis in Ovarian Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matthew Dean 14 In Vivo and Ex Vivo Analysis of Omental Adhesion in Ovarian Cancer . . . . . . . . Elizabeth I. Harper and Tyvette S. Hilliard 15 The Role of the Tumor Microenvironment in CSC Enrichment and Chemoresistance: 3D Co-culture Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael Bregenzer, Eric Horst, Pooja Mehta, Catherine Snyder, Taylor Repetto, and Geeta Mehta 16 Establishment of In Vivo Ovarian Cancer Mouse Models Using Intraperitoneal Tumor Cell Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sonam Mittal, Prachi Gupta, Pradeep Chaluvally-Raghavan, and Sunila Pradeep 17 Humanized Patient-Derived Xenograft Models of Ovarian Cancer . . . . . . . . . . . . Sarah B. Gitto, Erin George, Sergey Medvedev, Fiona Simpkins, and Daniel J. Powell Jr 18 Xenograft Models of Ovarian Cancer for Therapy Evaluation . . . . . . . . . . . . . . . . Mihaela Popa, Vibeke Fosse, Katrin Kleinmanns, Line Bjørge, and Emmet McCormack 19 Confocal Imaging of Single-Cell Signaling in Orthotopic Models of Ovarian Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dylan O. Matvey, Thomas S. C. Ng, and Miles A. Miller Correction to: Mass Cytometry for the Characterization of Individual Cell Types in Ovarian Solid Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

179

189 199

217

247

255

275

295

C1 317

Contributors CHRISTINA ANDREOU • Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai and Icahn Institute for Data Science and Genomic Technology, New York, NY, USA LISA BARROILHET • Department of Obstetrics and Gynecology, University of Wisconsin School of Medicine and Public Health, Madison, WI, USA; University of Wisconsin Carbone Cancer Center, University of Wisconsin School of Medicine and Public Health, Madison, WI, USA KRISTIN G. BEAUMONT • Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai and Icahn Institute for Data Science and Genomic Technology, New York, NY, USA LINE BJØRGE • CCBIO, Department of Clinical Science, University of Bergen, Bergen, Norway; Department of Obstetrics and Gynaecology, Haukeland University Hospital, Bergen, Norway MICHAEL BREGENZER • Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, USA JOANNA E. BURDETTE • Department of Pharmaceutical Sciences and Center for Biomolecular Sciences, University of Illinois at Chicago, Chicago, IL, USA MOLLY J. CARROLL • Department of Biomedical Engineering, University of WisconsinMadison, Madison, WI, USA PRADEEP CHALUVALLY-RAGHAVAN • Department of Obstetrics and Gynecology, Department of Physiology, Cancer Center, Medical College of Wisconsin, Milwaukee, WI, USA MATTHEW DEAN • Department of Animal Science, University of Illinois at UrbanaChampaign, Urbana, IL, USA MARK A. ECKERT • Department of Obstetrics and Gynecology/Section of Gynecologic Oncology, University of Chicago, Chicago, IL, USA ETHAN ELLIS • Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai and Icahn Institute for Data Science and Genomic Technology, New York, NY, USA WENDY J. FANTL • Department of Urology, Department of Obstetrics and Gynecology, Stanford Comprehensive Cancer Institute, Stanford University School of Medicine, Stanford, CA, USA KAITLIN C. FOGG • School of Chemical, Biological, and Environmental Engineering, Oregon State University, Corvallis, OR, USA VIBEKE FOSSE • CCBIO, Department of Clinical Science, University of Bergen, Bergen, Norway KATHERINE FUH • Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine and Alvin J. Siteman Cancer Center, St. Louis, MO, USA ERIN GEORGE • Ovarian Cancer Research Center, Division of Gynecology Oncology, Department of Obstetrics and Gynecology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA SARAH B. GITTO • Ovarian Cancer Research Center, Division of Gynecology Oncology, Department of Obstetrics and Gynecology, Department of Pathology and Laboratory

xi

xii

Contributors

Medicine, Abramson Cancer Center, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA; Center for Cellular Immunotherapies, University of Pennsylvania, Philadelphia, PA, USA VERONICA D. GONZALEZ • Baxter Laboratory for Stem Cell Biology, Department of Microbiology & Immunology, Stanford University School of Medicine, Stanford, CA, USA; 10X Genomics, Pleasanton, CA, USA PRACHI GUPTA • Department of Obstetrics and Gynecology, Medical College of Wisconsin, Milwaukee, WI, USA ELIZABETH I. HARPER • Integrated Biomedical Sciences Graduate Program, University of Notre Dame, Notre Dame, IN, USA; Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN, USA; Harper Cancer Research Institute, University of Notre Dame, Notre Dame, IN, USA TYVETTE S. HILLIARD • Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN, USA; Harper Cancer Research Institute, University of Notre Dame, Notre Dame, IN, USA ERIC HORST • Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, USA YING-WEN HUANG • Department of Urology, Stanford University School of Medicine, Stanford, CA, USA KATRIN KLEINMANNS • CCBIO, Department of Clinical Science, University of Bergen, Bergen, Norway PAMELA K. KREEGER • Department of Biomedical Engineering, University of WisconsinMadison, Madison, WI, USA; Department of Cell and Regenerative Biology, Department of Obstetrics and Gynecology, and University of Wisconsin Carbone Cancer Center, University of Wisconsin School of Medicine and Public Health, Madison, WI, USA KRISTYN S. MASTERS • Department of Biomedical Engineering and Department of Materials Science and Engineering, University of Wisconsin-Madison, Madison, WI, USA; Departments of Medicine, University of Wisconsin Carbone Cancer Center, University of Wisconsin School of Medicine and Public Health, Madison, WI, USA DYLAN O. MATVEY • Center for Systems Biology, Massachusetts General Hospital Research Institute, Boston, MA, USA EMMET MCCORMACK • CCBIO, Centre for Pharmacy, Vivarium, and Department of Clinical Science, University of Bergen, Bergen, Norway STEPHANIE M. MCGREGOR • Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, Madison, WI, USA SERGEY MEDVEDEV • Ovarian Cancer Research Center, Division of Gynecology Oncology, Department of Obstetrics and Gynecology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA GEETA MEHTA • Department of Biomedical Engineering, Materials Science and Engineering, Macromolecular Science and Engineering, Rogel Cancer Center, and Precision Health, University of Michigan, Ann Arbor, MI, USA POOJA MEHTA • Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI, USA HANNAH M. MICEK • Department of Biomedical Engineering, University of WisconsinMadison, Madison, WI, USA MILES A. MILLER • Center for Systems Biology, Massachusetts General Hospital Research Institute, Boston, MA, USA; Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

Contributors

xiii

SONAM MITTAL • Department of Obstetrics and Gynecology, Medical College of Wisconsin, Milwaukee, WI, USA ABIR MUKHERJEE • Department of Obstetrics and Gynecology/Section of Gynecologic Oncology, University of Chicago, Chicago, IL, USA MARY MULLEN • Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Alvin J. Siteman Cancer Center, Washington University School of Medicine, St. Louis, MO, USA THOMAS S. C. NG • Center for Systems Biology, Massachusetts General Hospital Research Institute, Boston, MA, USA; Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA HOLLIE NOIA • Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, and Alvin J. Siteman Cancer Center, Washington University School of Medicine, St. Louis, MO, USA ANDREA S. O’SHEA • Division of Gynecologic Oncology, Department of Obstetrics, Gynecology and Women’s Health, University of Minnesota Medical School, Minneapolis, MN, USA MANISH S. PATANKAR • Department of Obstetrics and Gynecology, University of WisconsinMadison, Madison, WI, USA MIHAELA POPA • CCBIO, Department of Clinical Science, University of Bergen, Bergen, Norway DANIEL J. POWELL JR. • Ovarian Cancer Research Center, Division of Gynecology Oncology, Department of Obstetrics and Gynecology, Department of Pathology and Laboratory Medicine, Abramson Cancer Center, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA; Center for Cellular Immunotherapies, University of Pennsylvania, Philadelphia, PA, USA SUNILA PRADEEP • Department of Obstetrics and Gynecology, Department of Physiology, Cancer Center, Medical College of Wisconsin, Milwaukee, WI, USA CARINE M. RENNER • Department of Biomedical Engineering, University of WisconsinMadison, Madison, WI, USA TAYLOR REPETTO • Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI, USA ANGELA RUSSO • Department of Pharmaceutical Sciences and Center for Biomolecular Sciences, University of Illinois at Chicago, Chicago, IL, USA ROBERT SEBRA • Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai and Icahn Institute for Data Science and Genomic Technology, New York, NY, USA; Sema4, a Mount Sinai venture, Stamford, CT, USA; Black Family Stem Cell Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA DAGNA SHEERAR • University of Wisconsin Carbone Comprehensive Cancer Center, University of Wisconsin-Madison, Madison, WI, USA FIONA SIMPKINS • Ovarian Cancer Research Center, Division of Gynecology Oncology, Department of Obstetrics and Gynecology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA CATHERINE SNYDER • Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI, USA ALEKSANDAR K. STANIC • Department of Obstetrics and Gynecology, University of WisconsinMadison, Madison, WI, USA

xiv

Contributors

JESSICA VAZQUEZ • Department of Obstetrics and Gynecology, University of WisconsinMadison, Madison, WI, USA MIKE R. VISETSOUK • Department of Biomedical Engineering, University of WisconsinMadison, Madison, WI, USA KATARZYNA ZAWIERACZ • Department of Obstetrics and Gynecology/Section of Gynecologic Oncology, University of Chicago, Chicago, IL, USA

Part I Clinical Perspectives

Chapter 1 Clinical Staging of Ovarian Cancer Andrea S. O’Shea Abstract Ovarian cancer may arise from any of the histologic portions of the ovary including the epithelium, stroma, or germ cells. Of these, high-grade serous carcinoma arising from the epithelium of the ovary is the most common type. The clinical management and prognosis of ovarian cancer depend upon the stage of the cancer. Cancer stage is the extent to which the cancer has spread from its site of origin. For ovarian cancer, staging guidelines are determined by FIGO, the Fe´de´ration Internationale de Gyne´cologie et d’Obste´trique. The stage of ovarian cancer is determined by performing surgery to remove the ovaries and other gynecologic organs as well as lymph nodes and other tissues where the cancer may have spread. The histologic specimens from this surgery provide information from which the stage can be determined. In more advanced cases, this surgery may also include procedures to remove other areas of cancer. The stage of ovarian cancer guides treatment and is also the most important factor in ovarian cancer prognosis. Most epithelial ovarian cancers are diagnosed in advanced stages and are treated with surgery and chemotherapy. Despite aggressive treatment, the survival of patients with advanced stage epithelial ovarian cancer remains low, and more effective diagnostic and therapeutic approaches are needed. Key words Ovarian cancer, Epithelial ovarian cancer, International Federation of Gynecology and Obstetrics (FIGO), Staging, Surgical staging

1

Introduction The stage of a cancer is the extent to which the cancer has spread from the site of origin. Stage is determined differently for different types of cancer, with each type of cancer having its own staging protocols and staging system. For ovarian cancer, staging guidelines are determined by the International Federation of Gynecology and Obstetrics, referred to as FIGO, the acronym of its French name, Fe´de´ration Internationale de Gyne´cologie et d’Obste´trique [1]. Accurate determination and reporting of the stage of ovarian cancer have both prognostic and treatment implications. It is also important for clinical trial eligibility and to ensure accurate communication surrounding all aspects of ovarian cancer. Ovarian

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_1, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

3

4

Andrea S. O’Shea

cancer is staged surgically, which means that the stage of ovarian cancer is established by performing surgery to determine the extent to which the cancer has spread. It is important to note that because the ovaries are located within the peritoneal cavity, a mass that is limited to the ovary and suspected of being cancerous is not biopsied preoperatively due to concern for seeding of the peritoneal cavity. Thus, surgery is performed with the suspicion of ovarian cancer and the diagnosis is confirmed intra-operatively after removal of the ovary with staging proceeding from there. Although generally applying to malignancies arising from the ovary itself, the term ovarian cancer can also refer to cancers arising from the fallopian tube and the peritoneum—the lining of the abdomen and abdominal viscera. These malignancies are collectively referred to under the umbrella term of ovarian cancer owing to a similar cell of origin, involvement of the ovary in nearly all cases, and identical treatment paradigms. Ovarian cancers are broadly categorized by the histologic portion of the ovary from which they arise: the epithelium (constituting the overlying mesothelial layer), the stroma, or the germ cells. The vast majority of ovarian malignancies have an epithelial origin, and epithelial ovarian cancer is the leading cause of death from gynecologic cancer in the United States and Europe [2]. Epithelial ovarian cancer includes the most common subtype, highgrade serous carcinoma, as well as the other subtypes listed in Table 1. Far less common than epithelial ovarian cancers are malignant tumors arising from the supporting stromal cells or the ovarian germ cells. These subtypes are referred to as malignant sex cord-stromal tumors or malignant germ cell tumors, respectively [2]. The complete list of ovarian cancer histologic subtypes, or histotypes, is shown in Table 1. Although these various types of ovarian cancers have differences in risk factors, patterns of spread, response to chemotherapy, tumor markers, and prognosis, the staging system used for all classifications of ovarian cancer is the same. For the purposes of Table 1 Subtypes of malignant ovarian tumors Epithelial ovarian cancers

Sex-cord stromal tumors

Germ cell tumors

High-grade serous Low-grade serous

Granulosa cell tumor

Dysgerminoma Immature teratoma

Clear cell Mucinous

Sertoli–Leydig cell tumor

Yolk sac tumor Embryonal carcinoma

Endometrioid Carcinosarcoma

Sex-cord tumor with annular tubules

Choriocarcinoma Mixed germ cell tumor

Ovarian Cancer Staging

5

this chapter, the term ovarian cancer will be used to refer to the most common histologic type, epithelial ovarian cancer, in all its subtypes, including high-grade serous carcinomas originating from the fallopian tube or peritoneum. When discussing ovarian cancer, it is important to denote the specific type of ovarian cancer, including the subtype and grade where applicable [3].

2

FIGO Staging At the time of this publication, the most recent staging for ovarian cancer was provided by FIGO in 2014 and is shown in Table 2 [4]. This staging system was also approved by the American Joint

Table 2 FIGO 2014 Staging of carcinoma of the ovary, fallopian tube, and primary peritoneal carcinoma Stage I: Tumor confined to ovaries/fallopian tubes IA

Tumor limited to 1 ovary/tube, capsule/serosa intact, no tumor on surface, negative washings

IB

Tumor involves both ovaries/tubes, otherwise like IA

IC

Tumor limited to 1 or both ovaries/tubes IC1 Surgical spill IC2 Capsule rupture before surgery or tumor on ovarian/serosal surface IC3 Malignant cells in ascites or peritoneal washings

Stage II: Pelvic extension or primary peritoneal cancer IIA

Extension and/or implant on uterus and/or tubes(ovaries) (not applicable to primary peritoneal cancer)

IIB

Extension to other pelvic intra-peritoneal tissues

Stage III: Confirmed spread to extra-pelvic peritoneum and/or metastasis to the retroperitoneal lymph nodes IIIA Positive retroperitoneal lymph nodes and/or microscopic metastasis beyond the pelvis IIIA1 Positive retroperitoneal lymph nodes only IIIA1(i)

Metastasis 10 mm

IIIA1(ii)

Metastasis >10 mm

IIIA2 Microscopic, extrapelvic (above the pelvic brim) peritoneal involvement  positive retroperitoneal lymph nodes IIIB Macroscopic, extrapelvic, peritoneal metastasis 2 cm  positive retroperitoneal lymph nodes. Includes extension to capsule of liver/spleen IIIC Macroscopic, extrapelvic, peritoneal metastasis >2 cm  positive retroperitoneal lymph nodes. Includes extension to capsule of liver/spleen Stage IV: Distant metastasis excluding peritoneal metastasis IVA Pleural effusion with positive cytology IVB Hepatic and/or splenic parenchymal metastasis, metastasis to extra-abdominal sites (including lymph nodes)

6

Andrea S. O’Shea

Committee on Cancer (AJCC), which provides a second staging system known as the Tumor–Nodes–Metastasis (TNM) staging system [5]. The TNM Staging System is based on the extent of the tumor (T), spread to lymph nodes (N), and the presence of metastases (M) [6]. Although for ovarian cancer the TNM stage is frequently included in pathology reports, clinically this information is synthesized and reported as the corresponding FIGO stage. One exception that may be encountered is in the case of a patient who does not undergo a full staging procedure, generally due to the omission of a lymph node dissection. Although in some cases still referred to by its FIGO stage, in this case the stage is more accurately reported as a TNM stage with an X denoting the lymph node portion of the staging. For example, a stage of T1aNXM0 for a tumor confined to one ovary (T1a) with no lymph node assessment (NX) and no metastases (M0).

3

Surgical Staging of Ovarian Cancer Staging surgery for suspected early stage ovarian cancer includes the removal of all pelvic gynecologic organs: the uterus and cervix (total hysterectomy), both fallopian tubes and ovaries (bilateral salpingo-oophorectomy), a pelvic and para-aortic lymph node dissection (lymphadenectomy), removal of the omentum (omentectomy), peritoneal biopsies, and collection of pelvic fluid or washings. For mucinous carcinoma of the ovary, the rate of lymph node metastases is low and routine lymphadenectomy may be omitted [7, 8]. For advanced ovarian cancer where disease is obviously present beyond the ovary, staging takes place as part of an initial debulking surgery. This is discussed in greater detail in the treatment portion of the chapter. This staging surgery is done for the dual purpose of determining the stage and removing the cancer. In some cases, surgery as the initial treatment for ovarian cancer may not be feasible or may not be in the best interest of the patient. In these cases, a diagnosis will be made through a biopsy of a metastatic site of cancer (often the omental mass/cake or an enlarged lymph node) or by ascitic fluid cytology. Biopsy is generally preferred due to increased sensitivity and ability to determine the subtype and grade by histology. The stage is then determined to the extent possible clinically using the histologic or cytologic diagnosis combined with physical examination and imaging findings. When reporting the stage of ovarian cancer, it should include the primary site if possible (i.e., ovary, fallopian tube, or peritoneum) which is determined pathologically, the histologic subtype, and the grade where applicable [3]. For ovarian cancer, a stage is only assigned at the initial diagnosis/appearance of the cancer, and

Ovarian Cancer Staging

7

the cancer is never re-staged if/when it recurs, instead being referred to as recurrent ovarian cancer regardless of the extent of recurrence. A few common pitfalls of ovarian cancer staging include bowel infiltration to the mucosa, hepatic and splenic parenchymal involvement, inguinal lymph node involvement, and umbilical or abdominal wall deposits, all of which are designated as Stage IVB rather than Stage IIIC despite their intra-abdominal or nearby location. Additionally, to make a diagnosis of Stage IVA disease, pleural cytology showing malignant cells within a pleural effusion is required; the diagnosis cannot be made from imaging alone [3]. The majority of ovarian cancers are diagnosed at an advanced stage, with only about 25% of ovarian cancers diagnosed at an early stage (Stage I or II disease) [3, 9]. Early stage tumors are more likely to be mucinous, clear cell, or endometrioid histology, whereas high-grade serous carcinoma most commonly presents at an advanced stage [3].

4

Ovarian Cancer Grade Another important component of ovarian cancer is the grade of the tumor. Grade is a measure of how abnormal cancer cells look microscopically as compared to normal tissue. In general, the closer tumor cells appear to normal tissue, the lower the grade. Typically, high-grade tumors have a more aggressive clinical course. Epithelial ovarian cancers are graded on either a two- or three-tier grading system depending upon the specific histologic subtype. Serous carcinoma is graded on a two-tier system as either high- or lowgrade. Mucinous and endometrioid carcinoma are graded on a three-tier system as Grade 1, 2, or 3. The remainder of epithelial ovarian cancers are regarded as high grade although this is not overtly specified when reporting the stage.

5

Treatment of Ovarian Cancer Stage and histologic subtype determine the specific treatment for ovarian cancer, but in most cases, treatment includes a combination of surgery and chemotherapy. For suspected early stage ovarian cancer (those without overt evidence of extra-ovarian disease), the initial treatment includes a staging surgery as noted above. If early stage disease is confirmed, additional treatment depends upon the specific stage and grade. For Stage IA and IB, low-grade ovarian cancers, treatment with chemotherapy after surgery has not been shown to improve patient outcomes [10, 11]. These patients have a 5-year overall survival rate of >95% in the absence of any additional

8

Andrea S. O’Shea

treatment [10]. For all other epithelial ovarian cancers, treatment with platinum-based chemotherapy is recommended following surgery [12]. For advanced-stage ovarian cancer where cancer is widely disseminated, the goal of surgery shifts to removal of all visible cancer, termed debulking or cytoreductive surgery. This debulking surgery is unique to ovarian cancer due to the way in which ovarian cancer metastasizes: spreading along the peritoneal surfaces without invasion until late in the disease process [13]. This surgical procedure includes a total hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and any additional procedures necessary to remove all visible cancer, which may include procedures such as bowel resection(s) or splenectomy (removal of the spleen). In patients with Stage III or IV disease who do not have enlarged lymph nodes, lymph node dissection is not performed during debulking surgery, as doing so has not been shown to improve patient survival and is associated with additional complications [14]. The term “optimal debulking” is applied to surgical procedures in which any remaining site of cancer at the conclusion of surgery is 1 cm. The size of residual disease deposits at the conclusion of surgery correlates inversely with patient survival [15]. In patients with advanced stage disease, surgery alone is not curative; thus, these patients must also be treated with chemotherapy after surgery. Chemotherapy should include a platinum-based agent, usually carboplatin, ideally in combination with a taxane, generally paclitaxel. Additional agents are continually being studied; however, to date none have been shown to improve overall survival over a platinum and taxane regimen [15, 16]. Each treatment course of chemotherapy is referred to as a cycle, and in the treatment of ovarian cancer, treatment consists of six cycles of chemotherapy with each cycle being 3 weeks in length. In certain patients, particularly those with extensive metastatic disease, chemotherapy may be given prior to surgery. This is termed neoadjuvant chemotherapy. For these patients, chemotherapy cycles are split with an interval surgery (termed interval debulking) in the middle of the six cycles. The goal of the interval debulking surgery is the same as upfront or initial debulking surgery, to remove the gynecologic organs and all visible diseases. The remainder of the chemotherapy is given after the surgery. This approach has been shown to be non-inferior to upfront surgery [17–19]. Although many variables affect ovarian cancer survival rates, the most important prognostic factors predicting long-term survival from ovarian cancer are the FIGO stage and the amount of disease remaining after debulking surgery [2]. Survival for ovarian cancer is inversely related to cancer stage with a 5-year relative survival of 89% for patients with Stage I disease, 70% for those with Stage II disease, 36% for those with Stage III disease, and 17% for those with Stage IV disease [20]. Likewise, survival is

Ovarian Cancer Staging

9

inversely related to residual disease after debulking surgery with suboptimal (>1 cm), optimal, and microscopic residual diseases having overall survival rates of 33, 40, and 68 months, respectively [15].

6

Conclusion Stage at diagnosis remains the most important prognostic factor in ovarian cancer, with greater than 70% of epithelial ovarian cancers diagnosed at an advanced stage and 95% of ovarian malignancies, there is an emphasis on epithelial tumors. Key words Ovarian cancer, Pathologic classification

1

Introduction The term “ovarian cancer” actually incorporates many diseases, most of which are entirely unrelated to one another, excepting of course that they occur within the ovary [1]. In some cases, such as germ cell tumors, the cell of origin unquestionably resides in the ovary proper; in many other types, the tumor actually stems from other organs. Indeed, most epithelial cancers, i.e., carcinoma, are thought to emerge from either the adjacent fallopian tube epithelium or from deposits of endometriosis. The approach to pathologic characterization of these tumors is rooted in traditional morphologic assessment by light microscopy, and importantly, the initial determination of pathologic classification is frequently limited to the relatively crude technique of a hematoxylin and eosinstained frozen section. With rare exception, only patients who initially present with extraovarian spread of their disease have tissue that is amenable to needle core biopsy. Without a known diagnosis prior to entering surgery—but with the need for such a diagnosis to determine which type of surgery to pursue—an intraoperative frozen section is often required to guide immediate surgical

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_2, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

11

12

Stephanie M. McGregor

management, and in current practice, this technique is limited to light microscopy. After the surgery has concluded, this basic assessment of morphology by light microscopy can be supplemented with ancillary studies as necessary, generally by immunohistochemistry (IHC). Microscopic features must be viewed in the context of both the clinical presentation and gross findings, including the other relevant organs, such as the fallopian tube and omentum (see Notes 1 and 2). Molecular studies on ovarian tumors are performed primarily to identify germline mutations or actionable mutations that can drive therapeutic decision-making and only rarely have a place in the diagnostic process.

2

Materials Materials for the gross evaluation, sampling, and processing of tissue, as well as those for slide preparation are housed within clinical laboratories, but they mirror those that are used for processing tissue from other species.

3

Methods

3.1 Specimen and Tissue Sampling

1. Ovaries are generally brought into the grossing area of surgical pathology laboratories immediately upon removal, often with a request for intraoperative frozen section in order to evaluate for malignancy in cases with high suspicion. Prior to slide preparation, the specimen must undergo thorough gross evaluation to determine what tissue should be evaluated histologically. Gross examination begins with careful inspection of the ovarian surface to search for rupture or tumor deposits, which impact the clinical stage in the event that a malignancy is identified (Fig. 1). At least one representative section (1–2 cm in greatest dimension) of the most solid-appearing tissue is selected. A slide is prepared immediately, and a diagnosis is relayed to the operating team to guide intraoperative decision-making, i.e., whether or not the patient will undergo immediate surgical staging, such as removal of the omentum and lymph nodes, or if only removal of the ovary is more appropriate at that time. 2. After the frozen section is performed, the remainder of the tissue can be sampled in the fresh state or placed into formalin for fixation prior to tissue sampling. In either case, the tissue must undergo formalin fixation prior to additional processing (see Note 3). 3. The internal aspects of the ovary may be solid or cystic, and it is important to note the composition of each, as well as the color and consistency. Cyst linings must be inspected for

Pathologic Classification of Ovarian Cancer

13

Fig. 1 Cystic ovary prior to opening in the pathology laboratory. The specimen is largely intact, but rupture is evident in the upper left of the photograph, where tumor is extruding through the ovarian serosa

excrescences, which are more likely to harbor malignancy than a smooth lining (Fig. 2). For complete evaluation, specimens should be sampled with one section per centimeter of the greatest tumor dimension, with the exception of mucinous tumors, for which there should be two sections per centimeter. All gross appearances of the tumor must be sampled. 4. Tissue sections should be no more than 2 mm in thickness and have area surrounding them within a cassette to ensure proper fixation and processing (Fig. 3). Inadequate fixation and processing can result in poor tissue quality for future applications. 5. Essentially all routine clinical work uses only formalin-fixed paraffin-embedded (FFPE) tissue; even tissue frozen for intraoperative evaluation is thawed and then fixed for further processing. Therefore, investigators with a need for fresh or frozen tissue need to have a prospective collection process that operates in conjunction with the clinical workflow and relevant research infrastructure within their organization. 3.2

Slide Preparation

1. For optimal results, tissue should be fixed in formalin for at least 6 and no more than 72 h prior to processing. Once tissue has been processed and embedded within paraffin, it is stable for years. 2. In routine clinical workflow, a single 4–5 μm slide is cut using a microtome and stained with hematoxylin and eosin according

14

Stephanie M. McGregor

Fig. 2 Cut surface of an ovarian low-grade serous carcinoma. A range of appearances are evident, including smooth and glistening cyst wall, cauliflower-like excrescences, and solid growth. Areas with solid growth are more likely to contain diagnostic carcinoma, whereas cauliflower-like excrescences are more characteristic of borderline tumors. Smooth cyst wall components are unlikely to contain architecture with the complexity of either borderline tumors or carcinoma. Despite these general trends, gross appearances can be misleading and must be complemented with microscopic interpretation

Fig. 3 Block of formalin-fixed and paraffin-embedded tissue. In this block, ink has been applied to the ovarian surface in order to facilitate histologic examination for potential surface involvement. The presence of ink also helps ensure that the tissue is oriented properly at the time of embedding into paraffin

Pathologic Classification of Ovarian Cancer

15

to standard methods. Additional slides of the same thickness will be cut for any ancillary studies that are necessary for diagnosis. Some amount of tissue is inevitably lost in the process of “facing” the block to achieve representation of all the tissues; how much remains in the block after sections have been taken is variable, but it will certainly be less than 2 mm. 3. Remaining tissue is archived according to regional guidelines. In the United States, guidelines put forth by the College of American Pathologists require that blocks are stored for at least 10 years, but many academic institutions keep them indefinitely. 3.3 Approach to Histologic Assessment

1. Essentially any type of cell can form a tumor in the ovary. Broad categories include epithelial tumors (which are referred to as carcinoma when malignant), germ cell tumors, and sex cordstromal tumors (cells thought to be derived from the stroma of the genital ridge that support growth of germ cells). Each of these categories contains both benign and malignant entities. In addition to these major categories, there can be other types of tumors in the ovary, such as sarcomas that originate from mesenchymal cells or lymphoma. 2. Histologic assessment incorporates multiple factors simultaneously, collectively encompassing the overall architectural complexity of the lesion and the types of cells within it. In assessing these features, the pathologist is asking whether the tumor is primary to the ovary or a metastasis from another site, and in either case, how it should be classified further. In cases with distinctive findings, morphology alone is sufficient. But in other cases, IHC may be necessary even to place a tumor within a broad category. For example, a tumor with rounded cells could be a carcinoma or a sex cord-stromal tumor with rounded cells mimicking carcinoma, a difference that could have clinical implications. Positive staining for keratins and absence of sex cord-stromal markers, such as inhibin, can support epithelial differentiation and therefore broad classification as carcinoma [2]. Then additional stains could be performed to classify the carcinoma further, if necessary [3]. Similarly, germ cell tumors generally express SALL4, which can be used to place a tumor into the germ cell category, and specific types can be further identified with additional markers [4, 5]. It is important to note that immunostains are generally not entirely sensitive or specific for any given entity, so they must always be interpreted in the context of morphology and clinical context.

16

Stephanie M. McGregor

3.4 Approach to Epithelial Tumors

1. Epithelial tumors of the ovary are classified according to the appearance of the epithelium proper and its architectural complexity, both of which are considerations for predicting how likely the tumor is to behave aggressively [1, 3, 6]. 2. Categories based on architectural complexity include cystadenoma, borderline tumor/atypical proliferative tumor, and carcinoma. Cystadenomas have the least complex architecture and are benign, whereas carcinoma demonstrates highly complex architecture and is frankly malignant. Borderline tumors/atypical proliferative tumors have intermediate features and are regarded as having lower malignant potential. These categories will be discussed in more detail below for each of the major types of epithelium. 3. The major types of epithelial tumors based on the appearance of the epithelium proper include serous, endometrioid, mucinous, and clear cell. Serous and endometrioid tumors are so-named because of their resemblance to the normal fallopian tube and endometrial epithelium, respectively; mucinous tumors have readily evident mucin within their cytoplasm, which can be similar in appearance to normal endocervical or gastrointestinal epithelium (Fig. 4). Clear cell tumors do not resemble normal epithelium, and as the name implies, typically have clear cytoplasm (see Subheading 3.8). 4. In addition to a descriptive term about the qualitative appearance, tumors are also frequently given a grade, which is intended to provide further risk stratification to aid in clinical decision-making. Grade is assigned based on how similar the tumor is to its corresponding normal tissue. Low-grade tumors generally have relatively small nuclei, low mitotic activity, and some recapitulation of normal structures, such as glands. In contrast, high-grade tumors tend to have large and irregular nuclei, often with prominent nucleoli and uneven chromatin, as well as brisk mitotic activity and architecture that deviates substantially from normal tissue.

3.5 High-Grade Serous Carcinoma

1. Morphologically, high-grade serous carcinoma (HGSC) consists of severely atypical cells, which are characterized by enlarged nuclei with prominent nucleoli and brisk mitotic activity (Fig. 5). Previously known as papillary serous carcinoma, it is common for this tumor to exhibit papillary growth, but it commonly demonstrates a variety of other growth patterns and in many cases actually has no detectable papillary component (Fig. 6). 2. If it is unclear that a tumor with serous morphology is high grade, IHC can be performed for p53 as a surrogate for the presence of TP53 mutation, which is a characteristic of HGSC

Pathologic Classification of Ovarian Cancer

17

Fig. 4 Representative images of normal epithelium. (Top left) Mucinous glandular epithelium of the endocervix with small basally located nuclei and abundant mucin (arrow). (Top right) Weakly active endometrium with columnar epithelium and intervening specialized endometrial stroma, which consists of plump spindle cells (arrow). (Lower left) Mixed epithelium of the fallopian tube includes secretory (left arrow) and ciliated cells (right arrow). (Lower right) Mucinous epithelium of the colon demonstrates interspersed goblet cells that contain large mucin droplets (arrow)

Fig. 5 High-grade serous carcinoma. The tumor demonstrates large and hyperchromatic nuclei, many with prominent nucleoli, which in some cases have a cherry-red appearance. Brisk mitotic activity is evident, including atypical mitotic figures (arrows)

18

Stephanie M. McGregor

Fig. 6 Representative growth patterns of high-grade serous carcinoma. (Top left) Papillary architecture with narrow and tightly packed fibrovascular cores containing congested vasculature. In this example, the cores are so compressed that the papillary architecture may be difficult to appreciate. (Top right) Predominant papillary and focal micropapillary architecture. Papillary architecture is distinguished by evident fibrovascular cores underlying the epithelium, while the micropapillary component consists of small tufts of cells without discernible fibrovascular cores. (Bottom left) Predominantly solid growth with focal cribriform pattern, characterized by small spaces in a background of solid growth. (Bottom right) Transitional pattern, characterized by broad fronds of tumor in a stratified pattern that resemble those seen in urothelial carcinoma, closely abutting one another in this example

(Fig. 7) [7]. In the absence of TP53 mutation, nuclear p53 staining is present in a minority of cells and with variably weak to moderate intensity (“wild type” pattern). Tumors with TP53 mutation can have one of several aberrant patterns, including strong and diffuse staining in at least 80% of tumor cells, a complete absence of staining (“null” pattern), or cytoplasmic staining. These staining patterns can be used to distinguish HGSC from low-grade serous carcinoma in difficult cases, as the latter has p53-independent pathogenesis and should have a wild-type pattern for p53 by IHC. Of note, the null pattern is only reliable when a wild-type pattern is evident in the surrounding or entrapped benign tissue (e.g., blood vessels and reactive fibroblasts).

Pathologic Classification of Ovarian Cancer

19

Fig. 7 Aberrant staining patterns for p53. (Top) Paired H&E and p53 immunostains demonstrate strong and diffuse staining in >90% of tumor cells. The wild-type pattern in the surrounding stroma illustrates the difference in intensity that is characteristic of this mutant pattern. (Bottom) Paired H&E and p53 immunostains demonstrate the null pattern for p53. The presence of wild-type staining in the surrounding stroma serves as an internal positive control to indicate that the stain is not a false negative for technical reasons

3. Though most other types of ovarian epithelial malignancy range from cystadenoma to carcinoma (discussed below), HGSC has no associated benign counterpart within the ovary. Instead, it is thought to originate, in at least most cases, from the fallopian tube epithelium [8–10]. These precursor lesions, known as serous tubal intraepithelial carcinoma (STIC), are thought to arise from the secretory epithelium and are usually within the fimbriated end of the tube rather than in the tube proper (Fig. 8). STIC lesions must exhibit cytologic atypia similar to that seen in HGSC. In the absence of associated HGSC or unequivocal morphology, IHC for p53 and Ki-67 is performed to establish the diagnosis. Stains are supportive when there is aberrant p53 staining, in any of the patterns described above, and elevated Ki-67 labeling, which acts as a surrogate for proliferative activity. Unfortunately, reproducibility between observers is difficult, and the threshold for what constitutes “elevated” labeling is inconsistent among various authors, but 10% is the prevailing threshold currently [11–13].

20

Stephanie M. McGregor

Fig. 8 Serous tubal intraepithelial carcinoma (STIC). Hematoxylin and eosinstained section of fimbriated end of the fallopian tube demonstrates abrupt transition between normal tubal epithelium and a STIC lesion. The normal epithelium is a single layer of small cells with well-developed cilia and without mitoses. In contrast, the STIC is crowded and disorganized, with enlarged, hyperchromatic nuclei that contain prominent nucleoli, and has readily apparent mitotic activity

Table 1 Criteria for assigning primary site in high-grade serous carcinoma Disease distribution

Assigned primary site Notes

STIC and/or mucosal invasion present Fallopian tube in fallopian tube with any amount of ovarian and/or extraovarian disease

Fallopian tube is also assigned as the primary site if the fallopian tube is incorporated into the tumor

Ovarian disease present without STIC or mucosal invasion in fallopian tube, with or without extraovarian disease

Ovary

Fallopian tubes must be identified grossly and fully examined histologically

Normal ovaries and fallopian tubes with presence of peritoneal disease

Peritoneum

A diagnosis of primary peritoneal HGSC cannot be made if the patient has had neoadjuvant chemotherapy Fallopian tubes must be identified grossly and fully examined histologically

No residual disease after neoadjuvant chemotherapy

Tubo-ovarian

4. With the understanding that HGSC likely originates in the fallopian tube even when it presents as an ovarian mass, a consensus approach has been developed for assigning primary site for these tumors (Table 1) [8]. Using this classification, approximately 80% of cases are actually classified as having originated within the fallopian tube, but they are often casually referred to as ovarian cancer in the clinical setting and archived tissue obtained prior to this shift in dogma will consistently be recorded as an ovarian primary [14, 15].

