Prostate Cancer

Methods in Molecular Biology 1786

Zoran Culig Editor

Prostate Cancer Methods and Protocols

Methods

in

M o l e c u l a r B i o lo g y

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

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

Prostate Cancer Methods and Protocols

Edited by

Zoran Culig Department of Urology, Innsbruch Medical University, Innsbruck, Austria

Editor Zoran Culig Department of Urology Innsbruch Medical University Innsbruck, Austria

ISSN 1064-3745     ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-4939-7843-4    ISBN 978-1-4939-7845-8 (eBook) https://doi.org/10.1007/978-1-4939-7845-8 Library of Congress Control Number: 2018942021 © Springer Science+Business Media, LLC, part of Springer Nature 2018 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express 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. Printed on acid-free paper This Humana Press imprint is published by the registered company Springer Science+Business Media, LLC part of Springer Nature. The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.

Preface Progress has  been made in the development of cellular and animal models for prostate cancer. In addition to cell lines which have been used over many years in research, patient-­derived xenograft models are available for studying specific questions in prostate cancer biology. Combination of various cellular and animal models is reasonable especially in research with novel drug substances for targeting oncogenes in prostate cancer. Experimental procedures in order to develop novel models are the focus of this chapter. Our authors also provided contributions on imaging, an area in which rapid progress is observable. Methodologies described in this book will allow readers to learn about modern analysis of steroid receptor function. Androgen receptor is a recognized target in castration therapy-­resistant prostate cancer, and use of the antiandrogen enzalutamide leads to a more efficient inhibition of androgen receptor function. Several methodological approaches described in this book could facilitate studies with novel drugs including antiandrogens and chemotherapeutics. The editor and the publisher appreciate the efforts of the authors, principal investigators, and their associates who provided details of their experimental work. Innsbruck, Austria

Zoran Culig

v

Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .     v Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   ix 1 Generation of Prostate Cancer Patient-Derived Xenografts to Investigate Mechanisms of Novel Treatments and Treatment Resistance. . . .     1 Hung-Ming Lam, Holly M. Nguyen, and Eva Corey 2 Methods to Study Angiogenesis in a Mouse Model of Prostate Cancer�������������    29 Ana-Rita Pedrosa, Alexandre Trindade, and António Duarte 3 Methodologies Applied to Establish Cell Cultures in Prostate Cancer���������������    55 Anne T. Collins 4 Protocols for Migration and Invasion Studies in Prostate Cancer ���������������������    67 Arjanneke F. van de Merbel, Geertje van der Horst, Jeroen T. Buijs, and Gabri van der Pluijm 5 Transplantable Animal Studies and Whole-Body Optical Imaging in Prostate Carcinoma�������������������������������������������������������������������������������������    81 Geertje van der Horst, Maaike van der Mark, Henry Cheung, and Gabri van der Pluijm 6 Protocols for Tissue Microarrays in Prostate Cancer Studies�����������������������������  103 Tatjana Vlajnic, Serenella Eppenberger-Castori, and Lukas Bubendorf 7 Functional Studies on Steroid Receptors ���������������������������������������������������������  117 Simon Schlanger and Hannelore V. Heemers 8 Protocols for Studies on TMPRSS2/ERG in Prostate Cancer���������������������������  131 Hubert Pakula, Douglas E. Linn, Daniel R. Schmidt, Marit Van Gorsel, Matthew G. Vander Heiden, and Zhe Li 9 Protocols for the Study of Taxanes Chemosensitivity in Prostate Cancer�����������  153 M. Luz Flores and Carmen Sáez 10 A Method for Prostate and Breast Cancer Cell Spheroid Cultures Using Gelatin Methacryloyl-Based Hydrogels ��������������������������������������������������� 175 Christoph Meinert, Christina Theodoropoulos, Travis J. Klein, Dietmar W. Hutmacher, and Daniela Loessner 11 Protocols for Studies on Genetically Engineered Mouse Models in Prostate Cancer��������������������������������������������������������������������������������������������� 195 Chris W. D. Armstrong, Oksana Lyubomska, Melissa J. LaBonte, and David J. J. Waugh 12 Protocols for Studies on Stromal Cells in Prostate Cancer���������������������������������  207 Damien A. Leach and Grant Buchanan 13 Techniques for Evaluation of AR Transcriptional Output and Recruitment to DNA������������������������������������������������������������������������������������������������������������� 219 Manqi Zhang, William C. Krause, and Irina U. Agoulnik

vii

viii

Contents

14 NMR-Based Prostate Cancer Metabolomics������������������������������������������������������� 237 Leslie R. Euceda, Maria K. Andersen, May-Britt Tessem, Siver A. Moestue, Maria T. Grinde, and Tone F. Bathen 15 Studies on Steroid Receptor Coactivators in Prostate Cancer ����������������������������� 259 Zoran Culig and Frédéric R. Santer Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .   263

Contributors Irina U. Agoulnik  •  Biomolecular Sciences Institute, FIU, Miami, FL, USA; Baylor College of Medicine, Houston, TX, USA; Department of Human and Molecular Genetics, Herbert Wertheim College of Medicine, Miami, FL, USA Maria K. Andersen  •  Department of Circulation and Medical Imaging, NTNU—The Norwegian University of Science and Technology, Trondheim, Norway Chris W. D. Armstrong  •  Centre for Cancer Research and Cell Biology (CCRCB), Queen’s University Belfast, Belfast, UK Tone F. Bathen  •  Department of Circulation and Medical Imaging, NTNU—The Norwegian University of Science and Technology, Trondheim, Norway Lukas Bubendorf  •  Institute of Pathology, University Hospital Basel, Basel, Switzerland Grant Buchanan  •  Divisions of Medicine and Surgery, The Basil Hetzel Institute for Translational Health Research, University of Adelaide, Adelaide, SA, Australia; Department of Radiation Oncology, Canberra Teaching Hospital, Canberra, Australia Jeroen T. Buijs  •  Department of Urology, Leiden University Medical Center, Leiden, The Netherlands Henry Cheung  •  Department of Urology, Leiden University Medical Center, Leiden, The Netherlands Anne T. Collins  •  YCR Cancer Research Unit, Department of Biology, University of York, York, UK Eva Corey  •  Department of Urology, University of Washington, Seattle, WA, USA Zoran Culig  •  Experimental Urology, Department of Urology, Medical University of Innsbruck, Innsbruck, Austria António Duarte  •  Faculdade de Medicina Veterinaria—ULisboa, Av. da Universidade Tecnica, Centro Interdisciplinar de Investigação em Sanidade Animal (CIISA), University of Lisbon, Lisbon, Portugal; Instituto Gulbenkian de Ciência, Oeiras, Portugal Serenella Eppenberger-Castori  •  Institute of Pathology, University Hospital Basel, Basel, Switzerland Leslie R. Euceda  •  Department of Circulation and Medical Imaging, NTNU—The Norwegian University of Science and Technology, Trondheim, Norway Marit Van Gorsel  •  Department of Biology, The Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA, USA Maria T. Grinde  •  Department of Circulation and Medical Imaging, NTNU—The Norwegian University of Science and Technology, Trondheim, Norway Hannelore V. Heemers  •  Department of Cancer Biology, Cleveland Clinic, Lerner Research Institute, Cleveland, OH, USA; Department of Urology, Cleveland Clinic, Glickman Urologic & Kidney Institute, Cleveland, OH, USA; Department of Hematology/Medical Oncology, Cleveland Clinic, Taussig Cancer Institute, Cleveland, OH, USA Geertje van der Horst  •  Department of Urology, Leiden University Medical Center, Leiden, The Netherlands