Pathologic Classification of Ovarian Cancer

21

5. In addition to STIC, which is considered grounds for identifying the fallopian tube as a primary site, there are less advanced presumed precursor lesions that are of no clinical significance in modern practice but hold academic interest as we try to understand the origin of HGSC. In the presence of morphology that is suggestive of a STIC, but for which there is not conspicuous mitotic activity or elevated Ki-67 labeling, a diagnosis of serous tubal intraepithelial lesion (STIL) is appropriate [12, 16]. Cases with neither cytologic atypia nor elevated Ki-67 but with aberrant p53 staining are classified as “p53 signature” lesions; in the clinical setting, these are identified only as an incidental finding in histologically normal tissue that has been stained for p53 to investigate adjacent sites of lesional tissue, but they do appear to be related to HGSC pathogenesis based on molecular studies [17]. Histologic prominence of secretory cells, known as secretory cell outgrowth (SCOUT), has also been described as a presumed precursor, but this is less established [10, 18]. 6. Many patients with HGSC undergo neoadjuvant chemotherapy followed by interval debulking, in which an attempt is made to remove all visible residual tumor. The amount of treatment effect varies among cases, but most cases have at least an intermediate response in which there is a reduction in tumor volume compared to that present before chemotherapy. The remaining tumor is usually associated with inflammation and fibrosis. In some cases, residual tumor may exist only in the form of single scattered cells (Fig. 9). Morphologic evidence of the response to therapy is included in the pathology report in the form of the chemotherapy response score (CRS, Table 2) [19]. The CRS has been widely adopted, but it is important to recognize that it is based only on the amount of omental

Fig. 9 High-grade serous carcinoma with treatment effect. Residual carcinoma cells (arrows) are present as single cells and small clusters with a wide range of nuclear size and shape, often with abundant eosinophilic cytoplasm

22

Stephanie M. McGregor

Table 2 Criteria for chemotherapy response score Score

Description

CRS1: No definitive or minimal response. Identified

Abundant tumor is present and either fibrosis and inflammation are absent or they are essentially present only within dense tumor

CRS2: Moderate response identified Tumor is easily identified with a fairly regular distribution throughout involved areas, but abundant inflammation and fibrosis are also present in excess of what would be expected without treatment CRS3: Marked response with no or minimal residual cancer

No more than scattered foci of tumor cells are present within a background of fibrosis and inflammation, with no nodules exceeding 2 mm in greatest dimension

disease and therefore does not necessarily reflect the amount of tumor that is present in the ovary or other organs. Of note, care must be taken when using tissue that has been exposed to chemotherapy, as this exposure has the potential to affect results. If chemotherapy-naı¨ve tissue is required but only limited FFPE tissue is needed, material remaining from the diagnostic biopsy prior to administration of neoadjuvant chemotherapy has the potential to be useful in select cases. 7. While it has no role in diagnosis, molecular testing for homologous recombination deficiency is often performed in patients without germline BRCA1/2 mutation, because these tumors are likely to respond to poly (ADP-ribose) polymerase (PARP) inhibitors [20]. Patients with germline BRCA1/2 mutation do not undergo this testing because they have presumed homologous recombination deficiency based on their germline mutation. 3.6 Low-Grade Serous Neoplasia

1. While HGSC is postulated to originate from the fallopian tube, and there is no intermediate lesion to confound interpretation, low-grade serous carcinoma (LGSC) is regarded as emerging from a cystadenoma to borderline tumor continuum within the ovary, which can be collectively referred to as low-grade serous neoplasia (LGSN). Cystadenomas are benign and characterized by cystic spaces lined by a single layer of epithelium lacking significant cytologic atypia or mitotic activity; broad papillae are frequently present, but there is not extensive branching (Fig. 10). Serous borderline tumors (SBT) are regarded as having low malignant potential and exhibit extensive hierarchical branching of papillary structures, with cell stratification and tufting, mild to moderate cytologic atypia in the form of moderately enlarged nuclei with conspicuous but not prominent nucleoli, and detectable mitotic activity (Fig. 11). LGSC is a

Pathologic Classification of Ovarian Cancer

23

Fig. 10 Serous cystadenoma. Large fronds are lined by a single layer of epithelium, which ranges from flattened to low columnar. The cells lack cytologic atypia and mitotic activity is not evident

Fig. 11 Serous borderline tumor (SBT). Hierarchical branching in SBT result in the appearance of detached tufts of epithelial cells around the surfaces of tumor papillae

frank malignancy that is distinguished from SBT by the presence of invasive nests of tumor cells, which frequently demonstrate reversed polarity, i.e., nuclei polarized toward the outside of the nests as opposed to the basally oriented nuclei of SBT (Fig. 12) [21, 22]. In some cases, the entire tumor has sufficient complexity to be regarded as carcinoma, but there is often a background of tumor that in isolation would be regarded as SBT; if the areas of carcinoma are 90% of cases of AGCT and can be used for diagnostic purposes in particularly difficult cases, but the finding of FOXL2 mutation is not entirely specific and must be interpreted in the context of all available clinical and pathologic findings; FOXL2 mutations are not characteristic of JGCT, occurring in 6 parameters per single cell. By contrast, mass cytometry can, with minimal signal overlap, distinguish isotopes separated by 1 atomic mass unit (amu) permitting simultaneous analysis of up to 60 parameters per single cell (Fig. 1). In the latest, third generation mass cytometer, the mass range for metal isotope detection lies between 75 and 209 amu [12, 13]. This has enabled the use of non-lanthanide metal isotopes within this mass range to be developed into reagents, not only for additional antibody tags, but for other cellular readouts: palladium

Mass Cytometry of Ovarian Solid Tumors

61

isotopes (102, 104, 105, 106, 108, 110 amu) for barcoding cells [14, 15], rhodium (103Rh), and cisplatin (a platinum (Pt)-based chemotherapeutic agent) for cell viability. The latter is measured with 195Pt, the most abundant Pt isotope [16, 17]. Iodine (127I) as iododeoxyuridine is used for cell cycle measurements [18], iridium (191/193Ir) as a DNA intercalator to denote nucleated single cells [17], and bismuth (209Bi) as an additional antibody tag [13] (Fig. 1). Although not used in our current ovarian mass cytometry studies, isotopes of cadmium, platinum, ruthenium, osmium, and tellurium have been used for antibody labeling and barcoding [19–22]. Antibody labeling occurs through covalent attachment to metalchelated polymers (MCPs). High binding affinities (10 16 M) between chelating groups on the polymers and metals ensure that there is no release of free metal. A functionalized terminal maleimide group on the MCP forms a covalent thioether bond with a sulfhydryl group on the antibody. The latter is produced by a partial reduction of the antibody which occurs predominantly on the Fc heavy chain [17, 23, 24]. MCPs are water soluble, and each polymer can accommodate up to 30 chelated metals. On average, four to six of these MCPs may be covalently attached to each antibody. Samples once stained with metal isotope-conjugated antibodies can be introduced into the mass cytometer. They are immediately nebulized, and the resulting single-cell droplets are introduced into an inductively coupled argon plasma where each droplet is vaporized, atomized, and ionized [6–8, 25, 26]. The resulting single cell-derived ion cloud passes through a high-pass filter to remove abundant ions derived from the introduced argon gas as well as cellular elements with a mass 89) enter an electrostatic quadrupole doublet (07) for focusing into an ion beam that enters an accelerator and time-of-flight (TOF) mass analyzer (08). Ions are separated based on mass:charge ratio, lighter ions reaching the detector (09) first. Ion counts are quantified, integrated for each cell, producing data in FCS format for downstream analysis (10)

well some of the sample processing steps (to be pointed out later in this chapter). While the overall design and implementation of a mass cytometry study for ovarian tumors is lengthy, convenient points at which the protocol can be halted, and samples stored are shown in Fig. 4. With appropriate modifications, these protocols for mass cytometry are applicable to single-cell studies of other solid tumors and tissues.

2

Materials All studies with human tumor tissue require IRB approval. The time-of-flight readout for mass cytometry is very sensitive to contaminants. Reagents of the highest analytical grade must always be used for sample processing. To keep metal contamination to a minimum, it is strongly recommended to use sterile plasticware and new glassware that has never been detergent washed particularly as many laboratory soaps contain high levels of barium. Double-filtered, distilled, deionized water for reagent preparation is essential to remove metal contaminants such as barium, lead, and mercury. To emphasize this, in sections below, this is also referred to as “CyTOF water.” Additionally, we make phosphate-buffered

Mass Cytometry of Ovarian Solid Tumors

63

Fig. 3 Workflow for mass cytometry study of ovarian tumors. Freshly resected, newly diagnosed ovarian tumors (01) were disaggregated into single-cell suspensions (02), stained with metal tagged antibodies against surface and intracellular epitopes (03) chosen to characterize both ovarian tumor cells and infiltrating immune cells, before introduction into the mass cytometer (04). Single-cell data (05) generated as FCS files was used for downstream analysis and identified previously unrecognized minority cell types (06). In the example shown, the identified cell type was correlated with early relapse (07) [3]

saline (PBS) in house. Our 10 stocks are diluted to 1 and filtered with 0.22 μm filters. At the start of each experiment, PBS should be filtered again. Storage of samples and reagents at temperatures indicated is critical. Before storage, disaggregated tumor cells are fixed rather than prepared for viable freezing. In our experience, fixation provides cleaner mass cytometry data. Our protocol has been successfully used for ovarian tumor masses resected from several anatomical sites.

64

Veronica D. Gonzalez et al.

Fig. 4 Stages at which the mass cytometry protocol can be halted for sample storage. Many reagents for mass cytometry studies can be prepared in advance and safely stored under specified conditions. Shown are the main stages of the protocol with time required for their completion. They can be performed up to and including the day of the mass cytometer run. Times will be dependent on size of tumor to be dissociated, number of antibodies to be conjugated, and number of samples to be barcoded and stained with antibody cocktails. As mentioned in the text, run time on the mass cytometer is based on the cell flow rate and needs to be optimized to maximize the resolution of intact single-cell events. For our ovarian studies, taking into account the large size of tumor cells, this was often reduced to 300–500 cells/s compared with 500 cells/s used for immune cell studies 2.1 Sample Collection

1. “Tumor transport medium”: Dulbecco’s Modified Eagles medium (DMEM)/F12 (1:1) high glucose, supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine, 1% MEM vitamins, 3% penicillin-streptomycin (PS), 0.6% gentamicin, 5 μg/mL transferrin, 12.5 μg/mL fetuin, 20 μg/mL insulin. 2. 50 mL Falcon tubes. 3. Template to record times and tumor characteristics (Table 1).

2.2 Tumor Dissociation into Single Cells

1. gentleMACS Dissociator (Miltenyi Biotech). 2. Tumor dissociation kit (Miltenyi). 3. gentleMACS C-tubes (Miltenyi).

Mass Cytometry of Ovarian Solid Tumors

65

Table 1 Template for recording parameters during tumor collection Sample ID

Date

Time of resection

Time received in laboratory

Appearance

Photograph: Yes/No

Weight

Time sample processing began

Viability (trypan blue after dissociation)

Number fixed cells per cryovial

Number frozen vials

Comments

Fig. 5 Disaggregated cells from freshly resected solid HGSC tumors. Images taken from two dissociated tumors showing a typical pattern of cells. Tumor, immune, and clumped cells are indicated

4. Microscope—used to examine extent of dissociation (Fig. 5). 5. Scissors, razor, forceps. 6. Vortex Mixer. 7. Red Blood Cell Lysis Buffer (Miltenyi Biotec). 8. Cell strainer (70 μm pore size). 9. Unsupplemented RPMI-1640 medium (no L-glutamine). RPMI-1640 medium will be referred to as “RPMI-1640.” 10. Complete RPMI-1640 supplemented with 10% FCS, 2 mM Lglutamine, and 1% penicillin-streptomycin (PS). 2.3 Cell Counting and Viability After Tumor Disaggregation

1. Hemocytometer. 2. Cell culture inverted microscope. 3. 0.4% trypan blue dissolved in sterile-filtered water.

66

Veronica D. Gonzalez et al.

2.4 Viability and Live–Dead Staining

1. There are three options from which the researcher can choose to determine live–dead staining by mass cytometry: in-house prepared cisplatin (reagents 2–5), commercially available cisplatin (reagent 6), and commercially available rhodium (reagent 7). Reagents 8 and 9 are required for all three options (see Note 1). 2. Cisplatin (Sigma). 3. DMSO ampules (Hybri-Max™, sterile-filtered, BioReagent, 99.7%). 4. For in-house preparations, make aliquots of a 250 mM cisplatin stock solution in DMSO to be stored at 80  C. Dissolve 25 mg cisplatin (Sigma-Aldrich) in 333 μL of DMSO (HybriMax, Sigma Aldrich). CAUTION: Cisplatin is a chemotherapeutic agent and highly toxic. Avoid contact with skin, eyes, and clothing. Fatal if swallowed. 5. For a fresh 50 μM cisplatin 2 working solution add 10 μL cisplatin stock solution to 50 mL unsupplemented RPMI-1640 (1 in 5000 dilution). Final concentration for cell incubation 25 μM. 6. Commercially available Cell-ID Cisplatin (Fluidigm): Natural Abundance Platinum supplied as a 5 mM stock solution in DMSO stored at 80  C (see Note 1). Prepare a fresh 10 μM Cell-ID Cisplatin 2 working solution by adding 50 μL CellID Cisplatin stock solution to 25 mL unsupplemented RPMI1640 (1 in 500 dilution). Final concentration for cell incubation 5 μM. 7. Cell-ID Intercalator 103Rh (Fluidigm) supplied as stocks of 500 or 2000 μM. Final concentration for incubating with cells 1 μM. Note: this reagent has a dual-purpose use for labeling dead cells or for labeling single nucleated cells. Here we use it for labeling dead cells. 8. Unsupplemented RPMI-1640 without L-glutamine. 9. Complete RPMI-1640, 10% fetal calf serum, 2 mM L-glutamine, 1% PS.

2.5 Fixation and Preparation for Freezing

1. Table-top refrigerated swinging bucket centrifuge. 2. 50-mL conical tubes. 3. 1 PBS sterile-filtered low-barium, calcium, and magnesiumfree made in-house with Milli-Q purified water. 4. Cell staining medium (CSM): for 500 mL, use sterile-filtered PBS with 2.5 g protease-free bovine serum albumin (BSA) fraction V (Sigma-Aldrich) and 100 mg sodium azide (0.5%). Store CSM in a Nalgene polypropylene bottle that has never been washed with detergent. Can be stored for up to 6 months at

Mass Cytometry of Ovarian Solid Tumors

67

4  C. CAUTION: Sodium azide is highly toxic. Avoid contact with skin, eyes, and clothing. Fatal if swallowed. 5. 16% Paraformaldehyde (PFA) ampules EM grade (Electron Microscopy Science). PFA should be filtered through a Millex PDVF Syringe filter (0.1 μm) and may be stored up to 2 weeks in the dark. For working solution, prepare 1.6% PFA solution in PBS (using filtered PFA and sterile-filtered low-barium, calcium, and magnesium-free phosphate-buffered saline). CAUTION: PFA is an irritant. Avoid contact with skin and eyes. Avoid inhalation. 2.6 Reagents for 20-Plex Palladium Bar-Coding

1. 96-well PCR plates. 2. 96-well foil covers. 3. 10 mM stock solutions of six palladium isotopes (102, 104, 105, 106, 108, 110) as nitrate salts (Trace Sciences International) in DMSO. 4. DMSO ampules (Hybri-Max™, sterile-filtered, BioReagent, 99.7%). 5. Barcode template (Fig. 6).

2.7 Sample Thawing and Barcoding

1. Table-top 96-well format refrigerated swinging bucket centrifuge. 2. Multi-channel pipettors for 10, 200, and 1000 μL (8-channel). 3. Vortex Mixer. 4. 96-well aspirator. 5. Aluminum 96-well block. 6. Cluster tube system composed of 96 1.2-mL polypropylene tubes arranged in a rack with microtitration plate format. Rack can be reused. 7. Cluster tubes, 1.2 mL polypropylene (loose). 8. FACS tubes, 5 mL. 9. 96-well deep well microplates polypropylene 2.0 mL nonsterile. 10. 100 palladium barcode plates. Each Pd isotope is at 100 μM. 11. Cells should be barcoded to the level of ~500 counts per single cell. It is therefore recommended to titrate the barcode. Prepare two FACS or cluster tubes with a mixture of three palladium isotopes each at 100 μM in DMSO. Tube 1 contains Pd 102, 103, 105 and Tube 2 contains Pd 106, 108, 110. 12. For commercially available barcoding, Cell-ID 20-Plex palladium barcoding kit (Fluidigm). 13. 1 PBS—sterile-filtered low-barium, calcium, and magnesium-free made in-house with distilled, deionized water.

68

Veronica D. Gonzalez et al.

Fig. 6 Sample barcoding with palladium isotopes for mass cytometry. Cells from a given sample are identified by addition of a unique combination of three of six palladium isotopes (“six-choose-three”). Up to 20 samples can be barcoded for subsequent sample pooling in this doublet-free scheme [14, 15]. (a) Individual palladium isotopes (3000 stock concentration) pipetted into the central grid within a 96-well PCR plate in the configuration shown. See text for subsequent pipetting steps. (b) The 20 possible unique triplet barcode combinations. (c) Sample layout after barcode assignment, before sample pooling. Blue shading denotes the presence of isotope and white denotes solvent (DMSO) (a, b)

14. Saponin (Sigma-Aldrich). Make a stock saponin solution of 2% in PBS and store at 4  C for up to 6 months. Working solution is 0.02% saponin made from a 1:10 dilution of the stock into PBS. 15. Cell staining medium (CSM). 16. Anchor samples are present on every barcode plate to control for batch effects. For our ovarian tumor cell study, three HGSC ovarian cell lines were included on every plate and were prepared as described in Subheadings 3.4 and 3.5. 2.8 Antibody Conjugation with Metal-Chelated Polymers

1. Many metal-conjugated antibodies are commercially available from Fluidigm. Studies (such as our ovarian study) often require antibodies for which a metal antibody conjugate does not exist. To give the researcher maximum flexibility, kits and protocols are available for in-house antibody conjugations.

Mass Cytometry of Ovarian Solid Tumors

69

2. Purified carrier-free antibodies (100–500 μg). 3. Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) solution, 0.5 M, pH 7.0 (Millipore Sigma or ThermoFisher). 4. NanoDrop spectrophotometer concentration.

for

measuring

antibody

5. Digital-control water bath at 37  C. 6. Table-top Microfuge. 7. Vortex Mixer. 8. Centrifugal Filter Unit: 3 kDa Amicon Ultra 0.5 mL V bottom. 9. Centrifugal Filter Unit: 50 kDa Amicon Ultra 0.5 mL V bottom. 10. Eppendorf Protein LoBind Tubes, volume 1.5 mL. 11. Pipettors, and appropriate aerosol barrier filter tips. 12. 1.5-mL microcentrifuge tubes with socket screwcap. 13. Maxpar X8 multi-metal antibody conjugation kit (Fluidigm) comprising MCP X8 polymer and a choice of up to ten metals per kit. Buffers included: R-Buffer, C-Buffer, L-Buffer, W-Buffer. 14. Trivalent Metal Lanthanide Solutions (50–100 mM stocks in L-Buffer). 15. Antibody Stabilizer PBS (Candor Biosciences). Add 0.2% sodium azide and store at 4  C. 2.9 Antibody Titrations

1. Table-top refrigerated swinging bucket centrifuge. 2. Shaker. 3. 1.2-mL polypropylene cluster tubes. 4. Antibodies to be titrated. 5. Human TruStain FcX (Fc receptor blocking solution, Biolegend) used when titrating antibodies against extracellular epitopes. 6. 1 PBS—sterile-filtered low-barium, calcium, and magnesium-free made in-house with Milli-Q purified water. 7. Cell staining medium (CSM). 8. 16% Paraformaldehyde (PFA) ampules EM grade (Electron Microscopy Science). 9. Ice-cold methanol. CAUTION: Methanol is flammable. Keep away from heat and avoid contact with skin and eyes and inhalation. 10. Cell-ID Intercalator-Iridium solution, 500 μM (Fluidigm) to identify nucleated cells. For working solution, prepare 1.6% PFA solution in PBS (using filtered PFA and sterile-filtered

70

Veronica D. Gonzalez et al.

low-barium, calcium, and magnesium-free phosphate-buffered saline). Dilute iridium solution 1:5000 into PFA solution to a final concentration of 0.1 μM prepared fresh for each experiment. Individual samples are incubated with 1 mL of the PFA-iridium solution and combined barcoded samples are incubated with 2 mL of this solution. 11. Generic template to record key parameters for titration, antibody clone, vendor, date of conjugation, metal, yield, concentration, other. 2.10 Antibody Staining of Individual or Bar-Coded Samples

1. Shaker. 2. Metal conjugated antibodies—prepare separate mixes of antibodies against surface proteins and intracellular proteins. Amount of each antibody predetermined from titrations as described above. 3. Cell staining medium (CSM). 4. Human TruStain FcX (Fc receptor blocking solution, Biolegend). 5. Ice-cold methanol. 6. Iridium intercalator working solution (see Subheading 2.8, item 10).

2.11 Sample Preparation for Loading into Mass Cytometer

1. Mass cytometer; CyTOF 2, or CyTOF 3 (Helios) (Fluidigm). 2. 5-mL Falcon round-bottom polystyrene test tubes with cell strainer snap cap. 3. EQ four element calibration beads with metal isotopes 140/142 Ce, 151/153Eu, 165Ho, and 175/176Lu (Fluidigm). 4. Cell staining medium (CSM). 5. CyTOF water.

3

Methods

3.1 Sample Collection

1. Ensure use of tumor samples for study is IRB-approved. 2. Minimize time between tissue resection and the start of tissue processing. Within 3–5 h is acceptable. Tissue should be fully immersed in “Transport Medium” in a container such as a 50-mL Falcon tube at 4–8  C. 3. A number of variables, such as size of tumor piece to be processed and overall familiarity with the protocol will impact processing time but allow ~6 h (Fig. 4). 4. Use template to record tumor characteristics and note key times from resection to freezing cells (Table 1) (see Note 2).

Mass Cytometry of Ovarian Solid Tumors

3.2 Tumor Dissociation into Single Cells

71

1. Photograph the tumor for macroscopic appearance. 2. Follow our modified Miltenyi protocol for soft tumor dissociation using the gentleMACS dissociator and enzyme dissociation kit for single-cell dissociation (see Note 3). 3. Prepare enzyme mix: 4.7 mL of DMEM, 200 μL of Solution 1, 100 μL of Solution 2, and 25 μL of Solution 3 into a gentleMACS C-Tube. 4. Remove fat, fibrous, and necrotic areas from the tumor sample. Cut the tumor (“mince”) into small pieces of 2–4 mm with scissors and/or razor using forceps to hold tissue. 5. Transfer the tissue pieces into the gentleMACS C-Tube containing the enzyme mix. 6. Tightly close C-Tube and attach it upside down onto the sleeve of the gentleMACS Dissociator (see Note 4). Ensure that the sample material is within the area of the rotor/stator. 7. Run the gentleMACS Program h_tumor_01. 8. After termination of the program, detach C-Tube from the gentleMACS Dissociator. 9. Incubate sample for 30 min at 37  C with continuous rotation using the MACSmix Tube Rotator. 10. Attach C-Tube upside down onto the sleeve of the gentleMACS Dissociator. 11. Ensure that the sample material is within the area of the rotor/ stator. 12. Run the gentleMACS Program h_tumor_02. 13. After termination of the program, detach C-Tube from the gentleMACS Dissociator. 14. Incubate sample for 30 min at 37  C under continuous rotation using the MACSmix Tube Rotator. 15. Attach C-Tube upside down onto the sleeve of the gentleMACS Dissociator. Ensure that the sample material is within the area of the rotor/stator. 16. Run the gentleMACS Program h_tumor_03. 17. Optional: If cells are stuck to walls of container and/or near the cap, perform a short centrifugation step (300  g for 3 min) to collect the sample material at the bottom of the tube. 18. Examine a small sample aliquot under the microscope to determine extent of dissociation. If dissociation is incomplete, recap tube and repeat program. Alternatively, there may be tumor chunks mixed with single cells, in which case the single cells could be aliquoted into a separate tube and only the chunks subjected to a repeat process.

72

Veronica D. Gonzalez et al.

19. Resuspend sample in 5 mL RPMI-1640 and apply the cell suspension to an appropriate cell strainer (70 μm) placed on a 50-mL tube. 20. Dissociated tissue can be removed from the closed C-Tube by pipetting through the septum-sealed opening in the center of the cap of the C Tube. Use 1000 μL pipette tips. 21. Wash cell strainer with 20 mL of complete RPMI-1640. 22. Centrifuge cell suspension at 300  g for 7 min. Aspirate supernatant completely. 23. Determine, by eye, if there is red blood cell contamination (dark red). To remove erythrocytes or dead cells, use Red Blood Cell Lysis Solution or perform a density gradient centrifugation step. For red blood cell lysis: (1) dilute one volume of cell suspension with 10 volumes of 1 Red Blood Cell Lysis Solution. (2) Vortex for 5 s and incubate for 2 min at room temperature. (3) Centrifuge at 300  g for 5 min (5 mL tubes) or 10 min (15 or 50 mL tubes) at room temperature. (4) Aspirate supernatant completely. (5). Resuspend the cell pellet in a volume of RPMI-1640. Aspirate supernatant completely. 24. Depending on size of cell pellet (with or without red blood cell lysis step) add 1–5 mL RPMI-1640 and count cells as described in Subheading 3.3. Desired concentration is ~5  106 cells/ mL. 3.3 Cell Counts and Viability After Tumor Disaggregation

1. Using hemocytometer, count cells and determine viability with trypan blue (see Note 5). 2. Resuspend cells from Subheading 3.2, step 23 in RPMI-1640. The presence of serum leads to high background as trypan blue binds strongly to serum proteins. Be sure to use unsupplemented medium. Cells from a dissociated tumor may be too concentrated and therefore require a two-step dilution. Usually, a concentration of up to 5  106 cells/mL is optimal. 3. Transfer 10 μL of cell suspension to microfuge tube, test tube or well of 96-well plate. 4. Add 10 μL 0.4% trypan blue to the tube/well and mix thoroughly. If larger volume of cells is used, add equal volume of trypan blue. 5. With the coverslip placed over the hemocytometer, fill each chamber with the trypan blue - cell suspension using a pipettor or Pasteur pipette tip. 6. Count cells within 3–5 min of mixing with trypan blue, as longer incubation periods will lead to cell death and reduced viability counts.

Mass Cytometry of Ovarian Solid Tumors

73

7. Count cells within the 1 mm center square and within the four 1 mm corner squares in chamber 1 of the hemocytometer and repeat for chamber 2. Count “blue” dead cells. Clean hemocytometer and repeat. 8. Calculate total number of cells from the formula (average cell count per square  dilution factor  104). Determine cell viability as the percentage of trypan blue positive cells out of the total number of cells counted. 9. If cell number 5  106 cells, resuspend in 5 mL. 10. Record cell number and viability. From our experience samples with viability 2 years without any, or only minimal, changes to epitope integrity. 3.6 Preparing Palladium Barcoding Plates

1. Barcoding individual samples and then combining and processing as one sample before antibody staining minimizes cell loss, variability arising from pipetting errors, variations in cell number between individual samples, sample carry-over and machine performance with the added advantages of minimizing cell-cell doublets and reducing reagent consumption [14, 15, 29, 30]. 2. The barcode scheme we developed labels each sample with three out of six palladium isotopes in a doublet-free scheme that allows barcoding of up to 20 samples per plate (see Note 7) (Fig. 6). 3. The in-house protocol provided differs from the published protocol [14, 15] in that it uses non-chelated palladium nitrate salts (Bjornson et al., in preparation).

Mass Cytometry of Ovarian Solid Tumors

75

4. To prepare barcoding plates in-house, thaw 10 mM stock solutions of each of the six palladium isotopes in DMSO (see Note 8). Add 198 μL of each stock into 462 μL DMSO to produce 660 μL of a 3 mM solution for each isotope (3000). 5. Pipette 60 μL of each isotope into a well, according to the barcode template, of the 96-well plate for a final volume of 180 μL per well for three isotopes. This creates a 1000 combination of three palladium barcodes per well. Each isotope is at a concentration of 1 mM. 6. Mix barcodes well by pipetting 10–20 times and avoid bubbles. 7. Use this plate to create more 1000 palladium barcode plates by pipetting 18 μL aliquots into a desired number of 96-well PCR plates. 8. Seal plates with foil adhesive and store at 80  C with desiccant to protect from moisture. Storage can be for at least a year. 9. To prepare plates with 100 working concentrations of barcode combinations, rapidly thaw one 1000 plate. 10. Using a multi-pipettor, add 162 μL DMSO to each well containing 18 μL of barcode (1:10 dilution to give 100 μM concentration of each bar code). Mix well, but careful to avoid bubbles, by pipetting 10–20 times with a P200 multi-pipettor set at 120 μL. 11. Prepare a batch of 100 working palladium barcode plates by aliquoting 12 μL of barcode, according to the template, into a desired number of 96-well PCR plates. 12. Store plates as in step 8. 3.7 Sample Thawing and Barcoding

1. When analyzing samples across multiple barcoded plates, it is critical to correct for any batch effects [31]. Therefore, we routinely use cisplatin or rhodium-treated fixed frozen “anchor” samples, prepared in large amounts according to steps in Subheadings 3.4 and 3.5, to include on every plate. For ovarian tumor studies, we used HGSC cell lines as anchor samples, while for characterizing tumor infiltrating immune cells, peripheral blood mononuclear cells served this role. Clearly the choice of “anchor” samples is study dependent. 2. We routinely barcode PFA-fixed cells using a “pre-permeabilization” protocol [14]. Rather than introducing barcodes into cells after methanol permeabilization, which can destroy subgroups of surface and intracellular epitopes, we perform transient permeabilization using a low concentration of saponin (0.02% final concentration). 3. The six-choose-three doublet-free barcoding scheme allows up to 20 samples per barcode plate (see Note 7). For >20 samples, barcode plates should be processed sequentially as the timings are critical especially adding the diluted barcode to cells.

76

Veronica D. Gonzalez et al.

4. The protocols provided for barcoding samples have been optimized for one to three million cells (PBMCs, cell lines and tumor cells) per sample in 1 mL PBS-0.02% saponin such that each barcode labels a single cell with ~500 counts per barcode. 5. Ovarian tumors have variable proportions of cells with a range of sizes (immune ~7 15 μm and tumor 20 30 μm) as well as variable amounts of cellular debris, all of which can affect the uptake of palladium barcodes. In addition, cell number and the method of cell fixation also affect uptake of barcodes. Therefore, it is advisable to titrate the barcode concentration for the cell number and conditions to be used in the main study. The optimal concentration is one corresponding to 500 counts of barcode per single cell. 6. To prepare cells for titration or main study, thaw fixed-frozen cells at room temperature (takes ~5–10 min) (Fig. 4). 7. Count cells and adjust volume with CSM for aliquoting 1–3  106 cells/1 mL into a labeled cluster tube on ice. 8. Wash once with 1 mL ice-cold CSM by centrifuging at 600  g for 5 min at 4  C. 9. Wash once with 1 mL ice-cold PBS by centrifuging at 600  g for 5 min at 4  C. 10. Wash once with 1 mL ice-cold PBS-0.02% saponin by centrifuging at 600  g for 5 min at 4  C. 11. Resuspend cells in ~60 μL ice-cold PBS-0.02% saponin. 12. Cells at this step can be used to titrate the barcodes or proceed to barcode samples in main study. 13. Perform a “coarse” titration using the 100x barcodes (100 μM) to ensure that barcoded cells will register ~500 counts per cell. If needed, adjustments to the concentration can be made for the main study when the 100 barcode plate is diluted. 14. Perform titration with two different mixtures of three palladium isotope nitrate salts. Working quickly, perform dilutions (1:50, 1:100 and 1:200) of each palladium mixture into ice cold PBS-0.02% saponin. Use a separate cluster tube for each dilution (six tubes total). Incubate cells (step 11) with palladium isotope mixtures at room temperature for 20 min. Wash cells twice with CSM. Combine cells for each dilution of the palladium isotope mixtures, so that each tube will have all six barcodes (three tubes total). This checks that a barcode is strongly bound to its assigned cell aliquot and does not fall off and contaminate another cell aliquot. Perform titrations in triplicate. Process for introduction into the mass cytometer (Subheading 3.12, steps 2–6).

Mass Cytometry of Ovarian Solid Tumors

77

15. For the main study, palladium barcoding can be performed either with in-house or commercially available plates from Fluidigm. Overall, the principles are the same. The researcher is referred to Fluidigm’s protocols if they choose to purchase ready-made barcode plates. The steps below are for barcoding with plates prepared in-house (Fig. 6). 16. Thaw barcoding plate with stock solutions of palladium nitrate isotopes (“barcoding reagent”) rapidly at room temperature on an aluminum 96-well block. 17. Use a multichannel pipettor to dilute 10 μL of 100 barcodes (in DMSO) from the first row of the barcoding plate into 1 mL of PBS-0.02% saponin aliquoted into a corresponding row of a 96-well deep-well 2-mL microplate on ice. Pipette up and down 3–4 times to ensure good mixing as the DMSO can freeze in the pipette tips. 18. Immediately transfer 900 μL of the diluted barcoding agent to the first row of PBS-washed cells aliquoted in cluster tubes and mix thoroughly by pipetting up and down. This step should be done 1 row at a time using a multi-channel pipettor. It is absolutely critical that the PBS-0.02% saponin is ice-cold and that this entire procedure is performed in less than 60 s. Warm PBS or taking too long will consistently reduce or eliminate the barcode signal. 19. Repeat steps 17 and 18 for each subsequent row of barcodes. 20. Allow cells mixed with barcodes to incubate at room temperature for 20 min. 21. During this time, Subheading 3.10).

make

up

antibody

cocktail

(see

22. Pellet cells by centrifuging at 600  g in a table-top refrigerated centrifuge for 5 min. Aspirate supernatant with 96-well aspirator. 23. Resuspend cell pellets in 1 mL of CSM with gentle vortexing and centrifuge at 600  g for 5 min. Remove supernatant and wash twice with 1 mL CSM by centrifuging at 600  g for 5 min. After removing supernatant from second wash, leave a residual volume of ~100 μL. 24. Add 50 μL of CSM to each tube and vortex at low speed. 25. Combine cells from cluster tubes into a single 5-mL FACS tube. This is done by using a P200 or P1000 pipettor or a multi-channel pipettor to combine each row with the adjacent row below it and then combining these into the FACS tube. 26. Once all cells have been combined, add 100 μL of CSM to each cluster tube to rinse off any remaining cells. Regardless of the method chosen, do NOT change the pipette tip, as the main

78

Veronica D. Gonzalez et al.

source of cell loss is from cells sticking to tubes and pipette tips. Avoid foaming as this will make combining cells into a single tube more difficult. If for some reason there is too much foaming or too great a volume to combine 20 samples into a FACS tube, perform an intermediate spin at 500  g for 2 min. Combining samples should yield a partially or completely full FACS tube depending on how many samples are combined (250 μL CSM per sample). 27. Count a 10 μL aliquot of cells prior to centrifuging. 28. Pellet combined samples in 5 mL FACS tube by centrifuging at 500  g for 5 min. 29. Resuspend cells in CSM with a volume that depends on the cell counts (step 27). 2–3  106 cells resuspend in 45 μL CSM. 3–6  106 cells resuspend in 90 μL CSM. 15–18  106 cells resuspend in 270 μL CSM. 30. Stain combined barcoded samples with antibody cocktails and process for introduction into mass cytometer according to steps described in Subheadings 3.10 and 3.11. 3.8 Antibody Conjugation with Metal-Chelated Polymers

1. There are two main reasons for performing in-house conjugations with MCPs: no commercial metal conjugate exists for a desired antibody (see Note 9) and/or an antibody is needed on another metal channel (see Note 10). 2. Initially, for a desired epitope, we perform a three-point titration with two to three different clones using fluorescencebased flow cytometry with appropriate positive and negative cell line controls. If an antibody looks promising, we proceed with conjugation to metal-chelated polymer (MCP). 3. This protocol is optimized to label 100 μg antibody at a starting concentration of 1 mg/mL (protocol from Fluidigm with slight modifications). 4. We use the X8 polymer that was developed to chelate lanthanides and is the MCP used most often (see Note 11). The X8 polymer is used to conjugate IgG isotypes and affinity purified polyclonal antibodies. It will not work for IgM antibodies. 5. It is critical that antibodies used for conjugation are “carrierfree” (no BSA, hydrolyzed protein, gelatin, or other) (see Note 12). Sodium azide (0.2%) used routinely does not pose any problems. It is always best to check with the vendor regarding the components of each antibody solution as in some cases a custom “carrier-free” order needs to be placed.