ix

x

Contributors

Dietmar W. Hutmacher  •  Queensland University of Technology (QUT), Brisbane, Australia; Australian Prostate Cancer Research Centre—Queensland, Translational Research Institute, Brisbane, QLD, Australia; George W Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, USA; Institute for Advanced Study, Technical University of Munich, Munich, Germany Travis J. Klein  •  Queensland University of Technology (QUT), Brisbane, QLD, Australia William C. Krause  •  Department of Cellular and Molecular Pharmacology, University of California, San Francisco, San Francisco, CA, USA Melissa J. LaBonte  •  Centre for Cancer Research and Cell Biology (CCRCB), Queen’s University Belfast, Belfast, UK Hung-Ming Lam  •  Department of Urology, University of Washington, Seattle, WA, USA Damien A. Leach  •  Divisions of Medicine and Surgery, The Basil Hetzel Institute for Translational Health Research, University of Adelaide, Adelaide, SA, Australia; Department of Surgery and Cancer, Imperial College London, London, UK Zhe Li  •  Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA Douglas E. Linn  •  Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA Daniela Loessner  •  Queensland University of Technology (QUT), Brisbane, QLD, Australia; Barts Cancer Institute, Queen Mary University of London, London, UK M. Luz Flores  •  Instituto de Biomedicina de Sevilla (IBIS), Hospital Universitario Virgen del Rocío, CSIC, Universidad de Sevilla, Seville, Spain Oksana Lyubomska  •  Centre for Cancer Research and Cell Biology (CCRCB), Queen’s University Belfast, Belfast, UK Maaike van der Mark  •  Department of Urology, Leiden University Medical Center, Leiden, The Netherlands Christoph Meinert  •  Queensland University of Technology (QUT), Brisbane, QLD, Australia Arjanneke F. van de Merbel  •  Department of Urology, Leiden University Medical Center, Leiden, The Netherlands Siver A. Moestue  •  Department of Clinical and Molecular Medicine, NTNU—The Norwegian University of Science and Technology, Trondheim, Norway; Department of Pharmacy, Faculty of Health Sciences, Nord University, Bodø, Norway Holly M. Nguyen  •  Department of Urology, University of Washington, Seattle, WA, USA Hubert Pakula  •  Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA Ana-Rita Pedrosa  •  Faculdade de Medicina Veterinaria—ULisboa, Av. da Universidade Tecnica, Centro Interdisciplinar de Investigação em Sanidade Animal (CIISA), University of Lisbon, Lisbon, Portugal Gabri van der Pluijm  •  Department of Urology, Leiden University Medical Center, Leiden, The Netherlands Carmen Sáez  •  Instituto de Biomedicina de Sevilla (IBIS), Hospital Universitario Virgen del Rocío, CSIC, Universidad de Sevilla, Seville, Spain; Department of Pathology, Hospital Universitario Virgen del Rocío, Seville, Spain Frédéric R. Santer  •  Experimental Urology, Department of Urology, Medical University of Innsbruck, Innsbruck, Austria

Contributors

xi

Simon Schlanger  •  Department of Cancer Biology, Cleveland Clinic, Lerner Research Institute, Cleveland, OH, USA Daniel R. Schmidt  •  Harvard Radiation Oncology Program, Boston, MA, USA May-Britt Tessem  •  Department of Circulation and Medical Imaging, NTNU—The Norwegian University of Science and Technology, Trondheim, Norway Christina Theodoropoulos  •  Queensland University of Technology (QUT), Brisbane, QLD, Australia Alexandre Trindade  •  Faculdade de Medicina Veterinaria—ULisboa, Av. da Universidade Tecnica, Centro Interdisciplinar de Investigação em Sanidade Animal (CIISA), University of Lisbon, Lisbon, Portugal; Instituto Gulbenkian de Ciência, Oeiras, Portugal Matthew G. Vander Heiden  •  Department of Biology, The Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA, USA Tatjana Vlajnic  •  Institute of Pathology, University Hospital Basel, Basel, Switzerland David J. J. Waugh  •  Centre for Cancer Research and Cell Biology (CCRCB), Queen’s University Belfast, Belfast, UK Manqi Zhang  •  Department of Chemistry and Biochemistry, Florida International University, Miami, FL, USA

Chapter 1 Generation of Prostate Cancer Patient-Derived Xenografts to Investigate Mechanisms of Novel Treatments and Treatment Resistance Hung-Ming Lam, Holly M. Nguyen, and Eva Corey Abstract Treatment advances lead to survival benefits of patients with advanced prostate cancer. These treatments are highly efficacious in a subset of patients; however, similarly to other cancers, after initial responses the tumors develop resistance (acquired resistance) and the patients succumb to the disease. Furthermore, there is a subset of patients who do not respond to the treatment at all (de novo resistance). Preclinical testing using patient-derived xenografts (PDXs) has led to successful drug development, and PDXs will continue to provide valuable resources to generate clinically relevant data with translational potential. PDXs demonstrate tumor heterogeneity observed in patients, preserve tumor-microenvironment architecture, and provide clinically relevant treatment responses. In view of the evolving biology of the advanced prostate cancer associated with new treatments, PDXs representing these new tumor phenotypes are urgently needed for the study of treatment responses and resistance. In this chapter, we describe methodologies used to establish prostate cancer PDXs and use of these PDXs to study de novo and acquired resistance. Key words Prostate cancer, Abiraterone, Enzalutamide, Resistance, Testosterone

1  Introduction Advanced prostate cancer is a heterogeneous disease and there is no curative treatment available at this time. Treatment advances have led to the approval of new agents that extend the median survival of patients with this disease for only 3–6 months [1, 2]. However, within the population there are patients with tumors that do not respond to these therapies exhibiting de novo resistance, and patients with tumors that initially respond but after the initial response develop acquired resistance. Patients demonstrating de novo or acquired resistance ultimately succumb to the disease. Preclinical research is needed to better understand the complex biology of advanced prostate cancer and molecular signatures of responses and resistance to therapies. With few exceptions, Zoran Culig (ed.), Prostate Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 1786, https://doi.org/10.1007/978-1-4939-7845-8_1, © Springer Science+Business Media, LLC, part of Springer Nature 2018

1

2

Hung-Ming Lam et al.

­ reclinical studies in advanced prostate cancer used cell lines to p generate xenografts and in many cases these studies provided results that were hardly replicable in clinical trials. Reasons for such discrepancies are that (1) the in vivo use of cell lines does not necessarily represent the high level of cellular heterogeneity observed in the original tumors in patients, (2) xenografts from cell lines do not retain the epithelial-stromal interaction and the tumor architecture of patient tumors, and (3) the design of preclinical or clinical trials that may not mirror each other. In order to generate clinically relevant data, preclinical research has to be focused on the generation and the use of patient-derived, clinically predictive models—patient-derived xenografts (PDXs). For studies of advanced prostate cancer, locally advanced and metastatic tumors obtained from patients represent a valuable source for PDXs establishment. To establish PDX models, fresh tissues harvested from patients are directly implanted into immunocompromised mice. Established PDXs retain the 3D tumormicroenvironment architecture of the original tumor and also replicate the intra-tumoral heterogeneity of the patient’s tumor. Our and others data show that the morphology and heterogeneity of PDXs are maintained over repeated passages [3, 4]. However, it is well recognized that prostate cancer PDXs are very difficult to establish with a very low take rate, and in many cases after initial growth, the PDXs fail to grow after the first transplant or are lost after serial passaging—leading to the high cost of developing and sustaining serially passaged models. Nonetheless, once established prostate cancer PDXs provide reliable preclinical models to study the biology of this disease and to evaluate treatment responses that predict tumor resistance and tumor sensitivity up to 97% and 90% accuracy, respectively [5]. However, one has to be aware that when designing preclinical studies and in order to generate clinically relevant results, it is absolutely critical to use multiple prostate cancer PDXs that exhibit the wide range of characteristics mimicking the notoriously heterogeneous character of this disease in patients. Despite the difficulty of prostate cancer PDX establishment, our group (LuCaP xenograft series; see Table 1) and other groups (see Table 2) have demonstrated success in establishing, characterizing, and conducting preclinical studies using prostate cancer PDXs [3, 4, 6–22]. To make these models available to the research community, The Movember Foundation—the leading global organization committed to changing men’s health—funded the GAP1 Global Prostate Cancer Xenograft Initiative, which focused on making tissue microarrays of all existing prostate cancer PDXs for molecular and histological characterization. It is well accepted that treatment resistance is associated with phenotypic changes of the tumors with alternative pathways being involved in tumor survival and progression. The emergence of new drug resistance urgently prompts the development of new PDXs

Prostate Cancer Patient-Derived Xenografts

3

Table 1 Sources and characteristics of LuCaP prostate cancer PDXs (PMID: 28156002) Number Patients

156

Samples implanted

277

Tumor sources Lymph node metastasis

8

Bone metastasis

2

Liver metastasis

5

TURP

4

Ascites

1

Bladder metastasis

1

Omental fat

1

Characteristics

# of PDXs

AS vs CR pairs

11

Mutated AR

8

AR v5,6,7

1

TMPRSS2/ERG

10

PTEN-negative

13

Rb Deletion

2

Osteoblastic

7

Neuroendocrine

4

that capture the biology underlying treatment responses and resistance of these most recent tumors. For example, PDXs of tumors from patients that do not respond to the most recent second-­ generation antiandrogens abiraterone and/or enzalutamide are needed to investigate mechanisms of de novo resistance, while PDXs of tumors from patients that originally responded but subsequently developed resistance are needed to investigate mechanisms of acquired resistance. In addition, with more vigorous inhibition of androgen receptor signaling, it appears that prostate cancers in some cases undergo trans-differentiation to neuronal phenotype that displays minimum to no androgen receptor signaling activity [3, 23]. Currently, very limited number of these neuronal-­like PDXs is available [3, 24–26], and very little is known about the biology behind and the potential treatment targets of this very

LCNEC

Prostate

Prostate

Prostate

Prostate

CRPC-rectal wall Pelvic exenteration

CRPC-­prostate

CRPC-­prostate

MDA-­PCa-­114-4

MDA-­PCa-­114-9

MDA-­PCa-­114-2

MDA-­PCa-­114-6

MDA-­PCa-­79

MDA-­PCa-­117-9

MDA-­PCa-­130

Cystoprostatectomy

Cystoprostatectomy

SCPC

MDA-­PCa-­114-­23 Rectal wall

Adenocarcinoma

Adenocarcinoma with NE differentiation

Adenocarcinoma

LCNEC

LCNEC

SCPC

MDA-­PCa-­114-­20 Rectal wall

Pelvic exenteration

SCPC

CRPC

MDA-­PCa-­114 series

MDA-­PCa-­114-­13 Bladder neck

Left hemipelvis

MDA-PCa-­118b

Tumor type

LCNEC

Left superior pubic ramus

MDA-PCa-­118a

Biopsy

Procedure

MDA-­PCa-­114-­11 Bladder neck

CRPC

MDA-­PCa-­118 series

Navone

Type and source

Xenograft name

Group

Table 2 Summary of existing prostate cancer PDXs in addition to the LuCaP series Mouse strain

Take rate

100%

Subcutaneous CB17 SCID 12%

Graft site

Yes

Yes

[11–13]

[4, 11, 12]

[9, 10]

T supplement? Refs.