Mass Cytometry of Ovarian Solid Tumors

79

6. Use aerosol barrier filter tips for all steps. Many steps are carried out in Amicon filter units. Care should be taken not to allow pipette tips to touch the delicate membrane or the antibody solutions. For the wash steps, use a P1000 pipettor to introduce buffers gently down the side of the filter device. 7. Perform loading of polymer with metal ions (steps 9 and 12) and partial reduction of antibody (steps 10, 11 and 13) in parallel. 8. All centrifugations are carried out at room temperature. Store antibodies at 4  C before and after conjugation. 9. Loading metal-chelating polymer: Retrieve tube with X8 polymer (stored at 20  C), thaw and label with the metal to be chelated. Centrifuge both the polymer and lanthanide metal solution for 10 s in a microfuge to make sure polymer is consolidated in the bottom of the tube. Add 95 μL of L buffer to the polymer tube. Pipette up and down to mix well. Add 5 μL of 50 mM lanthanide metal solution to the polymer tube. Mix well by pipetting before incubating Maxpar polymer tubes for 1 h at room temperature vortexing every 20 min. Proceed to step 12 once the hour is complete. 10. Antibody buffer exchange: In parallel with loading the chelating polymer with metal (step 9), prepare the antibody for partial reduction. Antibodies are supplied either in solution or lyophilized as 100 μg aliquots. Antibodies in solution can be diluted with PBS if needed; lyophilized aliquots are reconstituted in 100 μL PBS using the tube in which the antibody is supplied. Centrifuge antibody at 12,000  g for 5 min to spin out any aggregates. Remove a 2 μL aliquot and measure the absorbance at 280 nm against PBS as a blank and calculate the concentration in a NanoDrop spectrophotometer (set in IgG mode). This measurement both ensures that the amount of antibody received is ~100 μg and that any adjustment can be made for the correct concentration. Pipette antibody solution into a labeled 50 kDa Amicon filter unit. Bring total volume to 400 μL with R buffer and centrifuge 12,000  g for 8 min at room temperature. Discard flow-through. Final volume should be 20 μL before proceeding to next step. 11. Partial antibody reduction with TCEP solution: Freshly prepare 4 mM TCEP by diluting 8 μL of 0.5 M TCEP into 992 μL R-buffer. Add 100 μL of 4 mM TCEP solution to the washed antibody on the 50 kDa filter from step 10. Mix by light tapping. Incubate in a 37  C water bath for exactly 30 min. This time is critical to avoid reduction of all disulfide bonds which would compromise the integrity of the antibody. Proceed to step 14 once the 30 min is completed.

80

Veronica D. Gonzalez et al.

12. Clean-up metal-chelated polymer: After the polymer-metal incubation, add 200 μL of L-buffer. Mix and transfer the polymer-metal solution to a labeled 3 kDa Amicon filter unit and centrifuge at 12,000  g for 25 min at room temperature. Discard flow-through. Add 400 μL of C-buffer to 3 kDa Amicon filter unit with metal-chelated polymer and centrifuge at 12,000  g for 30 min at room temperature. Discard flowthrough. 13. Clean-up partially reduced antibody: After antibody incubation, add 300 μL of C-buffer to the partially reduced antibody in the 50 kDa filter unit. Mix by pipetting or vortexing. Centrifuge at 12,000  g for 8 min at room temperature. Discard flow through. Perform second wash by adding 400 μL of C-buffer to the antibody filter unit and centrifuge at 12,000  g for 8 min at room temperature. 14. Conjugating antibody with metal-chelated polymer: Resuspend metal-chelated polymer in 200 μL of C-buffer. Transfer resuspended metal-loaded polymer to the corresponding antibody in 50 kDa filter unit. Mix briefly by gentle pipetting. Incubate at 37  C in water bath for 90 min. Reaction time can be extended up to 2 h although it should be complete by 60 min. 15. Washing metal conjugated antibody: Add 200 μL of W buffer to the conjugated antibody in the 50 kDa filter unit. Pipette and vortex to mix. 16. Centrifuge at 12,000  g for 8 min at room temperature or until the volume is ~20 μL. Discard flow-through and wash an additional 2 times with 400 μL W buffer. 17. Recovery of metal-conjugated antibody: Add 50 μL W-buffer to the metal-conjugated antibody on the 50 kDa filter unit and pipette to mix and rinse the walls of the column. Invert the column into a new collection tube. Centrifuge at 1000  g for 2 min at room temperature. Add another 50 μL of W buffer to the to the metal-conjugated antibody on the 50 kDa filter unit, pipette to mix and rinse the walls of the column. Invert the column into the same collection tube. Spin at 1000  g for 2 min. 18. Quantification and storage: Measure the absorbance at 280 nm of a 2 μL aliquot of the conjugated antibody against a W-buffer blank on the NanoDrop spectrophotometer (set to IgG mode) and determine the recovery. The yield of a conjugated antibody depends on the clone but yields of 60–70% are expected. 19. Ideally, the antibody should be titrated (see Subheading 3.9) as soon as possible after conjugation in order to select the optimal concentration for storage. Usually, we dilute conjugated

Mass Cytometry of Ovarian Solid Tumors

81

antibodies in antibody stabilizer PBS to a concentration of 0.2, 0.3, or 0.4 mg/mL. These concentrations are high enough such that only 1 or 2 μL will be needed, keeping the volume of combined antibodies low, when they are added to samples in the study. 20. Conjugated antibodies can be stored at 4  C for at least a year. We have stored conjugated antibodies for much longer but always perform a pilot run to make sure there is no loss of signal. 3.9 Antibody Titrations

1. We routinely validate the specificity of both in-house and commercially available MCP-conjugated antibodies. 2. Antibody titrations are performed to determine the optimal antibody concentration that has maximal signal to noise ratio. 3. It is critical to select positive control cells known to express the epitope of interest and negative control cells known not to express the epitope of interest. 4. For our ovarian studies, we used HGSC cell lines (positive controls for tumor cells and negative controls for immune cells) and immune cell lines and peripheral blood mononuclear cells (negative controls for tumor cells and positive controls for immune cells). Cells were treated with cisplatin, fixed and frozen (Subheadings 3.4 and 3.5) and served as reproducible systems for titrating the antibodies used in our studies [2, 3, 32]. 5. Titrations are performed with 1  106 cells/100 μL CSM per 1.2-mL cluster tube for each antibody concentration. 6. When titrating antibodies against extracellular epitopes, add 5 μL Fc receptor blocking solution to cells in 95 μL CSM. This blocking solution is especially important when using immune cells for titration or staining (see Subheadings 3.10 and 3.11). It prevents unwanted staining from Ig Fc binding to Fc receptors present especially on monocytes, and NK cells. 7. We routinely titrate antibodies with a 6-point twofold serial dilution from 8 to 0.25 μg/mL. A “no-antibody” sample serves as a control for background. 8. Titrations are performed with a conjugated antibody before methanol treatment (for extracellular epitopes) or after methanol permeabilization (for intracellular epitopes). Regardless of whether the titration is with an antibody against an extracellular or intracellular epitope, cells are always permeabilized with methanol treatment for two reasons: (1) to determine whether the antibody bound to an extracellular epitope will perform well after methanol and (2) to introduce the DNA intercalator into cells. Antibody staining protocol described in Subheading 3.10.

82

Veronica D. Gonzalez et al.

9. The concentration of each metal-conjugated antibody selected for the cocktails is determined from the maximal signal to noise ratio obtained from titrations (see Note 13). 10. Fill out templates to record results of titrations and all relevant information for future reference, including amount of each antibody to use in the panel per sample, positive and negative controls used for titrations, vendor, clone, species, lot number, other. 3.10 Antibody Staining Individual Samples

1. Before initiating any study using precious clinical samples, it is essential to perform a pilot run, usually with a “throw-away” sample, and with the entire mass cytometry antibody panel. This is to ensure optimal performance and check for potential problems especially channel spillover (see Note 10) or incompatibility of reagents (see Note 14). 2. Prepare separately a 2 surface antibody cocktail and a 2 intracellular antibody cocktail. We perform sequential staining to ensure that the methanol permeabilization step does not destroy surface epitopes. For each cocktail, mix an amount of each metal-conjugated antibody (based on titration data and the desired number of samples to be stained) in CSM. The volume of each cocktail per sample is 50 μL (see steps 6 and 10 below). 3. Aliquot 1–3  106 cells into a 1.2-mL cluster tube and wash twice with 1 mL CSM by centrifuging at 600  g for 5 min at 4  C. 4. Resuspend cell pellet in 45 μL CSM. 5. Add 5 μL Fc receptor blocking solution, mix well and incubate for 10 min at room temperature. 6. Stain by adding 50 μL pre-mixed 2 surface antibody cocktail and incubate for 45 min at room temperature on a shaker at low speed. 7. Wash twice with 1 mL CSM by centrifuging at 600  g for 5 min at 4  C. 8. Permeabilize cells with 1 mL ice-cold 100% methanol for 20 min at 4  C. 9. Wash twice with 1 mL CSM by centrifuging at 600  g for 5 min at 4  C. 10. Add 50 μL CSM and 50 μL 2 pre-mixed intracellular antibody cocktail (including cPARP antibody) (see Note 15) for 1 h at room temperature on a shaker. 11. Wash twice with 1 mL CSM by centrifuging at 600  g for 5 min at 4  C.

Mass Cytometry of Ovarian Solid Tumors

83

12. Add 1 mL freshly prepared iridium DNA intercalator working solution. 13. Incubate for 1 h at room temperature or store at 4  C until ready for mass cytometry run. Our preference is to store no longer than 48 h although other researchers accept storage up to 1 week. 3.11 Staining Combined Barcoded Samples

1. Combined barcoded samples in CSM in a 5 mL FACS tube (from Subheading 3.7, step 24) are processed for staining with antibody cocktails. 2. Prepare 2 surface antibody and intracellular antibody panels separately with total volume dependent on cell number. Examples: 1–3  106 cells, the total volume for staining is 100 μL. 3–6  106 cells, the total volume for staining is 200 μL. 15–18  106 cells, the total volume for staining is 600 μL. 3. Add 5 μL Fc receptor blocking solution/100 μL final cell staining volume, mix well and incubate for 10 min at room temperature. 4. Stain with 2 pre-mixed surface antibody cocktail by incubating for 45 min at room temperature on a shaker. Volume of cocktail to add dependent on cell number. Examples: 1–3  106 cells in 45 μL CSM + 5 μL Fc Block, add 50 μL 2 antibody cocktail. 3–6  106 cells in 90 μL CSM + 10 μL Fc Block, add 100 μL 2 antibody cocktail. 15–18  106 in 270 μL CSM + 30 μL Fc Block, add 300 μL 2 antibody cocktail. 5. Wash twice with 4 mL CSM by centrifuging at 500  g for 5 min at 4  C. 6. Permeabilize cells with 2 mL ice-cold 100% methanol for 20 min at 4  C. 7. Wash twice with 4 mL CSM by centrifuging at 500  g for 5 min at 4  C. 8. Depending on cell number (see step 2), add, 50, 100, or 300 μL CSM to resuspend cell pellet. 9. Add 50, 100, or 300 μL of pre-mixed 2 intracellular antibody cocktail (including cPARP antibody) (see Note 15) and incubate for 1 h at room temperature on a shaker. 10. Wash twice with 4 mL CSM by centrifuging at 500  g for 5 min at 4  C.

84

Veronica D. Gonzalez et al.

11. Add 2 mL freshly prepared iridium DNA intercalator working solution. 12. Incubate for 1 h at room temperature or store at 4  C until ready for mass cytometry. Our preference is to store no longer than 48 h although other researchers accept storage up to 1 week. 3.12 Preparation for Sample Loading into Mass Cytometer After Antibody Staining

1. Wash once with CSM by centrifuging at 600  g for 5 min at 4  C. Use 1 mL for a single sample in a 1.2-mL cluster tube or 4 mL for combined barcoded samples in 5 mL FACS tube. 2. Wash twice with CyTOF water by centrifuging at 600  g for 5 min at 4  C. Use 1 mL for an individual sample or 4 mL for combined barcoded sample washes. 3. An additional wash is advisable to avoid high backgrounds and streaking. 4. Resuspend cell pellet at a concentration of 1–2  106 cells/mL with a 0.1 EQ bead solution diluted in CyTOF water. These are metal-embedded polystyrene normalization bead standards [33]. Normalization beads control for machine sensitivity that can vary between different mass cytometers and over time for an individual mass cytometer. The variability is primarily due to buildup of biological material and variations in plasma ionization over time. 5. Filter cells into Corning Falcon test tubes (5 mL) with cell strainer snap-cap immediately before loading into mass cytometer. This final filtering step helps to remove any cell clumps. 6. Load sample into mass cytometer with flow rate of 400–500 cells/s (see Note 16).

3.13 Initial Analysis steps

1. Data normalization from bead standards: Normalize data from either in-house or commercial EQ beads using a mathematical correction with step-by-step instructions that can be accessed with the links below. In the first two links, the data are normalized to the median signals from beads within the same experiment whereas the Fluidgm implementation normalizes to a universal standard: (1) https://github.com/nolanlab/ bead-normalization/wiki/Normalizing-FCS-Files—MATLAB implementation; (2) https://github.com/ParkerICI/pre messa—R implementation. (3) Fluidigm instrument software with normalization to a universal standard. 2. Debarcoding: After data normalization samples are debarcoded from one FCS (flow cytometry standard) file incorporating all the combined samples resulting in individual barcoded files for downstream analysis. A general overview is provided for debarcoding, but for step-by-step details, the reader is referred to the original publication [15] and the following

Mass Cytometry of Ovarian Solid Tumors

85

links that can be selected for the data analysis workplan: (1) https://github.com/nolanlab/single-cell-debarcoder— MATLAB implementation; (2) https://github.com/ ParkerICI/premessa—R implementation; (3) Fluidigm CyTOF software v6.7.1014. According to the barcoding scheme, each cell event will be positive for three palladium isotopes and negative for the other three. The debarcoding software assigns a barcode to a cell event by calculating the distance between three positive and three negative palladium channels, where the distance is always set between palladium isotopes with the third and fourth highest intensities. The software first provides users with a panel to visualize the barcoded events. Clear separations between positive and negative channels allow barcode assignment to be made to a cell with high confidence. However, there will be a variable number of events where the separation is minimal and intensity low, due to debris and separation minimal and intensity high due to cell doublets and or aggregates. Additionally, there may be events with intermediate distances making it difficult to assign them with a barcode. This was particularly true for the ovarian tumor cell data (and applicable to other dissociated tumors) and can be due to a wide range of cell sizes (~10 to >50 μm) [3]. To address these issues, the software provides two filters that can be tuned by the user. The first is the barcode separation filter which changes the separation between positive and negative channels while simultaneously showing the effect on event yield. The second Mahalanobis distance filter is for fine tuning of outliers not removed by the barcode separation filter. Debarcoded samples are saved as FCS files and named so that the barcode key is appended to the base file name. 3. Determination of data quality: After data normalization and debarcoding, each sample is gated to remove debris and cell doublets, dead cells, and apoptotic cells. If the data is of high quality, the output of these gates will be a high frequency of viable cells for further analysis (Fig. 7). In the case of our ovarian data, we used subsequent gating schemes to identify tumor cells [3] and tumor infiltrating immune cells [2]. Examples are presented showing the ranges in proportions of cell populations within disaggregated tumors (Fig. 7). These gating steps are performed using software from CellEngine (https://cellengine.com) or Cytobank (https://www.cyto bank.org/). Both the sites have many tools available for downstream analysis. The latter falls outside the scope of this review.

86

Veronica D. Gonzalez et al.

Fig. 7 Gated cell populations from disaggregated ovarian tumors. Plots from three different tumors showing range in percentages of gated populations across tumors. Normalized and debarcoded samples were gated for cells (1) and singlets (2). Dead cells (cisplatin+) (3) and apoptotic cells (cPARP+) (4) were both gated out. Resulting viable cell populations (5) were gated for CD45+ immune cells (encircled) and CD45 tumor cells (rectangle). Cells in the tumor cell compartment were gated from the CD45 population with FAP and CD31 (to remove stromal and angiogenic cells, respectively). Gated tumor and immune compartments are ready for downstream analysis

4

Notes 1. The decision to use in-house versus commercially available cisplatin is one of preference. If cost is an issue, the in-house cisplatin preparation is cheaper. The commercial formulation (5 μM final concentration) calls for a 5-min incubation which may be preferable to the 1-min incubation with the in-house preparation (25 μM) which was optimized as reported in our original publication [16]. Platinum has five naturally occurring stable isotopes, and for the cisplatin protocols, we use the 195 amu channel as a readout for the most abundantly occurring platinum isotope. However, there may be situations in which mono-isotopic cisplatin solutions are required. 194Pt, 195 Pt, 196Pt, and 198Pt cisplatin solutions are available from Fluidigm. As noted in the introduction, purified platinum isotopes may also be used for labeling antibodies in which case monoisotopic platinum or rhodium can be used to measure cell viability. Conditions using a monoisotopic preparation of cisplatin for determining cell viability are slightly different and

Mass Cytometry of Ovarian Solid Tumors

87

pilot titrations are advisable to determine the optimal concentration. 2. An efficient clinical team, with strong support from the surgeon can help ensure timely processing of resected tumor tissue, thus optimizing the number of viable single cells. 3. Ovarian tumors processed as described produce viable suspensions of tumor and immune cells. Stromal cells are largely destroyed, and imaging is the technology of choice for their study (Bedia, Donoso, Gonzalez, Howitt, and Fantl, unpublished). 4. At the time of our ovarian tumor study, we were in possession of a gentleMACS Dissociator. However, the gentleMACS Octo Dissociator can simultaneously process eight samples and includes heaters for 37  C enzyme incubations, enabling a more streamlined process. 5. The presence of dead cells within a sample can be a serious confounding factor for data analysis and interpretation. Recording the number of cells and percentage of dead cells after tumor disaggregation is critical for mass cytometry studies. When preparing samples for mass cytometry, this critical issue is addressed in three ways. The first is to assess cell viability after tumor dissociation and before further processing. For this we use trypan blue to determine the viability of a small aliquot (10 μL) of sample. An automated cell counter will not be able to distinguish trypan blue-positive cells, and for this reason, we prefer to use a hemocytometer. Importantly, a hemocytometer allows visual inspection of a sample to determine the extent of dissociation and whether fine-tuning any steps of the dissociation process can increase the percentage of viable single cells. In our experience, more extensive “mincing” can reduce clumps. Images are shown for two dissociated tumors (Fig. 5). Maximal viability is desirable for mass cytometry studies, but values >60% are acceptable. In our ovarian tumor study median viability was 74% with a range of 62–90%. Samples with poor viability should not be included as they leak cellular contents and compromise data from the remaining cells. Viability of diseased tissue is variable, adding to the challenges of preparing single-cell suspensions from a solid tumor. 6. Most samples will contain a percentage of dead cells at the time of introduction into the mass cytometer, and it is essential to enumerate their frequency after a mass cytometry run and before data analysis. Cisplatin, a platinum-containing chemotherapeutic agent, is used as a viability stain [16]. Cisplatin enters dead cells through their compromised plasma membranes and reacts rapidly with protein functional groups such as R-SH or R-S-CH3 to form platinum-S bonds. The Pt bonds

88

Veronica D. Gonzalez et al.

are resistant to sample fixation and to all subsequent mass cytometry processing steps, making cisplatin a very robust agent for labeling, and detecting dead cells. The welldocumented DNA damaging property of Pt does not compromise the data as it manifests itself after a much longer exposure time measured in days [34]. Although cisplatin is a mixture of five naturally occurring platinum (Pt) isotopes, it is best detected in the 195-channel which corresponds to its most abundant isotope. There may be studies where the use of monoisotopic cisplatin is the viability stain of choice (see Note 1). It is important to note that platinum-based chemotherapy is used for the treatment of many tumors including ovarian cancer. To study tumors treated with neoadjuvant platinum-based therapy, we use an alternative viability stain, Cell-ID Intercalator-103Rh [17]. It can be detected in the 103-channel of the mass cytometer. 7. The six palladium isotopes have atomic masses between 102 and 110 amu, and do not overlap with lanthanide readouts. They are well-suited for their use as barcoding agents (Fig. 1). The “six-choose-three” doublet filtering scheme allows up to 20 samples to each be uniquely barcoded with three palladium isotopes [15]. Cell doublet readouts can arise through incomplete enzymatic/mechanical dissociation and also when single cells pass coincidently through the mass analyzer. Any event that produces a signal of 4 palladium isotopes is registered as a cell doublet and is removed from the data file [15]. 8. The main advantage of preparing barcode plates in-house is the cost effectiveness of producing large batches of plates. Optimal preparation of barcode plates is with a robot. Critically, manual preparation requires reproducible accurate pipetting with a multichannel pipettor. An overview of the main steps will be provided here, but the reader is referred to Zunder et al. for additional details [15]. Palladium barcode plates (at 1000 to conserve freezer space) are prepared according to a template (Fig. 6). They are configured across wells C4-F9 of 96-well PCR plates, resulting in a 5  4 grid and then sealed to protect from moisture during long-term storage (at least 3 years) at 80  C. Plates with barcode concentrations 100 are prepared from the 1000 plates. Concentrations of palladium isotopes for barcoding are optimized from titrations [15]. 9. As the number of mass cytometry users across the international research community continues to increase, so has the availability of commercially produced metal-conjugated antibodies. Nevertheless, researchers still need the option of designing specialized antibody panels to address their study needs. This is especially true for non-immune cell studies. Selecting a panel

Mass Cytometry of Ovarian Solid Tumors

89

Fig. 8 Exemplary antibody panel for interrogating ovarian tumor cells by mass cytometry. Antibody panel used for our ovarian tumor study for in-depth interrogation of single cells. Up to 45 tagged antibodies can be used simultaneously. Antibodies against tumor, immune, intra- and extracellular epitopes are represented. Selection of antibodies for individual panels will vary from study to study (Reproduced from [3] with permission)

of proteins that address the biology of a tissue under study is key and will determine antibody selection and the requirement for in-house conjugation when no commercially available metal conjugates exist. The antibody panel used for our ovarian study is shown in Fig. 8. When selecting an antibody for in-house conjugation, a general rule-of-thumb is that antibodies that perform well in traditional fluorescence-based flow cytometry will also perform well in mass cytometry. If a desired antibody has not been specified by the vendor for use in flow cytometry, it could nevertheless be suitable for mass cytometry, and the lack of mention may merely reflect that its performance in flow cytometry has not been tested. If there are limited choices for a desired antibody, it is always worth evaluating its suitability for mass cytometry.

90

Veronica D. Gonzalez et al.

10. Several considerations need to be taken into account for metal assignment to a given antibody. Metal isotopes used for mass cytometry are enriched up to 96–99%. Nevertheless, the sensitivity of the mass analyzer can result in signal spillovers into other isotope channels. These are small (1 L is common), it may be impractical to save all of the fluid. Fluid volumes from patients treated with neoadjuvant chemotherapy may be much smaller. 3. Thaw aliquots of ascites on ice. When thawing a large aliquot, aliquot any remaining ascites into 1 mL aliquots prior to refreezing, so that most samples are frozen/thawed twice. 4. Do not aliquot more than 45 mL of ascites into the tube as it will expand once frozen and may crack the tube, resulting in a leak. 5. The coding system will be set by the IRB regulations of each institution and may not allow the use of identifying information, including date of collection. Careful record keeping is essential to utilize the sample across multiple experiments for comparative analysis or to obtain patient outcome data, where allowed. 6. Quantification of the cytokines and growth factors in the ascites is necessary to fully characterize the ovarian cancer tumor microenvironment. Samples collected by this protocol are appropriate for such analysis. Enzyme-linked immunosorbent assays (ELISAs) provide quantitative and precise measurement of soluble factors present in ascites [4]. Additionally, multiplex immunoassays have been utilized to measure cytokine levels in

102

Hannah M. Micek et al.

ascites and present a higher-throughput option for characterization [13]. There are commercially available multiplex immunoassays and ELISAs specific to a myriad of relevant cytokines, chemokines, and growth factors that may be used to profile the composition of ascites. As it has been shown that there is significant variability in the levels of soluble factors between patients, it is recommended that multiple patients are profiled to obtain a representative picture [13]. Furthermore, it is important to consider patient disease stage in analyses, as increased levels of factors such as IL-8 and IL-6 have been linked to increased tumorigenicity [14, 15]. 7. Incorporation of ascites fluid into cell culture models is a facile approach to recapitulate the soluble tumor microenvironment in vitro. Ascites can be added to tumor cell cultures or co-cultures of tumor cells and other relevant cell types such as macrophages or mesothelial cells to study the impact of soluble factors in the ascites on outcomes such as adhesion [4]. A 10% (v/v) dilution of ascites in the culture medium is recommended for in vitro experiments to maintain proper cell media nutrients for viability. 8. For example, if you plan to freeze ascites cells, perform aggregate size distribution analysis, embed aggregates for histological analysis, and analyze cell populations with flow cytometry, there are four applications and the cell pellets should be combined and reconstituted in 48 mL of HI FBS. Once the cells are resuspended and well mixed, transfer 12 mL of this cell suspension to four different conical tubes. Each conical tube will be used for one of the downstream cellular applications (Fig. 1c). 9. When counting cells, avoid counting platelets that may be present. These cell fragments are non-nucleated and much smaller than the other cell types in the ascites. 10. The cellular fraction in ascites is complex and can include multiple cell types (e.g., tumor cells, immune cells, mesothelial cells) as well as both single cells and aggregates of cells. Characterization of the composition of the cellular fraction may provide a better understanding of the metastatic process or provide prognostic markers. Flow cytometry of common cell types in the ascites including tumor cells (CD326+), immune cells (CD45+), and mesothelial cells (MSLN+) can be performed to profile cell types in the ascites (Fig. 4). 11. Depending on aggregate size, a recommended objective to use is 5. If you take multiple images in one well, ensure that the capture area of the image does not overlap, or some aggregates may be double counted. While images can be captured with brightfield, phase optics may improve automated image analysis.

Ascites

103

Fig. 4 Representative flow cytometry plots of cells isolated from ascites. (a) Selection of cell population from debris based on FSC-A vs. SSC-A. (b) Selection of single cells based on FSC-A vs. FSC-H. (c) Selection of live cells using a fixed cell-compatible viability dye. (d) CD45 (immune cells) and CD326 (tumor cells) expression of cells. (e) CD326 and MSLN (mesothelial cells) expression of cells

12. With the assumption that the aggregates are spherical, this size range includes aggregates with a diameter of 40 μm and greater. Forty micrometer diameter aggregates should be the smallest aggregates that will be trapped with the 40-μm cell strainer and so define the lower bound of the size analysis. If you would like to determine sizes of individual aggregates, this can be approximated from: Area ¼ π ðradiusÞ2 Note that scaling for images can be set in Analyze: Set Scale in FIJI. The specific scaling factor to convert micron to pixels is dependent on the objective and binning used during image capture. 13. A macro script can be created in ImageJ to perform all the analysis steps with one click. 14. It is useful to confirm that aggregates were embedded in the agarose using a light microscope. Hold the microcentrifuge tube on the stage and adjust focus until cells come into focus.

104

Hannah M. Micek et al.

References 1. Ahmed N, Stenvers KL (2013) Getting to know ovarian cancer ascites: opportunities for targeted therapy-based translational research. Front Oncol 3:256. https://doi.org/10. 3389/fonc.2013.00256 2. Szender JB, Emmons T, Belliotti S, Dickson D, Khan A, Morrell K, Khan A, Singel KL, Mayor PC, Moysich KB, Odunsi K, Segal BH, Eng KH (2017) Impact of ascites volume on clinical outcomes in ovarian cancer: a cohort study. Gynecol Oncol 146(3):491–497. https://doi. org/10.1016/j.ygyno.2017.06.008 3. Kim S, Gwak H, Kim HS, Kim B, Dhanasekaran DN, Song YS (2016) Malignant ascites enhances migratory and invasive properties of ovarian cancer cells with membrane bound IL-6R in vitro. Oncotarget 7 (50):83148–83159. https://doi.org/10. 18632/oncotarget.13074 4. Carroll MJ, Fogg KC, Patel HA, Krause HB, Mancha AS, Patankar MS, Weisman PS, Barroilhet L, Kreeger PK (2018) Alternatively-activated macrophages upregulate mesothelial expression of P-selectin to enhance adhesion of ovarian cancer cells. Cancer Res 78(13):3560–3573. https://doi.org/ 10.1158/0008-5472.CAN-17-3341 5. Soritau O, Tomuleasa CI, Pall E, Virag P, Fischer-Fodor E, Foris V, Barbos O, Tatomir C, Kacso G, Irimie A (2010) Enhanced chemoresistance and tumor sphere formation as a laboratory model for peritoneal micrometastasis in epithelial ovarian cancer. Romanian J Morphol Embryol 51(2):259–264 6. Frankel A, Buckman R, Kerbel RS (1997) Abrogation of taxol-induced G2-M arrest and apoptosis in human ovarian cancer cells grown as multicellular tumor spheroids. Cancer Res 57(12):2388–2393 7. Al Habyan S, Kalos C, Szymborski J, McCaffrey L (2018) Multicellular detachment generates metastatic spheroids during intra-abdominal dissemination in epithelial ovarian cancer. Oncogene 37(37):5127–5135. https://doi.org/10. 1038/s41388-018-0317-x 8. Raghavan S, Ward MR, Rowley KR, Wold RM, Takayama S, Buckanovich RJ, Mehta G (2015) Formation of stable small cell number three-

dimensional ovarian cancer spheroids using hanging drop arrays for preclinical drug sensitivity assays. Gynecol Oncol 138(1):181–189. https://doi.org/10.1016/j.ygyno.2015.04. 014 9. Gencoglu MF, Barney LE, Hall CL, Brooks EA, Schwartz AD, Corbett DC, Stevens KR, Peyton SR (2018) Comparative study of multicellular tumor spheroid formation methods and implications for drug screening. ACS Biomater Sci Eng 4(2):410–420. https://doi.org/ 10.1021/acsbiomaterials.7b00069 10. Latifi A, Luwor RB, Bilandzic M, Nazaretian S, Stenvers K, Pyman J, Zhu H, Thompson EW, Quinn MA, Findlay JK, Ahmed N (2012) Isolation and characterization of tumor cells from the ascites of ovarian cancer patients: molecular phenotype of chemoresistant ovarian tumors. PLoS One 7(10):e46858. https://doi.org/10. 1371/journal.pone.0046858 11. Collard JG, van Beek WP, Janssen JW, Schijven JF (1985) Transfection by human oncogenes: concomitant induction of tumorigenicity and tumor-associated membrane alterations. Int J Cancer 35(2):207–213. https://doi.org/10. 1002/ijc.2910350211 12. Koh CM (2013) Preparation of cells for microscopy using ‘cell blocks’. Methods Enzymol 533:249–255. https://doi.org/10.1016/ B978-0-12-420067-8.00018-0 13. Matte I, Lane D, Laplante C, Rancourt C, Piche A (2012) Profiling of cytokines in human epithelial ovarian cancer ascites. Am J Cancer Res 2(5):566–580 14. Plante M, Rubin SC, Wong GY, Federici MG, Finstad CL, Gastl GA (1994) Interleukin-6 level in serum and ascites as a prognostic factor in patients with epithelial ovarian cancer. Cancer 73(7):1882–1888. https://doi.org/10. 1002/1097-0142(19940401)73:73.0.co;2-r 15. Huang S, Robinson JB, Deguzman A, Bucana CD, Fidler IJ (2000) Blockade of nuclear factor-kappaB signaling inhibits angiogenesis and tumorigenicity of human ovarian cancer cells by suppressing expression of vascular endothelial growth factor and interleukin 8. Cancer Res 60(19):5334–5339

Chapter 6 Multispectral Staining and Analysis of Extracellular Matrix Carine M. Renner, Mike R. Visetsouk, Pamela K. Kreeger, and Kristyn S. Masters Abstract Multiplexed immunofluorescent (IF) techniques enable the detection of multiple antigens within the same sample and are therefore useful in situations where samples are rare or small in size. Similar to standard IF, multiplexed IF yields information on both the location and relative amount of detected antigens. While this method has been used primarily to detail cell phenotypes, we have recently adapted it to profile the extracellular matrix (ECM), which provides technical challenges due to autofluorescence and spatial overlap. This chapter details the planning, execution, optimization, and troubleshooting to use multiplexed IF to profile the ECM of human fallopian tube tissue. Key words Serous tubal intra-epithelial carcinoma, STIC, Ovarian cancer

1

Introduction Limited sample availability and size are major challenges in the analysis of many types of clinical specimens. Different technologies have arisen to gather more information from scarce samples [1], but many of these approaches either require fresh specimens, or they focus on profiling cellular markers and are not amenable to analyzing the contents of the extracellular space—i.e., the extracellular matrix (ECM). The ECM performs critical functions in tissue homeostasis and pathology, and changes in its composition and distribution can provide important insight on disease pathogenesis. Immunohistochemistry (IHC) and immunofluorescence (IF) are commonly used to acquire information about both the location and the amount of different ECM components in fixed or frozen tissue samples. However, standard IHC/IF allows for detection of only 2–3 antigens per section, thus necessitating numerous tissue sections if one wants to characterize the ECM environment. Multiplexed IF has been developed to increase the amount of information gained from a single specimen [2]. Multiplexed staining can

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_6, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

105

106

Carine M. Renner et al.

recognize up to 30 markers in a single sample [3], although typical applications and commercial systems offer detection of 6 markers, plus a nuclear stain [2, 4]. Such multiplex techniques are being increasingly employed in cancer biology research and diagnostic pathology [5–7]. Multiplexed IF shows promise as a valuable technique for characterizing the ECM and may prove particularly useful in the case of limited samples [8]. Multiplexed IF is similar to many IHC/IF protocols, as formalin-fixed, paraffin-embedded (FFPE) tissue sections must be deparaffinized, subjected to antigen retrieval, blocking steps, incubation with primary and secondary antibodies, mounting, and imaging. To do multiplexed imaging, samples are stripped and treated with each antibody pair/detection fluorophore in sequence. To interpret these images correctly, staining must be optimized for antibody dilutions/timings, as well as order of application, and a spectral library of single stained slides must be developed to calibrate fluorescent signal intensity and specificity. While multiplexed IF has been performed for many different markers and tissue types, its application to analyze the ECM is accompanied by several challenges due to the diffuse nature of the ECM and the extensive co-localization of individual components. In this report, we describe methods for planning, executing, and troubleshooting the use of multiplexed IF, with a focus on considerations for using this technique to detect ECM components. This work was performed in the context of profiling the ECM of human fallopian tubes and serous tubal intraepithelial carcinoma (STIC) samples https://doi.org/10.1369/00221554211061359. STIC regions are exceptionally small, comprised of only hundreds of cells [9], and thus are not feasible to analyze using many standard techniques.

2

Materials All solutions should be prepared using ultrapure water (18.2 MΩ cm). Reagents may be stored at room temperature unless otherwise stated. Local waste disposal regulations must be followed to safely dispose of materials.

2.1

Tissue Sections

1. Appropriate IACUC or IRB approvals must be obtained to collect the tissue of interest. 2. FPPE tissue blocks should be cut into 5 μm sections and mounted on glass slides.

2.2

Deparaffinization

1. Glass slide holding jars. 2. Safeclear II (a Xylene substitute).

Multispectral Staining of ECM

107

3. Ethanol at 100%, 90%, 70%, and 50% dilutions. Dilutions are prepared by mixing ultrapure water with pure (200 proof; 100%) ethanol to achieve the desired concentration. 4. Neutral buffered formalin (10%): dilute 40% concentrate 1:4 with ultrapure water. 2.3 Wash and Slide Preparation

1. 25 mM Tris–HCl, pH 7.4 + Tween (0.05%) (TBST). Let sit for at least 10 min before use. 2. Avidin block (see Note 1). Store at 4
C. 3. Biotin block (see Note 1). Store at 4
C.

2.4

Antigen Retrieval

1. Opal-7 Color Manual IHC kit (Akoya Biosciences). 2. Working AR6 buffer solution: dilute 10 AR6 buffer stock from Opal kit 1:10 with ultrapure water. Make fresh. 3. Polypropylene slide jars. 4. Microwave (1000–1200 W) with adjustable power settings.

2.5

Staining

1. Hydrophobic wax pen. 2. Humidified chamber. 3. Antibody diluent/block, supplied in the Opal kit. Store at 4
C as directed by manufacturer. 4. Polymer HRP Ms + Rb from Opal kit, typically used at supplied concentration. 5. 1 Amplification Diluent, supplied in the Opal kit. 6. Opal fluorophores, supplied in Opal kit. Reconstitute in 75 μL dimethyl sulfoxide before use. Store at 20
C. These will be diluted immediately before use. 7. DAPI working solution: Add 1 drop of Opal kit supplied DAPI solution to 500 mL TBST. Make as needed, not suitable for storage. 8. Primary antibody for generation of spectral library, diluted to desired concentration in Opal antibody diluent/block. Diluted antibody cannot be reused or stored. See Note 2. 9. Primary antibodies specific to ECM of choice, diluted to desired concentrations in Opal antibody diluent/block. Diluted antibody cannot be reused or stored.

2.6 Coverslip Mounting

1. Glass coverslips (1 mm thickness). See Note 3.

2.7 Microscopy and Analysis

1. Nuance multispectral imaging system (Perkin Elmer).

2. Glycerol-based liquid mountant. See Note 4.

2. Nuance software (v3.0.2). 3. FIJI/ImageJ software (2.0.0, https://imagej.net/Fiji).

108

3

Carine M. Renner et al.

Methods All steps should be carried out at room temperature unless specified otherwise.

3.1 Generation of the Spectral Library

1. For each fluorophore to be detected in the final multiplex, a separate 5 μm section of FFPE tissue mounted on a glass slide should be used. An example spectral library generation plan is displayed in Table 1 (see Notes 5 and 6). 2. Mix the AR6 buffer working solution as specified in Subheading 2.4, item 2. 3. Deparaffinization: Place slides in glass holding jars and deparaffinize as follows: (a) 100% ethanol two times for 3 min each. (b) 90% ethanol for 3 min. (c) 70% for 3 min. (d) 50% ethanol for 3 min. (e) Two rinses in ultrapure water for 1 min each. (f) Fix in 10% neutral buffered formalin for 20 min. (g) Two rinses in ultrapure water for 1 min each. (h) Rinse in AR6 buffer working solution for 2 min. 4. Antigen retrieval: Transfer slides to a polypropylene slide jar and fill the jar completely with AR6 buffer working solution. The jar should be filled all the way to account for volume that will evaporate during the microwave process. Loosely cover the jar with a lid and microwave at 100% power just until it boils, about 1 min 10 s when cold; do only one jar at a time. Microwave for an additional 15 min at 20% power. Tissue should remain submerged in buffer at all times; buffer may be

Table 1 Example staining plan to generate a spectral library Slide

Primary antibody

Opal fluorophore

1

Stat3 1:200

520 1:100

2

Stat3 1:200

540 1:200

3

Stat3 1:200

570 1:100

4

Stat3 1:200

620 1:200

5

Stat3 1:200

650 1:100

6

Stat3 1:200

690 1:100

7

None

DAPI only

Multispectral Staining of ECM

109

refilled throughout the microwave process as necessary. Remove slides from the microwave and allow them to cool to room temperature, 15–30 min. Rinse in ultrapure water twice. Repeat for additional jars. See Note 7. 5. Slide preparation and blocking: Outline tissue sample with hydrophobic barrier pen and rinse slides in TBST for 2 min. Cover samples in the supplied Antibody Diluent/Block and incubate in a closed humidified chamber overnight at 4
C, or 30 min at room temperature. 6. Avidin/Biotin block: Drain slides, but do not rinse, and dab around tissue sections with a Kimwipe without touching the tissue. Cover specimen completely with avidin solution and incubate for 15 min at room temperature, followed by washing 2  3 min in TBST. Cover specimen completely with biotin solution and incubate at room temperature for another 15 min, followed by washing 2  3 min in TBST. 7. Primary antibody incubation: Dilute primary antibody in Antibody Diluent/Block to achieve desired concentration. See Note 8. Incubate sections in primary antibody solution for 1–2 h at room temperature. Rinse 2  3 min in TBST. 8. Secondary-HRP incubation: Remove slides from TBST and drain off excess TBST by touching a Kimwipe to the liquid near the tissue section without touching the section itself. Incubate slides in Polymer HRP Ms + Rb for 10 min at room temperature. This step is time sensitive—it is important to have incubation length as equal as possible between slides. Rinse 2  3 min in TBST at room temperature with agitation. See Note 9. 9. Opal signal generation: Remove slides from TBST and drain off excess TBST by touching a Kimwipe to the liquid near the tissue section without touching the section itself. Dilute Opal Fluorophore stock in 1 Amplification Diluent to make the working solution as desired and pipette working solution onto each slide. Diluted fluorophores should be used immediately and are not suitable for reuse or storage. Recommended starting dilution is 1:100 and may require optimization (see Notes 6 and 8). Incubate slides in the fluorophore solution for 10 min, followed by washing 2  3 min in TBST. Rinse in ultrapure water 2  3 min and then rinse in AR6 buffer for 3 min. 10. In spectral library generation, a slide not previously stained with any fluorophore should be stained with DAPI; the slide should be processed as described in steps 1–8. Prepare the DAPI working solution as specified in Subheading 2.5, item 7 and apply to slide for 5 min in a humidity chamber. Wash slides for 2 min in TBST and then for 2 min in ultrapure water.

110

Carine M. Renner et al.

11. Mounting: Remove excess liquid from the specimen carefully with Kimwipe, without directly touching the tissue. Apply roughly one drop of glycerol-based liquid mountant onto a coverslip and drop it gently onto the tissue. Do not move the coverslip around once it is on the tissue. Avoid generating bubbles in the mountant when mounting coverslips. Allow for coverslips to dry onto the slide overnight in a dark place. 3.2 Imaging the Spectral Library

1. Open the Nuance imaging system and generate a protocol that sets up the filters that will be used during the imaging process. Save the protocol for future use. This process will set up the microscope and software for prompts to change filters during the imaging process. If there is a previously created protocol, load the protocol to start imaging. 2. Locate the area of tissue desired for imaging. Select “Autoexpose Mono” and “Autoexpose Cube” to set automatic exposures. Exposure levels may be entered manually if desired. 3. Select “Acquire Cube” to capture an image. Image each slide/ fluorophore by switching through the filters on the microscope as prompted. Save each image as a “cube.”

3.3 Computing the Spectral Library

1. Open the Nuance software and load a cube captured from the spectral library slides. 2. In the “spectra tab,” pick out representative colors for each fluorophore and label them accordingly. Include an additional channel and label it “Autofluorescence.” 3. For each individual fluorophore, select the appropriately labeled channel in the “spectra tab” for that fluorophore. Zoom into an area of strong fluorophore signal and select a few pixels in that area. Repeat selection multiple times until the signal curve stops changing. There should only be a single peak in the curve when possible (see Note 10). Signal intensity should be between 1000 and 4980 units (see Note 11). The result of this step is the “known spectrum.” 4. Select in the spectra tab the channel designated for autofluorescence (AF). Using the images captured for the fluorophore signal in step 3, select areas of tissue where there is a strong autofluorescent signal. The result of this step is the “mixed spectrum.” 5. Select the “manual compute spectra” button and select the fluorophore signal as the “known spectrum” and the autofluorescence you imaged on that same slide as the “mixed spectrum.” Press “manual compute spectra.” Ideally, there should only be a single peak in the resulting spectral curve, as seen in the individual curves in Fig. 1. See Note 10.

Multispectral Staining of ECM

111

Fig. 1 Example of a functional spectral library, generated using the methods outlined in Subheadings 3.1, steps 1–9

6. Select “transfer to spectrum” for the specific fluorophore signal and then “transfer to library.” This step sets the definition of the true fluorophore signal, with the autofluorescent signal subtracted out. 7. The computing spectra step (steps 2–6) must be done individually for each fluorophore, making sure to reselect autofluorescence specific to each fluorophore slide. 8. EXAMPLE: In order to create a spectral curve for the 520 fluorophore, select a few pixels of fluorophore signal at a time until the signal curve stops changing. This is the “known” spectrum. On the same image used to generate the “known” 520 signal, find an area of autofluorescence and select pixels to generate an autofluorescent signal. This is the “mixed” spectrum. To then compute the spectral library curve for 520, go to

112

Carine M. Renner et al.

“manual compute spectra” and select the 520 fluorophore signal as the “known spectrum” and the autofluorescence you imaged specific to 520 as the “mixed spectrum” and press “manual compute spectra.” Transfer the computed spectrum to the library. 9. After all spectra are computed, the spectral library should look similar to the results shown in Fig. 1. Ideally, all spectra should have single peaks of similar intensity that are all higher than the autofluorescence spectrum but lower than 5000. 10. Save as a “spectral library” protocol once all spectra have been manually computed. 3.4 Multispectral Detection of ECM Proteins

1. Tissue preparation and staining: A multiplexing plan should be generated that accounts for desired multiplexing priorities (see Note 12), staining order (see Note 13), and fluorophore mixing (see Note 14). For this example, eight common ECM components were detected across two different slides of fallopian tube tissue. Antibody dilutions and the order of staining within each slide are presented in Table 2 (see Note 15). These combinations of primary antibody and fluorophore will likely require optimization for other tissues and staining targets (see Note 8). 2. Starting with the first antigen to be detected on each slide, follow Subheadings 3.1, steps 2–9. 3. For the detection of each additional target antigen on a slide, restart the protocol at Subheading 3.1, step 4. 4. Once all the targets have been stained, proceed to DAPI counterstain, Subheading 3.1, step 10. 5. If staining/stripping spans several days, cover the tissue in antibody diluent/block and store in a closed humidified chamber at 4
C during breaks. 6. Mount using the same process defined in Subheading 3.1, step 11.

3.5 Acquisition and Processing of Multispectral ECM Images

1. Image at the same magnification used to generate the spectral library. Follow imaging steps listed in Subheading 3.2. Figure 2 shows an example composite image of a STIC and its individual channels. 2. Load protocol for appropriate spectral library and then load desired cube. 3. The loaded cube will contain the raw composite cube, which has to be “unmixed” to see the individual channels. Under the spectra tab, click on “unmix.” Channels can be selected and deselected to choose which ones to unmix. Individual channel images can be saved by “save image” ! “save all as displayed.” Composite images can be saved by “save image” ! “save.”

Multispectral Staining of ECM

113

Table 2 ECM component multiplexes Slide A multiplex

Slide B multiplex

Primary antibody

Opal fluorophore

Primary antibody

Opal fluorophore

Hyaluronic Acid 1:50

650 1:200

Coll IV 1:50

650 1:200

Perlecan 1:200

570 1:200

Laminin 1:400

570 1:200

Fibronectin 1:500

690 1:200

Coll I 1:1000

690 1:300

Versican 1:500

620 1:200

Coll III 1:1000

620 1:200

P53 1:1000

520 1:200

P53 1:1000

520 1:200

DAPI

DAPI

Fig. 2 Multispectral composite image and individual fluorophore channels following staining of a STIC section for four different ECM components and the tumor marker P53. Scale bar ¼ 200 μm

4. To adjust composite image appearance, open “Display Control” and select “Adjust” for an individual channel. Under “Custom,” there are Min Clip and Max Clip values. True signal will have Min and Max values greater than 0.00 (signals that are just noise will have Min and Max equal to 0.00). These values can be adjusted to dim down false signal. 5. Images may be exported by individual channel as a 16 bit Tiff. Unmix the composite image, select “Save Image” ! “Save All AS (Raw Data File).” The resulting exported image is in grayscale. 3.6 Image Quantification

1. Open DAPI and individual fluorophore image files in FIJI imaging software. 2. Create a region of interest (ROI) using the DAPI image for a slide. Use the free-hand selection tool to outline the ROI. The ROI can be saved using the ROI Manager Tool (Analyze ! Tools ! ROI Manager).

114

Carine M. Renner et al.

Fig. 3 Example of ROI outlined areas in a STIC section. Scale bar ¼ 200 μm

3. In the case of fallopian tube ECM, ROIs for the total tissue (minus blood vessels), the epithelia, and the local stroma were drawn (Fig. 3). The local stroma was defined as stroma within 5 μm of the epithelial area. 4. Select “Set Measurements” from the Analyze menu and select “mean gray value.” Then, select “Measure” from the Analyze menu. 5. For subsequent fluorophore images taken from the same slide, go back to the saved ROIs and repeat analysis.

4

Notes 1. Avidin and biotin solutions used in the images produced in this chapter were purchased from Vector Labs (SP-2001). As the concentrations of these solutions are proprietary, users are advised to optimize this variable if they utilize a different avidin or biotin solution. 2. Spectral library generation is an imaging calibration step, where the antibody used does not need to match any of your antigens of interest for your eventual experiment. The primary antibody used in spectral library generation should yield a strong, clear, punctate signal. Due to the diffuse nature of ECM staining, targeting an ECM protein to generate the spectral library can produce the problems seen in Fig. 4. Specifically, some fluorophores (650, 690) exceeded the maximum desired intensity, while the remaining (520, 540, 620) had weaker signals than autofluorescence. 3. The same coverslips must be used between the spectral library and the experimental slides. Coverslips of varying thicknesses will yield inaccurate signal intensities when unmixing images. 4. The same mountant must be used between spectral library slides and all future slides; different mountants will affect the fluorophore signal and autofluorescence differently. 5. A spectral library must be generated before tissue can be stained and imaged for the antigens of interest. The spectral library functions as a calibration tool for fluorescent signal intensity and specificity and is necessary for accurately

Multispectral Staining of ECM

115

Fig. 4 (a, b) An example of a spectral library generated using an ECM protein as the primary antibody. The diffuse staining of the primary antibody prevented the signals in 520 and 620 from being stronger than the autofluorescent signal shown by the black line. Additionally, DAPI and fluorophores 650 and 690 stained significantly stronger than the other fluorophores, creating too large of a discrepancy in signal intensity. These two issues both rendered the spectral library unusable

unmixing images prior to analysis of multiplex slides. Spectral libraries are generated specific to magnification used during imaging and tissue type. In the process of creating a spectral library, a single primary antibody is paired with every Opal fluorophore to be used during the eventual multiplex staining. The primary antibody selected for generation of a spectral library does not have to match any of the antibodies to be used in actual analysis. Rather, it is recommended to select an antigen that is seen as a positive control for the most prominent cell type in your sample. The primary antibody selected for the spectral library should yield a clear and punctate signal. In the optimization described here, Stat3 was used for staining fallopian tube tissue [10].

116

Carine M. Renner et al.

6. For any fluorophores with a signal >5000, staining should be performed using a lower concentration (higher dilution) of that fluorophore. A new slide should be used, and Subheadings 3.1–3.3 followed. Signal values above 5000 may also result from incorrect definition of known spectra or incorrect selection of autofluorescence signal. For example, in Table 1 fluorophores 540 and 570 were diluted more because the 1:100 dilution generated a signal stronger than the suggested upper limit. 7. Adequate microwave treatment is necessary not only for antigen retrieval and antibody–antigen complex stripping, but also for the reduction of autofluorescence. The blocking steps in Subheadings 3.1, steps 5 and 6 function to moderate autofluorescence during the staining process, and the spectral library computation algorithm works to minimize autofluorescence acquired during imaging. Any of these steps may be adjusted and optimized if autofluorescence becomes a concern. Microwaves may differ in their power, and the timing for this step may therefore need to be adjusted accordingly. Blocking steps may also be extended. Lastly, the image areas selected for autofluorescence calculation in Subheading 3.3, step 4 may influence image processing. It is recommended to select areas of high autofluorescence to eliminate as much autofluorescent signal as possible during computation of the spectral library. 8. For each primary antibody and fluorophore pair, there is a unique and delicate balance between generating a detectable signal and generating a signal that is too strong and will bleed through into other channels. Each primary antibody concentration needs to be optimized for the specific fluorophore it will be paired with. Starting with the manufacturer-recommended dilution, stain with different dilutions of primary antibody to determine what dilution is needed to achieve a detectable stain. Secondary antibody and fluorophore dilutions may then be adjusted in order to generate a signal that minimizes background while avoiding saturation and channel bleed-through. If fluorophore bleed-through occurs at the minimum fluorophore concentration needed for antigen detection, then one should return to optimizing the primary antibody concentration and repeat the above steps. Once secondary concentration, fluorophore concentration, and staining order are optimized, final adjustments to primary antibody concentrations may be needed to intensify or diminish signal. 9. The concentration of the secondary HRP complex may be adjusted to diminish nonspecific background staining. To do so, perform the multispectral staining protocol on tissue without using a primary antibody. Deparaffinize, perform antigen retrieval, and block according to Subheadings 3.1, steps 3–9,

Multispectral Staining of ECM

117

skipping the primary antibody step. Use different dilutions of secondary complex (e.g., undiluted, 1:1, 1:2, 1:3). Any fluorophore may be used. Do not stain with DAPI. 10. Multiple peaks may be the result of an error in selection of pixels or may be due to the unavoidable autofluorescence of the tissue sample. If multiple peaks are observed, try selecting different pixels that appear cleaner. Selecting additional areas of tissue a few pixels at a time may also help consolidate a curve. If these steps fail to correct the issue, there is the possibility that autofluorescence could be diminished as suggested in Note 6, but multiple peaks may also be due to unavoidable tissue characteristics. 11. Intensity values for spectra should be between 1000 and 4980 and similar across all fluorophores. Viewing the curves on an “un-normalized” scale is best for checking the quality of the spectra. 12. There are several aspects to consider when deciding which combination of antigens to detect on the same tissue sample. The user should first consider what the goal of the multispectral staining will be. If the goal is to determine and visualize co-localization of specific proteins, then the priority would be to include those antibody stains on the same slide. If co-localization is not of importance, then consider separating out proteins that are in close proximity to one another in order to create composite images in which each protein can be visualized and appreciated instead of overlapping in a mix of colors. 13. Antigens exhibiting the weakest staining signals should be first in the staining order; conversely, antigens with the strongest signal should be last. Additionally, antigens exhibiting weak staining should be paired with fluorophores whose channels exhibit the least autofluorescence. Weaker antibodies often work best when paired with 570 and 620. Stronger antibodies may be paired with fluorophores in the spectral range where autofluorescence is commonly the greatest (generally 520 and 540). Other strong antibodies can be paired with 650 and 690, as these tend to be weaker fluorophores with less bleed through to neighboring channels. An additional consideration for order would be to place antigens requiring more retrieval toward the end of the staining order. 14. To minimize fluorophore bleed-through, design a staining plan that will keep neighboring channels on separate slides. For example, do not stain with fluorophore 520 and 540 on the same slide; instead, pair 520 with either 570 or 620 and/or 650 or 690. Fluorophore concentrations may also be adjusted to minimize bleed through between neighboring channels. If

118

Carine M. Renner et al.

significant bleed-though is encountered, fluorophore dilution may be increased; however, this change may also affect antigen detection and signal quality. 15. For the specific ECM proteins detected here, HA was the weakest stain for Slide A and thus placed first. Fibronectin and versican were both strong and thus placed toward the end. Col IV was stained first on Slide B due to spatial discrepancies with Col III. Col I and Col III were stained toward the end as they were strong signals. P53 was stained last on both slides as it was the strongest signal, with clear cellular localization. Spatial distribution of antibodies appeared to significantly affect staining for Col IV. With Col III and Col IV both being parts of the basement membrane, and Col III being a very abundant protein in the fallopian tube, it was discovered that Col IV would fail to stain when succeeding Col III. However, when placed before Col III in staining order, the less abundant Col IV protein generated a strong signal. It is theorized that the abundant adherence of antibody + signal generating complex to Col III blocked Col IV antibodies and signaling complexes from sufficiently accessing Col IV, thus failing to produce a signal. References ˇ olovic´ M, Krstic´ D, 1. Van Gool A, Corrales F, C Oliver-Martos B, Martı´nez-Ca´ceres E, Jakasa I, Gajski G, Brun V, Kyriacou K, BurzynskaPedziwiatr I, Wozniak LA, Nierkens S, Pascual Garcı´a C, Katrlik J, Bojic-Trbojevic Z, Vacek J, Llorente A, Antohe F, Suica V, Suarez G, t’Kindt R, Martin P, Penque D, Martins IL, Bodoki E, Iacob BC, Aydindogan E, Timur S, Allinson J, Sutton C, Luider T, Wittfooth S, Sammar M (2020) Analytical techniques for multiplex analysis of protein biomarkers. Exp Rev Proteom 17(4):257–273. https://doi. org/10.1080/14789450.2020.1763174 2. Stack EC, Foukas PG, Lee PP (2016) Multiplexed tissue biomarker imaging. J Immunother Cancer 4:9. https://doi.org/10.1186/ s40425-016-0115-3 3. Gerdes MJ, Sevinsky CJ, Sood A, Adak S, Bello MO, Bordwell A, Can A, Corwin A, Dinn S, Filkins RJ, Hollman D, Kamath V, Kaanumalle S, Kenny K, Larsen M, Lazare M, Li Q, Lowes C, McCulloch CC, McDonough E, Montalto MC, Pang Z, Rittscher J, Santamaria-Pang A, Sarachan BD, Seel ML, Seppo A, Shaikh K, Sui Y, Zhang J,

Ginty F (2013) Highly multiplexed single-cell analysis of formalin-fixed, paraffin-embedded cancer tissue. Proc Natl Acad Sci U S A 110(29):11982–11987. https://doi.org/10. 1073/pnas.1300136110 4. Stack EC, Wang C, Roman KA, Hoyt CC (2014) Multiplexed immunohistochemistry, imaging, and quantitation: a review, with an assessment of Tyramide signal amplification, multispectral imaging and multiplex analysis. Methods 70(1):46–58. https://doi.org/10. 1016/j.ymeth.2014.08.016 5. Esbona K, Yi Y, Saha S, Yu M, Van Doorn RR, Conklin MW, Graham DS, Wisinski KB, Ponik SM, Eliceiri KW, Wilke LG, Keely PJ (2018) The presence of cyclooxygenase 2, tumorassociated macrophages, and collagen alignment as prognostic markers for invasive breast carcinoma patients. Am J Pathol 188(3): 559–573. https://doi.org/10.1016/j.ajpath. 2017.10.025 6. Parra ER, Uraoka N, Jiang M, Cook P, Gibbons D, Forget MA, Bernatchez C, Haymaker C, Wistuba II, Rodriguez-Canales J (2017) Validation of multiplex

Multispectral Staining of ECM immunofluorescence panels using multispectral microscopy for immune-profiling of formalin-fixed and paraffin-embedded human tumor tissues. Sci Rep 7(1):13380. https:// doi.org/10.1038/s41598-017-13942-8 7. Liu WL, Wang LW, Chen JM, Yuan JP, Xiang QM, Yang GF, Qu AP, Liu J, Li Y (2016) Application of multispectral imaging in quantitative immunohistochemistry study of breast cancer: a comparative study. Tumour Biol 37(4):5013–5024. https://doi.org/10.1007/ s13277-015-4327-9 8. Bhaumik S, Boyer J, Banerjee C, Clark S, Sebastiao N, Vela E, Towne P (2020) Fluorescent multiplexing of 3D spheroids: analysis of biomarkers using automated immunohistochemistry staining platform and multispectral imaging. J Cell Biochem 121:4974. https:// doi.org/10.1002/jcb.29827

119

9. Visvanathan K, Vang R, Shaw P, Gross A, Soslow R, Parkash V, Shih Ie M, Kurman RJ (2011) Diagnosis of serous tubal intraepithelial carcinoma based on morphologic and immunohistochemical features: a reproducibility study. Am J Surg Pathol 35(12):1766–1775. https://doi.org/10.1097/PAS. 0b013e31822f58bc 10. Saini U, Suarez AA, Naidu S, Wallbillich JJ, Bixel K, Wanner RA, Bice J, Kladney RD, Lester J, Karlan BY, Goodfellow PJ, Cohn DE, Selvendiran K (2018) STAT3/PIAS3 levels serve as “early signature” genes in the development of high-grade serous carcinoma from the fallopian tube. Cancer Res 78(7): 1739–1750. https://doi.org/10.1158/ 0008-5472.CAN-17-1671

Chapter 7 Quantitative Analysis of the Extracellular Matrix by Immunoblot Kaitlin C. Fogg Abstract Knowing the global and local concentration of extracellular matrix proteins provides critical information for building tissue engineered constructs of healthy and diseased tissue. Here we describe a method of integrating quantitative dot blot along with immunohistochemistry on the same patient sample in order to calculate the overall protein concentration as well as the concentration in a region of interest. Ke ywords Dot blot, FFPE, IHC, Collagen, ECM

1

Introduction The extracellular matrix (ECM) density impacts a wide array of cell behaviors, including proliferation, migration, and survival [1]. Thus when developing scaffolds for tissue engineering applications, it is beneficial to know the range of ECM concentrations present in healthy and diseased tissue. Formalin-fixed paraffinembedded (FPPE) tissue is one of the most common types of clinical samples as FPPE blocks can be stored for decades and hold a wealth of information [2]. FFPE samples are most commonly used for immunohistochemistry (IHC), a laboratory technique that yields information on the spatial distribution and relative expression of a protein of interest. Furthermore, protein has been successfully extracted from FFPE blocks and used for Western blot and mass spectrometry [3]. Western blot is performed on samples that have been separated by molecular weight by electrophoresis and yields relative expression of a protein of interest. However, it is size constrained as to how many samples can be run on one gel, making it difficult to evaluate multiple samples and a standard curve simultaneously. Mass spectrometry yields relative abundance of all proteins present in the sample, but again does not yield a concentration. Quantitative dot blot works in a similar manner to Western

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_7, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

121

122

Kaitlin C. Fogg

blot; however, due to the high protein binding capacity of the polyvinylidene difluoride (PVDF) membrane, there is no need for the electrophoresis step, and the sample can be directly spotted onto the membrane. Furthermore, as there is no electrophoresis step, the PVDF membranes are not size constrained by electrophoresis equipment and a standard curve can be made on the dot blot alongside the samples. This yields a standard curve that can be used to calculate the concentration instead of the relative abundance. By integrating a quantitative dot blot with quantitative IHC from the same FFPE sample, the global and local concentration of extracellular matrix (ECM) proteins can be quantified. The protocol in this chapter is specific for collagen type I based on our experience using it to quantify type I collagen in benign and metastatic omentum [4]; however, this process can be applied to a wide range of proteins. For example, it has also been used to quantify collagen type I and collagen type III in ovary sections [5]. This information was then used to build in vitro models with biologically relevant concentrations of the corresponding ECM components. Additionally, Yu et al. recently extracted protein from FFPE samples for quantitative dot blot of HER2 in breast cancer samples [6], thus it is not limited to ECM proteins and could be used to develop cellular models with appropriate levels of proteins of interest.

2 2.1

Materials FFPE Blocks

1. One 5 μm section from each FFPE block (see Note 1). 2. Two slides with 5 μm sections of the same tissue type for antibody controls (see Note 2). 3. A paraffin curl of 20–60 μm (depending on available sample) of each FFPE sample in a 1.5-mL microcentrifuge tube (see Note 3).

2.2 Collagen Type I Standard

1. 10 mg/mL collagen type I (see Note 4). 2. Positive displacement pipette and pipette tips. 3. 10
PBS. 4. 0.1N NaOH. 5. 1
PBS. 6. Cell culture incubator (37  C, 5% CO2, 95% relative humidity).

2.3 Protein Extraction

1. Standards prepared in Subheading 3.1. 2. Formalin-fixed paraffin-embedded (FFPE) curls. 3. 1.5-mL microcentrifuge tubes. 4. Safeclear II Xylene Substitute.

Quantitative Analysis of ECM

123

5. Ethanol (EtOH) dilutions in distilled H2O: 100%, 70%, 50%. 6. Protein extraction buffer (PEB): 20 mM Tris–HCl in distilled H2O, 2% SDS, pH ¼ 8. 7. Orbital shaker or tube shaker. 8. Centrifuge that can go up to 16,000
g and can be set to 4  C. 2.4 Quantitative Dot Blot

1. Protein extraction buffer (PEB): 20 mM Tris–HCl in distilled H2O, 2% SDS, pH ¼ 8. 2. 1.5-mL centrifuge tubes. 3. PVDF membranes. 4. Methanol. 5. Petri dishes. 6. Tris-buffered saline (TBS): 250 mM Tris base/Tris–HCl, 27 mM KCl, 1.37 M NaCl, pH ¼ 7.4. 7. Forceps. 8. Orbital shaker. 9. Oven that can be set to 37  C. 10. Blocking buffer: TBS + 0.1% Tween + 1% goat serum. 11. Primary antibody. 12. TBST: TBS + 0.1% Tween. 13. Horseradish peroxidase (HRP) secondary antibody. 14. Chemiluminescent detection reagents, do not mix until ready to use. 15. Aluminum foil. 16. Kim wipes. 17. Plastic wrap. 18. Chemiluminescent imager.

2.5 Fluorescent Immunostaining

1. Twelve Coplin jars for every ten slides. 2. SafeClear II Xylene Substitute. 3. Fume hood. 4. Waste bottle. 5. Forceps. 6. Paper towels. 7. Kim wipes. 8. Ethanol (EtOH) dilutions in distilled H2O: 100%, 90%, 70%, 50%. 9. Tris-buffered saline (TBS): 250 mM Tris base/Tris–HCl, 27 mM KCl, 1.37 M NaCl, pH ¼ 7.4. 10. Water bath set to 90  C.

124

Kaitlin C. Fogg

11. Universal Antigen Retrieval Solution. 12. Peroxidase-Antiperoxidase (PAP) Pen, a hydrophobic barrier pen for marking slides for immunohistochemistry applications. 13. TBS-B: TBS + 1% BSA. 14. Antibody diluting solution: TBS-B + 1% goat serum. 15. Humidifying chamber (see Note 5). 16. Primary antibody (same as used for the dot blot). 17. Fluorescently tagged secondary antibody. 18. Aluminum foil. 19. TBS-T: TBS + 0.1% Tween. 20. Coverslips. 21. Glycerol-based liquid mounting media/sealant with DAPI. 22. Slide box. 23. Fluorescent microscope with tiling abilities. 24. FIJI software.

3

Methods The overall process consists of four main steps: preparing the FFPE tissue block, performing a dot blot, performing fluorescent IHC, then quantifying them to calculate the global and local collagen concentrations. A visual overview of the process is provided in Fig. 1.

3.1 Preparing a Standard

1. Combine the following in a 1.5-mL microcentrifuge tube in the following order: 10 μL 10
PBS, 10 μL 0.1N NaOH, and 80 μL 10 mg/mL collagen type I to generate 100 μL of 8 mg/ mL collagen I. 2. Use positive displacement pipette to mix thoroughly (see Note 6). 3. Incubate at 37  C for 1 h to gel. 4. Add 1 mL 1
PBS, then remove all liquid before proceeding to protein extraction (Subheading 3.2).

3.2 Protein Extraction

1. Incubate FPPE curls and standards in 1.5-mL SafeClear for 20 min at room temperature with gentle shaking on orbital or tube shaker. 2. Centrifuge at 16,000
g for 5 min. 3. Remove SafeClear supernatant (place in a waste container, not down the drain).

Quantitative Analysis of ECM

125

Fig. 1 Schematic depicting the overall process of calculating the global and local collagen concentrations from FFPE samples. FFPE formalin-fixed paraffin-embedded, IHC immunohistochemistry, MFI median fluorescent intensity. (Created with BioRender.com)

4. Repeat steps 1–3 two additional times (for a total of three incubations) or until paraffin wax is gone. 5. Add 1 mL 100% ethanol to each sample. 6. Incubate 2 min at room temperature. 7. Centrifuge at 16,000
g for 3 min. 8. Remove supernatant. 9. Repeat steps 5–8 with a second 2-min incubation of the same ethanol concentration. 10. Repeat steps 5–9 for 70% ethanol. 11. Repeat steps 5–9 for 50% ethanol. 12. Resuspend tissue pellet in 50 μL PEB. The protein is now in the liquid. 13. Centrifuge at 16,000
g for 20 min at 4  C. 14. Remove supernatant (this now contains your protein) and pipette into new labeled microcentrifuge tube. 15. Proceed to Dot Blot or store at 80  C until further use. 3.3

Dot Blot

1. Prepare 1000 μL of 1 mg/mL Collagen I solution in a new 1.5mL centrifuge tube by adding 937.5 μL PEB to 62.5 μL of the collagen 1 solution from Subheading 3.2.

126

Kaitlin C. Fogg

Table 1 Preparation of a standard curve ranging from 1 to 0.008 mg/mL Concentration (mg/mL)

Collagen solution (μL)

PEB (μL)

1.00

100 μL 1 mg/mL

0.0 μL PEB

0.500

50 μL 1 mg/mL

50.0 μL PEB

0.250

50 μL 0.5 mg/mL

50.0 μL PEB

0.125

50 μL 25 mg/mL

50.0 μL PEB

0.063

50 μL 125 mg/mL

50.0 μL PEB

0.031

50 μL 063 mg/mL

50.0 μL PEB

0.016

50 μL 0.031 mg/mL

50.0 μL PEB

0.008

50 μL 0.016 mg/mL

50.0 μL PEB

2. In new microcentrifuge tubes, prepare a standard curve via 1:1 serial dilution with PEB as detailed in Table 1. 3. Place PVDF membrane in 100% methanol in a petri dish for 1 min, until the membrane is translucent. 4. Using forceps, place the membrane in distilled water in a petri dish for 2–3 min, until the membrane floats. 5. Remove the membrane from the water, do not blot excess MeOH or water from the membrane (see Note 7). 6. Carefully pipette 1 μL sample or standard onto the membrane (dot the standards and samples in triplicate on the PVDF membrane). 7. Pipette the samples in a grid pattern, making note of layout and cutting a notch in the top right corner to orient the blot later (see Fig. 2 for an example layout). 8. Allow the membrane to dry at 37  C in an oven for 5 min (this sets the protein into the membrane). 9. Rinse membrane for 5 min in TBS
3 (NO TWEEN) with gentle shaking on an orbital shaker. 10. Block with blocking buffer for 1 h at room temperature with gentle shaking on an orbital shaker. 11. Dilute primary antibody in blocking buffer according to recommended concentrations for Western blotting (see Note 8). 12. Incubate in primary antibody overnight at 4  C on an orbital shaker. 13. Collect the primary antibody into a 50-mL conical tube to save it (can be used for ~1 month if stored at 4  C).

Quantitative Analysis of ECM

127

Fig. 2 Example layout for a dot blot. Gray dots indicate the standards and their corresponding concentrations from high (dark gray) to low (light gray). Blue dots indicate the samples. Standards and samples are arranged in triplicate horizontally. The dotted line indicates the top of the PVDF membrane

14. Rinse membrane for 5 min in TBST on the orbital shaker. Repeat for a total of 3 washes. 15. Dilute secondary antibody (make sure it is an HRP-conjugated secondary, NOT a fluorescent antibody) in blocking buffer at manufacturer recommended dilution for chemiluminescence (see Note 8). 16. Incubate blots protected from light for 1 h at room temperature with gentle shaking on orbital shaker. 17. Pour off secondary solution (this cannot be reused). 18. Rinse membrane for 5 min in TBST three times on orbital shaker. 19. Rinse membrane for 5 min in TBS (no Tween) three times on orbital shaker. 20. Prepare 5 mL of chemiluminescent substrate by combining 2.5 mL of each of the two components. 21. Remove all TBS and add 2–5 mL of chemiluminescent substrate so that the PVDF membrane is saturated. Protect from light with aluminum foil and let incubate for 5 min in the dark. 22. Remove excess liquid without touching the membrane by using plastic forceps to carefully lift it and dab off the excess liquid on a Kim Wipe. 23. Place the membrane face up on a piece of plastic wrap, and then cover with a second piece and seal (make sure there are no bubbles, see Note 9).

128

Kaitlin C. Fogg

24. Image the blot immediately on a chemiluminescent imager, making sure the standards and samples are visible while reducing background. 3.4

Deparaffinization

1. Place slides with 5 mm sections (samples and controls from Subheading 2.1) in Coplin jars with 100 mL SafeClear per jar in the fume hood. 2. Incubate for 5 min. 3. Remove slides with forceps, dip up and down ten times, then remove excess liquid by blotting the side of the slides on paper towels. 4. Put back into the same Coplin jar of SafeClear. 5. Incubate for 5 min. 6. Dispose of SafeClear in a waste bottle in a fume hood. 7. Add 100 mL 100% EtOH solution to new Coplin jar. 8. Add slides to the jar, and incubate for 2 min. 9. Repeat steps 7 and 8 with a second 2-min incubation in a new Coplin jar of the same ethanol concentration. 10. Repeat steps 7–9 for 70% ethanol. 11. Repeat steps 7–9 for 50% ethanol. 12. Wash slides in a new Coplin jar of 100 mL TBS-T for 5 min. 13. Repeat step 12 in a new Coplin jar.

3.5

Antigen Retrieval

1. Prepare a water bath at 90  C (NOT boiling). 2. Add 100 mL Universal Antigen Retrieval Solution to a Coplin jar. 3. Add slides, and incubate for 10 min in water bath at 90  C. 4. Incubate 10 min on bench to cool. 5. Wash slides in 100 mL TBS-B in a new Coplin jar for 5 min at room temperature. 6. Repeat step 5 in a new Coplin jar.

3.6

Immunostaining

1. One slide at a time, let slide drip dry on a large Kim Wipe. When the slide is dry, use the PAP pen to make a circle around your sample. Let the wax dry for approximately 1 min (until the wax is opaque and solid rather than clear and liquid). 2. Add 400 μL TBS-B within the PAP circle to block excess protein-binding sites. 3. Place all slides in a humidifying chamber. 4. Incubate 1–2 h at room temperature. 5. Dilute primary antibody in Antibody Diluting Solution. 6. Drain TBS-B by tapping slide on a Kim Wipe—do not touch Kim Wipe to the tissue.

Quantitative Analysis of ECM

129

7. Add 400 μL primary antibody solution to each slide within the Pap circle (for secondary only control add Antibody Dilution Solution). As noted in Subheadings 2 and 3, this primary antibody must be the same as the one used for the dot blot (see Note 2 for more details on antibody controls, see Note 10 on optimizing primary and secondary antibodies). 8. Place all slides in a humidifying chamber. 9. Incubate for 4 h at room temperature or overnight at 4  C. 10. Add 100 mL TBS-B to a Coplin jar. 11. Drain the primary antibody solution by tapping slide on a Kim Wipe—do not touch Kim Wipe to the tissue. 12. Place the slides in the Coplin jar with TBS-B and incubate for 2 min. 13. Carefully dump out the TBS-B into the sink while holding the slides in and refill with 100 mL of fresh TBS-B. Alternatively, place slides in a new Coplin jar of 100 mL TBS-B. 14. Repeat steps 12 and 13 three times for a total of four washes. 15. Dilute secondary antibody in Antibody Diluting Solution and wrap the tube in aluminum foil to protect from light. 16. Add 400 μL secondary antibody solution to slides within the Pap circle (for primary only control add Antibody Dilution Solution). 17. Place slides in the humidifying chamber. 18. Cover the humidifying chamber in aluminum foil to protect from light. 19. Incubate 1 h at room temperature in the humidifying chamber. 20. Drain the secondary antibody solution by tapping slide on a Kim Wipe—do not touch Kim Wipe to the tissue. 21. Place the slides in the Coplin jar with 100 mL TBS-B, cover with aluminum foil, and incubate for 2 min. 22. Carefully dump out the TBS-B into the sink while holding the slides in and refill with 100 mL of fresh TBS-B. Alternatively, place in a new Coplin jar of 100 mL TBS-B. 23. Repeat steps 21 and 22 two times for a total of three washes. 24. Drain slides on Kim Wipe. 25. Add 20 μL mounting media (1 drop) to coverslip. 26. Flip coverslip and carefully add to slide. 27. Incubate overnight at room temperature in the dark (in a drawer or covered in aluminum foil). 28. Once cover slips are sealed, store at 4  C in the dark by placing them in a slide box.

130

3.7

Kaitlin C. Fogg

Imaging

1. Using a 10
objective, take a tiled image for each slide of the entire piece of tissue in both the DAPI and the channel for the fluorophore corresponding to the secondary antibody. 2. Make sure the tiled image has stitched together and saved. 3. Use the same exposure settings (intensity, time) for every slide.