Group

CRPC

Bladder wall

MDA-­PCa-­155 series

MDA-­PCa-­155-2

MDA-­PCa-­146-­20 Bladder wall

MDA-­PCa-­146-­17 Bladder wall

MDA-­PCa-­146-­10 Bladder wall

Pelvic exenteration

Cystoprostatectomy

SCPC

SCPC

Adenocarcinoma with NE differentiation

MDA-­PCa-­180-­30 Seminal vesicle

CRPC

Adenocarcinoma with NE differentiation

MDA-­PCa-­180-­21 Bladder wall

MDA-­PCa-­146 series

Adenocarcinoma with NE differentiation

Cystoprostatectomy

MDA-­PCa-­180-­18 Bladder wall

CRPC

MDA-­PCa-­180 series

Adenocarcinoma

Adenocarcinoma with NE differentiation

Prostate

MDA-­PCa-­170-4

MDA-­PCa-­180-­14 Bladder wall

Prostate

MDA-­PCa-­170-1

Cystoprostatectomy

Tumor type

Adenocarcinoma with NE differentiation

CRPC

MDA-­PCa-­170 series

Procedure

MDA-­PCa-­180-­11 Bladder wall

Type and source

Xenograft name

Graft site

Mouse strain

Take rate T supplement? Refs.

(continued)

Wang

Group

Table 2 (continued)

SCPC SCPC Adenocarcinoma

MDA-­PCa-­155-­12 Bladder wall

MDA-­PCa-­155-­16 Prostate

MDA-­PCa-­149-1

LTL313H

LTL313D

LTL313C

LTL313BR

LTL313B

CR line of LTL313B generated in the mouse

Primary prostate cancer

LTL313 series

LTL313A

Biopsy

Primary prostate cancer

LTL311

Biopsy

Biopsy

LTL (Living Primary prostate Tumor cancer Laboratory) 310

CRPC-ureter

SCPC

Bladder wall

Tumor type

MDA-­PCa-­155-9

Procedure

Type and source

Xenograft name

Mouse strain

Renal capsule NOD-SCID

Graft site

Take rate

Yes

http://www. livingtumorlab. com/PDC_ Prostate.html

[3]

T supplement? Refs.

Jamieson

Group

Primary prostate cancer

LTL331 series

Lymph node metastasis

Primary prostate cancer

Primary prostate Prostatectomy cancer, treatment naïve

Primary prostate Prostatectomy cancer, treatment naïve

LTL412

LTL418

PCa1

**LTL220M

PCSD1 (Prostate Cancer San Diego 1)

**LTL221N

CRPC-femur

Penile metastasis

*LTL370

**LTL220N

Ureteral metastasis

*LTL352

Hemiarthroplasty

Prostatectomy

Surgery to remove metastasis

Surgery to remove metastasis

CR line of LTL331 generated in the mouse

Surgery to remove metastasis

Prostatectomy

Procedure

LTL331R

LTL331

Type and source

Xenograft name

NE

NE

Tumor type

Mouse strain

Take rate

Yes

Yes

Subcutaneous

100%

No

[8]

[15]

[14]

T supplement? Refs.

Intra-femoral (RAG2)−/− 66– No γC−/− 100%

Graft site

(continued)

Procedure

Metastatic prostate cancer-­femur

BM18

Mouse strain

Take rate

72%

Renal capsule NU/NU, NOD-­ SCID Orthotopic

Localized Prostatectomy prostate cancer (grade 7–9)

Loda

Subcutaneous

56%

Subcutaneous (RAG2)−/− γC−/−

Subcutaneous SCID

Adenocarcinoma Subcutaneous NU/NU

10%

27%

Renal capsule NOD-SCID 100%

Graft site

Adenocarcinoma Subcutaneous NU/NU with NE differentiation

Tumor type

Localized prostate Prostatectomy cancer (grade 4) , treatment naïve

Bone surgery (intramedullary fixation for a pending fracture)

Biopsy

CRPC-­prostate

UCRU-PR-4

Prostatectomy

Localized TURP prostate cancer (poorly differentiated)

Dissociated cells

Localized Prostatectomy prostate cancer (moderate grade)

Type and source

UCRU-PR-2

Xenograft name

Pheel

Williams

Risbridger

Group

Table 2 (continued)

Yes

Yes

Yes

Yes

Yes

[7]

[6]

[17]

[16]

[19]

T supplement? Refs.

Primary prostate cancer, treatment naïve

CRPC-­prostate

CRPC-­prostate

CRPC-­scrotal skin

Primary prostate TURP cancer, treatment naïve

Primary prostate cancer, treatment naïve

Primary prostate cancer

Metastasis-­bone

Primary prostate cancer

PC310

PC324

PC339

PC374

*PC346

*PC346B

PC82

PC133

PC135

Graft site

Mouse strain

Adenocarcinoma with NE transdifferen­tiation upon castration of mice

Adenocarcinoma Subcutaneous NRMI/NU, NU/NU with NE transdifferen­tiation upon castration of mice

Tumor type

Take rate Yes/no

[21, 22]

[20]

[20, 25]

[20, 24]

T supplement? Refs.

CRPC castration-resistant prostate cancer, LCNEC large-cell neuroendocrine carcinoma, NE neuroendocrine, SCPC small-cell prostate carcinoma, T testosterone, TURP transurethral resection of the prostate; * or ** denotes model from different tisseus from the same patient

TURP

Surgery to remove metastasis

TURP

TURP

Prostatectomy

Surgery to remove metastasis

Lymph node metastasis

PC295

Weerden

Procedure

Type and source

Xenograft name

Group

10

Hung-Ming Lam et al.

aggressive disease. In this chapter, we will provide methods to establish PDXs and methods how to address de novo and acquired drug resistance.

2  Materials 2.1  Chemicals and Kits

1. Sterile 0.9% sodium chloride for injection. 2. Ketaset (Ketamine, 100 mg/mL). 3. Xyla-Ject (Xylazine, 20 mg/mL). 4. 1× GIBCO Dulbecco’s Phosphate Buffered Saline (DPBS), stored at 4 °C. 5. Gentamicin (40 mg/mL), injection, USP, stored at 20–25 °C. 6. Betadine surgical scrub, 7.5% povidone iodine. 7. 70% ethanol. 8. BrdU (5-bromo-2′-deoxy-uridine), 97%. 80 mg/kg solution is prepared in sterile saline just before injection. 9. Buprenex (Buprenorphine 0.3 mg/mL). 10. Architect Total PSA Assay (Abbott Laboratories, Abbott Park, IL).

2.2  Surgical and Injection Materials

1. 4/0 Coated visorb polyglycolic acid suture. 2. 13-Gauge cancer implant needle. 3. Deltaphase isothermal pad. 4. Deltaphase operating board. 5. Microdissecting forceps, 4″, serrated. 6. Tungsten carbide iris scissors, 4.5″. 7. Autoclip applier, 9 mm autoclips, and autoclip remover. 8. Sterile petri dish. 9. Rib-back carbon steel surgical blade, 22, sterile. 10. Calibrated micropipet. 11. Tailveiner restrainer. 12. 26-Gauge needle. 13. 1.0 mL Syringe. 14. Feeding needle, 20 gauge 1 ½ in. 15. Electric clippers. 16. Sterile gauze.

2.3  Mice

Xenotransplantation of human cancer tissues is mostly performed in immunocompromised mice and multiple strains of mice are available.