3.8

Analysis

1. Quantify the intensity of the dots for the samples and standards from the dot blot. This can be done in FIJI or the software connected to the chemiluminescent imaging equipment. 2. Use standard curve to generate formula: ½Col Isample,dot blot ¼ m  ðintensityÞ þ b 3. Calculate the mass of collagen (in mg) in each sample (the volume of Protein Extraction Buffer ¼ 0.2 mL) Col ITissue ¼ ½Col Isample,dot blot  Volume of Protein Extraction Buffer 4. Using the IHC images of the whole tissue section and the thickness of the paraffin section used for the dot blot, calculate the volume of the tissue used for the dot blot in mL. Tissue Volume ¼ Tissue Area  Thickness of Dot Blot Paraffin Section 5. Calculate the global collagen concentration: ½Col ITissue ¼

Col ITissue Tissue Volume

6. Measure the median fluorescent intensity (MFI) of the tiled tissue image ¼ MFITissue 7. Measure the MFI of regions of interest (see Note 11 for suggested regions of interest) ¼ MFILocal 8. Calculate local collagen concentrations at regions of interest: ½Col ILocal ¼

4

MFILocal  ½Col ITissue MFITissue

Notes 1. FFPE blocks may require institutional review board (IRB) approval; alternatively FFPE may be acquired from either a commercial source or a university tissue bank that acts as an honest broker.

Quantitative Analysis of ECM

131

2. Make sure to have at least three biological replicates (FFPE blocks) per condition. For each FFPE block, request one slide per protein of interest, as well as two additional slides of the same type of tissue for the two negative controls: (1) primary antibody with no secondary antibody, (2) secondary antibody with no primary control. It is critical this is the same type of tissue to account for background noise. 3. Try and have this curl as close as possible to the slides that were cut for IHC. 4. This protocol is specifically for investigating the local and global concentrations of collagen type I. However, this protocol can be easily adapted for other types of ECM proteins, by adding a known amount of that type of ECM protein to the type I collagen gel or to another natural or synthetic polymer of your choice. 5. Humidifying chambers can be made by taping down 5 mL pipettes with the tips broken off in a square 500 cm2 polystyrene dish or Pyrex baking dish with a lid. Place dampened Kim Wipes down around the edges and slides across the pipette rods to hold the slides out of the water. 6. When mixing high concentration collagen type I, it is important to mix slowly and move the collagen solution with the positive displacement pipette from bottom to top as you mix— this will prevent bubbles. 7. PVDF is hydrophobic and wetting it with methanol makes it possible for the aqueous protein solution to bind to the membrane. If the membrane dries out and turns opaque, this means that it is no longer pre-wetted, and the protein solution may not bind as well. To make things go faster, all the samples and standards should be prepared, and the grid layout established before the PVDF membrane is wetted with methanol. 8. Primary and secondary antibodies need to be optimized before testing clinical samples, typically using the standards. Our experience suggests that concentrations suggested for Western blot with chemiluminescence are a good starting point. To detect collagen I, we have utilized a 1:1000 dilution of a collagen I antibody (Abcam, ab34170) and a 1:2000 dilution of the HRP-conjugated secondary [4]. 9. To avoid bubbles in the saran wrap, get your hand wet with water and then sprinkle it on the lab bench top. Then, carefully place the saran wrap down on the wet bench top and use a credit card or ID to gently smooth it out. Place the dot blot on top of the smooth saran wrap, and carefully fold it over so that it lays down smooth over the blot and then forms a seal. Use scalpel or box cutter to remove the excess saran wrap, and then

132

Kaitlin C. Fogg

smooth out any remaining bubbles gently with the credit card/ ID you used earlier. 10. Primary and secondary antibodies need to be optimized prior to staining all of the FFPE samples. This should be done using the same type of tissue that you are intending to use for your protein quantification studies. A good starting dilution can be found by looking at the manufacturer’s recommendation for IHC. 11. Regions of interest can include epithelial vs. stromal cells, proximal to a tumor vs. distal from the tumor, keratinocyte rich vs. fibroblast rich, etc. References 1. Muncie JM, Weaver VM (2018) The physical and biochemical properties of the extracellular matrix regulate cell fate. Curr Top Dev Biol 130:1–37 2. Kokkat TJ, Patel MS, McGarvey D, LiVolsi VA, Baloch ZW (2013) Archived formalin-fixed paraffin-embedded (FFPE) blocks: a valuable underexploited resource for extraction of DNA, RNA, and protein. Biopreserv Biobank 11 (2):101–106 3. Scicchitano MS, Dalmas DA, Boyce RW, Thomas HC, Frazier KS (2009) Protein extraction of formalin-fixed, paraffin-embedded tissue enables robust proteomic profiles by mass spectrometry. J Histochem Cytochem 57 (9):849–860

4. Fogg KC, Renner CM, Christian H, Walker A, Marty-Santos L, Khan A et al (2020) Ovarian cells have increased proliferation in response to heparin-binding epidermal growth factor as collagen density increases. Tissue Eng A 26 (13–14):747–758 5. Fleszar AJ, Walker A, Porubsky V, Flanigan W, James D, Campagnola PJ et al (2018) The extracellular matrix of ovarian cortical inclusion cysts modulates invasion of fallopian tube epithelial cells. APL Bioeng 2(3):031902 6. Yu G, Zhang W, Zhang Y, Lv J, Wu S, Sui X et al (2020) Developing a routine lab test for absolute quantification of HER2 in FFPE breast cancer tissues using Quantitative Dot Blot (QDB) method. Sci Rep 10(1):12502

Part III Cell Isolation

Chapter 8 Multiparameter Single-Cell Characterization of Ovarian Intratumor Heterogeneity Kristin G. Beaumont, Christina Andreou, Ethan Ellis, and Robert Sebra Abstract Cancer is a complex disease rooted in heterogeneity, which is the phenomenon of individual cells, tissues, or patients having distinct phenotypic and/or genetic characteristics. Observed divergent disease etiology is likely rooted, at least in part, in tumor heterogeneity and the classification of distinct and important subpopulations of cells within the tumor and its associated microenvironment has remained a technical challenge. Standard next-generation sequencing of bulk tumor tissue provides an overall average genetic profile of the sample, and masks contributions from individual cells and minor populations of cells, particularly in heterogeneous samples. Only with the advent of single-cell analysis and sequencing technologies has it become possible to characterize key contributions of cellular subpopulations in order to more comprehensively characterize disease. This chapter describes a method to generate linked phenotypic and genotypic data at single-cell resolution using a real-time single-cell resolved platform. Specifically, the example method provided here is used to link cellular growth kinetics and expression of a prognostic marker protein, CA-125, in cells derived from ovarian cancer patients with their single-cell genomic profiles, but the method is translatable to other cell types and phenotypes of interest. Key words Single-cell analysis, Single-cell DNA Seq, Single-cell RNA Seq, Tumor heterogeneity, Ovarian cancer

1

Introduction Cancer is a complex disease rooted in heterogeneity, which is the phenomenon of individual cells, tissues, or patients having distinct phenotypic and/or genetic characteristics. This is particularly true for cancers such as ovarian cancer, which has the worst prognosis of any of the gynecological cancers and is comprised of several different categories of tumors, each with its own molecular biology, phenotype progression, and prognosis [1–3]. Better understanding the complexity of this disease is needed to inform more robust diagnosis and treatment because such divergent disease etiology is likely rooted, at least in part, in tumor heterogeneity as well as heterogeneity in the tumor microenvironment. The classification

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_8, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

135

136

Kristin G. Beaumont et al.

of distinct and important subpopulations of cells within the tumor and its associated microenvironment has remained a technical challenge. Next-generation sequencing (NGS) of bulk tumor/tumor microenvironment tissue has yielded some exciting results, including the discovery of certain driver mutations and variants of unknown significance. Using common NGS approaches, genetic material is extracted from bulk tissue and sequenced, which provides an overall average genetic profile of the sample, but masks contributions from individual cells and minor populations of cells, particularly in heterogeneous samples. Only with the advent of single cell sequencing technologies has it become possible to characterize the key contributions of cell subpopulations in order to identify new therapeutic targets, disease biomarkers and understand modes of treatment for disease, where single-cell sequencing characterizes and reports the genome or transcriptome of each individual cell in a population, allowing the identification of subpopulations of cells that are genetically similar or distinct [4–7]. One limitation of single-cell sequencing as it currently exists is that it is challenging to link complex phenotypes (such as cell growth or dynamic protein expression) with cell-specific genotypic or transcriptomic profiles. This chapter describes a method to generate linked phenotypic and genotypic data at single-cell resolution using a real-time single-cell resolved platform. We are specifically interested in understanding the expression of CA-125, which is a known prognostic marker in ovarian cancer, at the single cell level and whether expression of this marker correlates with other known cancer cellular genotypes or phenotypes [8]. This data is especially useful in understanding the strengths and limitations of using CA-125 in a prognostic capacity. Thus, the example method provided here is used to link cellular growth kinetics and CA-125 expression in cells derived from ovarian cancer patients with their single cell genomic profiles, but the method is translatable to other cell types and phenotypes of interest. Moreover, single exported cells can be subjected to plate-based single cell RNA Seq rather than single-cell DNA Seq if desired.

2

Materials

2.1 Generation of Single-Cell Suspension

1. 50-mL conical tubes. 2. Supplemented cell culture media: Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. 3. Tissue scalpel. 4. GentleMACS C-tube. 5. GentleMACS Octo Dissociator with Heaters.

Single Cell Characterization of Intratumor Heterogeneity

137

6. Human Tissue Dissociation Kit, including Enzymes H, R and A, solubilized as directed (Miltenyi Biotec). 7. 70-μm cell strainer. 8. 40-μm cell strainer. 9. Red blood cell lysis solution (1). 10. Trypan Blue solution 0.4%. 11. Disposable hemocytometers. 2.2 Beacon Chip Preparation and Cell Loading

1. OptoSelect 3500 Chip and Chip Wetting Solution (Berkeley Lights Inc). 2. Syto24 live cell nuclear dye. 3. DNA LoBind 1.5-mL tubes. 4. Pluronic F-127, 0.2-μm filtered (10% solution in water). 5. Bovine fibronectin. 6. CA-125 primary antibody (Clone M11). 7. Goat anti-Mouse IgG (H + L) cross-absorbed secondary antibody, Texas Red-x.

2.3 Single-Cell DNA Sequencing Preparation

1. Qiagen REPLI-g Single Cell Kit. 2. Qiagen buffer EB. 3. Absolute ethanol (200 Proof), Molecular Biology Grade. 4. Nuclease-free water. 5. Agencourt AMPure XP, 60 mL. 6. Magnetic separator for 96-well plates. 7. DNA Broad Range Qubit Assay. 8. DNA 12000 kit (Agilent).

2.4 Single-Cell DNA Sequencing

1. Ion AmpliSeq™ Cancer Hotspot Panel v2. 2. Ion AmpliSeq™ library kit 2.0. 3. Ion Xpress™ barcode adapters 1-96 Kit. 4. Ion Library TaqMan™ Quantitation Kit. 5. Ion GeneStudio™ S5 System. 6. Ion 540™ Kit-Chef. 7. Ion 540™ Chip Kit.

3

Methods

3.1 Generation of Single-Cell Suspension

1. Obtain tumor tissue as quickly as possible following surgical removal (see Note 1). 2. Working quickly, remove any fat-containing sections using a tissue scalpel. Mince the remaining tissue into 2–4 mm pieces.

138

Kristin G. Beaumont et al.

3. Add no more than 1 g of tissue to the C-tube used with the GentleMACS system and add 200 μL of Enzyme H, 100 μL of Enzyme R and 25 μL of Enzyme A from the human tumor kit. 4. Disrupt tissue using the “37C_h_TDK_1” protocol for soft human tumors on the GentleMACS system, as recommended for ovarian tumors (see Note 2). 5. Pass suspension through a 70-μm cell strainer, noting that suspensions containing a lot of debris or a high density of cells may clog the strainer, resulting in inefficient cleanup. To ameliorate this, dilute the suspension by 10–100 in supplemented cell culture media prior to filtering (see Note 3). 6. Repeat step 5 with a 40-μm cell strainer. 7. Centrifuge cell suspension at 300  g for 15 min at 4  C to pellet cells. 8. If cell pellet contains any evidence of red blood cell (RBC) contamination (either a red band at the top of the pellet or a pinkish tint to the pellet), subject it to a 2 min RBC lysis in 1 mL of 1 Red Blood Cell Lysis Solution. 9. Centrifuge cell suspension at 300  g for 15 min at 4  C and re-examine the pellet for signs of RBC contamination. If needed, repeat step 8. If no further RBC lysis is needed, resuspend cells in 1 mL of supplemented cell culture media. 10. Assess cell viability using Trypan Blue staining on a disposable hemocytometer. Target is >80% viable cells. 11. Assess debris content of suspension—suspensions containing a lot of debris will clog the Beacon chip and cause the experiment to fail. If samples contain substantial debris, refilter or otherwise clean up to prevent this from happening (see Note 4). When suspension is sufficiently debris-free, resuspend cells at 1–2  106 cells/mL. 12. Alternatively, banked patient-derived cells or cell lines can be used (see Note 5). If using banked cells or cell lines, they should be detached from the culture surface, and disaggregated by the method that is typically utilized for passaging of that particular cell type (bypassing steps 3–6). Once disaggregated, the cell suspension should be centrifuged as in step 7 and counted and evaluated as in steps 10 and 11. 3.2 Beacon Chip Preparation and Cell Loading

1. As an overview (see Fig. 1 for summary), the subsequent steps describe the isolation of single tumor cells on a nanofluidic single-cell handling platform, followed by phenotypic characterization of growth (in our case, over 6 days in culture) and CA-125 expression. Following characterization of any desired phenotypes, cells are exported off of the platform one by one, to tracked individual wells of 96-well plates for plate-based

Single Cell Characterization of Intratumor Heterogeneity

139

Fig. 1 Overview of experimental design

single-cell sequencing. Overall, the whole process takes approximately a week to execute start to finish (from tissue acquisition through sequencing) where (a) cell loading, penning, and phenotyping, (b) cell export, (c) library preparation, (d) sequencing, and (e) data transfer each require approximately a day, and any incubation (for cell growth, etc.) adds time to the process. 2. Prepare an OptoSelect 3500 chip with wetting solution (see Note 6 for chip description and Note 7 for wetting details), to enable cell penning and flush three times, each with 250 μL of supplemented cell culture media. 3. Pre-stain cells with Syto24, a live cell nuclear stain in a DNA LoBind tube, prior to loading in order to enable accurate cell counting of adherent cells. Incubate cells for 15 min at room temperature in 1 mL of supplemented cell culture media containing a 170 nM solution of Syto24 (to minimize viability impact) and 0.1% F-127. 4. Gently centrifuge cells (300  g for 3 min), remove the staining solution, being careful not to disturb the cells at the bottom of the tube. Add 750 μL of supplemented cell culture media. 5. Gently mix, centrifuge (300  g for 3 min) and remove the staining solution, being careful not to disturb the cells at the bottom of the tube. 6. Resuspend cells at 1–2  106 cells/mL in 30 μL of supplemented cell culture media with 0.1% F-127 added. If using multiple cell lines, or cells from multiple patients, prepare all cell suspensions and load into select regions of the chip (see Note 8). 7. Gently resuspend cells immediately prior to loading, by slowly pipetting 20 μL of the cell suspension with a P100 pipettor three times in a 1.5-mL centrifuge tube. Import using the “Load” operation in Cell Analysis System software and Small Volume Import using default parameters (see Note 9), where

140

Kristin G. Beaumont et al.

Fig. 2 Picture showing export needle submerged in 1.5-mL tube containing cell suspension, for loading cells onto the optofluidic chip

you will load cells directly from the 1.5 mL DNA LoBind tube by immersing Beacon’s export needle down to the bottom of the tube (see Fig. 2). 8. Pen cells using an optimized optoelectronic positioning (OEP) voltage and 5 μm/s cage speed to start, increasing or decreasing voltage and speed as necessary to ensure that cells pen at the lowest possible input voltage (see Note 10). 9. Once penning is complete, flush the chip three times, each with 250 μL of culture media and set chip temperature to 37  C. 3.3 Phenotypic Characterization of Cells

1. Culture cells at 37  C in 5% CO2-buffered supplemented cell culture media containing 10 μg/mL bovine fibronectin, perfused at 0.01 μL/s. After 24 h, switch perfusion media to 5% CO2-buffered supplemented cell culture media without fibronectin. Replace bulk media every 48 h to reduce likelihood of contamination. 2. Culture cells on chip for desired period of time, acquiring time lapse images as needed (see Note 11). 3. To assess CA-125 cell surface expression at a desired timepoint, import a dilution of 1:5 CA-125 primary antibody in

Single Cell Characterization of Intratumor Heterogeneity

141

Fig. 3 Cluster of ovarian cancer cells stained with Syto24 for viability assessment and anti-CA-125 for CA-125 expression evaluation

supplemented cell culture media, followed by incubation of 60 min to allow diffusion of the antibody into the pens (see Note 12). 4. Flush the unbound primary out of the chip with three flushes of 250 μL of supplemented cell culture media at 1 μL/s. 5. Perform secondary antibody staining by importing goat antimouse secondary antibody conjugated with TxRed dye at 20 μg/mL dilution in supplemented cell culture media, followed by a 60 min incubation to allow diffusion of the secondary antibody into the pens. 6. Flush the unbound secondary antibody out of the chip with three flushes of 250 μL of supplemented cell culture media at 1 μL/s. If secondary antibody remains (as evidenced by background fluorescent signal within the chip pens), repeat flushing cycles until background signal is sufficiently low for image analysis. 7. If desired, additional Syto24 dye (at 50 μM) can be loaded and incubated for 30 min to enable staining of cell nuclei. 8. Capture and analyze brightfield and fluorescent images (for our system exposure times were FITC 200 ms; TxRED 7500 ms) to obtain protein expression levels for each pen and cell (see Note 13 for analysis suggestions). See Fig. 3 for example. 9. While our desired phenotypic metrics were cell growth and CA-125 expression, many other phenotypes from different cell types are accessible via this method (see Note 14).

142

Kristin G. Beaumont et al.

3.4 Single-Cell DNA Sequencing Preparation

1. For cells of interest (and controls), export onto a 96-well plate and store at 20  C (see Note 15). 2. Perform whole-genome amplification (WGA) using the multiple displacement amplification (MDA) method (in this case with the QIAgen REPLI-g Single Cell Kit). 3. After WGA chemistry is completed and polymerase is inactivated, perform a 1 AMPure XP bead-based purification on the amplification products. Briefly, add 50 μL of beads to each WGA reaction and incubate at room temperature while shaking on a vortex mixer at 2000 rpm for 10 min. 4. Place reactions on a magnetic separation device for 5 min or until solution is clear, then remove and discard the supernatant. 5. Prepare fresh 80% ethanol, by diluting absolute ethanol in nuclease-free water. 6. While the tubes remain on the magnetic separation device, wash beads twice with 100 μL 80% ethanol, without disturbing the beads. Carefully remove wash solution and discard. 7. Remove reactions from the magnet and briefly centrifuge the plate (using a pulse spin) to collect any residual ethanol from the tube walls, which is then removed and discarded. 8. Air-dry beads for 2 min. 9. Add 55 μL of elution buffer (EB) to each reaction and mix thoroughly to resuspend the beads. Vortex at 2000 rpm at room temperature for 2 min, then spin down and place on the magnetic separation device for 1 min or until the solution is completely clear. 10. Remove the eluate and store in a clean tube. Assess quantity and the quality of each WGA reaction product using the DNA Broad Range Qubit Assay and the 2100 Bioanalyzer, using the DNA 12000 kit, respectively (see Note 16). Dilute samples to 10 ng/μL in EB buffer.

3.5 Single-Cell DNA Sequencing

1. Prepare libraries for sequencing. In this example, we used 20 ng of DNA from each sample as input for generating Ion Torrent libraries. 2. Use the Ion AmpliSeq library kit 2.0 in conjunction with the Ion Ampliseq Cancer Hotspot Panel v2, to generate 207 amplicons covering 50 oncogenes. 3. Barcode the resulting amplicons using the IonXpress Barcode Adapters 1-96 kit and purify the final libraries using AMPure XP beads. 4. Quantify final libraries using the Ion Library Taqman Quantification kit and dilute to 50 pM. Template on an Ion Chef Instrument and sequence on the Ion Genestudio S5 System

Single Cell Characterization of Intratumor Heterogeneity

143

using 520 flows per chip at default parameters with the Ion 540 Chip Kit and Ion 540 Chef Kit. 5. Perform basecalling and alignment to the Ampliseq Hotspot v2 reference using Torrent Software Suite (TSS 5.0.4). Variants can be identified from the aligned reads using Torrent Variant Caller (TVC 5.0.4.0) running on the “Somatic- Low Stringency” workflow.

4

Notes 1. Our work was conducted under Mount Sinai IRB #10-1166. Tissue should be stored by operating team in a 50-mL conical tube containing 40 mL supplemented cell culture media on wet ice, and it is important to proceed to the following steps as quickly as possible to minimize tissue storage time. If overnight storage of tissue is unavoidable, store tubes on ice at 4  C and carefully assess cell viability after dissociation the following day—cells must be viable for single-cell sequencing and cell viability following tissue storage is highly variable and sampledependent. Minimum tissue weight required is highly dependent on source and quality; viable cell recovery should be assessed for every sample. 2. Optimization is required for each tumor/tissue type—both kit and dissociation program should be optimized, and if desired, cell yield and viability are not obtained, custom protocols should be considered. 3. Cell straining is facilitated by rinsing strainer with 1 mL of culture media prior to cell filtration and 15–20 mL of culture media after cells have passed through the filter. 4. Figure 4 shows a sample containing unsuitable amounts of debris, as well as an ideal single-cell suspension. If further filtration does not remove debris, optimization of live cell enrichment methods (with magnetic beads or sorting instrumentation) may be required. 5. If using frozen cells, they should be thawed to maximize viable cell recovery. Recovery in culture is recommended prior to use on Beacon. 6. The OptoSelect 3500 chip is a silicon-based nanofluidic consumable device, containing 3500 individual slots (“nanopens”) into which single cells, or groups of cells, can be isolated for analysis on the Beacon single-cell handling platform. The nanopens are fluidically isolated, such that transport in and out of the pens is by diffusion only. Cells are moved into pens (“penned”) using Optoelectronic positioning, which is a method by which cages of light are automatically generated

144

Kristin G. Beaumont et al.

Fig. 4 Left panel shows a single-cell suspension containing unsuitable amounts of cellular debris, whereas right panel shows an ideal single-cell suspension

to encompass single cells of interest, and gently drag them into the desired pens via creation of a dielectric force gradient. 7. Optoselect chip wetting is an automated process on the Beacon single-cell handling platform, wherein the chip is flooded with excess wetting reagent, heated at 50  C for 15 min, and then flushed with copious amounts of water. 8. Loading multiple cell lines on a single chip as part of the same experiment is a useful approach for adding internal controls and mitigating experimental variation. This can be accomplished by penning cells in cycles—as an example, cell type 1 can be loaded and penned into one region of the chip, followed by rinsing (three times, each with 250 μL culture media) of unpenned cells. This would then be followed by loading and penning of cell type 2 and subsequent rinsing of unpenned cells. This can be repeated several times until all 3500 pens are occupied or all cell types have been loaded to the desired cell number. It should be noted that acquisition of brightfield images between loading cycles helps to track cell type location. 9. Default parameters include leading volume of air bubble, package volume of cells and lagging volume of air bubble. Default cell package volume is 5 μL, but this can be optimized for each experiment. 10. Suggested starting voltage is 4.7 V—penning speed should not exceed 8 μm/s. 11. We acquired images every 2 hours for timelapse over 6 days in culture. 12. Other antibodies can readily be substituted—optimization will be required prior execution, as the required staining concentration on the OptoSelect chip is usually higher than in other culture environments.

Single Cell Characterization of Intratumor Heterogeneity

145

13. The fluorescence intensity can be analyzed and quantified on the software that comes as part of the Beacon platform (Assay Analyzer which is one module within Cell Analysis System package). Alternatively, since images are acquired as tiff files on the Beacon platform, these can be exported and analyzed using ImageJ/Fiji or any other image analysis tool that is desired. 14. We have worked with a variety of ATCC-sourced cell lines (including MCF7, MDA-MB-231, OVCAR3, and MT4 among others) as well as patient-derived cells (peripheral blood mononuclear cells (PBMCs) and others) to characterize phenotypes like cell migration, mitochondrial activity, cell survival/cell death, and cell infection. As long as phenotypes can be captured using the fluorescence microscopy available on the platform (4–10 magnification with brightfield, DAPI, FITC, Texas Red, or Cy5 imaging channels), then they can be analyzed using this method and linked to single cell genomic or transcriptomic data. 15. Export can be of single cells in pens of interest or, alternatively, small groups of cells, depending on desired data. When doing single-cell analysis by this method, a useful positive control is groups of cells (we used groups of 10, 100, and 1000 cells) to serve as a “bulk” analysis comparison. If plate is processed immediately after export, single-cell RNA Seq can also be performed at this step using plate-based low input library preparation methods (such as SMART-Seq). It is important to note that the pen number of the cell (which is associated with its phenotypic information) is retained at this stage (i.e., the experiment is designed to export cells by pen number into wells of a 96-well plate, so it is always defined which cell has been exported to which well). 16. Expected yield for each reaction is between 400 and 900 ng/μ L and the minimum input is 0.2 ng/μL.

Acknowledgments The authors wish to acknowledge Dr. John Martignetti and his research group for their assistance with ovarian tissue samples. References 1. Meinhold-Heerlein I, Hauptmann S (2014) The heterogeneity of ovarian cancer. Arch Gynecol Obstet 289(2):237–239. https://doi.org/10. 1007/s00404-013-3114-3 2. Schwarz RF, Ng CK, Cooke SL, Newman S, Temple J, Piskorz AM et al (2015) Spatial and

temporal heterogeneity in high-grade serous ovarian cancer: a phylogenetic analysis. PLoS Med 12(2):e1001789. https://doi.org/10. 1371/journal.pmed.1001789 3. Kohn EC, Ivy SP (2017) Whence high-grade serous ovarian cancer. American Society of

146

Kristin G. Beaumont et al.

Clinical Oncology educational book. American Society of Clinical Oncology. Annual meeting. 37:443–448. https://doi.org/10.14694/ edbk_174718 4. Winterhoff BJ, Maile M, Mitra AK, Sebe A, Bazzaro M, Geller MA et al (2017) Single cell sequencing reveals heterogeneity within ovarian cancer epithelium and cancer associated stromal cells. Gynecol Oncol 144(3):598–606. https:// doi.org/10.1016/j.ygyno.2017.01.015 5. Izar B, Tirosh I, Stover EH, Wakiro I, Cuoco MS, Alter I et al (2020) A single-cell landscape of high-grade serous ovarian cancer. Nat Med 26 (8):1271–1279. https://doi.org/10.1038/ s41591-020-0926-0

6. Winterhoff B, Talukdar S, Chang Z, Wang J, Starr TK (2019) Single-cell sequencing in ovarian cancer: a new frontier in precision medicine. Curr Opin Obstet Gynecol 31(1):49–55. https://doi.org/10.1097/gco. 0000000000000516 7. Ortega MA, Poirion O, Zhu X, Huang S, Wolfgruber TK, Sebra R et al (2017) Using singlecell multiple omics approaches to resolve tumor heterogeneity. Clin Transl Med 6(1):46. https://doi.org/10.1186/s40169-017-0177-y 8. Scholler N, Urban N (2007) CA125 in ovarian cancer. Biomark Med 1(4):513–523. https:// doi.org/10.2217/17520363.1.4.513

Chapter 9 Culturing Primary Human Mesothelial Cells Mary Mullen, Hollie Noia, and Katherine Fuh Abstract Mesothelial cells line the serosal cavities and associated organs. In order to metastasize to distant organs, ovarian tumor cells must first attach and then clear the mesothelial cells. Therefore, human primary mesothelial cells (HPMCs) are necessary to effectively study ovarian cancer metastases. Here, we describe methods to obtain HPMCs from human omentum and to culture in vitro. Key words Mesothelial cells, Primary mesothelial cells, Culture, Human omentum, Ovarian cancer

1

Introduction Mesothelial cells line the three serosal cavities (pleural, pericardial, and peritoneal) as well as the organs within these cavities. These cells provide a smooth, frictionless surface to protect moving organs and tissue from damage [1]. They are flattened, squamous-like cells that rest in a monolayer on a basement membrane supported by extracellular matrices (ECM) and stroma. Between individual mesothelial cells are tight junctions which maintain serosal function and allow for the secretion of surfactant, proteoglycans, and glycosaminoglycans [1, 2]. While mesothelial cells are derived from the mesoderm and express the mesenchymal markers vimentin and desmin, these cells also express epithelial cell markers such as cytokeratins [3]. Accordingly, they exhibit a changing phenotype consistent with an epithelial-to-mesenchymal transition (EMT). This dynamic phenotype allows these cells to participate in their many different functions [2, 4] including maintaining serosal integrity, antigen presentation, transport of fluid and cells across serosal cavities, inflammation and tissue repair, coagulation and fibrinolysis, and tumor cell adhesion, growth, and clearance [1, 2, 5, 6]. Primary human mesothelial cells (HPMC) are important in studying cancer metastases, particularly ovarian cancer metastases

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_9, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

147

148

Mary Mullen et al.

which is characterized by superficial invasion and intraperitoneal spread. During ovarian cancer spread, primary cancer cells detach and disseminate intraabdominally to sites of metastases, particularly the omentum, peritoneum, and bowel serosa [7]. These sites are lined by mesothelial cells. In order for ovarian cancer cells to efficiently form metastases, they must clear the mechanical barrier formed by mesothelial cells and expose the underlying basement membrane and stroma. Previous research suggests that there is significant crosstalk between mesothelial cells, stromal cells, and ovarian cancer cells which facilitates the unique metastases demonstrated in this disease [8, 9]. In order to understand the ovarian cancer disease model, HPMC cultures are necessary. Here we present methods to culture HPMCs from human omentum. In our experiments, omentum was obtained at the time of surgery. HPMCs are isolated from the omentum and grown in vitro. The omentum is digested in multiple steps in order to ensure isolation of HPMCs. The isolated cells proliferate for an average of 2 weeks. These HPMC cultures presumably maintain the phenotypic traits consistent with in vivo mesothelial cells allowing for accurate ovarian cancer models.

2

Materials Prepare and store all reagents at room temperature (unless otherwise indicated). 1. Dulbecco’s phosphate-buffered saline (DPBS), sterile. 2. Trypsin: 0.05% trypsin–EDTA, stored at 4  C. 3. Cell Culture Medium: Dulbecco’s Modified Eagle’s Medium (DMEM), 20% fetal bovine serum (FBS), 1% penicillinstreptomycin, 1% non-essential amino acid, 2% MEM vitamin solution (100), stored at 4  C. 4. Normal saline. 5. 50-mL conical centrifuge tubes. 6. T75 flasks. 7. Disposable forceps. 8. Disposable 10 blade scalpel. 9. 15-cm culture plate. 10. 10-mL and 1000-μL pipette. 11. Centrifuge to accommodate 50-mL conical centrifuge tubes. 12. Rotator capable of being placed in an incubator. 13. Dimethyl sulfoxide (DMSO) (for freezing cells).

Culturing Primary Human Mesothelial Cells

3

149

Methods Carry out all procedures at room temperature unless otherwise specified. These experiments should be performed in a laminar flow hood unless otherwise specified. Figure 1 summarizes the steps of the methods described below.

3.1 Tissue Procurement

1. Prior to obtaining patient tissue, it is necessary to obtain appropriate Institutional Review Board approval as well as consent from participants. 2. Omental samples can be obtained from any patient undergoing intra-abdominal surgery for benign pathology. 3. We recommend obtaining an 8  12 cm piece of omentum; however, a minimum size of 4  4 cm of omentum will be sufficient. A full omentectomy does not need to be performed. 4. The omental sample should be transported from the operating room to the lab in normal saline or DPBS. It can be transported at room temperature. 5. Omentum obtained from the operating room ideally should be used immediately. If it is not, it should be transferred to Cell Culture Medium and stored at 4  C. Omentum should be stored for no more than 24 h prior to culturing.

3.2

HPMC Isolation

1. Warm trypsin, Cell Culture Medium, and PBS in 37  C water bath. 2. Transfer 10 mL of DPBS to a 15-cm plate. 3. Transfer up to an 8  12 cm piece of omentum to the 15-cm plate (see Notes 1 and 2). 4. Cut with the scalpel and forceps into 3-mm3 segments. 5. Using the forceps, transfer the omental segments into 50-mL conical centrifuge tubes until the volume of tissue reaches approximately 20 mL in each tube (see Note 3). 6. Add 10 mL of trypsin, followed by 10 mL of DPBS to each tube, shake quickly for 1 min to mix, and place on a rotator at 10 RPM in a 37  C incubator for 15 min. 7. Centrifuge at 120 RCF for 3 min. At this point, the conical tube will contain the remaining omental tissue, supernatant, and a pellet at the bottom of the tube. The pellet is the digested omental cells. 8. Using the forceps, place the undigested omentum of each tube into a new 50-mL conical centrifuge tube, keeping the pellet undisturbed at the bottom of the original tubes. Set the new tubes aside as they will be used for the second wash (Subheading 3.3).

150

Mary Mullen et al.

Fig. 1 Summary of the steps to isolate human primary mesothelial cells (HPMCs) from patient omentum. (1) Omentum is isolated during surgery and (2) diced into 3-mm3 pieces that are (3) transferred to individual conical tubes. (4) Trypsin and PBS are added to the omental pieces for digestion. (5) After digestion and centrifugation, large tissue pieces are removed and transferred to (6) new conical tubes for secondary digestion. (7) After removal of tissue pieces, the supernatant is removed, and (8) the cells are resuspended in 1 mL culture medium

Culturing Primary Human Mesothelial Cells

151

9. Aspirate supernatant with careful attention to not disturb pellet at the bottom, repeat for all tubes (see Note 4). 10. Resuspend all cell pellets with 1 mL Cell Culture Medium each. 11. Combine all resuspended pellets with the 1000-μL pipette into one centrifuge tube (see Note 5). 12. Pipette 1 mL of resuspended cells with the 1000-μL pipette into a T75 flask (see Note 6), repeat for all resuspended cell pellets. 13. Add 15 mL of Cell Culture Medium to each T75 flask. 14. Incubate T75 flasks in 37  C incubator. 3.3 Second Wash (HPMC)

1. Collect the tubes containing omentum from step 8 of Subheading 3.2. You will repeat the steps of Subheading 3.2 with this omentum as summarized below. Add 10 mL of trypsin followed by 10 mL of PBS to each tube containing omentum, shake quickly for 1 min to mix, and place on rotator at 10 RPM in a 37  C incubator for 30 min. 2. Centrifuge at 120 RCF for 3 min. Again at this point the tube will contain remaining omental tissue, supernatant, and a pellet at the bottom of the tube. The pellet is the digested omental cells. 3. Using the forceps, transfer the remaining omental segments into a new 50-mL conical centrifuge tube keeping the pellet undisturbed at the bottom of the original tube (see Note 7). These omental samples will be discarded. 4. Aspirate supernatant with careful attention to not disturb pellet at the bottom, repeat for all tubes. 5. Resuspend all cell pellets with 1 mL of Cell Culture Medium each. 6. Pipette 1 mL of resuspended cells with the 1000-μL pipette into a T75 flask, repeat for all resuspended cell pellets. 7. Add 15 mL of Cell Culture Medium to each T75 flask. 8. Incubate T75 flasks in 37  C incubator.

3.4 Maintenance of Growing HPMCs

1. Change media for the first time when cells are 30% confluent or greater or at 2 weeks (see Note 8). Then, change media every 3–4 days. 2. Freeze when cells are 100% confluent (see Note 9) (Fig. 2). Notes regarding thawing and appropriate passage numbers are below (see Notes 10 and 11).

152

Mary Mullen et al.

Fig. 2 Mesothelial cell culture. Confluent HPMCs at 10 magnification. Once the cells are confluent, they have the characteristic polygonal appearance of HPMCs

4

Notes 1. In our lab, human omental samples were obtained from consenting patients undergoing surgery for benign gynecologic disease. 2. Use blunt dissection to remove excess fat or debris from the omental sample. 3. Usually fills about 2–8 conical tubes if using an 8  12 piece of omentum. 4. There will be a “seal” at the top of the supernatant. Use the aspirator tip to break this seal prior to aspirating the supernatant. 5. Mix the pellets together for a more homogenous mixture and less variability from one flask to the next. 6. One tube of processed omentum will result in one T75 flask of cells. 7. To discard omentum add equal parts 10% bleach solution to 50-mL conical centrifuge tubes and caps, and discard in hazardous biological waste according to EHS guidelines for hazardous pathological waste. 8. Allow at least 3 weeks to grow before discarding. 9. It is possible that there are omental fibroblasts mixed with the HPMCs (Fig. 3). Preferably there should be 0–5% of mixed cells. We will freeze up to 10% of mixed cells, but will discard

Culturing Primary Human Mesothelial Cells

153

Fig. 3 Mesothelial cell culture contaminated with fibroblasts. HPMCs contaminated with >10% fibroblast cells at 10 magnification. The fibroblast cells have a distinct appearance from the HPMCs as indicated in the image

anything that is over 10% mixed as this population of cells would not purely represent HPMCs. 10. A 10-cm plate should be frozen in Cell Culture Medium supplemented 10% DMSO in 1 mL aliquots. 11. Cells should be thawed in a 37  C water bath. They should then be added to 10 mL of Cell Culture Medium and placed into a 15-mL centrifuge tube. Spin cells at 120 RCF for 3 min, aspirate supernatant, resuspend cell pellet into 10 mL of Cell Culture media, place into a 75 cm2 cell culture flask and placed in the 37  C incubator. These cells should not be passaged more than 2–3 times. References 1. Mutsaers SE (2004) The mesothelial cell. Int J Biochem Cell Biol 36(1):9–16. https://doi. org/10.1016/S1357-2725(03)00242-5 2. Mutsaers SE (2002) Mesothelial cells: their structure, function and role in serosal repair. Respirology 7(3):171–191. https://doi.org/ 10.1046/j.1440-1843.2002.00404.x 3. Ferrandez-Izquierdo A, Navarro-Fos S, Gonzalez-Devesa M, Gil-Benso R, LlombartBosch A (1994) Immunocytochemical typification of mesothelial cells in effusions: in vivo and in vitro models. Diagn Cytopathol 10 (3):256–262. https://doi.org/10.1002/dc. 2840100313 4. Foley-Comer AJ, Herrick SE, Al-Mishlab T, Preˆle CM, Laurent GJ, Mutsaers SE (2002) Evidence for incorporation of free-floating

mesothelial cells as a mechanism of serosal healing. J Cell Sci 115(Pt 7):1383–1389 5. Jones LM, Gardner MJ, Catterall JB, Turner GA (1995) Hyaluronic acid secreted by mesothelial cells: a natural barrier to ovarian cancer cell adhesion. Clin Exp Metastasis 13(5):373–380. https://doi.org/10.1007/bf00121913 6. Cunliffe WJ, Sugarbaker PH (1989) Gastrointestinal malignancy: rationale for adjuvant therapy using early postoperative intraperitoneal chemotherapy. Br J Surg 76(10):1082–1090. https://doi.org/10.1002/bjs.1800761030 7. Lengyel E (2010) Ovarian cancer development and metastasis. Am J Pathol 177(3):1053–1064. https://doi.org/10.2353/ajpath.2010.100105 8. Rieppi M, Vergani V, Gatto C, Zanetta G, Allavena P, Taraboletti G, Giavazzi R (1999)

154

Mary Mullen et al.

Mesothelial cells induce the motility of human ovarian carcinoma cells. Int J Cancer 80 (2):303–307. https://doi.org/10.1002/(sici) 1097-0215(19990118)80:23.0.co;2-w 9. Kenny HA, Chiang CY, White EA, Schryver EM, Habis M, Romero IL, Ladanyi A, Penicka

CV, George J, Matlin K, Montag A, Wroblewski K, Yamada SD, Mazar AP, Bowtell D, Lengyel E (2014) Mesothelial cells promote early ovarian cancer metastasis through fibronectin secretion. J Clin Invest 124 (10):4614–4628. https://doi.org/10.1172/ jci74778

Chapter 10 Isolation of Normal and Cancer-Associated Fibroblasts Katarzyna Zawieracz and Mark A. Eckert Abstract Cancer-associated fibroblasts (CAFs) play important roles in regulating tumor progression, metastasis, and response to therapies. Accurately modeling the interplay between cancer cells and the tumor microenvironment (TME) requires the use of primary cells from patient samples. Here we describe methods for the isolation of both primary CAFs and fibroblasts from omental tissue using a combination of mechanical dissociation and enzymatic digestion. Primary cells can be used for functional and mechanistic studies and may be safely cryopreserved. Key words Cancer-associated fibroblasts, Fibroblasts, Cancer, Ovarian cancer, Primary cells, Isolation, Stroma, Tumor microenvironment, Mesothelial cells, Metastasis, Omentum

1

Introduction High-grade serous ovarian cancers (HGSOC) have a high proportion of stroma, with cancer-associated fibroblasts (CAFs) composing a large proportion of cells in the tumor microenvironment (TME) [1–3]. Functionally, CAFs are involved in the promotion of ovarian cancer growth and invasion by secreting a variety of chemokines, cytokines, and growth factors such as CCL5, IL-6, IL-8, HBEGF, and TGFB [4–6]. In addition, CAFs increase the aggressiveness of cancer cells by releasing factors that may induce epithelial–mesenchymal transition (EMT) and enhance metastatic abilities [7, 8]. Cancer cells can regulate the reprogramming of normal fibroblasts to CAFs by secreting TGFB and PDGF as well as by modulating the expression of specific miRNAs in normal fibroblasts [9, 10]. Due to their central roles in regulating the behavior of multiple cell types in tumors, much attention has been directed toward the isolation, validation, and experimental investigation of primary CAFs from both patient samples as well as mouse models. Fibroblasts are nonepithelial cells likely derived from the embryonic mesoderm [11]. In normal tissues, fibroblasts can be

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_10, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

155

156

Katarzyna Zawieracz and Mark A. Eckert

identified as single cells embedded in the extracellular matrix of the interstitial space. They are relatively easy to culture, as they can be derived from most human or mouse tissues after digestion and can be grown under a range of cell culture conditions [11, 12]. There are several hypotheses regarding the origins of CAFs, including epigenetic reprogramming of normal fibroblasts by cancer cells or the differentiation of tissue-resident mesenchymal stem cells [2, 11–13]. It is likely that the origins of CAFs may be cancerand tissue-type specific. It is also important to note that both normal fibroblasts and CAFs are heterogeneous populations of cells with subpopulations characterized by distinct molecular markers and disparate functions [11, 13, 14]. Identification of specific markers of CAFs and normal fibroblasts has enabled the characterization of their populations in vivo and facilitated studies of the bi-directional communication between CAFs and tumor and other cells [1, 3, 15]. The phenotypic heterogeneity of CAFs is mirrored by the wide range of markers used to characterize them, including ACTA2 (actin alpha 2, smooth muscle), vimentin, FAP (fibroblast activation protein alpha), S100A4 (also known as FSP1—fibroblast-specific protein-1), PDGFRA, PDGFRB, and DDR2 (discoidin domain receptor tyrosine kinase 2) [5, 6, 13, 14, 16]. Of these, S100A4 can be reliably used as a marker to detect quiescent, non-proliferating fibroblasts [11]. None of these markers are specific to CAFs but, when used in combination, they can serve to reliably identify the CAF population [5]. The combination of relatively simple isolation from tissues and reliable markers for live-cell sorting has enabled studies of the tumor microenvironment (TME) using these primary cells [5, 11, 13, 16–23]. Following their isolation, CAFs can be readily used to isolate DNA, RNA, and proteins to achieve a snapshot of the genomic, transcriptomic, and proteomic state of the tumor stroma. Moreover, patient-derived CAFs, along with mesothelial cells and components of extracellular matrix, can be used to construct co-culture models that recapitulate the tumor-stroma crosstalk [16–18]. Those models, developed for ovarian and other cancers, enable studies examining the effectiveness of therapies in highthroughput screening and preclinical studies [16–22]. Most protocols for the isolation of CAFs and normal fibroblasts require initial mechanical disruption of the tissue followed by digestion with trypsin, hyaluronidase, collagenase type 3, or other enzymes (Fig. 1) [1, 3, 16, 24, 25]. Here we present an optimized method to isolate CAFs as well as normal omental fibroblasts from omentum. This general protocol can be readily adapted for selection of fibroblasts by beads coated with antibodies against EpCAM to deplete epithelial tumor cells or FSP1 to positively enrich for fibroblasts [23, 26]. Both CAFs and fibroblasts can be cryopreserved for future use.

Isolation of Normal and Cancer-Associated Fibroblasts

157

Fig. 1 Isolation of CAFs from ovarian cancer tissue. This protocol uses a combination of initial mechanical dissociation with subsequent enzymatic dissociation using trypsin, hyaluronidase, and collagenase

2

Materials All reagents and tools used for the isolation protocol must be sterile. Store reagents according to manufacturer’s guidelines. The quality of fetal bovine serum (FBS) can vary between different lots of the product; try to use the same lot for your series of experiments to facilitate reproducibility.

2.1 Isolation of CAFs and Normal Fibroblasts

1. Class II biological safety cabinet. 2. Plastic petri dishes, 100 mm or 200 mm, with lid. 3. 15-ml and 50-ml propylene conical tubes. 4. Tissue culture flasks (T-25, T-75, and T-175). 5. Forceps with straight tip. 6. Sharp surgical scissors with straight tip. 7. Scalpel, 4 cm in length, size #10. 8. Orbital agitator capable of maintaining 37
C. 9. Tissue culture centrifuge capable of at least 500 rcf. 10. PBS (phosphate-buffered saline): autoclave prepared solution (see Note 1) and use only in sterile conditions with aseptic technique. 11. 0.25% Trypsin, 0.1% EDTA in HBSS without calcium, magnesium, and sodium bicarbonate.

158

Katarzyna Zawieracz and Mark A. Eckert

12. Fibroblast growth media: RPMI 1640, 20% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Store at 4
C for up to 1 month. 13. Dissociation media: Add 0.5 mg/ml of hyaluronidase and 3 mg/ml of collagenase type 3 to 20 ml of PBS. Adjust volumes based on the tissue sample size and number of tissues being processed. Dissociation media should be prepared immediately before use. 2.2 Cryopreservation of Primary Cells

1. Cell freezing container. 2. Freezing media: 10% DMSO (dimethyl sulfoxide) in FBS. Filter the freezing media with a 0.22-μm filter or 0.22-μm syringe filter. Store the media at 4
C for up to 1 month. 3. PBS: Dilute the 10 stock 1:10 with deionized water and autoclave. Use only in sterile conditions with aseptic technique. 4. 0.25% Trypsin, 0.1% EDTA in HBSS without calcium, magnesium, and sodium bicarbonate. 5. Fibroblast growth media: RPMI 1640, 20% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Store at 4
C for up to 1 month. 6. Tissue culture centrifuge capable of at least 500 rcf. 7. 0.4% Trypan Blue Solution. 8. Countess™ II FL Automated Cell Counter or other cell counting system of choice. 9. Countess™ Cell Counting Chamber Slides or other cell counting system of choice. 10. Cryogenic storage vials. 11. 80
C freezer. 12. Vapor phase cryogenic storage system.

2.3 Cryorecovery of Primary Cells

1. Class II Biological Safety Cabinet. 2. 15-ml and 50-ml propylene conical tubes. 3. Tissue culture flasks (T-25, T-75, and T-175). 4. Tissue culture centrifuge capable of at least 500 rcf. 5. Fibroblast growth media: RPMI 1640, 20% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Store at 4
C for up to 1 month. 6. Bead bath (alternatively, use a water bath).

Isolation of Normal and Cancer-Associated Fibroblasts

3

159

Methods All procedures should be performed at room temperature unless otherwise noted. Patient-derived samples should be treated as potentially infected with human pathogens and therefore work with these samples should always be performed in a biological safety cabinet (BSC) type II (or higher). Due to the risk of human-derived samples carrying infectious agents or mycoplasma, it is advised to keep a separate workplace for these samples or to thoroughly disinfect the workplace before and after the isolation procedure. All reagents and tools used should be sterile and opened only in the BSC with aseptic technique. Cells are incubated in a humidified incubator with 5% CO2 at 37
C. CAFs and normal fibroblasts can be cryopreserved and recovered with standard techniques (Subheadings 3.2 and 3.3). All tissues should be procured with Institutional Review Board (IRB) approval and de-identified. High-grade serous ovarian cancer metastatic omental tumor tissues are routinely collected during initial debulking surgeries. These tissue specimens may also be collected from interval debulking surgeries in the context of neoadjuvant chemotherapy. Adjacent normal omental tissue may also be collected from these surgeries. Tissues should be kept fully immersed in PBS or normal saline at room temperature before beginning dissociation as soon as possible (within 2 h). Total tissue available from a surgery can vary widely based on the extent of metastasis.

3.1 Isolation of CAFs and Normal Fibroblasts

1. Place the specimens of human omentum in a 100 or 200 mm plastic petri dish half-filled with sterile PBS. Adjust the size of dish to the size of processed tissue. Use sterile forceps to hold the tissue and wash it several times with PBS to remove blood. 2. Mince a portion of omentum 2–3 cm3 in size using sterile scissors and forceps until the tissue pieces are 3–5 mm3. A sterile scalpel can be used as well. For isolation of CAFs, care should be taken to avoid areas composed entirely of adipocytes (adjacent normal omentum), as determined by stiffness of tissue or appearance (Fig. 2a). Normal omental fibroblasts can be reliably isolated with modifications to the same protocol (see Note 2). 3. Place the minced omental tissue into a 50-ml conical tube filled with 10 ml of sterile PBS and 10 ml of 0.25% Trypsin, 0.1% EDTA. Place the conical tube on an orbital shaker at 37
C for 30 min (after this step, mesothelial cells can also be isolated, see Note 3). 4. For CAF isolation, continue with the tissue digestion by discarding the supernatant and adding 20 ml of dissociation media

160

Katarzyna Zawieracz and Mark A. Eckert

Fig. 2 Morphology of normal fibroblasts and CAFs. (a) Gross image of omental tumor from patient with highgrade serous ovarian cancer demonstrating extensive tumor involvement on the right side of the specimen while the left side has areas of adjacent normal tissue characterized by the presence of yellow adipocyte-rich tissue. Tissue is in 15 cm plate; scale bar ¼ 2 cm. (b) With brightfield microscopy, CAFs are much larger than normal fibroblasts and possess abundant stress fibers. Scale bar ¼ 100 μm

to the conical tube with the remaining tissue and place the conical tube back on the orbital shaker at 37
C for 6 h. 5. Allow any remaining intact tissue pieces to settle to the bottom of the tube and carefully remove supernatant containing dissociated cells to a new 50-ml conical tube. Discard remaining tissue. Centrifuge the solution containing cells at 500 rcf for 5 min. Wash pellet twice with 20 ml of fibroblast growth medium, as prepared in Subheading 2.1, item 12. 6. Resuspend fibroblasts in fibroblast growth medium and plate in a T-175 plastic culture dish (Fig. 2b; see Notes 4 and 5). Allow 1–2 days of recovery before assessing cellular morphology with microscopy.

Isolation of Normal and Cancer-Associated Fibroblasts

161

7. When cultures have reached 80–100% confluence and are actively proliferating (many mitotic figures are visible), it is time to subculture as fibroblasts are not a contact inhibited cell type (see Note 6). At this stage, cells can also be counted and frozen for future use (see Subheading 3.2). 3.2 Cryopreservation of Primary Cells

1. Once the cells have reached confluence (in T-75 or T-175 tissue culture flasks), they can be further expanded or frozen for future use (see Note 7). 2. Pre-warm the cell freezing container and cell freezing media to room temperature. 3. Wash the fibroblasts in the tissue culture flask with PBS, aspirate the PBS and add 0.25% Trypsin, 0.1% EDTA to them and return to the 37
C incubator until all cells have detached from the flask (typically less than 5 min). Inactivate the trypsin by adding at least a fivefold excess of fibroblast growth medium, prepared as in Subheading 2.1, item 12. 4. Count the cells in the suspension using your method of choice that incorporates a marker of cellular viability (e.g., trypan blue). 5. Calculate the volume of cell suspension that would contain 2  106 cells and multiply the volume by the number of vials for cryopreservation. Aliquot the calculated volume with around 10% excess added (see Note 8). 6. Centrifuge the solution with cells in suspension at 500 rcf for 5 min. Aspirate the media above the cell pellet. 7. Resuspend the cell pellet with the pre-warmed freezing media. Calculate the volume of media such that 1 ml of freezing media is added for each vial for cryopreservation. 8. Gently aliquot the cells resuspended in freezing media to cryovial tubes (1 ml per tube). Save one vial for STR profiling and mycoplasma testing. In case the cells are very precious, prepare one additional vial for thawing immediately after freezing to ensure the cryopreservation was successful (see Note 9). 9. Place the cryovial tubes in the cell freezing container stored at room temperature. Place the container with cells in the 80
C freezer. Leave the box in the freezer for at least 4 h before placing the cryovials in long-term storage in liquid nitrogen or short-term storage in a 80
C freezer.

3.3 Cryorecovery of Primary Cells

1. Remove the cryovial containing primary cells from storage in liquid nitrogen vapor phase and place it in a bead bath. 2. In a biological safety cabinet: prepare a 15-ml propylene conical tube filled with 9 ml of pre-warmed fibroblast growth media, prepared as in Subheading 2.1, item 12.

162

Katarzyna Zawieracz and Mark A. Eckert

3. Rapidly thaw the cryovial with cells as fast as possible by agitating it in the bead bath. When the majority of the cryovial contents is in liquid form (but not completely thawed), move to the biological safety cabinet and transfer the cells to the conical tube with pre-warmed media using a serological pipette. 4. Centrifuge the cell suspension at 500 rcf for 5 min to remove the DMSO. Aspirate the media from above the cell pellet. 5. Resuspend the cell pellet in fresh fibroblast growth media (5 or 10 ml, as prepared in Subheading 2.1, item 12) and plate cells in a T-25 or T-75 flask, depending on the size of cell pellet.

4

Notes 1. While autoclaving the containers with PBS, follow the guidelines provided by the manufacturer. General rules applicable to autoclaving are: do not fill the whole container with the autoclaved liquid, use containers compatible with autoclaving conditions (high pressure and temperature), and do not seal the cap of the container tightly. 2. To isolate normal omental fibroblasts use benign omentum from patients without cancer. 3. At this point, mesothelial cells can be plated. Centrifuge the solution with cells in suspension at 500 rcf for 5 min. Wash the pellet twice with fibroblast growth medium, as prepared in Subheading 2.1, item 12. Plate cells with fibroblast growth medium in a T-175 tissue culture flask. Mesothelial cells can be more reliably isolated from benign omentum and have a cobblestone appearance when confluent [16]. 4. Adjust the amount of media for plating and the flask used to the number of cells present after the isolation. If the cell pellet is barely visible, cells should be plated in a T-25 flask in 3 ml of media. For a large pellet (approximately 0.5 ml cell pellet volume), cells should be plated in a T-175 in 20 ml of media. 5. Plating cells in the dish of this size would yield a confluent monolayer of fibroblasts in ~4–6 days (about 2  107 cells). For experiments or preparing 3D co-culture models, use the cells at early passages (1–3) to minimize de-differentiation or changes in the initial phenotype of isolated fibroblasts. 6. Isolated primary fibroblasts can be cultured and reliably used in experiments until passages 5–10, assuming cells retain expression of fibroblast markers and proliferative abilities. Miltenyi magnetic separation or FACS can be used to remove immune (CD45 positive) or tumor (EpCAM positive) cells. Although there is no single marker for cancer-associated fibroblasts, they

Isolation of Normal and Cancer-Associated Fibroblasts

163

can be characterized by high expression of α-SMA, vimentin, FAP, PDGFRα, desmin, DDR2, and S100A4. Endothelial contamination can be assessed by staining for CD31 (PECAM-1) or vWF. Mesothelial cells express cytokeratin-8, calretinin, and HBME. Immune cells can be identified by expression of CD45 and epithelial tumor cells by EpCAM or CA-125 (MUC-6). 7. Make sure that before freezing cells are actively proliferating and are not confluent. 8. It is a good practice to prepare one additional vial of cryopreserved cells in order to perform validation by STR profiling and test for mycoplasma contamination with commercially available PCR tests. 9. When cryopreserving cells for future use, it is a good practice to leave one T-75 (or other size) flask of cells in culture in case cryopreservation is unsuccessful. In the case of a particularly valuable sample, it is good practice to thaw one of the freshly frozen samples to determine post-cryopreservation viability.

Acknowledgments We would like to thank all of the members of the Lengyel Ovarian Cancer Research Laboratory at the University of Chicago for their helpful comments and feedback. In particular, we thank Dr. Abir Mukherjee for his assistance. M.A.E. was supported by the Ovarian Cancer Research Alliance (OCRA) Liz Tilberis Early Career Award 650339. References 1. Curtis M, Kenny HA, Ashcroft B, Mukherjee A, Johnson A, Zhang Y, Helou Y, Batlle R, Liu X, Gutierrez N, Gao X, Yamada SD, Lastra R, Montag A, Ahsan N, Locasale JW, Salomon AR, Nebreda AR, Lengyel E (2019) Fibroblasts mobilize tumor cell glycogen to promote proliferation and metastasis. Cell Metab 29:141–155.e9. https://doi.org/ 10.1016/j.cmet.2018.08.007 2. Dasari S, Fang Y, Mitra AK (2018) Cancer associated fibroblasts: naughty neighbors that drive ovarian cancer progression. Cancers 10(11):406. https://doi.org/10.3390/ cancers10110406 3. Eckert MA, Coscia F, Chryplewicz A, Chang JW, Hernandez KM, Pan S, Tienda SM, Nahotko DA, Li G, Blazenovic I, Lastra RR, Curtis M, Yamada SD, Perets R, McGregor SM, Andrade J, Fiehn O, Moellering RE,

Mann M, Lengyel E (2019) Proteomics reveals NNMT as a master metabolic regulator of cancer-associated fibroblasts. Nature 569: 723–728. https://doi.org/10.1038/s41586019-1173-8 4. Wu RC, Wang P, Lin SF, Zhang M, Song Q, Chu T, Wang BG, Kurman RJ, Vang R, Kinzler K, Tomasetti C, Jiao Y, Shih IM, Wang TL (2019) Genomic landscape and evolutionary trajectories of ovarian cancer precursor lesions. J Pathol 248:41–50. https://doi. org/10.1002/path.5219 5. Chen X, Song E (2019) Turning foes to friends: targeting cancer-associated fibroblasts. Nat Rev Drug Discov 18:99–115. https://doi. org/10.1038/s41573-018-0004-1 6. Alkasalias T, Moyano-Galceran L, ArsenianHenriksson M, Lehti K (2018) Fibroblasts in the tumor microenvironment: shield or spear?

164

Katarzyna Zawieracz and Mark A. Eckert

Int J Mol Sci 19(5):1532. https://doi.org/10. 3390/ijms19051532 7. Wang L, Zhang F, Cui JY, Chen L, Chen YT, Liu BW (2018) CAFs enhance paclitaxel resistance by inducing EMT through the IL6/JAK2/STAT3 pathway. Oncol Rep 39: 2081–2090. https://doi.org/10.3892/or. 2018.6311 8. Zhao L, Ji G, Le X, Luo Z, Wang C, Feng M, Xu L, Zhang Y, Lau WB, Lau B, Yang Y, Lei L, Yang H, Xuan Y, Chen Y, Deng X, Yi T, Yao S, Zhao X, Wei Y, Zhou S (2017) An integrated analysis identifies STAT4 as a key regulator of ovarian cancer metastasis. Oncogene 36: 3384–3396. https://doi.org/10.1038/onc. 2016.487 9. Mitra AK, Zillhardt M, Hua Y, Tiwari P, Murmann AE, Peter ME, Lengyel E (2012) MicroRNAs reprogram normal fibroblasts into cancer-associated fibroblasts in ovarian cancer. Cancer Discov 2:1100–1108. https://doi. org/10.1158/2159-8290.CD-12-0206 10. Pang W, Su J, Wang Y, Feng H, Dai X, Yuan Y, Chen X, Yao W (2015) Pancreatic cancersecreted miR-155 implicates in the conversion from normal fibroblasts to cancer-associated fibroblasts. Cancer Sci 106:1362–1369. https://doi.org/10.1111/cas.12747 11. Kalluri R (2016) The biology and function of fibroblasts in cancer. Nat Rev Cancer 16: 582–598. https://doi.org/10.1038/nrc. 2016.73 12. Huang H, Brekken RA (2020) Recent advances in understanding cancer-associated fibroblasts in pancreatic cancer. Am J Physiol Cell Physiol 319:C233–C243. https://doi. org/10.1152/ajpcell.00079.2020 13. Sahai E, Astsaturov I, Cukierman E, DeNardo DG, Egeblad M, Evans RM, Fearon D, Greten FR, Hingorani SR, Hunter T, Hynes RO, Jain RK, Janowitz T, Jorgensen C, Kimmelman AC, Kolonin MG, Maki RG, Powers RS, Pure E, Ramirez DC, Scherz-Shouval R, Sherman MH, Stewart S, Tlsty TD, Tuveson DA, Watt FM, Weaver V, Weeraratna AT, Werb Z (2020) A framework for advancing our understanding of cancer-associated fibroblasts. Nat Rev Cancer 20:174–186. https://doi.org/10.1038/ s41568-019-0238-1 14. Franco OE, Shaw AK, Strand DW, Hayward SW (2010) Cancer associated fibroblasts in cancer pathogenesis. Semin Cell Dev Biol 21: 33–39. https://doi.org/10.1016/j.semcdb. 2009.10.010 15. Ireland LV, Mielgo A (2018) Macrophages and fibroblasts, key players in cancer chemoresistance. Front Cell Dev Biol 6:131. https://doi. org/10.3389/fcell.2018.00131

16. Kenny HA, Krausz T, Yamada SD, Lengyel E (2007) Use of a novel 3D culture model to elucidate the role of mesothelial cells, fibroblasts and extra-cellular matrices on adhesion and invasion of ovarian cancer cells to the omentum. Int J Cancer 121:1463–1472. https://doi.org/10.1002/ijc.22874 17. Shan T, Chen S, Chen X, Lin WR, Li W, Ma J, Wu T, Cui X, Ji H, Li Y, Kang Y (2017) Cancer-associated fibroblasts enhance pancreatic cancer cell invasion by remodeling the metabolic conversion mechanism. Oncol Rep 37: 1971–1979. https://doi.org/10.3892/or. 2017.5479 18. Horie M, Saito A, Yamaguchi Y, Ohshima M, Nagase T (2015) Three-dimensional co-culture model for tumor-stromal interaction. J Vis Exp 15(5):353–364. https://doi.org/10. 3791/52469 19. Langhans SA (2018) Three-dimensional in vitro cell culture models in drug discovery and drug repositioning. Front Pharmacol 9:6. https://doi.org/10.3389/fphar.2018.00006 20. Chitty JL, Skhinas JN, Filipe EC, Wang S, Cupello CR, Grant RD, Yam M, Papanicolaou M, Major G, Zaratzian A, Da Silva AM, Tayao M, Vennin C, Timpson P, Madsen CD, Cox TR (2020) The miniorgano: a rapid high-throughput 3D coculture organotypic assay for oncology screening and drug development. Cancer Rep 3:e1209. https://doi.org/10.1002/cnr2.1209 21. Kenny HA, Lal-Nag M, Shen M, Kara B, Nahotko DA, Wroblewski K, Fazal S, Chen S, Chiang CY, Chen YJ, Brimacombe KR, Marugan J, Ferrer M, Lengyel E (2020) Quantitative high-throughput screening using an organotypic model identifies compounds that inhibit ovarian cancer metastasis. Mol Cancer Ther 19:52–62. https://doi.org/10.1158/ 1535-7163.MCT-19-0052 22. Truskey GA (2010) Endothelial cell vascular smooth muscle cell co-culture assay for high throughput screening assays for discovery of anti-angiogenesis agents and other therapeutic molecules. Int J High Throughput Screen 2010:171–181. https://doi.org/10.2147/ IJHTS.S13459 23. Sharon Y, Alon L, Glanz S, Servais C, Erez N (2013) Isolation of normal and cancerassociated fibroblasts from fresh tissues by Fluorescence Activated Cell Sorting (FACS). J Vis Exp:e4425. https://doi.org/10.3791/ 4425 24. Castello-Cros R, Cukierman E (2009) Stromagenesis during tumorigenesis: characterization of tumor-associated fibroblasts and stromaderived 3D matrices. Methods Mol Biol 522:

Isolation of Normal and Cancer-Associated Fibroblasts 275–305. https://doi.org/10.1007/978-159745-413-1_19 25. Proia DA, Kuperwasser C (2006) Reconstruction of human mammary tissues in a mouse model. Nat Protoc 1:206–214. https://doi. org/10.1038/nprot.2006.31 26. Costa A, Kieffer Y, Scholer-Dahirel A, Pelon F, Bourachot B, Cardon M, Sirven P, Magagna I,

165

Fuhrmann L, Bernard C, Bonneau C, Kondratova M, Kuperstein I, Zinovyev A, Givel AM, Parrini MC, Soumelis V, VincentSalomon A, Mechta-Grigoriou F (2018) Fibroblast heterogeneity and immunosuppressive environment in human breast cancer. Cancer Cell 33:463–479.e10. https://doi.org/10. 1016/j.ccell.2018.01.011

Chapter 11 Isolation of Primary Normal and Cancer-Associated Adipocytes from the Omentum Abir Mukherjee Abstract The omentum is the metastatic site for intra-abdominal cancers such as colon, stomach, and ovarian (where it is the primary site for metastasis). Adipocytes are the primary cell type of the omentum, and they aid in cancer cell proliferation, migration, and invasion. Therefore, systematic characterization of adipocyte–cancer cell interactions will help in understanding the metastatic spread of intra-abdominal cancer. Here, a detailed mechanical-enzymatic digestion method describes the isolation of both normal and cancerassociated adipocyte from omental tissues. Key words Omentum, Primary adipocytes, Collagenase, Cancer-associated adipocytes, Ovarian cancer, Metastasis

1

Introduction The omentum is a visceral adipose depot with endocrine tissue-like functions [1]. It is approximately 20  15  2 cm in size [2] and covers the bowel like a curtain (Fig. 1a). In addition to fat storage, it plays a significant role in obesity and insulin resistance [3]. The omental tissue, enveloped by mesothelial cells, is primarily composed of white adipocytes and is interspersed with other stromal cells. Distinct from subcutaneous depots, the omentum has secondary lymphoid organs called the “milky spots” and carries out immune surveillance/defense in the peritoneal cavity [4, 5]. These milky spots are arranged around a glomeruli-like network of capillaries and are comprised of B and T lymphocytes, macrophages, mast cells, and other stromal cells [6]. The omentum is the primary site for ovarian cancer (OVCA) metastasis. Cytokines secreted by the omental adipocytes aid in the “homing” of the OVCA cells to this tissue [7]. The cancer cells induce lipolysis in adipocytes, and as a consequence, the adipocyte diameter (size) reduces with increased omental colonization

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_11, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

167

168

Abir Mukherjee

Fig. 1 Isolation of human primary normal and cancer-associated adipocytes. (a) Full-length normal greater omentum. (b) Cancerous omentum from an ovarian cancer patient. Cancerous areas are outlined in green, and adjacent adipose tissue is outlined in yellow. (c) Minced omental tissue (described in Subheading 3.1, step 5). (d) Estimation of tissue volume after centrifugation (described in Subheading 3.1, step 6). (e) Digestion of the omental tissue in a rotatory shaker. (f) Isolated primary omental adipocytes, floating on top

[8]. OVCA cells subsequently upregulate a lipid transporter protein CD36 and a chaperon protein fatty acid-binding protein 4 (FABP4) to enhance lipid oxidation for energy [7, 9, 10]. Collagens are the major extracellular matrix proteins (ECM) in both normal and cancerous omentum. However, because of

Isolation of Adipocytes

169

activated stroma in the cancerous omentum [11, 12], there is a significant increase in fibronectin levels in the ECM of the omentum [8]. Regardless of the changes in ECM, digestion of the tissue with a crude collagenase solution is sufficient to release primary adipocytes. Typically, collagenase I preparation is a mixture of collagenase (isolated from Clostridium histolyticum, a grampositive bacteria) and proteases (such as clostripain and trypsin). Clostridium histolyticum generates two types of collagenase [13, 14] and is superior to its mammalian counterpart, as it cleaves the collagen helix at multiple sites [15, 16]. Therefore, for most tissue digestion protocols, collagenase I from Clostridium histolyticum is preferred. Herein, the method describes the use of collagenase I to isolate adipocytes from both normal and cancer tissue.

2

Materials All procedures involving tissue digestion must be carried out in tissue culture hood (class II biosafety hoods) with proper aseptic techniques. All solutions should be pre-warmed to 37  C, and all tools should be sterile before use.

2.1 Tissue Procurement

1. An institutional review board (IRB)-approved protocol is required to collect human omental tissue. Omentum is collected from donors/patients undergoing surgeries for either cancer- or non-cancer-related conditions, as per approved (IRB) protocol. 2. After resection, keep the tissue moist at all times (put 10–20 ml PBS in sample containers while transferring the tissue from the operating room to pathology and subsequently to the lab). Dry and/or ischemic tissue reduces the yield of viable adipocytes. 3. Ideally, the time between tissue resection and the adipocyte isolation should be kept as short as possible (10–15 min).

2.2 General Materials

1. Phosphate-buffered saline (PBS 1), calcium- and magnesium-free. 2. Syringes. 3. 0.22- and 0.8-μm syringe filters. 4. Plastic culture dishes—100 or 150 mm. 5. Scalpels. 6. Forceps. 7. Plastic Erlenmeyer flasks with vented cap. 8. Autoclavable funnel—100 mm diameter. 9. Nylon mesh—250 μm pore size, 50% open area.

170

Abir Mukherjee

10. Individually wrapped, sterile, serological 10, 50 ml; conical tubes—15 and 50 ml.

pipettes—5,

11. Class II biosafety cabinet. 12. Rotary incubator maintained at 37  C. 13. Rocking shaker. 14. Incubator maintained at 37  C and 5% carbon dioxide. 15. Calcein-AM and/or Bodipy 493/503. 2.3

Solutions

1. Adipocyte culture media: Prepare a 1:1 ratio of Dulbecco’s modified eagle media (DMEM with 4.5 g/L glucose, L-glutamine, sodium pyruvate) and Ham’s F12 nutrient mixture. Add 1% penicillin and 1% streptomycin to the media (see Note 1). Weigh 0.5 g of fatty acid free, fraction V (see Note 2) bovine serum albumin (BSA, 0.1%, w/v final concentration) and add to a 20 ml aliquot of culture media in a conical tube. Gently rock the mixture on a rocking shaker to facilitate dissolution (see Note 3). Filter the BSA solution using a syringe filter (0.22 μm) to sterilize and add to 500 ml culture media. 2. Adipocyte digestion solution: Prepare 0.2% (w/v) collagenase I solution in adipocyte culture media. Gently rock the mixture to dissolve. Filter through a 0.22-μm syringe filter to sterilize.

2.4 Nylon Mesh-Lined Funnel

1. To separate adipocytes from the digested tissue sample, prepare a nylon mesh-lined funnel. Cut nylon mesh (250 μm pore size, 50% open area) to appropriate dimensions (for example, 25 cm  25 cm) and insert into a funnel (100 mm diameter) that can be autoclaved. 2. Hold mesh in place with a metal paper clip. Autoclave the mesh-lined funnel and maintain sterility until use.

3

Methods The methods below describe the isolation of adipocytes from human omental tissue, with or without tumors. Of note, primary adipocytes (compared to differentiated 3T3L1 cells) are sensitive to culture conditions (avoid temperature and carbon dioxide level fluctuations, shaking while in culture, and repeated pipetting).

3.1 Isolation of Adipocytes from Benign Tissue

1. Wipe down the tissue culture hood and all the reagent tubes/ bottles and appliances with 70% ethanol. 2. Take tissue from the sample container and put it in an appropriate dish. Using a forceps, hold the tissue in place and pipette PBS onto the tissue to rinse. Wash several times (approximately 10 ml/per wash) to remove the blood and mesothelial cells (see Note 4). Alternately, add PBS to the dish containing the tissue and gently scrape off mesothelial cells using a scalpel.

Isolation of Adipocytes

171

3. Transfer the tissue to a 10- or 15-cm dish (Fig. 1a). 4. Add a sufficient amount of PBS to keep the tissue moist (see Note 5). Using scalpels, remove any cauterized tissue, large veins (to reduced endothelial cells), and highly desmoplastic areas devoid of adipocytes. 5. Mince tissue using scalpels (or scissors) into small (mm size) pieces (so that it can be pipetted using a 50-ml pipette (Fig. 1c). 6. Add a sufficient amount of PBS (to form a slurry) and pipette out minced tissue (using a 50 ml pipette) into 50-ml tube(s) to collect the entire minced tissue. Centrifuge the tube(s) at 300  g for 5–10 s and note the volume of the tissue (without PBS) (Fig. 1d). Remove the PBS using a pipette (the minced adipose tissue will be suspended in the PBS) and transfer/ decant minced tissue into an Erlenmeyer flask (see Note 6). 7. Add an equal volume of adipocyte digestion solution as the volume of the tissue in the flask. Cap the flask and digest tissue in a rotary incubator maintained at 80–100 rpm and 37  C for 1 h (see Note 7, Fig. 1e). 8. After digestion, decant the flask’s contents and let the adipocytes pass through the nylon mesh (described in Subheading 2.4, see Note 8). The filtrate contains both the stromal vascular fraction and adipocytes. The undigested tissue (left in the mesh) is a good source for omental fibroblasts (see Note 9). 9. Add adipocyte culture media to the filtrate to dilute the collagenase. Invert the tube(s) 3–4 times to wash the adipocytes and centrifuge at 200  g for 5 min. Remove the media with a 10-ml serological pipette. 10. Repeat step 9 until the media is primarily devoid of red blood cells (the color of the media is not red anymore, Fig. 1f). 11. Adipocytes will separate from the media and float on top (Fig. 1f). Remove the adipocyte fraction and transfer it into a fresh 15- or 50-ml tube for staining (step 12) or cell culture vessel (steps 13 and 14). 12. If desired, the primary adipocytes can be stained with calceinAM and Bodipy 493/503 to check for viability and purity of the adipocytes, respectively. Stain primary adipocytes in suspension by adding calcein-AM (at a final concentration of 1–5 μM) and/or Bodipy 493/503 (at a final concentration of 2–5 μg/ml) for 20 min. Centrifuge adipocytes at 200  g for 5 min, remove media containing dye using a pipette and replace with PBS or adipocyte culture media (Fig. 2). 13. Primary omental adipocytes are terminally differentiated; therefore, they do not proliferate. However, they can be

172

Abir Mukherjee

Fig. 2 Lipid content and viability of primary human omental adipocytes. Neutral lipid content: (a) Bright field and (b) Bodipy 493/503 staining of isolated adipocytes (described in Subheading 3.1, step 12). Viability estimation: (c) Bright field and (d) Calcein AM staining of adipocytes (described in Subheading 3.1, step 12)

maintained in culture for 3–5 days either in adipocyte culture media or in media used to culture fibroblasts (see Notes 9 and 10). 14. To generate adipocyte conditioned media, culture the adipocyte in adipocyte culture media (at a 1: 5 adipocyte packed cell volume to culture media volume). After 24–48 h, separate the adipocytes from the media by centrifugation at 200  g for 5 min and subsequently remove adipocytes using a pipette. Filter the conditioned media using a 0.8-μm syringe filter to remove any remaining adipocytes or cell debris. 3.2 Isolating Cancer-Associated Adipocytes

1. Dissect out and remove areas with visible and palpable tumors (Fig. 1b). Proceed with steps 1 through 5 (as described above) to digest the adjacent fat tissue and isolated primary cancerassociated adipocytes (see Note 11).