Prostate Cancer Patient-Derived Xenografts

11

1. NU/NU nude mouse (Crl:NU-Foxn1nu) (see Note 1). 2. NMRI/NU nude mouse (Crl:NMRI-Foxn1nu) [27] demonstrates enhanced xenotransplantation take rate [28]. 3. Fox Chase SCID mouse (CB17/Icr-Prkdcscid/IcrIcoCrl) is a severely combined immune deficient (SCID) mutant mouse that is deficient for T and B lymphocytes [29]. The mice exhibit normal NK cell and myeloid cell functions [30–32]. Up to 25% of CB17 SCID mice are “leaky” (i.e., spontaneously produce both T and B cells) [33, 34] and these mice in some occasions develop T-cell lymphoma. 4. Hairless SCID mouse, SHO™ Mouse (Crl:SHO-PrkdcscidHrhr) has been recently developed by Charles River Laboratories. This SCID mouse might be beneficial to use when one considers in vivo imaging because of the autofluorescence and luminescence of CB17 SCID mice are high. 5. Rag2 knock-out mouse (B6.129S7-Rag1tm1Mom/J) displays a SCID phenotype which completely lacks functional T and B cells and does not display a leaky phenotype [35]. 6. BALB/c Rag2 and the X-linked Il2rg double knock-out mouse (Rag2tm1Flv Il2rgtm1Flv) lacks T, B, and NK cells [36]. 7. Fox Chase SCID beige mouse (CB17.Cg-PrkdcscidLystbg-J/Crl) is a SCID mouse (without functional T and B cells) with 10 log PE fluorescence) cells (Fig. 7b) from the lineage negative population (≤102 log PE-Cy7 fluorescence) (Fig.  7a). Cells are collected in appropriate collection medium according to the desired following use for the cells (Fig. 1) (see Subheading 2.2).

3.9  RT-PCR Analysis of EC and Vascular Mural Cell Populations or Whole Prostate Snap Frozen

1. For ECs and vSMCs specific analysis, samples are collected at the endpoint of each experiment and prepared for FACS sorting (as described in Subheading 3.8). ECs and mural cells are sorted directly into the lysis buffer of the RNeasy Micro Kit. 2. For whole prostate analysis, tumor samples are collected at the endpoint of each experiment to an Eppendorf and snap frozen for RNA extraction in isopentane cooled at −80 °C by liquid nitrogen (Fig. 1).

Neo-angiogenesis Analysis A

51

B 5

LinComp-PE-A :: CD146

Comp-PE-A :: CD146

10

104

103 0

-103

10

5

10

4

CD146+ CD31+

103 0

-103 -103

103 105 104 0 Comp-PE-Cy7-A :: Lin

3

-10

3

4

10 10 10 0 Comp-FITC-A :: CD31

5

Fig. 7 FACS plots of the Lineage- and ECs and vSMCs. FACS plots showing the (a) lineage negative population (Lin-PE Cy7) from which the (b) ECs (Lin− (cd45− ter119−) cd31+) and vSMCs (Lin− (cd45− ter119−) cd146+ cd31−) were sorted and isolated for posterior specific gene transcription analysis

3. Total RNA is isolated according to manufacturer’s protocol using the RNeasy Micro Kit and the RNeasy Mini Kit, for the ECs and vSMCs isolated or whole prostate, respectively. 4. A total of 100 ng RNA per reaction (ECs and vSMCs) and 400 ng per reaction (whole prostate) is used to generate cDNA with the SuperScript III First Strand Synthesis Supermix Q RT-­PCR Kit. 5. Relative quantification real-time PCR analysis is performed using Sybergreen Fastmix ROX dye (Primer pair sequences used are described in Table 3). The housekeeping gene β-actin is used as endogenous control. 6. Results are presented as fold change relative to a Ctrl sample (which is always attributed the value 1) [14] (see Note 10).

4  Notes 1. If the prostatic tumor is large, it should appear immediately after entering the abdominal cavity (Fig. 2b I). We recommend in this cases to proceed with the dissection very similarly to what was described previously. Retract the fat pads (that should be laterally to the tumor) (Fig. 2b II–IV) to expose the testis and ductus deferens (Fig. 2b IV). Then proceed by looking for the bladder (which in very large tumors can be ­displaced from its normal location) and pulling it up (Fig. 2b V–VI). Proceed with cutting the ductus deferens (Fig. 2b VII), and the urethra (Fig. 2b VIII) separating the testis and the connective tissue from the rest of UGS.

52

Ana-Rita Pedrosa et al.

Table 3 Primer pair sequence list Gene

Forward sequence (5′–3′)

Reverse sequence (5′–3′)

Pecam

CAAGCAAAGCAGTGAAGCTG

TCTAACTTCGGCTTGGGAAA

Vegf-a

GGAGAGCAGAAGTCCCATGA

ACACAGGACGGCTTGAAGAT

Vegf-r1

GACCCTCTTTTGGCTCCTTC

CAGTCTCTCCCGTGCAAACT

Vegf-r2

GGACTCTCCCTGCCTACCTC

CGGCTCTTTCGCTTACTGTT

Vegf-r3

CGAAGCAGACGCTGATGATA

CCCAGGAAAGGACACACAGT

Pdgfr-β

TGATGAAGGTCTCCCAGAGG

AGGAGATGGTGGAGGAAGTG

Pdgf-b

CCTCGGCCTGTGACTAGAAG

TTTCGGTGCTTGCCTTTG

Tek

CCCCTGAACTGTGATGATGA

CTGGGCAAATGATGGTCTCT

Ang-1

CCATTTCGAGACTGTGCAGAT

CCCATTCACATCCATATTGC

Ang-2

CCTGGAGGTTGGACTGTCAT

CCCAGCCAGTACTCACCATC

Hif1α

GCCTTAACCTGTCTGCCACT

GGAGCCATCATGTTCCATTT

TGTTACCAACTGGGACGACA

GGGGTGTTGAAGGTCTCAAA

Control primers β-Actin

2. One can choose to mount up a Petri dish with black background (e.g., using black resin or Indian ink) to allow better contrast and facilitate the microdissection of the mouse prostate, or alternatively this also can be done using a dissecting microscope or stereoscope. If the prostatic tumor is large, proceed in a similar fashion as described previously, but pay extra attention, because the normal structures may be harder to identify (Fig. 3b). Always use the seminal vesicles and the bladder as a reference (Fig. 3b I). Proceed with separating the bladder and the seminal vesicles from the prostate, as described previously (Fig. 3b II–V). Transfer the prostate into a new 10 cm Petri dish containing fresh Dissecting Media. In the end always use the entrance of the urethra as a reference point if division of the prostate is desired (Fig. 3b VI–VIII). 3. Mice can respond differently to avertin. Therefore if the indicated dosage proves insufficient to drive a complete loss of consciousness, evidenced by complete loss of response to toe or tail pinch, an additional 0.05–0.1 ml of avertin solution can be administered. Preferably always use fresh (less than 1 month old) avertin 2.5% solution due to degradation of tribromoethanol which can lead to hepatotoxic and nephrotoxic effects. The mice should remain anesthetized for 30 m to 1 h and fully recover after 1–2 h.

Neo-angiogenesis Analysis

53

4. This step should be performed inside the laminar flux chamber to avoid the toxic vapors of xylene. 5. Put a little bit of mounting medium, like Entellan, with the help of a plastic pipette in the periphery of the slide and gently cover with the coverslip in a way that no air bubbles are formed between the two. 6. Some antibodies work better in a 3% H2O2 methanol solution, so if your antibody is not working using 3% H2O2 in distilled water try using 3% H2O2 in methanol. 7. For the negative Ctrls of the stainings, use appropriate isotype IgGs (rabbit or goat—see Table 1) as primary antibodies followed by appropriate secondary antibodies using the same exact conditions in the staining used for the tested antibody. 8. Reaction occurs when a brown color begins to appear, and the time may vary between abs (1–10 min). It is important to be consistent in relation to incubation time between different slides using the same ab. 9. The Hypoxyprobe™-1 Plus Kit brings solid pimonidazole HCl (which has to be diluted according to the manufacturer’s instructions), the primary (FITC conjugated to mouse IgG1 monoclonal antibody—FITC Mab1), and the secondary abs (rabbit anti-FITC conjugated with horseradish peroxidase). 10. When performing a relative transcription analysis on whole prostatic tissues (not when ECs are selected and isolated) between two mouse groups in which the variant (e.g., induction of an over-expression of a specific gene) causes changes in the number of vessels present, the fold changes in mRNA levels of other analyzed target genes should be normalized to Pecam mRNA levels, to compensate for variations in vascular density. References 1. Jemal A, Bray F, Center MM et al (2011) Global cancer statistics. CA Cancer J Clin 61:69–90 2. World Health Organization (2014) World cancer report, 2014, WHO report. WHO, Geneva 3. Taylor RA, Toivanen R, Risbridger GP (2010) Stem cells in prostate cancer: treating the root of the problem. Endocr Related Cancer 17:R273–R285 4. Hayward SW, Rosen MA, Cunha GR (1997) Stromal-epithelial interactions in the normal and neoplastic prostate. Br J Urol 79(Suppl 2): 18–26 5. Russo G, Mischi M, Scheepens W et al (2012) Angiogenesis in prostate cancer: onset, progression and imaging. BJU Int 110:E794–E808

6. Borre M, Offersen BV, Nerstrøm B et al (1998) Microvessel density predicts survival in prostate cancer patients subjected to watchful waiting. Br J Cancer 78:940–944 7. Bono AV, Celato N, Cova V et al (2002) Microvessel density in prostate carcinoma. Pros Cancer Pros Dis 5:123–127 8. Papetti M, Herman IM (2002) Mechanisms of normal and tumor-derived angiogenesis. Am J Physiol Cell Physiol 282:C947–C970 9. Greenberg NM, DeMayo F, Finegold MJ et al (1995) Prostate-cancer in a transgenic mouse. Proc Natl Acad Sci U S A 92:3439–3443 10. Cunha GR, Donjacour AA, Cooke PS et al (1987) The endocrinology and developmental biology of the prostate. Endocr Rev 8:338–362

54

Ana-Rita Pedrosa et al.