Isolation of Adipocytes

173

2. In step 6, instead of centrifugation for 5–10 s, centrifuge at 300  g for 5 min. 3. Follow steps 7 through 11 to obtain cancer-associated adipocytes.

4

Notes 1. The addition of penicillin and streptomycin significantly reduces bacterial contamination in adipocytes prepared from tissues obtained from the gross pathology room. 2. BSA fraction V or fatty acid free BSA is used to reduce/eliminate contaminating lipid species found in regular BSA. This is necessary for lipid signaling and/or lipidomic studies. 3. Avoid vortexing or vigorously shaking BSA solution to prevent frothing. 4. To isolate omental mesothelial cells, centrifuge the first PBS wash (before tissue is minced) at 300  g for 5 min, suspend the cell pellet in RPMI supplemented with 10% FBS, and culture cells in T-175 flasks in an incubator maintained at 37  C with 5% carbon dioxide. Of note, no RBC lysis is carried out before plating. Add additional media to the flask after 3 days without removing the old media. After 5 days, change to new media. Allow the cells to grow till confluency before plating for experiments. 5. Excess PBS makes it challenging to mince the tissue, so use just enough PBS to keep the tissue moist. 6. After centrifugation in Subheading 3.1, step 6, the tissue gets compacted; therefore it is easier to pipette out the PBS and decant the minced tissue than to remove the tissue directly with a pipette. 7. Digestion time is dependent on the size and the collagen content of the tissue. Therefore, increase digestion time in Subheading 3.1, step 7 in case of incomplete digestion after 1 h. In addition, pipetting (once every 20 min) with a 50-ml pipette facilitates digestion. 8. The undigested tissue might obstruct the flow-through and delay filtration of adipocytes. In such an event, remove the undigested tissue rather than waiting for the entire (slow dripping) solution to flow through. This will reduce the exposure of the adipocytes to the digestion solution beyond digestion time and decrease cell death. 9. Prepare a 10 fibroblast digestion solution (20 ml) composed of 10 ml PBS, 10 ml penicillin/streptomycin, 300 mg collagenase type III (dialyzed lyophilized powder), and 47.6 mg

174

Abir Mukherjee

hyaluronidase (dialyzed lyophilized powder). Sterilize using a 0.22-μm syringe filter. Dilute the solution to 1 using fibroblast culture media (DMEM, supplemented with 10% FBS, 1 MEM non-essential amino acids, and 1 MEM vitamins). To isolate fibroblasts, digest the residual tissue from step 8 using fibroblast digestion solution (1) for 6 h to overnight. After digestion, centrifuge at 300  g for 5 min and suspend the cell pellet in fibroblast culture media and grow the cells in T175 flasks in an incubator maintained at 37  C with 5% carbon dioxide. 10. Adipocyte viability of greater than 90% assessed by calcein-AM staining is indicative of successful isolation. 11. The method for isolating cancer-associated adipocytes is similar to normal adipocytes. Both methods use collagenase I to digest the omentum. Removal of visible and palpable tumors will prevent contamination by cancer cells. Cancer-associated fibroblasts and mesothelial cells can be isolated from cancerous omental tissue following the same protocol described (above) for normal omentum. References 1. Nieman KM, Romero IL, Van Houten B, Lengyel E (2013) Adipose tissue and adipocytes support tumorigenesis and metastasis. Biochim Biophys Acta 1831(10):1533–1541. https:// doi.org/10.1016/j.bbalip.2013.02.010 2. Lengyel E (2010) Ovarian cancer development and metastasis. Am J Pathol 177 (3):1053–1064. https://doi.org/10.2353/ ajpath.2010.100105 3. Collins D, Hogan AM, O’Shea D, Winter DC (2009) The omentum: anatomical, metabolic, and surgical aspects. J Gastrointest Surg 13 (6):1138–1146. https://doi.org/10.1007/ s11605-009-0855-1 4. Rangel-Moreno J, Moyron-Quiroz JE, Carragher DM, Kusser K, Hartson L, Moquin A, Randall TD (2009) Omental milky spots develop in the absence of lymphoid tissueinducer cells and support B and T cell responses to peritoneal antigens. Immunity 30 (5):731–743. https://doi.org/10.1016/j. immuni.2009.03.014 5. Hall JC, Heel KA, Papadimitriou JM, Platell C (1998) The pathobiology of peritonitis. Gastroenterology 114(1):185–196. https://doi. org/10.1016/s0016-5085(98)70646-8 6. Krist LF, Eestermans IL, Steenbergen JJ, Hoefsmit EC, Cuesta MA, Meyer S, Beelen RH (1995) Cellular composition of milky spots in the human greater omentum: an immunochemical and ultrastructural study.

Anat Rec 241(2):163–174. https://doi.org/ 10.1002/ar.1092410204 7. Nieman KM, Kenny HA, Penicka CV, Ladanyi A, Buell-Gutbrod R, Zillhardt MR, Romero IL, Carey MS, Mills GB, Hotamisligil GS, Yamada SD, Peter ME, Gwin K, Lengyel E (2011) Adipocytes promote ovarian cancer metastasis and provide energy for rapid tumor growth. Nat Med 17(11):1498–1503. https:// doi.org/10.1038/nm.2492 8. Pearce OMT, Delaine-Smith RM, Maniati E, Nichols S, Wang J, Bohm S, Rajeeve V, Ullah D, Chakravarty P, Jones RR, Montfort A, Dowe T, Gribben J, Jones JL, Kocher HM, Serody JS, Vincent BG, Connelly J, Brenton JD, Chelala C, Cutillas PR, Lockley M, Bessant C, Knight MM, Balkwill FR (2018) Deconstruction of a metastatic tumor microenvironment reveals a common matrix response in human cancers. Cancer Discov 8(3):304–319. https://doi.org/10.1158/ 2159-8290.CD-17-0284 9. Ladanyi A, Mukherjee A, Kenny HA, Johnson A, Mitra AK, Sundaresan S, Nieman KM, Pascual G, Benitah SA, Montag A, Yamada SD, Abumrad NA, Lengyel E (2018) Adipocyte-induced CD36 expression drives ovarian cancer progression and metastasis. Oncogene 37(17):2285–2301. https://doi. org/10.1038/s41388-017-0093-z 10. Mukherjee A, Chiang CY, Daifotis HA, Nieman KM, Fahrmann JF, Lastra RR, Romero

Isolation of Adipocytes IL, Fiehn O, Lengyel E (2020) Adipocyteinduced FABP4 expression in ovarian cancer cells promotes metastasis and mediates carboplatin resistance. Cancer Res 80 (8):1748–1761. https://doi.org/10.1158/ 0008-5472.CAN-19-1999 11. Kenny HA, Chiang CY, White EA, Schryver EM, Habis M, Romero IL, Ladanyi A, Penicka CV, George J, Matlin K, Montag A, Wroblewski K, Yamada SD, Mazar AP, Bowtell D, Lengyel E (2014) Mesothelial cells promote early ovarian cancer metastasis through fibronectin secretion. J Clin Invest 124(10):4614–4628. https://doi.org/10. 1172/JCI74778 12. Eckert MA, Coscia F, Chryplewicz A, Chang JW, Hernandez KM, Pan S, Tienda SM, Nahotko DA, Li G, Blazenovic I, Lastra RR, Curtis M, Yamada SD, Perets R, McGregor SM, Andrade J, Fiehn O, Moellering RE, Mann M, Lengyel E (2019) Proteomics reveals NNMT as a master metabolic regulator of cancer-associated fibroblasts. Nature 569 (7758):723–728. https://doi.org/10.1038/ s41586-019-1173-8

175

13. Matsushita O, Jung CM, Katayama S, Minami J, Takahashi Y, Okabe A (1999) Gene duplication and multiplicity of collagenases in Clostridium histolyticum. J Bacteriol 181 (3):923–933. https://doi.org/10.1128/JB. 181.3.923-933.1999 14. Yoshihara K, Matsushita O, Minami J, Okabe A (1994) Cloning and nucleotide sequence analysis of the colH gene from Clostridium histolyticum encoding a collagenase and a gelatinase. J Bacteriol 176(21):6489–6496. https://doi.org/10.1128/jb.176.21.64896496.1994 15. French MF, Bhown A, Van Wart HE (1992) Identification of Clostridium histolyticum collagenase hyperreactive sites in type I, II, and III collagens: lack of correlation with local triple helical stability. J Protein Chem 11(1):83–97. https://doi.org/10.1007/BF01025095 16. French MF, Mookhtiar KA, Van Wart HE (1987) Limited proteolysis of type I collagen at hyperreactive sites by class I and II Clostridium histolyticum collagenases: complementary digestion patterns. Biochemistry 26 (3):681–687. https://doi.org/10.1021/ bi00377a004

Part IV Model Systems

Chapter 12 Isolation of Fallopian Tube Epithelium for Assessment of Cilia Beating Frequency (CBF) Angela Russo and Joanna E. Burdette Abstract The fallopian tube epithelium (FTE) plays a critical role in reproduction and the genesis of ovarian cancer. The FTE columnar cells present with hair-like structures named “cilia” that are required for normal FTE function. Impairment of ciliary motion can lead to infertility, and it is influenced by hormonal signaling and endocrine disrupting compounds. Studying how cilia beating changes in response to these compounds is critical for understanding FTE physiology and pathology. In this protocol, we describe methods for isolating human fallopian tube epithelium, oviduct (murine equivalent of fallopian tube) epithelium, and ovaries. In addition, we describe methods for imaging and measuring cilia beating frequency using highresolution time-lapse imaging. Key words Fallopian tube epithelium, Ovulation, Cilia beating frequency, Ovary, Oviduct

1

Introduction The human fallopian tubes play a critical role in reproduction. The fallopian tubes are the site of fertilization and are also involved in the transport of the sperm, egg, and embryo. Tubal secretions are rich in glycoproteins and growth factors that regulate fertilization and embryo development [1–3]. Ciliary movement is pivotal to the fallopian tube function [2], and when ciliary motion is disrupted through trauma to the tubal epithelium by sexually transmitted infections, pelvic inflammatory disease, or environmental insults, fertility is also affected [4–8]. Ovarian hormones regulate the tubal epithelium during the menstrual cycle [9, 10] and impact the tubal epithelial structure and the expression of cilia genes [2, 11]. Estrogen stimulates ciliogenesis and increases cilia beating frequency while high levels of progesterone cause deciliation and decrease beating frequency [12, 13]. Elevated testosterone also regulates cilia gene expression and motion in human fallopian tube epithelium [14]. Human

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_12, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

179

180

Angela Russo and Joanna E. Burdette

fallopian secretory epithelial cells are a source of origin of highgrade serous ovarian cancer, and the outgrowth of the secretory population, as compared to the ciliated cells, is thought to be an early event in tumorigenesis [15, 16]. However, the relationship between ciliogenesis and ovarian cancer development is not known. Some studies suggest that improper cilia function may impede clearance of the fallopian tube epithelium, causing accumulation of oxidative stress [17] that may lead to DNA damage and tumorigenesis. Therefore, studying cilia movements is pivotal to understand fallopian tube normal function and pathogenesis. Herein we describe a method for isolation and ex vivo study of fallopian tube epithelium.

2 2.1

Materials Cell Culture

1. Mice CD1 or any strain of interest [N ¼ 6 per condition/ treatment, (see Note 1)]. 2. Fallopian tube tissues from patients (N ¼ 6 per condition/ treatment). 3. 70% ethanol.

2.2

Media

1. Ovary and oviduct dissection media: Leibovitz media with 1% penicillin/streptomycin. 2. Murine oviduct epithelial (MOE) cells media: α-MEM with ribonucleosides, deoxynucleosides, supplemented with 2 mM L-glutamine, 10% fetal bovine serum, 2 μg/ml epidermal growth factor (EGF), ITS (5 μg/ml insulin, 5 μg/ml transferrin, 5 ng/ml sodium selenite), 1 mg/ml gentamicin, and 18.2 μg/ml β-estradiol, 0.1% penicillin/streptomycin solution [18]. 3. Fallopian tube dissecting media: DMEM/F-12 medium supplemented with 1% penicillin/streptomycin, 50 mg/l gentamicin, 1% FBS. 4. Growth media: 50% αMEM, 50% F-12, supplemented with 0.3% BSA, 1 mg/ml bovine fetuin, 5 μg/ml insulin, 5 μg/ml transferrin and 5 μg/ml selenium, 10 mIU/ml folliclestimulating hormone (FSH), 1% penicillin/streptomycin solution, 50 mg/l gentamicin (see Note 2).

2.3 Equipment and Plates

1. Spinning disk inverted microscope, with a 100 silicone oil objective, incubation chamber, camera, and software capable of capturing fast time lapse images. 2. Matek P35G-1.5-14-C dishes. 3. ImageJ-Fiji Software.

Cilia Beating Frequency

181

4. Laminar flow hood. 5. Temperature-controlled centrifuge. 6. Dissecting microscope. 7. Inverted tissue culture microscope to check tissue health and cilia pulsing. 8. Coverslips. 9. 10-cm plates (to dissect tissue in). 10. 12-well plate. 11. Millicell 0.4 μm pore. 12. Tweezers. 13. Surgical scissors. 14. Scalpel.

3

Method

3.1 Isolation of Fallopian Tube Epithelium

1. Obtain human fallopian tubes from the hospital (see Note 3) after approval by the institutional review board (IRB) of the university of interest. 2. Wash the fallopian tubes in 70% ethanol and place in 10-cm dish containing 10 ml of pre-warmed fallopian tube dissecting media (n ¼ 6 tissues from six different patients in total, obtained at different times and processed separately, are recommended for sufficiently powered studies, although more may be necessary depending on the perturbation of interest). 3. Open the fallopian tube by inserting scissors in the tube. Opening the tube will expose the lumen. Once the lumen is exposed, the epithelium will become visible and can be removed with tweezers and surgical scissors (Fig. 1). 4. Cut the epithelium from the muscular wall and connective tissue, which is the compact and hard tissue that was originally the external part of the tube. 5. Make sections of the epithelium of similar size, about 2  2 mm, using scissors. 6. Flatten the small sections and position them on 0.4-μm pore Millicell inserts transwells (3 biopsies per well). 7. Place the transwells in 12-well plates containing 600 μl of growth media to create an air–liquid interface (no liquid is added inside the transwell). The FTE preparation will look really transparent at this time (Fig. 1). The day after check under an inverted microscope for cilia pulsing using a 10 or 20 objective to make sure that the tissue has healthy epithelium. Treatments can be added in the bottom well, such as

182

Angela Russo and Joanna E. Burdette

Fig. 1 Isolation of human fallopian tube epithelium (hFTE). (a) The tube is opened to expose the epithelium. (b) The epithelium is excised from the connective tissue. (c) The epithelium is cut into small 2  2 mm sections. (d) The epithelium is transferred into transwells

hormones or small molecule inhibitors of the pathway of interest (see Note 4). Cilia beating can be measured as described in Subheadings 3.3 and 3.4. 8. Every other day, remove 200 μl of growth media and replace with 200 μl of fresh media. The collected media can be used for ELISA, Western blots or frozen. 9. Maintain in a humidified atmosphere containing 5% CO2 at 37  C. 3.2 Isolation of Murine Ovaries and Oviducts

1. Obtain IACUC approval for animal studies. 2. Sacrifice female mice at 16 days of age (n ¼ 6 mice per condition or treatment). 3. Make an incision at the base of the abdomen and remove the reproductive tract (Fig. 2). 4. Isolate murine oviducts and ovaries using a dissecting scope and carefully remove, bursa, uterus, and fat pads [19]. 5. Make a cut between the ovary and the oviduct and slide the ovary out of the bursa, which is a membranous sac surrounding the ovary. 6. Separate the uterus from the oviduct. The oviduct is connected to the uterus but looks convoluted as compared to the uterus, which is straight.

Cilia Beating Frequency

183

Fig. 2 Isolation of mouse oviduct and ovary. (a) Murine reproductive system extending from the cervix to the kidney. (b) Isolated reproductive tract with oviduct, ovary, uterus, and fat pad under the dissecting microscope. (c) Image of oviduct and ovary isolated. (d) Flattened epithelium isolated from the oviduct

7. Transfer the ovaries and the oviducts into a 10-cm dish containing 12 ml of ovary and oviduct dissection media. 8. Cut the ovaries in half and put two halves into a transwell (see Note 5). Transfer the transwells into wells of a 12-well plate containing 600 μl of growth media to create an air–liquid interface (no liquid is added inside the transwell). 9. Cut the oviduct longwise, flatten and move into a transwell (one oviduct per transwell). Transfer the transwells into wells of a 12-well plate containing 600 μl of MOE media to create an air–liquid interface (no liquid is added inside the transwell). The oviduct preparation will look really transparent at this time (Fig. 2). This tissue will bear cilia beating that can be measured as described in Subheadings 3.3 and 3.4. 10. The ovaries and the oviduct can be positioned on the same transwell or separate transwells (see Note 6). When together, secretion from one tissue will influence the other tissue response. Treatments can be added in the bottom well, such as hormones or small molecule inhibitors of the pathway of interest (see Note 4). 11. Maintain in a humidified atmosphere containing 5% CO2 at 37  C. 3.3 Capturing Cilia Beating

1. Turn on the spinning disk inverted microscope and camera as indicated by the manual (see Note 7). 2. Remove neutral density (ND) filters from transmitted light path.

184

Angela Russo and Joanna E. Burdette

3. Select the 100 silicon oil objective and bright field. 4. Turn on incubated chamber and place lens heater on objective. Equilibrate to 5% CO2 and 37  C. 5. Add silicone oil on the objective and transfer the oviduct or human fallopian tube epithelium sample from the transwell to Mattek dish, add a drop of media and position a coverslip on top of the sample. Add a weight to flatten the tissue (see Note 8). 6. Add silicone immersion oil on the objective and place the Mattek dish with the sample on stage. 7. Focus to detect the epithelium. 8. Perform Kohler alignment for proper condenser alignment in an area away from the sample. 9. Select the microscope capture settings for diascopic imaging. 10. Crop the camera field of view to 100  100 pixels. In camera setting select 2–3 ms/frame. 11. Acquire images using fast time lapse (memory capture). Capture a total of 5000 frames. 3.4

Measuring CBF

1. Use image acquisition software to open the image and determine frames per second from the number of frames captured and elapsed time (we obtain these values by clicking image properties). 2. Close the image acquisition software and open Fiji/ImageJ (can be downloaded for free at https://imagej.net/Fiji). 3. Open the same image file previously opened with the acquisition software by using Fiji and a tool bar will appear. 4. From the tool bar, select the fifth box from the left, which is a line, and drop down to select segmented line. 5. Draw a segmented line midway between the base and the ends of the section of cilia. Double click to end the segment (Fig. 3). 6. From the top of screen, go to Image and select “stacks” first, and then select “reslice.” 7. A window will appear. Click ok for the following settings: output spacing to 1000, avoid interpretation, rotate 90 . 8. A kymograph will be generated. Zoom the kymograph to View 100% to increase the size of the picture. 9. Draw a single line on the kymograph from one peak to the next peak of the waveform (Fig. 3). 10. From Analyze, select Measure and report the length of the line that is the distance between two peaks. 11. Cilia beating frequency (CBF) is calculated as follows: frames per second/length of the line between two peaks.

Cilia Beating Frequency

185

Fig. 3 Kymograph generation and measurements. (a) Images showing cilia. (b) Magnification of cilia with selection at the base of the cilia to generate the kymograph. (c) Kymograph showing the cilia waves over time. The line drawn between two consecutive peaks of a waveform is shown in yellow. (d) Measurement of the line in (c) obtained by selecting Analyze and Measure. In a red box is shown the length that is the crucial value required for CBF calculations

12. Record the CBF and analyze four additional movies (5 total) from separate fields for the same patient specimen. Average the CBF values. 13. Tissue sections from different patients are cultured, treated, and analyzed the same way to obtain independent experiments for statistical analysis based on the experimental design (comparisons of interest, etc.).

4

Notes 1. For our studies, we use 16-day-old mice because we want pre-pubertal mice that we induce to ovulate in vitro. However, older mice can be used to address different questions. 2. Growth media can be made in advance with all ingredients except fetuin and FSH. Store at 4  C for up to 1 month. Fetuin and FSH are added fresh to an aliquot of the media when needed. 3. Human fallopian tubes are transported on ice from the surgery site, and the epithelium can be isolated the same day or kept at 4  C for up to 1 day. The epithelium should feel very soft;

186

Angela Russo and Joanna E. Burdette

however, some patients may have a dry and hard epithelium that is not good for isolation and may not be suitable for cilia beating. Having the surgical indication may help to identify samples that will have viable and healthy epithelium. However, depending on the scientific question, normal versus diseased fallopian tube may be chosen. 4. Estrogen (1 nM) increases CBF, whereas progesterone (10 nM) and testosterone (2 nM) reduce CBF. Basal CBF observed with 0.1 nM estrogen stimulation is about 20 Hz [13, 14]. 5. Ovaries isolated from mice need to be used the same day. The ovaries are cut in half to favor hormone release and exposure to media, and transferred from dissecting media into a transwell containing growth media in the bottom well. 6. When co-culturing ovary and fallopian tube or oviducts, keep in mind that the ovary is viable for about 8–10 days. After that time, the structure of the ovary collapses and does not look very tight. The fallopian tube cultured alone should remain viable for 14–20 days. The murine oviduct and ovary can be kept in MOE media for experiments lasting less than 24 h. 7. After treatments of the epithelium with hormones or small molecules, the tissue is transported to the spinning disk microscope for capturing videos on a heat block pre-heated at 37  C. Cilia movements depend on temperature and therefore the tissue needs to be kept at 37  C before and during image capturing. 8. We have found that a washer works well to weigh down the tissue.

Acknowledgments We thank the Northwestern University Center for Advanced Microscopy (supported by NCI CCSG P30 CA060553 awarded to the Robert H. Lurie Comprehensive Cancer Center) for their assistance with the cilia imaging. We also acknowledge the support from R01CA240301, UH3ES029073 and CA240423. References 1. Hunter RH, Einer-Jensen N, Greve T (2005) Somatic cell amplification of early pregnancy factors in the fallopian tube. Ital J Anat Embryol 110(2 Suppl 1):195–203 2. Lyons RA, Saridogan E, Djahanbakhch O (2006) The effect of ovarian follicular fluid

and peritoneal fluid on Fallopian tube ciliary beat frequency. Hum Reprod 21(1):52–56 3. Buhi WC, Alvarez IM, Kouba AJ (2000) Secreted proteins of the oviduct. Cells Tissues Organs 166(2):165–179

Cilia Beating Frequency 4. Bouyer J, Coste J, Shojaei T, Pouly JL, Fernandez H, Gerbaud L et al (2003) Risk factors for ectopic pregnancy: a comprehensive analysis based on a large case-control, population-based study in France. Am J Epidemiol 157(3):185–194 5. Cooper MD, McGraw PA, Melly MA (1986) Localization of gonococcal lipopolysaccharide and its relationship to toxic damage in human fallopian tube mucosa. Infect Immun 51(2):425–430 6. Cooper MD, Rapp J, Jeffery-Wiseman C, Barnes RC, Stephens DS (1990) Chlamydia trachomatis infection of human fallopian tube organ cultures. J Gen Microbiol 136(6):1109–1115 7. McGee ZA, Jensen RL, Clemens CM, TaylorRobinson D, Johnson AP, Gregg CR (1999) Gonococcal infection of human fallopian tube mucosa in organ culture: relationship of mucosal tissue TNF-alpha concentration to sloughing of ciliated cells. Sex Transm Dis 26(3):160–165 8. McGee ZA, Johnson AP, Taylor-Robinson D (1981) Pathogenic mechanisms of Neisseria gonorrhoeae: observations on damage to human fallopian tubes in organ culture by gonococci of colony type 1 or type 4. J Infect Dis 143(3):413–422 9. Critoph FN, Dennis KJ (1977) The cellular composition of the human oviduct epithelium. Br J Obstet Gynaecol 84(3):219–221 10. Novak J (1928) Die menstruation und ihre sto¨rungen. Springer, Berlin. 2 p. l., 93, 1 p. p 11. Lyons RA, Saridogan E, Djahanbakhch O (2006) The reproductive significance of human Fallopian tube cilia. Hum Reprod Update 12(4):363–372

187

12. Critoph FN, Dennis KJ (1977) Ciliary activity in the human oviduct. Br J Obstet Gynaecol 84(3):216–218 13. Zhu J, Xu Y, Rashedi AS, Pavone ME, Kim JJ, Woodruff TK et al (2016) Human fallopian tube epithelium co-culture with murine ovarian follicles reveals crosstalk in the reproductive cycle. Mol Hum Reprod 22(11):756–767 14. Jackson-Bey T, Colina J, Isenberg BC, Coppeta J, Urbanek M, Kim JJ et al (2020) Exposure of human fallopian tube epithelium to elevated testosterone results in alteration of cilia gene expression and beating. Hum Reprod 35(9):2086–2096 15. Kim J, Park EY, Kim O, Schilder JM, Coffey DM, Cho CH et al (2018) Cell origins of highgrade serous ovarian cancer. Cancers 10(11):433 16. Zhang S, Dolgalev I, Zhang T, Ran H, Levine DA, Neel BG (2019) Both fallopian tube and ovarian surface epithelium are cells-of-origin for high-grade serous ovarian carcinoma. Nat Commun 10(1):5367 17. Coan M, Rampioni Vinciguerra GL, Cesaratto L, Gardenal E, Bianchet R, Dassi E et al (2018) Exploring the role of fallopian ciliated cells in the pathogenesis of high-grade serous ovarian cancer. Int J Mol Sci 19(9):2512 18. King SM, Quartuccio SM, Vanderhyden BC, Burdette JE (2013) Early transformative changes in normal ovarian surface epithelium induced by oxidative stress require Akt upregulation, DNA damage and epithelial-stromal interaction. Carcinogenesis 34(5):1125–1133 19. Endsley MP, Moyle-Heyrman G, Karthikeyan S, Lantvit DD, Davis DA, Wei JJ et al (2015) Spontaneous transformation of murine oviductal epithelial cells: a model system to investigate the onset of fallopianderived tumors. Front Oncol 5:154

Chapter 13 Ex Vivo Ovarian Culture to Model the Initial Metastasis in Ovarian Cancer Matthew Dean Abstract Being able to accurately model metastasis is an important tool in cancer research. Several in vitro and ex vivo models have been developed to model metastasis from the ovary to the omentum, the most frequent metastatic site after leaving the ovary. However, the recent discovery that high-grade serous ovarian cancer (HGSOC) can originate in the fallopian tube and then metastasize to the ovary has necessitated the development of assays that can quantify the adhesion of tumor cells to the ovary. Here we describe a protocol for accessing the adhesion of fluorescent cells to mouse ovaries. This assay can be used to investigate the role of ovarian function, hormones, and adhesion molecules in metastasis of cancer cells originating in the fallopian tube to the ovary, an important step in the progression of HGSOC. Key words Ovarian cancer, Metastasis, Colonization, Ovary, Attachment

1

Introduction Metastasis is an important step in progression of any cancer [1]. After reaching a metastatic site, tumor cells must first attach to the tissue. After attachment, the tumor cells alter the local microenvironment by invading the underlying tissue, recruiting cancer-associated fibroblasts, and remodeling the extracellular matrix [2]. Highlighting the importance of metastasis in ovarian cancer patient survival, when ovarian cancer patients are diagnosed at Stage I/II, the survival rate is >70%. Unfortunately, most cases are diagnosed at Stage III/IV, when the survival rate is only 20–40%. Preclinical research indicates that physical adhesion of cancer cells to target organs may be a potential therapeutic target [3–5]. Thus, it is important to develop experimental models that mimic this first metastatic step of any cancer that can be used to evaluate the role of specific molecules or biological factors in adhesion of cancer cells.

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_13, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

189

190

Matthew Dean

Historically, it was assumed that all cases of epithelial ovarian cancer originated in the ovarian surface epithelium. After leaving the ovary, tumor cells colonized many peritoneal organs, with the omentum being one of the most common sites [5]. However, it is now clear that many cases of high-grade serous ovarian cancer (HGSOC) originate in the fallopian tube epithelium. Evidence for a fallopian tube origin of HGSOC includes transcriptomic and proteomics similarities between HGSOC tumors and the fallopian tube epithelium [6, 7], the reduced frequency of HGSOC after salpingectomy [8], and the multiple transgenic mouse models of oviduct-derived ovarian cancer [9, 10]. Even when HGSOC originates in the fallopian tube, colonization of the ovary is a frequent and important step in disease progression [11, 12]. In patients, large tumors usually form in the ovary, not the fallopian tube, which is the reason for the historical diagnosis of “ovarian” cancer. In mice, injecting tumorigenic cells into the bursa results in tumorigenesis, while injecting the same number of cells into the peritoneal space does not [11]. The reasons that fallopian tube-derived tumor cells preferentially colonize the ovary are likely multifaceted, but rupture of the ovary during ovulation and secretion of hormones and signaling factors by the ovaries and tumor cells have been implicated [9, 11, 13, 14]. Accurate ex vivo and in vivo models that accurately and efficiently recapitulate attachment to the ovary are important tools to increase our understanding of metastasis from the fallopian tube to the ovary. To that end, we recently developed an ex vivo adhesion assay that allows investigators to quantify the adhesion of cancer cells to mouse ovaries. Factors that can be manipulated in this system include the status of the ovary (e.g., cyclic or postmenopausal), presence of relevant hormones, or function of various proteins through the addition of inhibitors or by comparing different cell lines. Using this assay, we have shown that tumor cells attach to the wound produced during ovulation, that Wnt4 produced by the tumor cells increases ovarian attachment, and that multicellular tumor spheroids can attach to the ovary [9, 11, 15].

2 2.1

Materials Mouse Ovaries

1. Mouse ovaries. Strain, age, reproductive status, etc. can be varied depending on the needs of the experiment (see Note 1). 2. 50-mL falcon tubes. 3. Leibovitz’s L-15 media with 1
penicillin-streptomycin. 4. 70% ethanol or isopropanol. 5. Surgical scissors. 6. Tweezers.

Modeling Attachment of Tumor Cells to the Ovary

191

7. Forceps. 8. Scalpel and blades. 9. Dissecting microscope. 10. 10-cm dishes. 11. Dissecting hood. 2.2

Cell Culture

1. Monoclonal fluorescent cell line of interest. Cells constitutively expressing a fluorescent protein (e.g., GFP or RFP) are preferred (see Note 2). However, cells can be tagged with commercially available CellTracker™ dyes (Subheading 2.3). If you are comparing the attachment of two different cell lines, it needs to be confirmed that the cells are equally bright, so that differences in observed attachment is not due to one cell line being more visible. 2. Appropriate cell culture media. 3. Trypsin/EDTA. 4. Phosphate-buffered saline (PBS). 5. Cell culture-treated flasks or plates.

2.3 Labeling Cells with CellTracker™ Dye

1. Non-fluorescent cell line of interest, if fluorescent cell line is not available (see Note 2). 2. CellTracker™ Dye. Dyes with different excitation and emission characteristics are available. If using CellTracker™ Dye, use a dye appropriate for your microscope. For example, Red CMTPX Dye has an excitation/emission spectra of 577/602 nm. 3. Serum-free media. 4. Sterile PBS. 5. Dimethyl sulfoxide (DMSO).

2.4 Cell Attachment to Ovaries

1. Orbital shaker.

2.5 Counting and Imaging Cells Attached to Ovaries

1. Inverted fluorescent microscope attached to computer with digital camera.

2. 37  C incubator.

2. Microfuge tubes. 3. 24-well plates. 4. Cell media. 5. 1-mL tuberculin syringes. 6. Computer with ImageJ (https://imagej.nih.gov/ij/) installed.

192

3

Matthew Dean

Methods

3.1 Isolation of Mouse Ovaries

1. Sacrifice mice using CO2 inhalation and cervical dislocation per approved IACUC guidelines. If the experiment involves comparing the attachment of ovaries in different states (e.g., cycling and postmenopausal), collect each type of ovary in a separate falcon tube (see Note 1). 2. Collect ovaries using a sterile technique (see Note 3). Wet abdominal area of each mouse with 70% ethanol and exteriorize the reproductive tract. 3. Remove the ovaries as quickly as possible. It is preferable to quickly remove ovaries with excess tissue, e.g., by removing it with the adjacent fat pat and part of the uterus. Lift the ovary using the adjacent fat pad and tweezers, so that the ovary is not damaged. Cut the reproductive tract through the uterus. Extraneous tissue can be removed later under a dissecting microscope (Fig. 1a, b). 4. Place tissue in a 50-mL falcon tube containing pre-warmed L-15 media containing 1
penicillin-streptomycin. Place the ovary in the media as quickly as possible to minimize the chance of contamination. 5. Transport ovaries to the lab.

Fig. 1 Steps to collect ovaries. (a) After exteriorizing the reproductive tract, lift the ovary using the adjacent fat pad so the ovary is not damaged. Cut the uterus with sterile scissors, moving the ovary, oviduct, part of the uterus, and any miscellaneous fat (b) into L-15 media. In a sterile hood and with the help of a dissecting microscope, the ovary can be removed from the bursa and other extraneous tissue. (c) It is easy to damage the ovary by picking it up with tweezers. Therefore, move the ovary to a 1.5-mL microfuge tube by pushing it onto the side of a scalpel blade as shown

Modeling Attachment of Tumor Cells to the Ovary

3.2 Removal of Ovaries from the Bursa

193

1. Transfer contents of falcon tubes to 10-cm dishes. 2. If reusing the same tools that were used to remove the ovaries, move them to a new falcon tube containing fresh 70% ethanol. 3. Under a dissecting microscope in a biosafety cabinet, carefully remove each ovary from the bursa and trim any extraneous tissue. Be careful not to accidentally damage the ovary, as cells will preferentially attach to any areas with exposed extracellular matrix. 4. Add 900 μL of complete media appropriate for the cell line you are using 1.5-mL microfuge tubes. 5. Move one ovary to each microfuge tube. Be careful to not damage the ovaries while moving them. An efficient way to move them is to use tweezers to push them onto the side of a scalpel blade (Fig. 1c). 6. If the experiment includes the addition of hormones or inhibitors, they can be added to each tube at this stage. Note the final volume of each tube will be 1 mL. 7. Store the microfuge tubes containing the ovaries in a 37  C incubator while collecting the cells (below). The ovaries are stable for at least 3 h.

3.3 Collecting Cells Stably Expressing a Fluorescent Protein

1. Aspirate media from each cell line and then rinse with 1
PBS. 2. Collect cells with 1
trypsin. After cells detach, collect with media and count. 3. Centrifuge cells at 160
g for 10 min and resuspend cells at 300,000 cells/mL. 4. Add 100 μL of cell suspension (i.e., 30,000 cells) to each microfuge.

3.4 Collecting and Labeling Cells with Fluorescent CellTracker™ Dye

1. Using monoclonal cells that stably express a fluorescent protein is easiest (see Note 2). However, if fluorescent cells are unavailable, cells can be temporarily tagged with CellTracker™ Dye. 2. Dissolve CellTracker™ in DMSO at an appropriate stock concentration as recommended by the manufacture. This should be optimized before performing an experiment (see Note 2). 3. Remove culture media from cells. 4. Wash with 1
PBS. 5. Collect cells with trypsin, centrifuge at 160
g for 10 min, and resuspend at 300,000 cells/mL in the FBS-free media. 6. Add stock CellTracker™ dye (final concentration varies by specific CellTracker™ dye being used. Check manufacturer’s recommendation and optimize before performing assay). 7. Incubate at 37  C for 30 min. 8. Centrifuge at 160
g for 10 min and aspirate media with CellTracker™ dye.

194

Matthew Dean

Fig. 2 Schematic of 1.5-mL tube containing one mouse ovary and fluorescently tagged ovarian cancer cells. Tube is incubated at 37  C overnight on an orbital shaker to provide an opportunity for cells to attach to the ovary

9. Resuspend in 5 mL of 1
PBS. Centrifuge at 160
g for 10 minutes and aspirate PBS. Repeat 4–5 times to remove all CellTracker™ dye. 10. After the final centrifugation, resuspend cells in media appropriate for the cell line at 300,000 cell/mL. 11. Add 100 μL of cell suspension (i.e., 30,000 cells) to each tube. 3.5 Attachment of Fluorescent Cells to Ovaries

1. Each tube should now contain one ovary, 30,000 fluorescently tagged cells, and any additional treatments as appropriate (Fig. 2). 2. Lay microfuge tubes on their side on a rocker shaker at 37  C. The rotations per minute (r.p.m.) can be optimized for each cell line (see Note 4). 3. Incubate for 24 h.