11. Berquin IM, Min Y, Wu R et al (2005) Expression signature of the mouse prostate. J Biol Chem 280:36442–36451 12. Kaplan-Lefko PJ, Chen T-M, Ittmann MM et al (2003) Pathobiology of autochthonous prostate cancer in a pre-clinical transgenic mouse model. Prostate 55:219–237

13. Gratton JP, Lin MI, Yu J et al (2003) Selective inhibition of tumor microvascular permeability by cavtratin blocks tumor progression in mice. Cancer Cell 4:31–39 14. Pedrosa A-R, Trindade A, Carvalho C et al (2015) Endothelial Jagged1 promotes solid tumor growth through both pro-­angiogenic and angiocrine functions. Oncotarget 6:24404–24423

Chapter 3 Methodologies Applied to Establish Cell Cultures in Prostate Cancer Anne T. Collins Abstract This chapter focuses on primary cultures of the human malignant prostate. Current abilities to isolate and culture stem cells, transit-amplifying cells, and secretory luminal cells are described. Advantages and limitations of this model system are also discussed. Key words Primary culture, Prostate, Cancer, Feeders, Epithelial cells, Stromal cells

1  Introduction The prostate gland is conspicuous because of its propensity for developing disease; prostate cancer is the most frequently diagnosed cancer in men and is excelled only by lung cancer as the leading cause of cancer-related mortality among men. Prostate cancer is treated by surgery or radiation when confined to the gland and with androgen deprivation therapy, when it is no longer confined to the prostatic capsule. However, relapse is inevitable from hormone sensitive to castrate-resistant disease. The treatment of advanced prostate cancer remains disappointing and is hindered by the lack of relevant preclinical models to improve development of effective therapeutic strategies. Indeed, the limitations of current preclinical models are increasingly cited as a key cause of the low success rate of oncology drug development [1]. Traditionally, preclinical models of prostate cancer are cell lines cultivated in monolayer or xenografts derived from them. Unlike other solid tumors, few prostate cell lines are available and as such do not represent the heterogeneity and complexity of this disease. A key consideration is the length of time these cell lines have been in culture, undergoing extensive adaptation and selection. It is perhaps not surprising that there is little correlation with clinical response.

Zoran Culig (ed.), Prostate Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 1786, https://doi.org/10.1007/978-1-4939-7845-8_3, © Springer Science+Business Media, LLC, part of Springer Nature 2018

55

56

Anne T. Collins

Primary cell culture provides a model system that allows a greater number of patients to be studied and thus more likely reflects the diversity of this disease. The development of an improved in vitro system (replacing serum with growth factors and hormones) has permitted a number of basic and translational studies [2]. However, clonal growth can only be achieved with the use of fibroblast feeder cells. This technique also allows the expansion of subpopulations of cells, such as stem cells [3, 4]. The use of primary cultures is certainly becoming more common, but there are still challenges, such as maintenance of a fully differentiated, androgen-responsive phenotype and identification of cell surface markers with which to isolate cancer cells from nonmalignant epithelial cells. Careful histological examination of each biopsy must be undertaken to select for malignant cells, but most biopsies are a mixture of normal and malignant cells and with continuous culture, normal cells will outgrow the malignant cells [5, 6]. This chapter aims to illustrate the methods of processing, sorting, and establishing cultures from malignant primary prostate, and the challenges that this poses.

2  Materials 2.1  Primary Epithelial Growth Medium

To 500 ml of keratinocyte serum-free medium (KSFM) add 100 nM R1881, 2.5 μg of human recombinant epidermal growth factor (EGF), 25 mg of bovine pituitary extract (BPE), and 1% of 200 nM l-glutamine. Both EGF and BPE are supplied with the basal medium and have a shelf life of 6 months when stored at −20 ° C. Keratinocyte basal medium has a shelf life of 6 months, at +4 ° C. l-Glutamine is maintained as a stock solution, at −20 °C and is stable for 24 months at this temperature (see Note 1).

2.2  Stem Cell Medium

To 500 ml of complete KSFM add the following supplements: 2 ng/ml leukemia inhibitory factor (LIF), 2 ng/ml stem cell factor (SCF), 100 ng/ml cholera toxin (CT), 1 ng/ml granulocyte macrophage colony-stimulating factor (GM-CSF).

2.3  Stromal Growth Medium

RPMI 1640, 10% Fetal calf serum (FCS), 1% (200 nM), and 100 nM R1881.

2.4  Serum-Free Stromal Growth Medium

Dulbecco’s MEM/Nutrient Mix F12 1:1 basal media supplemented with 2.5 μg/ml insulin, 2.5 μg/ml transferrin, 1 μM dexamethasone, 5 μg/ml selenium, 10 ng/ml EGF, 25 μg/ml bovine pituitary extract, 100 nM R1881, and 25 μg/ml heparin.

2.5  Trypsin/EDTA Solution (T/E)

0.05% trypsin/EDTA (w/v). pH 7.4. A stock (0.5%) solution can be purchased from InVitrogen. Keep at −20 C. Dilute 1:10 before use.

l-glutamine

Primary Culture of Prostate Cancer Cells

57

2.6  Phosphate-­ Buffered Saline (PBS): Dulbecco’s Phosphate-Buffered Saline

KCl (0.2 g/l), KH2PO4 (0.2 g/l), NaCl (0.8 g/l), Na2HPO4·7H2O (2.16 g/l). Filter sterilize (0.22 μm bottle filter).

2.7  Transport Medium

To RPMI 1640 add 5% fetal calf serum (FCS), 100 nM R1881, 100 U/ml penicillin, and 100 μg/ml streptomycin. Add aprotinin at 10,000 IU/ml (20 μl/10 ml of transport medium) if the tissue cannot be processed immediately.

2.8  Collagenase-­ Digestion Medium

Dissolve collagenase Type 1 (Worthington collagenase, Lorne Diagnostics, St Edmunds, UK) in 7.5 ml of transport medium. Stir to dissolve and filter sterilize through a 0.22 μm filter. Use fresh. The final concentration of collagenase should be 1000 IU in a volume of 7.5 ml/g of tissue.

2.9  Complete RMPI 1640

RMPI 1640 supplemented with 10% FCS.

2.10  Freezing Medium

To 70 ml of RPMI 1640, add 20 ml of FCS and finally 10 ml dimethylsulfoxide (DMSO). Filter (0.2 μm), aliquot, and store at −20 °C.

2.11  STO Feeder Cells

STO cells (mouse embryonic fibroblasts) are cultured in DMEM, 10% FCS, and 2 mM Glutamine.

2.12  MACS Cell Buffer

PBS, 2 mM EDTA, and 0.5% FCS.

2.13  Milteny Equipment

QuadroMACS Separator or MidiMACS Separator (Miltenyi Biotec), MACSmix Tube Rotator (Miltyenyi Biotec).

3  Methods 3.1  Establishment of Primary Cell Culture from Prostate Cancer Biopsies

1. Tissue (see Note 2 for tissue sources) should be transported from the hospital theatre in transport medium and processed the same day. Place a small piece of tissue (3 × 3 × 1 mm) in a plastic holder with OCT (optimum cutter temperature compound). Snap freeze in a Dewar containing liquid nitrogen. Place in a plastic bijou, label and store at −80 °C. For histology, place a small piece of tissue in 10% phosphate-buffered formaldehyde. Store overnight and then transfer into 70% ethanol. Fixed tissue can be stored indefinitely at room temperature. 2. Weigh out fresh collagenase. The final concentration of collagenase should be 1000 IU in a total volume of 7.5 ml/g of

58

Anne T. Collins

tissue. For example, if the specific activity of collagenase is 148 U/mg = 1000 × 0.001/148 = 0.00675 g of collagenase/g of tissue. Dissolve the collagenase in a total volume of 7.5 ml RMPI1640. Filter (0.2 μm filter). Store at 4 °C until use. Double the volume if >1 g of tissue is used. 3. Aspirate transport medium from the sample tube, wash with PBS, and place the tissue in a 10 cm petri dish. Add the collagenase solution and mince tissue using fine forceps and a scalpel, or scissors if preferred. Make sure the tissue is chopped finely (2 × 2 mm3). 4. Transfer into a 125 ml Erlenmeyer flask (small sterile conical flask with sterile black screw-top lid). Incubate at 37 °C overnight in an orbital shaker, at 80 rpm. The tissue should be incubated in collagenase for a maximum of 17 h. Store digest at 4 °C if you cannot proceed promptly on Day 2. 5. Decant the digest into a universal tube and triturate by repeat pipetting with a 5 ml pipette and then pass through a 21g blunt needle (Monoject Blunt Cannula 21G) (see Note 3). Spin at 380 × g for 3 min to sediment the cells. Aspirate the supernatant and resuspend the cell pellet in 10 ml of PBS. 6. Proceed to separate stroma from epithelial cells if you are working with large volumes of tissue (e.g., 1 g or more) otherwise follow the protocol from 10 to 14. 7. Centrifuge at 800 rpm for 1 min. Epithelial cells will pellet and the supernatant will contain single stromal cells (see Note 4). 8. Collect the clumps of epithelial cells (organoids) using a Pasteur pipette. The pipette is rinsed in PBS before collecting the organoids to avoid tissue sticking to the wall of the pipette. 9. Repeat until all the organoids are collected. Stromal cells (supernatant) can be cultured separately (see Subheading 3.3). 10. Pellet the epithelial cells by centrifugation at 380 × g for 3 min and resuspend in 5 ml of complete KSFM. 11. Transfer the cell suspension in complete KSFM to collagen-­ coated petri dishes (100 mm). 12. Incubate at 37 °C in a 95% air/5% CO2, humidified incubator. 13. After 3–4 days, clumps of epithelial cells should have attached [7]. Remove the medium and feed with fresh growth medium. 14. Refeed with growth medium every 3–4 days until cells are semi-confluent. At this time subculture or freeze. 3.1.1  Subculture

1. To subculture remove medium, wash once with PBS and add trypsin/EDTA solution (1 ml per 100 mm dish). 2. Incubate at 37 ° C for approximately 3 min or until the cells have rounded up.

Primary Culture of Prostate Cancer Cells

59

3. Resuspend cells in complete RPMI with vigorous pipetting. Transfer the suspended cells to a centrifuge tube and rinse the plate with PBS and add to the centrifuge tube. 4. Spin cells in a centrifuge at 380 × g for 3–5 min. 5. Discard the supernatant and resuspend in complete KSFM. 6. Passage the cells at a ratio of 1:2 onto collagen-coated dishes (see Notes 5 and 6). 7. Incubate at 37 °C and refeed growth medium every 3–4 days. It is important to subculture when cells are semi-confluent and actively dividing. 3.1.2  Freezing and Thawing Cells

1. Prepare cells from semi-confluent plates as described under subculture (Subheading 3.1.1; steps 1–4). 2. Resuspend cells in freezing medium to yield a concentration of 5 × 106 cells per 1 ml. 3. Transfer 1 ml of cell suspension to sterile cryovials. 4. Place cryovials in a polystyrene cold box and transfer to −80 °C overnight. 5. Transfer to liquid nitrogen for long-term storage. 6. To thaw cells, remove cryovials from liquid nitrogen storage and immediately place in a 37 °C water bath to thaw. Agitate to hasten thawing. 7. Wipe the outside of the cryovial with 70% ethanol before transferring cells (drop-wise) into a centrifuge tube containing 5 ml of FCS. 8. Spin in a centrifuge at 380 × g for 3 min. 9. Discard supernatant and resuspend cells in 5 ml of complete KSFM. 10. Transfer cells to collagen-coated petri dishes (100 mm). 11. Feed every 3–4 days until semi-confluent.

3.2  Establishment of Primary Cell Culture from Stem, and Transit-Amplifying Cells

1. Follow steps 1–9 of Subheading 3.1 if you are working with large volumes of tissue (e.g., 1 g or more). Otherwise follow steps 1–5 of Subheading 3.1. 2. Pellet the cells at 380 × g for 3 min and resuspend in 1× trypsin (5 ml). 3. Incubate at 37 °C for 30 min in an orbital shaker at 80 rpm. The cells will become stringy and fibrous in trypsin. 4. Stop the trypsin with complete RMPI1640 medium (5 ml). Shake vigorously before centrifugation (380 × g for 3 min) and resuspend pellet in 15 ml of complete RPMI 1640. 5. Pass the sample through a cell strainer (70 μm) and collect in a 50 ml collecting tube.

60

Anne T. Collins

6. Cells can be cultured from the cell clumps that remain on the cell strainer. Otherwise, wash the strainer with 5 ml of complete growth medium and plate onto a collagen-coated 100 mm petri dish. Proceed with steps 12–14 of Subheading 3.1. 7. Block a type 1 collagen plate with heat-inactivated 0.3% BSA (in PBS) for 1 h at 37 °C. 8. To collect the α2β1integrinhi cells, incubate the cells that have passed through the cell strainer for 5–20 min at 37 °C. Collect the non-adherent cells by vigorously washing in PBS until the entire non-adherent fraction is collected. Check using a microscope. 9. Collect the rapidly adherent cells (α2β1integrinhi fraction) by incubating with 1× trypsin for up to 10 min. Wash plate vigorously with PBS and pool together collecting any remaining cells. 10. Pellet the α2β1integrinhi cells and resuspend in MACS buffer. 11. Proceed to CD133 selection either using MACS selection [7] or flow cell sorting. 12. Resuspend the CD133 cells in SCM and plate onto collagen-­ coated dishes (100 mm) in the presence of irradiated (STO) feeder cells (5 × 105 STOs per 100 mm dish). See Subheading 3.2.1 for the preparation of feeder cells and Note 7. 13. The STOs should be freshly prepared and confluent if clonal growth is to be achieved. 14. Wash the plate every second day with PBS and refeed. 15. Once colonies are apparent (Fig. 1), STOs are not required and the culture can be subcultured at a ratio of 1:2 as stipulated in Subheading 3.1. See Note 8. 3.2.1  Preparation of Feeder Cells

1. STO cells (mouse SIM embryonic fibroblasts) can be purchased from the European Collection of Cell Cultures (ECACC). 2. Trypsinize a semi-confluent flask of cells and resuspend in complete DMEM in a universal collecting tube. 3. Either irradiate (60 Gy) or treat with mitomycin C. 4. Dissolve 2 mg of mitomycin C by adding 2 ml of PBS to the vial. 5. Add 1 mg to 100 ml of culture medium. 6. Add 10 ml of this solution to a T75 flask of STOs for 3 h. 7. Remove the mitomycin C, wash cells 3× with PBS, trypsinise and resuspend in stem cell medium. 8. Irradiated and mitomycin C treated STOs can be stored at 4 °C for 5 days.

3.3  Establishment of Stromal Cell Cultures

1. Spin the supernatant collected in Subheading 3.1 step 9 at 380 × g for 3 min. 2. Resuspend in either complete RPMI 1640 or serum-free stromal growth medium and plate onto T75 flasks.

Primary Culture of Prostate Cancer Cells

61

Fig. 1 Phase-contrast micrograph of a human epithelial colony after 20 days growth in stem cell medium

3. Do not change the medium until cells begin to attach. This process can take up to 2 weeks. 4. Once cells are apparent, wash the flask with PBS and refeed with complete stromal growth medium. 5. Subculture once the cells are semi-confluent. Stromal cells are not density-dependent and can be passaged at a ratio of 1:5 or more. 3.4  Isolation and Culture of Secretory Luminal Cells

1. Follow steps 1–5 of Subheading 3.1. 2. Pellet the epithelial cells at 380 × g for 3 min and resuspend in 1× trypsin (5 ml). 3. Incubate at 37 °C for 30 min in an orbital shaker at 80 rpm. The cells will become stringy and fibrous in trypsin. 4. Stop the trypsin with complete RMPI1640 medium (5 ml). Shake vigorously before centrifugation (380 × g for 3 min) and resuspend pellet in 15 ml of complete RPMI 1640. 5. Pass the sample through a cell strainer (70 μm) and collect in a 50 ml collection tube. 6. Slowly add 15 ml of lymphocyte separating medium (MP Biomedical) to the bottom of the tube. 7. Spin, with the brake off, at 530 × g for 30 min at room temperature. 8. Collect cells at the interphase, with a wetted Pasteur pipette. 9. Transfer cells to a new 50 ml tube and add the equivalent volume of complete RPMI 1640 medium. 10. Spin at 530 × g for 15 min (with the brake on) to pellet the cells. 11. Discard the supernatant and resuspend the cells in cold MACS buffer. 12. Spin at 380 × g for 3 min and resuspend the pellet in 80 μl of cold MACS buffer.