3.6 Counting Cells Attached to Each Ovary

1. The next day, add 1 mL of media to new 1.5-mL microfuge tubes. 2. Cut the tip of a 1-mL pipette tip large enough that the ovaries can easily pass through. Use this tip to pipette the ovary out of each microfuge tube. Move the ovary to a new microfuge tube with 1 mL of media in order to separate ovaries from non-attached cells. Repeat this step three times to eliminate cells that are not attached to the ovaries. If, upon visualizing the ovaries (Subheading 3.7), there are many fluorescent cells in the well that are not attached to the ovary, you may need to wash the ovaries more. 3. With a 1-mL pipette and cut tip, move each ovary to a well of a 24-well plate. 4. Using an inverted fluorescent microscope, visualize each ovary.

Modeling Attachment of Tumor Cells to the Ovary

195

5. Carefully count the number of cells observed on the surface of each ovary. Ovaries can be visualized at 4–10
magnification (see Note 5). However, the example images (Subheading 3.7) need to be at a low enough magnification that the entire ovary can be captured in a single image. 6. Using a tuberculin syringe, flip the ovary over and count the number of cells attached to the other size. Add the number of cells counted on each side to obtain a total number for each ovary. This is an important step as the number of cells attached to each side of the ovary can vary (see Note 6). 3.7 Generating Images of Cells Attached to the Ovary

1. Capture representative images of the ovary. Use a magnification low enough to capture entire ovary in one image, usually 2.5–4
. Capture fluorescent images of each ovary using correct filters for the fluorescent tag. Then carefully capture a Brightfield image showing the silhouette of the ovary. It is important that the ovary does not move between images. Removing most of the fluid from the well, leaving just enough to keep the ovary wet, will help prevent the ovary from moving between images. 2. Dispose of ovaries in biohazard per standard laboratory protocols. 3. Open both the fluorescent image and the Brightfield image using ImageJ. 4. Select the fluorescent Image ! Adjust ! Threshold.

image.

Then

click

5. Adjust the threshold values for hue, saturation, and brightness, so that only pixels positive for fluorescence and above background are selected. Record the values for hue, saturation, and brightness, so the same values can be used in every image from an experiment. If the cells are tagged with red fluorescence, it may be useful to change the threshold color to white. 6. Click “select” and press Control + C to copy the selected pixels. 7. Select the Brightfield image. Press Control + V to paste the selected pixels onto the Brightfield image. 8. Save the new Overlay image as a different image, as to not overwrite the original Brightfield image (Fig. 3).

4

Notes 1. Mouse ovaries. On the day of the experiment, mouse ovaries will be needed. For our previous experiments, we used ovaries from CD-1 mice between 16 and 18 days of age. At this age, the ovaries are consistent in structure, lacking large follicles, or corpus lutea. This is ideal for experiments testing the effects of

196

Matthew Dean

Fig. 3 Example of Brightfield image (left) of mouse ovary with GFP-labeled cells attach. The GFP-labeled cells are visible under fluorescence (center). The pixels positive for GFP have been copied and pasted on top of the Brightfield image to produce the overlay image (right)

specific hormones or inhibitors on ovarian adhesion. We have also used a scalpel blade to wound the convex surface of the ovary in order to mimic the wound produced during ovulation without the hormonal changes associated with ovulation. If wounding the ovarian surface, it is important to practice so a small and consistent cut can be achieved every time. Ovaries from other strains of mice, transgenic mice, or from treated mice should be compatible with this assay, but the assay would need to be optimized accordingly. For example, it may be desirable to use ovaries from mice on different days of the reproductive cycle, from mice induced to superovulate, or from mice treated with 4-vinylcyclohexene diepoxide (VCD) to induce menopause. Our experience suggests that a minimum of four to six ovaries should be used per treatment condition. 2. Fluorescently tagged cell lines. You need to ensure that enough cells are available the same day as the mouse ovaries are ready. It is usually best to plate the cells in a large flask or plate so that there will be enough cells, even if the cells are less confluent than expected. Grow cells under normal conditions, in a humidified chamber at 37  C. Cells must be fluorescently tagged, so that they can be visualized when attached to the mouse ovary. Typically, it is better to generate cells that stably express a fluorescent protein, such as red fluorescent protein (RFP). Conversely, cells can be labeled with a fluorescent CellTracker™ Dye. If using Tracker Dye, the protocol should be optimized beforehand based on the manufacturer’s instructions to maximize the fluorescent signal. The final concentration of the CellTracker™ dye and the duration of time the cells are exposed to the CelllTracker™ needs to be optimized to ensure the cells are bright enough to see with a fluorescent microscope but does not kill the cells. If comparing the adhesion of two different cell lines, it is important to confirm that both cells lines are equally bright using flow cytometry before performing the experiment. If one cell line is significantly

Modeling Attachment of Tumor Cells to the Ovary

197

brighter than the other, observed differences could be due to how many cells can be visualized, not how many are attached to each ovary. 3. Ovaries must remain sterile. The abdomen of the mouse should be wetted with 70% ethanol or isopropanol to disinfected the area and the minimize hair loss. Surgical tools should be kept in 70% ethanol or isopropanol when not in use and it is best to wear a facemask. The L-15 media should contain 1
penicillinstreptomycin. Do not open the tube with media unless necessary. 4. Ideally between 10 and 100 cells per ovary allows for accurate quantification. If too many or too few cells attach to each ovary, quantification can be difficult. The r.p.m. and number of cells per tube can be optimized for each cell line. We have found that 40 r.p.m. with 30,000 cells per tube usually allows an appropriate number of cells to attach to each ovary, but this may vary with different cell lines. 5. The number of cells attached to each ovary should be counted as the ovary is visualized. The cells can be counted using 2.5–10
objective. At lower magnification (e.g., 2.5 or 4
), the cells attached to one entire side of the ovary can be counted without moving the ovary. At a higher magnification (e.g., 10
) only part of the ovary can be visualized, and hence the ovary will need to be moved as the cells are counted. However, a higher magnification can make it easier to see cell lines that are not very bright and to differentiate individual cells when many cells are close together. 6. It is important to count cells on both sides of the ovary. The number of cells per side can vary due to slight damage while isolating the ovary. More cells tend to attach near the hilus because this represented a “wound” on the ovarian surface that cannot be avoided. References 1. Klein CA (2013) Selection and adaptation during metastatic cancer progression. Nature 501(7467):365–372. https://doi.org/10. 1038/nature12628 2. Davidson B, Trope CG, Reich R (2014) The role of the tumor stroma in ovarian cancer. Front Oncol 4(104):1–11. https://doi.org/ 10.3389/fonc.2014.00104 3. Sawada K, Mitra AK, Radjabi AR et al (2008) Loss of E-cadherin promotes ovarian cancer metastasis via α5-integrin, which is a therapeutic target. Cancer Res 68(7):2329–2339. https://doi.org/10.1158/0008-5472.CAN07-5167

4. Steeg PS (2016) Targeting metastasis. Nat Rev Cancer 16(4):201–218. https://doi.org/10. 1038/nrc.2016.25 5. Tan DSP, Agarwal R, Kaye SB (2006) Mechanisms of transcoelomic metastasis in ovarian cancer. Lancet Oncol 7(11):925–934. https:// doi.org/10.1016/S1470-2045(06)70939-1 6. Marquez RT, Baggerly KA, Patterson AP et al (2005) Patterns of gene expression in different histotypes of epithelial ovarian cancer correlate with those in normal fallopian tube, endometrium, and colon. Clin Cancer Res 11(17):6116–6126. https://doi.org/10. 1158/1078-0432.CCR-04-2509

198

Matthew Dean

7. Coscia F, Watters KM, Curtis M et al (2016) Integrative proteomic profiling of ovarian cancer cell lines reveals precursor cell associated proteins and functional status. Nat Commun 7:12645. https://doi.org/10.1038/ ncomms12645 8. Falconer H, Yin L, Gro¨nberg H, Altman D (2015) Ovarian cancer risk after salpingectomy: a nationwide population-based study. J Natl Cancer Inst 107(2):dju410. https://doi. org/10.1093/jnci/dju410 9. Russo A, Czarnecki A, Dean M et al (2018) PTEN loss in the fallopian tube induces hyperplasia and ovarian tumor formation. Oncogene 37(15):1976–1990 10. Perets R, Wyant GA, Muto KW et al (2013) Transformation of the fallopian tube secretory epithelium leads to high-grade serous ovarian cancer in Brca;Tp53;Pten models. Cancer Cell 24(6):751–765. https://doi.org/10.1016/j. ccr.2013.10.013 11. Dean M, Jin V, Russo A et al (2019) Exposure of the extracellular matrix and colonization of

the ovary in metastasis of fallopian tube derived cancer. Carcinogenesis 40(1):41–51. https:// doi.org/10.1093/carcin/bgy170 12. Labidi-Galy SI, Papp E, Hallberg D et al (2017) High grade serous ovarian carcinomas originate in the fallopian tube. Nat Commun 8(1):1093. https://doi.org/10.1038/ s41467-017-00962-1 13. Dean M, Davis DA, Burdette JE (2017) Activin A stimulates migration of the fallopian tube epithelium, an origin of high-grade serous ovarian cancer, through non-canonical signaling. Cancer Lett 391:114–124. https://doi. org/10.1016/j.canlet.2017.01.011 14. Yang-Hartwich Y, Gurrea-Soteras M, Sumi N et al (2014) Ovulation and extra-ovarian origin of ovarian cancer. Sci Rep 4:6116. https://doi. org/10.1038/srep06116 15. Dean M, Jin V, Bergsten TM et al (2019) Loss of PTEN in fallopian tube epithelium results in multicellular tumor spheroid formation and metastasis to the ovary. Cancers 11(6):884. https://doi.org/10.3390/cancers11060884

Chapter 14 In Vivo and Ex Vivo Analysis of Omental Adhesion in Ovarian Cancer Elizabeth I. Harper and Tyvette S. Hilliard Abstract In vivo and ex vivo analyses of omental adhesion in ovarian cancer (OvCa) are necessary to understand the dynamics of OvCa metastasis. Here we describe methods to analyze OvCa omental adhesion, including in vivo and ex vivo fluorescent imaging, advanced microscopy, and histological techniques. The use of fluorescently tagged OvCa cells allows for omental tumor visualization and quantification in adhesion and tumor studies. Additionally, advanced microscopy modalities allow for visualization and multiplexed analysis of OvCa omental adhesion. Second harmonic generation microscopy permits the visualization and analysis of omental collagen, specifically the tumor-associated collagen signature that forms as the tumor progresses. Scanning electron microscopy is used for the observation of microscopic details between OvCa cells and the omentum, such as tunneling nanotubes or microvilli. Histological methods are used to investigate several intratumoral properties including visualizing tumor structure using hematoxylin and eosin stain; quantifying collagen with Masson’s trichrome stain; analyzing collagen structure with a collagen hybridizing peptide; and identifying a number of markers including, but not limited to proliferation, immune cell types, adhesion molecules, and fibroblasts with immunohistochemistry. Both the in vivo and ex vivo imaging modalities and subsequent analysis can provide insight into the interaction of metastasizing OvCa cells and the omentum. Key words Omentum, Ovarian cancer, Second harmonic generation microscopy, Scanning electron microscopy, Immunohistochemistry

1

Introduction Ovarian cancer (OvCa) is the deadliest gynecologic cancer, largely due to most patients being diagnosed after the tumor cells have spread to distant metastatic sites. OvCa metastasizes to multiple sites throughout the peritoneal cavity with an affinity to the omentum, an endocrine organ rich in adipocytes, immune cells, and extracellular matrix (ECM). The ECM is composed mainly of collagens I and III and is covered by a basement membrane and

Pamela K. Kreeger (ed.), Ovarian Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 2424, https://doi.org/10.1007/978-1-0716-1956-8_14, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

199

200

Elizabeth I. Harper and Tyvette S. Hilliard

mesothelial cell monolayer [1]. This chapter discusses several methods, described below, to analyze omental adhesion of ovarian tumor cells using murine models, microscopy, and histology. 1.1 Murine In Vivo Models

In vivo models are integral for analyzing the interaction between OvCa cells and the omentum. To model OvCa metastasis, OvCa cells are injected intraperitoneally (IP) to allow for circulation throughout the peritoneal cavity which mimicks exfoliation of tumor cells from the primary tumor of the ovary or fallopian tube [2]. There are several syngeneic mouse models that can be used to study OvCa omental adhesion. For the purpose of this chapter, the C57Bl/6 mouse strain with ID8Trp53
/
-RFP cells, a syngeneic mouse ovarian surface epithelial tumor cell line with a Red Fluorescent Protein (RFP) tag, will be discussed unless otherwise described (Fig. 1; see Note 1; [3, 4]).

Fig. 1 (a) Young (Y; 3–6 months) and aged (A; 20–23 months) mice injected with ID8-RFP cells intraperitoneally, then sacrificed at the study endpoint, 8 weeks. Peritoneal cavity was opened, and fluorescently imaged. (b) Individual organs that have been dissected from mice at the end of the tumor study, 8 weeks [3]

Analysis of Ovarian Cancer Omental Adhesion

1.2 Microscopy: Second Harmonic Generation Microscopy

201

Second harmonic generation (SHG) microscopy is a nonlinear imaging modality for fibrillar collagen (Fig. 2a, b). SHG can be used in conjunction with two photon excitation fluorescence (TPEF); however, SHG is preferred over TPEF for collagen visualization since it does not require exogenous tags. Additionally, SHG does not require an extensive fixation process like other imaging modalities and can image collagen embedded in tissues without

Fig. 2 Second harmonic generation (SHG) microscopy of the translucent region of a normal murine omentum at 25 (a) and 100 (b). The open spaces, or fenestrations, are a normal part of omental anatomy, which give the omentum its characteristic look as the abdominal “filter.” SHG with two photon emission fluorescence (TPEF) microscopy to image collagen remodeling around an OvCa tumor at 3 weeks post-injection, in both a 2D slice of the tumor (c) and 3D composite of the entire z-stack (d), where gray is collagen, red is RFP-tagged OvCa cells, and green is the collagen hybridizing peptide with Alexa Fluor 647 (psuedocolored)

202

Elizabeth I. Harper and Tyvette S. Hilliard

requiring any decellularization process. Multiple analyses can be conducted using SHG microscopy that can provide vital information on the interaction of ovarian tumor cells and the omentum. Images obtained using SHG can be used to identify the tumorassociated collagen signature (TACS) which is defined by linearization and alignment of collagen fibers and the deposition of collagen near the tumor [5]. Additionally, anisotropy, a measurement of fiber alignment, can be determined using the FibrilTool plugin in ImageJ (see ref. 6). Furthermore, SHG combined with fluorescence microscopy can be used to visualize fluorescent tumor cell adhesion and colonization to the omentum. The addition of a collagen hybridizing peptide (CHP) can be used to visualize how ovarian tumor cells alters the organization of omental collagen (Fig. 2c, d). The CHP binds to denatured collagen or anywhere the triple helix of the collagen molecule has unraveled. There are several reasons that collagen unravels, including heat or mechanical stress; however, the primary cause in and around tumors is enzymatic digestion [1]. The CHP is commercially available conjugated to biotin, 5-FAM, sCy7.5, and Cy3 (see Note 2). 1.3 Microscopy: Scanning Electron Microscopy

Scanning electron microscopy (SEM) uses a highly focused beam of electrons to measure sample topography resulting in highresolution images at high magnification. The interaction between ovarian tumor cells and the microenvironment can be investigated by fixing the tissue after an adhesion experiment (in vivo or ex vivo) and imaging with SEM. While requiring significantly more tissue preparation than SHG, SEM imaging allows for a much higher resolution and magnification that can reveal omental structure including microvilli and the interaction between adhered OvCa cells and the omentum (Fig. 3).

1.4

Often, the tumor microenvironment plays an important role in the intratumoral landscape. Crosstalk between metastatic sites and tumor cells impacts how the tumor grows, what the tumor expresses, and whether or not the metastatic tumor is established. The best way to get a snapshot of the interior of the tumor is through histology. To get a general overview of omental tumors, hematoxylin and eosin (H&E) staining can be used to stain cell nuclei purple (hematoxylin) and cell cytoplasm pink (eosin), allowing for analysis of cell morphology and the ability to distinguish between tumor cells and normal cells in the microenvironment (Fig. 4a). Masson’s trichrome stains the nuclei and the cytoplasm similar to H&E while also staining connective tissue in a bright blue, allowing for visualization and analysis of collagen (Fig. 4b). The omentum is a very collagenous organ, and collagen is an important structural component of solid tumors. To further investigate collagen, a collagen hybridizing peptide (CHP) can be used to identify denatured collagen, or collagen that is being remodeled

Histology

Analysis of Ovarian Cancer Omental Adhesion

203

Fig. 3 Scanning electron microscopy (SEM) images of a normal murine omentum at 1000 (a) and 5000 (b), and a mouse omentum (red arrows) covered in OvCa cells (yellow arrows), fixed and imaged at the end of a 24-h ex vivo adhesion experiment at 1000 (c) and 5000 (d)

by the tumor cells (Fig. 4c). Lastly, immunohistochemistry can be used to stain a wide range of markers in the cancer cells and surrounding tissue, identifying proliferating cells, immune cells, fibroblasts, and much more (Fig. 4d). The possibilities are limited only by the availability of antibodies to the intended target. 1.5

Ex Vivo Adhesion

In addition to in vivo models, ex vivo assays can also be very useful in exploring the adhesion of OvCa cells to the omentum. Instead of performing experiments in mice, the omentum is dissected out of the mice and then experimentally manipulated. To study adhesion of OvCa cells to the omentum, tissues are pinned to silicone-coated dishes and incubated with fluorescently tagged OvCa cells (Fig. 5). Additionally, this assay can be done with the peritoneum, as shown

204

Elizabeth I. Harper and Tyvette S. Hilliard

Fig. 4 ID8 murine omental tumors harvested at 8 weeks post-injection, formalin-fixed, paraffin-embedded, sectioned, and stained with various histological stains. (a) Hemotoxylin and eosin (H&E) stain, marking cell nuclei (purple) and cell cytoplasm (pink). (b) Masson’s trichrome stain, marking cell nuclei (purple), cell cytoplasm (pink), and collagen (blue). (c) Collagen hybridizing peptide–biotin conjugate (CHP-B) that stains denatured collagen molecules (brown) and counterstained with hemotoxylin (purple). (d) Immunohistochemistry for PCNA, a proliferation marker that stains the nuclei of proliferating cells (brown) and counterstained with hemotoxylin (purple). All images are at 10 magnification, insets are 100

Fig. 5 Workflow for ex vivo adhesion assay. Peritoneal tissues (peritoneum or omentum) are removed and pinned on silicone-coated wells. Fluorescent-labeled cells are added and incubated for 1.5–2 h. Fluorescent cells are imaged and quantified or tissues are fixed for SEM imaging. Adapted from ref. [7]

Analysis of Ovarian Cancer Omental Adhesion

205

in Fig. 5, or primary human tissue samples. The amount of OvCa cell adherence can be imaged using fluorescence microscopy and subsequently quantified using ImageJ cell counter. The use of these methods to study OvCa omental adhesion are outlined below. Some of the methods described are applicable in analyzing both murine and primary human samples.

2

Materials

2.1 In Vivo Adhesion and Tumor Study

1. Mice: Several mouse strains can be used (see Note 3). 2. ID8 culture medium: Dulbecco’s Modified Eagle Medium with 4% fetal bovine serum, 1% penicillin/streptomycin, 1% insulin-transferrin-selenium. 3. Phosphate-buffered solution (PBS), pH 7.4. 4. Trypsin. 5. Tissue culture plates (100 mm and 150 mm). 6. Pipette controller, pipettes (micro and serological), and pipette tips. 7. Dissection tools: Forceps (straight and curved) and scissors (straight and curved). 8. 3-mL syringes with 25G  5/800 needle. 9. 10 mL syringes with 18G  5/800 needle. 10. In vivo fluorescent imager. 11. 10% Neutral Buffered Formalin.

2.2 Second Harmonic Generation Microscopy

1. Omental samples from in vivo studies in Subheading 2.1. 2. Multiphoton confocal microscope equipped with tungsten and halogen visible light sources and four lasers: argon 458 nm, 488 nm, 515 nm; HeNe 543 nm; red diode 635; and a Mai Tai DeepSea titanium-sapphire 690–1040 nm, and 25 XLPlanN, 1.05 na water objective. 3. Collagen hybridizing peptide biotin conjugate (CHP-B): Final solution should be 40 mM CHP-B and 10 mM StreptavidinAlexa Fluor 647 in PBS. Combine CHP-B and PBS, heat at 80  C for 5 min to denature any dimers or trimers, and cool on ice for 30–60 s. Add the Streptavidin-Alexa Fluor 647conjugate (CHP-Alexa Fluor). Incubate in the dark for 30 min (see Note 4). 4. Streptavidin-Alexa Fluor 647 conjugate. 5. Microscope coverslips 22  50 mm. 6. Micro-fine insulin syringe, 1 mL. 7. Phosphate-buffered saline (PBS), pH 7.4.

206

2.3

Elizabeth I. Harper and Tyvette S. Hilliard

Histology

1. Fixed, processed, and sectioned omental samples from Subheading 2.1. 2. Microscope slides (superfrost plus 25  75  1 mm). 3. Lab drying oven. 4. Slide racks and containers. 5. Hematoxylin and eosin staining kit (hemotoxylin, eosin, acid alcohol, basic buffer or bluing reagent). 6. Xylene. 7. Ethanol (EtOH, varies depending on assay—100%, 95%, 90%, 80%, 70%). 8. Avidin Biotin Complex immunoperoxidase detection kit (blocking serum, biotinylated secondary antibody, avidin/ streptavidin reagent, biotin-labeled horseradish peroxidase (HRP) reagent). 9. 10 mM sodium citrate. 10. 3% H2O2. 11. Hydrophobic pen. 12. PBS, pH 7.4. 13. Serum (varies depending on primary antibody). 14. Primary antibodies (vary, can be biotinylated). 15. Secondary antibodies (varies depending on primary antibody, HRP-conjugated or biotinylated). 16. 3,30 -Diaminobenzidine (DAB) peroxidase substrate kit (DAB concentrate, peroxidase buffer/diluent). 17. Masson’s trichrome stain kit (Bouin’s fluid, Weigert’s hematoxylin, Biebrich scarlet/acid fuschin solution, phosphomolybdic/phosphotungstic acid, analine blue, and 1% acetic acid). 18. Collagen hybridizing peptide biotin conjugate (CHP-B). 19. Cytoseal mounting medium. 20. Microscope coverslips (22  22 mm or 24  50 mm). 21. Microscope or slide scanner microscope.

2.4 Ex Vivo Adhesion Assay

1. Mice: See Subheading 2.1, item 1 (see Note 5). 2. SYLGARD 184 Silicone Elastomer Kit (Elastomer and curing solutions). 3. Degassing chamber. 4. 24-well tissue culture plates. 5. Entomological pins. 6. Ovarian cancer cells (varies; see Notes 3 and 5).

Analysis of Ovarian Cancer Omental Adhesion

2.5 Scanning Electron Microscopy

207

1. Omental samples from in vivo (Subheading 2.1) or ex vivo (Subheading 2.4) adhesion studies. 2. Cacodylate buffer: Stock solution of 0.4 M, working solution of 0.1 M. Mix cacodylic acid-sodium salts with ultrapure water to create stock solution. Working solution should be a pH of 7.2–7.4, adjust as necessary with HCl or NaOH. 3. Primary fixative: 2% glutaraldehyde, 2% paraformaldehyde in 0.1 M cacodylate buffer, pH 7.3. 4. Osmium tetroxide. 5. EM microwave vacuum. 6. Critical point dryer. 7. Ethanol (EtOH: 20%, 50%, 70%, 90%, 100%). 8. SEM specimen stubs. 9. Carbon imaging sticker. 10. Conductive silver paint. 11. Toothpicks. 12. Dissection microscope. 13. Sputter coater. 14. Scanning electron microscope.

3

Methods

3.1 In Vivo Adhesion and Tumor Study Analysis

1. Plate fluorescently tagged tumor cells on 150-mm tissue culture dishes. Allow cells to grow to ~80% confluence. Plate enough plates to have 1 plate/mouse plus a few extra plates to account for cell death as you prepare cells for injection (see Note 6). 2. Trypsinize, collect, and count cells. Resuspend cells in warmed, sterile PBS at 10 million cells/mL. Inject 1 mL of tumor cells per mouse IP with a 3 mL syringe and 25G  5/800 needle (see Note 7). 3. For short-term in vivo adhesion studies, after 24 h euthanize mice, dissect and image the omentum and other peritoneal organs using a standard fluorescence imaging system to visualize adhered cells (see Note 8). 4. For a tumor study, sacrifice and dissect mice 5.5 weeks post injection when using ID8Trp53
/
-RFP cells. Tumor studies are timed at a humane endpoint, before there is excessive ascites or significant weight loss. For CHP/SHG analysis (Subheading 3.2), wait for 3 weeks until injection/sacrifice. 5. Dissect mouse by making an incision down the midline and sides of the ventral parietal peritoneum. Pull the ventral skin

208

Elizabeth I. Harper and Tyvette S. Hilliard

away and expose the peritoneal cavity. For tumor studies, if ascites fluid is present, remove with a 10-mL syringe and 18G needle prior to exposing the peritoneal cavity. 6. Scan the abdominal organs, including the omentum, in situ using a fluorescence imaging system (see Note 9; Fig. 1a). 7. Dissect the omentum and other peritoneal organs and image (see Note 8; Fig. 1b). Omentum can be placed in formalin for histology or scanning electron primary fixative for later analysis (see Subheadings 3.3–3.5 and 3.7 respectively). 8. Tumor cell adhesion measured by RFP signal intensity or fluorescent area can be quantified in ImageJ (see Note 9, see ref. 2). 9. There are several ways to analyze the images in ImageJ: Area percentage analysis can be used to measure the fluorescent tumor area. In ImageJ set threshold by Image ! Adjust ! Threshold ! Set value. Use the rectangle tool to select the analysis area. Analyze ! Analyze Particles, then copy and paste results table containing area into Excel. Raw integrated density (RID) for measuring fluorescent tumor intensity can also be measured by setting the threshold selecting Image ! Adjust ! Threshold ! Set value. Then select Analyze ! Set Measurements and check the Integrated Density box. Use the rectangle tool to select the analysis area. Analyze ! Analyze Particles, then copy and paste into Excel (see Note 10). 3.2 Second Harmonic Generation Microscopy: In Vivo Assays

1. Initiate a tumor study as described in Subheading 3.1, steps 1 and 2 and allow tumor cell adhesion for 3 weeks post injection of RFP-tagged OvCa cells. 2. To visualize denatured omental collagen, inject 100 μL/mouse IP of CHP-Alexa Fluor compound in live mice and incubate for 3 h prior to euthanization and dissection. 3. Dissect omentum and wash gently with PBS. Spread the omentum out on a coverslip so the translucent area is visible and flat. Cover with another coverslip and image on a multiphoton microscope. 4. Image collagen using the RXD1 detector emission filter (420–260 nm) for SHG signal (Fig. 2a, b). For fluorescence, use one or multiple detector emission filters including RXD2 detector emission filter (495–540 nm; unused), RXD3 detector emission filter (380–560 nm; RFP-tagged OvCa cells), RXD4 detector emission filter (575–630 nm; Alexa Fluor 647; Fig. 2c, d). 5. Using low resolution (e.g., 512  512 and 8.0 μs/pixel), scan the tissue to find an area of interest. Adjust the correction collar on the objective, as needed, to find the brightest setting. Using

Analysis of Ovarian Cancer Omental Adhesion

209

the fine focus, set the start and finish position (1 μm steps) to take a z-stack image. Once start and finish positions are set, take a high resolution image (e.g., 1024  1024) (see Note 11). 6. There are several ways to analyze the images in ImageJ including fluorescent tumor area and fluorescent tumor intensity as described in Subheading 3.1, step 9. Additionally, anisotropy, a measure of omental collagen alignment, can be measured using the FibrilTool plugin in ImageJ (see ref. 6). 3.3 Histology: Hematoxylin and Eosin Assay

1. Harvest tumors from Subheading 3.1, fix in 10% formalin for at least 24 h (see Note 12), process and embed in paraffin. Section tumor tissues at a width of 5 μm and place on microscope slides (see Note 13). 2. Place slides in oven for 15 min at 65  C. 3. Place slides in a slide rack in xylene three times for 2 min. 4. Rehydrate for 1 min each in decreasing amounts of ethanol (EtOH) (two times in 100% EtOH, one time in 95% EtOH). 5. Place slides in ddH2O for 1 min. 6. Apply hematoxylin for 30 s to 3 min. 7. Rinse slides with flowing tap water for 2 min. 8. Place slides in acid alcohol for 45 s. 9. Rinse slides in tap water for 1 min. 10. Place slides in basic buffer wash for 45 s. 11. Rinse slides in tap water for 1 min. 12. Place slides in 95% EtOH for 1 min. 13. Place slides in eosin for 30 s. 14. Place slides in 100% EtOH three times for 1 min. 15. Place slides in xylene two times for 1 min and one time for 2 min. 16. Mount coverslips immediately, let slides dry overnight. 17. Slides are analyzed by a pathologist to determine tumor characteristics including morphology, distribution of cells, and general layout (Fig. 4a).

3.4 Histology: Trichrome Staining

1. Prepare slides as described in Subheading 3.3, step 1. 2. Place slides in oven for 1 h at 65  C. 3. Place slides in a slide rack in xylene one time for 10 min, then two times for 5 min. 4. Rehydrate samples for 3 min each in decreasing amounts of EtOH (two times in 100% EtOH, one time each 90% EtOH and 80% EtOH). 5. Place in ddH2O for 3 min.

210

Elizabeth I. Harper and Tyvette S. Hilliard

6. Put slides in Bouin’s fluid overnight (see Note 14). 7. Rinse slides in ddH2O (see Note 15). 8. Mix Weigert’s hematoxylin components 1:1, and apply to slides for 5–12 min, depending on desired darkness of the stain. 9. Rinse slides under running water for 2 min. 10. Apply Beibrich scarlett/acid fuschin solution to the slides for 10–15 min. 11. Wash in distilled water 3–5, until water runs clear. 12. Differentiate in phosphomolybdic/phosphotungstic acid for 15 min (see Note 16). 13. Without rinsing, change to analine blue solution for 10 min. 14. Wash slides three times in ddH2O. 15. Apply 1% acetic acid for 5 min. 16. Dehydrate in 90% EtOH two times for 2 min, 100% EtOH two times for 2 min. 17. Place slides in xylene three times for 5 min. 18. Mount slides, let dry overnight, and image on microscope or slide scanner (Fig. 4b). 19. Collagen content can be quantified using image analysis software, or slides can analyzed by a pathologist (see Note 17) to determine tumor characteristics including morphology, distribution of cells, and fibrosis (see Note 18). 3.5 Histology: Immunohistochemistry

1. Prepare slides as described in Subheading 3.4, steps 1–5 (see Note 19). 2. Antigen retrieval: boil in 10 mM sodium citrate for 30 min (see Note 20). Allow slides to cool for 20 min. 3. Rinse slides with PBS for 5 min. 4. Place slides in 3% H2O2 for 10 min. 5. Wash slides with ddH2O2 for 3 min. 6. Rinse slides in PBS for 5 min. 7. Use hydrophobic pen to encircle tissue and block with serum for 1 h (see Note 21). 8. Incubate slides with antibody (concentration variable depending on antibody) or CHP-B peptide (0.5 μM) overnight (see Note 22). 9. Rinse slides three times for 5 min with PBS. 10. Apply HRP-conjugated secondary antibody or use ABC kit for biotinylated antibodies or CHP-B (see Note 23). 11. Develop with DAB for 1–10 min (see Note 24). 12. Counterstain with hematoxylin.

Analysis of Ovarian Cancer Omental Adhesion

211

13. Dehydrate with increasing amounts of EtOH (70%, 80%, 90%, 100%) for 3 min each, and with xylene for three times for 5 min, in reverse order of the rehydration steps. 14. Apply coverslip, let mounting medium dry. Image and analyze (see Note 17, Fig. 4c, d). 3.6 Ex Vivo Adhesion Fluorescence Assay

1. Prior to starting procedure, coat a 24-well dish with silicone, using the SYLGARD 184 Silicone Elastomer Kit at a 1:10 Curing Agent:Elastomer dilution. Add approximately 0.5 mL of solution into each well and place dishes to cure in degassing chamber overnight to avoid bubbles in the silicone (see Note 25). 2. Sacrifice mice and dissect out omentum (see Note 8). Wash omentum gently with PBS and add PBS to the silicone-coated dish. Trim pancreas as necessary for organ to lie flat, but leave enough to place pins. 3. Using forceps, pin the tissue in place (see Note 26). Pin around the edges of the omentum, using the pancreas and the fat band to keep the translucent area open and flat but not stretched (see Note 27). 4. Trypsinize fluorescently tagged OvCa cells and resuspend at 4  105 cells/mL, with 2.5 mL/well. Add cells to the pinned omentum and incubate 1.5–2 h at 37  C (Fig. 5; see Note 28). 5. Remove media and wash quickly and gently with ice-cold PBS three times. 6. Remove pins as they were inserted. Place tissue on a microscope slide so the side facing up during the assay is facing the objective on the fluorescence microscope being used. Add a coverslip to the top. Tissues waiting to be imaged should remain on ice (see Note 29). 7. Image 9 random fields for each omental tissue sample at 10. 8. Count the number of cells adhered using ImageJ cell counter plug in. 9. For further analysis, place tissues in SEM prefix for later imaging (see SEM Subheading 3.7).

3.7 Scanning Electron Microscopy: In Vivo or Ex Vivo Adhesion Assays

1. Place samples (from in vivo studies or ex vivo assay) in primary fixative solution immediately (see Note 30). Store in primary fixative for at least 1 h, preferably overnight at 4  C (see Note 31). For multiple samples, use a 24-well plate to avoid tissue folding. 2. Wash samples three times for 20 min each in 0.1 M Cacodylate Buffer. 3. Fix with 1% OsO4 in 0.1 M Cacodylate Buffer. Stock solution is generally 2–4% (see Note 32).

212

Elizabeth I. Harper and Tyvette S. Hilliard

4. Place in microwave vacuum chamber for a pre-programmed cycle of 1 min off, 80 s on, 3 min off, 40 s on. Ensure vacuum is on prior to starting. 5. Rinse samples quickly three times with 0.1 M Cacodylate Buffer. 6. Rinse quickly once with ddH2O. 7. Wash three times with ddH2O for 5 min each wash, with rocking. 8. Dehydrate the samples stepwise with EtOH, microwaving at each step at 100 W for 40 s. Start with 20% EtOH, then 50%, then 70%, then 90%, then use 100% EtOH three times (see Note 33). 9. Critical point dry (CPD) the samples. Load samples into cartridges while submerged in 100% EtOH. Put loaded cartridges into CPD machine and fill with 100% EtOH up to the overflow hole. Replace the top of the machine and screw on tightly. Run the samples through the CPD program (see Note 34). Store samples at 37  C until ready to image, to prevent moisture retention. 10. Mount the dried samples onto SEM specimen stubs with carbon imaging sticker. Use conductive silver paint to link the sample to the stubs, making sure there is a connection between the metal plate and the sample (see Note 35). Do not get silver paint on any part of the tissue to be imaged but paint everywhere around the tissue (see Note 36). 11. Sputter coat the samples with iridium at a thickness of 5 nm for conductivity. 12. Image samples on a scanning electron microscope at multiple magnifications, focusing on the junctions between the OvCa cells and the omentum (Fig. 3).

4

Notes 1. RFP was chosen because it has a long wavelength that can penetrate farther into organic tissue than shorter wavelengths, allowing the detection of the signal through the peritoneal wall as a way to monitor tumor progression with in vivo imaging. 2. Since 5-FAM, sCy7.5 and Cy3 both have bleed-through to the RFP channel, the biotin conjugate (CHP-B) can be used with Alexa Fluor 647 conjugated to streptavidin (CHP-Alexa Fluor) to overcome the spectral overlap. The biotin–streptavidin bond is one of the strongest documented non-covalent interactions in nature, so it can withstand a short incubation period in the

Analysis of Ovarian Cancer Omental Adhesion

213

mouse peritoneal cavity prior to imaging. Importantly, each streptavidin molecule has four binding domains for biotin, so a 1:4 molar ratio must be used [8]. 3. For syngeneic studies some examples include C57Bl/6 mice used with RFP-tagged ID8 (with or without genetic modification) mouse ovarian surface epithelial tumor cells [4, 9] or FVB/NJ mice used with RFP-tagged oviductal epithelium tumor cells FVB/N-PTENshRNA/KRASG12V [10]. If using human RFP-tagged OvCa cells (e.g., OVCAR3, OVCAR5) immunocompromised mice must be used such as athymic nude mice (nu/nu). 4. The conjugated CHP-B is stable at room temperature for 1–2 h. The heating step is to denature dimers/trimers of the peptide, so the longer the solution sits on the bench top the more likely it will redimerize/retrimerize so it is best used immediately to avoid reheating. Additionally, it is best to use the CHP-Alexa Fluor directly after the 30 min incubation as reheating could have an effect on the biotin/streptavidin bond. 5. As this is a short assay, it is fine to use immunocompetent mice even if not using a syngeneic cell line (i.e., human OvCa cell line with C57Bl/6 omentum). 6. The number of cells injected and the time to tumor development is variable by cell line, depending on aggressiveness of cell line, and should be accurately determined prior to starting any experiment. Number of mice/experiment varies based on experiments; for ex vivo or SHG studies, 4–6 mice/cohort is generally suitable; for larger in vivo tumor studies, 10–15 mice/cohort is more apt. You want to use mice that are mature (>3 months old) but not too old (