62

Anne T. Collins

13. Add 20  μl of biotin-antibody cocktail (Lineage depletion kit—human). 14. Incubate for 10 min at 4 °C in MACSmix Tube Rotator. 15. Add 4 ml of MACS buffer. 16. Spin at 380 × g for 5 min at 4 °C and discard the supernatant. 17. Resuspend cells in 80 μl of MACS buffer. 18. Add 40  μl of FcR blocking reagent, 40 μl of anti-CD31 microbeads (CD31 human MicroBead kit), and 40 μl of anti-biotin microbeads (Lineage depletion kit—human). 19. Incubate for 10 min at 4 °C in MACSmix Tube Rotator. 20. Add 4 ml of MACS buffer and spin at 380 × g for 5 min at 4 °C. 21. Discard the supernatant and resuspend the cells in 500μl of MACS buffer. 22. Proceed to magnetic separation following the manufacturer’s instructions. 23. Collect the flow-through (Lin−/CD31− fraction) and resuspend in 80 μl of MACs buffer. 24. Add 20  μl of biotin-CD24 antibody and incubate for 15 min at 4 °C on a MACSmix Tube Rotator. 25. Add 4 ml of MACS buffer and spin at 380 × g for 5 min at 4 °C. 26. Discard the supernatant and resuspend cells in 80 μl of MACS buffer. 27. Add 20μl of anti-biotin microbeads and incubate for 15 min at 4 °C on a MACSmix Tube Rotator. 28. Add 4 ml of MACS buffer to the cells and spin at 380 × g for 5 min at 4 °C. 29. Discard the supernatant and proceed to MACS selection following the manufacturer’s instructions. 30. The cells retained in the column will be the CD24+ fraction. 31. To increase the purity of the CD24+ fraction, elute the column and pass over another MS column. 32. Resuspend in 1 ml of SCM and add to a collagen-coated plate or process for downstream analysis. 33. Luminal cells are dependent upon androgens and are cultured in the presence of 10 nM androgen (R1881, or DHT) (see Note 9). 3.5  Epithelial Characterization

Characterization of the cultured cells can be carried out using a panel of antibodies (Table 1). Patterns of cytokeratin expression serve as valuable markers for the epithelial phenotypes observed in the normal and malignant prostate. Luminal cells consistently express the primary cytokertins (numbers 8, 18, and 19) characteristic

Primary Culture of Prostate Cancer Cells

63

Table 1 Antibodies used for detection of cytokeratins, clusters of differentiation, prostate-specific antigen, androgen receptor, and p63

Antibody (clone)

Specificity

Epithelial phenotype

AE1

Cytokeratin 5

Basal

LL002

Cytokeratin 14

Basal

CY-90

Cytokeratin 18

Luminal

DB105

CD44

Basal

32D12

CD24

Luminal

28IA4

PSA

Luminal

AR-441

AR

Luminal

4A4

P63

Basal

of simple epithelia [8], while basal cells express mainly large molecular weight forms (numbers 4–7, 10, 11, 14, and 15) [9]. Epithelial cells intermediate between basal and luminal cells are also present in the normal and malignant prostate and make up the majority of cells in primary culture [6]. The androgen receptor is present in most cells, but is weakly expressed and is localized perinuclearly, even in the presence of androgens in the medium. Prostate-specific antigen (PSA) is only expressed by rare cells and most notably if cultures are allowed to become confluent. CD44 and CD24 are cell surface markers which label the basal and luminal cells, respectively. Prostate cancer is characterized by loss of basal cells. However, primary cells from malignant prostate are characteristically intermediate as they express markers associated with basal and luminal cells. It is therefore difficult to exclude the possibility that normal cells are present in the culture. One marker that can distinguish between normal and malignant prostate is expression of the fusion transcript TMPRSS2:ERG (transmembrane protease, serine 2:v-ets erythroblasosis virus E26 oncogene homolog fusion product). Cultures can be screened using an RT-PCR based approach as a further test for malignancy. The presence of the fusion is found in approximately half of all men with prostate cancer [10]. In our hands AMACR does not discriminate between normal and malignant cells in primary culture. 3.6  Sphere Formation

Like most cells propagated as a monolayer in vitro prostate epithelia are proliferating very rapidly and are most likely in a state of regeneration rather than steady state. In vitro the differentiation program can best be obtained by culturing the epithelial cells on collagen rafts or in Matrigel together with stromal

64

Anne T. Collins

Fig. 2 (a) Prostatic-like acini derived from a patient with a Gleason 9 tumor (Day 14). (b) Cross section of acini stained with high molecular weight keratins (clone 34βE12). 103 malignant epithelial cells grown in monolayer were seeded into Matrigel together with primary stromal cells in the presence of 10 nM DHT

cells and androgens [11]. Figure 2 demonstrated that epithelial cells grown in 3D and in the presence of stroma, androgens and Matrigel induce prostatic-like acini.

4  Notes 1. l-Glutamine: l-Glutamine is not as stable as other amino acids in culture and is extremely sensitive to storage temperatures, pH, and age. Degradation increases with temperature and can lead to the buildup of ammonia. Avoid freeze-thaw cycles and if a stock solution is maintained, thaw the stock solution once, aliquot into smaller volumes, and freeze. l-Glutamine may precipitate during the thawing process. Place the aliquot in a 37 °C water bath and gently swirl during thawing. Use immediately once it has thawed. 2. Cancerous tissue can be obtained from a channel transurethral resection of the prostate (cTURP) or a radical prostatectomy. Depending upon the stage of disease, targeted needle biopsies (from patients undergoing radical prostatectomy) to remove malignant and normal tissue can be carried out without compromising diagnosis (14g needle gauge). A normal area of the prostate can also be biopsied for subsequent culture. 3. The tissue digest should pass through a 21g needle easily. The finer the tissue is chopped (on Day 1) the better the yield, and the easier it will pass through the needle.

Primary Culture of Prostate Cancer Cells

65

4. Collagenase digests prostatic tissue into clumps or organoids of epithelial cells and predominately single stromal cells. For high Gleason grade cancers, epithelial cells cannot be isolated from the stromal fraction by differential centrifugation. Omit this step and proceed to primary culture of the epithelial cells following collagenase digest. 5. It is important that the cells are seeded at high density for optimal growth. As a rule, 2 × 105 cells per 100 mm collagencoated dish. Lowering the density will result in differentiation. Feeder cells can be used if clonal growth is required. 6. Colony forming efficiencies average 80% for secondary cultures from prostate cancer biopsies (if subcultured at a ratio of 1:2) and decrease with subsequent subculture. Generally, cultures from prostate cancer biopsies undergo 25–28 population doublings before becoming senescent [12]. 7. STO cells are used routinely as feeder layers prepared by irradiation or mitomycinic treatment for maintenance of teratocarcinoma stem cells in the undifferentiated state. They are also used to maintain prostate stem cells [4]. 8. To derive colonies from rare populations, the petri dish should be confluent (with STOs) but not overcrowded. It is also important to get rid of cells that have not plated as dying/ dead cells will affect colony growth. A confluent layer of STOs is necessary UNTIL epithelial colonies are apparent. 9. To maintain luminal cells in primary culture, it is important to supplement the growth media with androgens [13]. However, they are inevitably lost in primary culture, due to overgrowth with highly proliferative intermediate cells [6]. References 1. Ellis LM, Fidler IJ (2010) Finding the tumor copycat. Therapy fails, patients don’t. Nat Med 16:974–975 2. Peehl DM, Wong ST, Stamey TA (1988) Clonal growth characteristics of adult human prostatic epithelial cells. In Vitro 24:530–536 3. Hudson DL, O’Hare M, Watt FM, Masters JRW (2000) Proliferative heterogeneity in the human prostate: evidence for epithelial stem cells. Lab Investig 80:1243–1250 4. Collins AT, Habib FK, Maitland NJ, Neal DE (2001) Identification and isolation of human prostate epithelial stem cells based on α2β1-­ integrin expression. J Cell Sci 114:3865–3872 5. Birnie R, Bryce SD, Roome C et al (2008) Gene expression profiling of human prostate cancer stem cells reveals a pro-inflammatory

phenotype and the importance of extracellular matrix interactions. Genome Biol 9:R83. https://doi.org/10.1186/gb-2008-9-5-r83 6. Robinson EJ, Neal DE, Collins AT (1998) Basal cells are progenitors of luminal cells in primary cultures of differentiating human prostatic epithelium. Prostate 37:149–160 7. Richardson GD, Robson CN, Lang SH, Neal DE, Maitland NJ, Collins AT (2004) CD133, a novel marker for human prostatic epithelial stem cells. J Cell Sci 117:3539–3545 8. Brawer MK, Peehl DM, Stamey TA, Bostwick DG (1985) Keratin immunoreactivity in the benign and neoplastic human prostate. Cancer Res 45:3663–3667 9. Wernert N, Seitz G, Achstätter T (1987) Immunohistochemical investigation of differ-

66

Anne T. Collins

ent cytokeratins and vimentin in the prostate from the fetal period up to adulthood and in prostate carcinoma. Pathol Res Pract 182:617–626 10. Tomlins SA, Rhodes DR, Perner S et al (2005) Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310:644–648 11. Lang SH, Stark M, Collins A, Paul AB, Stower MJ, Maitland NJ (2001) Experimental pros-

tate epithelial morphogenesis in response to stroma and three dimensional matrigel culture. Cell Growth Differ 12:631–640 12. Collins AT, Berry PA, Hyde C, Stower MJ, Maitland NJ (2005) Prospective identification of tumorigenic prostate cancer stem cells. Cancer Res 65:10946–10951 13. Cunha GR, Bigsby RM, Cooke PS, Sugimura Y (1985) Stromal-epithelial interactions in adult organs. Cell Differ 17:137–148

Chapter 4 Protocols for Migration and Invasion Studies in Prostate Cancer Arjanneke F. van de Merbel, Geertje van der Horst, Jeroen T. Buijs, and Gabri van der Pluijm Abstract Prostate cancer is the most common malignancy diagnosed in men in the western world. The development of distant metastases and therapy resistance are major clinical problems in the management of prostate cancer patients. In order for prostate cancer to metastasize to distant sites in the human body, prostate cancer cells have to migrate and invade neighboring tissue. Cancer cells can acquire a migratory and invasive phenotype in several ways, including single cell and collective migration. As a requisite for migration, epithelial prostate cancer cells often need to acquire a motile, mesenchymal-like phenotype. This way prostate cancer cells often lose polarity and epithelial characteristics (e.g., expression of E-cadherin homotypic adhesion receptor), and acquire mesenchymal phenotype (for example, cytoskeletal rearrangements, enhanced expression of proteolytic enzymes and other repertory of integrins). This process is referred to as epithelial-to-mesenchymal transition (EMT). Cellular invasion, one of the hallmarks of cancer, is characterized by the movement of cells through a three-dimensional matrix, resulting in remodeling of the cellular environment. Cellular invasion requires adhesion, proteolysis of the extracellular matrix, and migration of cells. Studying the migratory and invasive ability of cells in vitro represents a useful tool to assess the aggressiveness of solid cancers, including those of the prostate. This chapter provides a comprehensive description of the Transwell migration assay, a commonly used technique to investigate the migratory behavior of prostate cancer cells in vitro. Furthermore, we will provide an overview of the adaptations to the Transwell migration protocol to study the invasive capacity of prostate cancer cells, i.e., the Transwell invasion assay. Finally, we will present a detailed description of the procedures required to stain the Transwell filter inserts and quantify the migration and/or invasion. Key words Prostate cancer, Migration, Invasion, Transwell migration assay, Transwell invasion assay

1  Introduction 1.1  Clinical Problem of Prostate Cancer

Prostate cancer is amongst the most commonly diagnosed cancers in men and represents the second cause of cancer death [1]. The development of distant metastases and therapy resistance represent major clinical problems in the management and treatment of prostate cancer patients. Initially, the treatment of organ-confined

Zoran Culig (ed.), Prostate Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 1786, https://doi.org/10.1007/978-1-4939-7845-8_4, © Springer Science+Business Media, LLC, part of Springer Nature 2018

67

68

Arjanneke F. van de Merbel et al.

prostate cancer (i.e., surgical removal and/or radiotherapy) is very effective in the majority of the prostate cancer patients. However, 20–30% of all prostate cancer patients will relapse [2]. These patients will subsequently be treated with androgen deprivation therapy, resulting in an initial regression of the tumor. However, these tumors will invariably grow, resulting in the development of castration-resistant prostate cancer (CRPC) and the formation of distant metastases in 70–80% of the CRPC patients [3, 4]. 1.2  Migration and Invasion of Prostate Cancer Cells

The most critical and life-threatening event in a patient with organ-­ confined prostate cancer is the acquisition of an invasive and migratory phenotype by the tumor cells. Migration is defined as the movement of cells across the body, and includes single cell migration and collective migration [5]. During single cell migration, individual cells migrate through a tissue by undergoing epithelial-to-mesenchymal transition (EMT) [6]. EMT is a biological process involved in regulating developmental events such as primitive streak formation and gastrulation [7]. However, EMT is also involved in adult pathology, including fibrosis [8] and cancer progression [9]. Several different processes are required for EMT, including activation of transcription factors (a.o. SNAI1, SNAI2, ZEB1, ZEB2, Twist) and altering the expression of specific cell-surface proteins (e.g., downregulation of E-cadherin, upregulation of vimentin and alpha-smooth muscle actin (α-SMA)) [9, 10]. In addition, it has been shown that for EMT reorganization of cytoskeletal proteins, production of ECM-degrading enzymes, and changes in the expression of specific microRNAs are required (e.g., miR-200 family) [9, 11, 12]. In prostate cancer, the ability of prostate cancer cells to undergo EMT is associated with an increased invasiveness and metastasis formation [9, 13, 14]. In contrast to single cell migration, collective cell migration is characterized by the coordinated movement of a cluster of cells [15]. During collective cell migration, a cluster of cells follows the front leader cell, which pulls the follower cells in the same direction. Importantly, cells retain their physical intracellular connections (e.g., E-cadherin-mediated adherens junctions). Furthermore, cells coordinate their cytoskeletal organization and their response to migratory signals, enabling a cluster of cells to respond as a single unit [5]. Recently, it has been shown that the coordination of collective migration is mediated by interactions between Rac1 small GTPase, angiomotin-Rich1 complex and the Merlin tumor-­ suppressor protein [16]. Collective cell migration is possible across two dimensions (i.e., as a monolayer of cells migrates across a tissue surface) and across three dimensions (as a cluster of cancer cells that invades neighboring stroma). Interestingly, collective cell migration is known to be involved in cancer and metastases formation [5, 17].

In Vitro Migration and Invasion Assays in Prostate Cancer

69

Besides the capacity of prostate cancer cells to migrate, invasion of the neighboring tissue is an important part of the metastatic cascade. Invasion is defined as the movement of motile cells through a three-dimensional matrix (i.e., the basement membrane and stroma) while remodeling the direct environment [18]. Invasion is involved in both developmental processes and pathological conditions, e.g., wound healing and metastatic disease [19, 20]. Since both migration and invasion are involved in prostate cancer invasiveness and metastases, examining the migratory and invasive capacity of prostate cancer cells in vitro represents an important method in the cancer field. Currently, different techniques are employed to study the migration and invasion of prostate cancer cells in vitro. In the section below, we describe the most commonly used assays to study prostate cells migration and invasion. 1.3  Migration Assays 1.3.1  Transwell Migration Assay

A commonly used method to study migration of prostate cancer cells in vitro is the Transwell migration assay, also called Boyden chamber assay. During this method, prostate cancer cells are seeded on top of a permeable filter insert [21]. This filter insert is positioned in a well of a culture plate and is placed in direct contact to medium supplemented with chemoattractants (e.g., serum or soluble growth factors) (Fig. 1a). Prostate cancer cells are allowed to migrate for a fixed period of time, after which the migration is stopped by fixing and staining the migrated cells on the filter inserts. Subsequently, cellular migration is quantified. The Transwell migration assay is a commonly used established technique in the field of prostate cancer research and has been performed with several established prostate cancer cell lines, including PC-3, DU-145, LNCaP, and C4-2B cells [22–26]. A major advantage of the Transwell migration assay is that the technique is relatively easy to perform. Moreover, the Transwell migration assay is very versatile; several adaptations to the protocol are possible depending on the research question. For instance, the effect of a growth factor on the migration can be investigated by the addition of growth factor of interest to either the upper or the lower chamber [27]. The effect of a gene of interest on the migration can be investigated with genetically modified prostate cancer cells (i.e., overexpression or knockdown/knockout of a gene) [26]. The Transwell migration assay cannot be used to assess cellular migration in real time (endpoint measurements), thus representing a limitation of this technology. 1. Setup of the Migration Assay One day prior to setting up the Transwell migration assay, the prostate cancer cells are serum-starved overnight. In this way, the prostate cells respond better to the chemoattractant gradient across the chamber. The next day, the serum-starved prostate cancer cells are seeded on top of the filters in starvation medium. The filter

70

Arjanneke F. van de Merbel et al.

A

B TRANSWELL

TRANSWELL

MIGRATION

INVASION ASSAY

ASSAY

Transwell

Filter

insert

coated with Filter

matrix

Medium

C C4

D

Migration

C4-2

C4-2B4

E

Migration **

number of migratory cells / field

% Area migratory cells

15

10

5

0

C4

C 4-2

C 4-2B

200 ****

150 **** 100 50 0

C4

C4-2

C4-2B

Fig. 1 Overview of the Transwell migration and Transwell invasion assay. Schematic overview of the (a) Transwell migration assay and (b) Transwell invasion assay. (c) Representative black and white images of the filter Transwell filter inserts (5× magnification) of a Transwell migration assay performed on C4, C4-2, and C4-2B4 prostate cancer cells. These images were used for quantification by using ImageJ (d) and by manually counting the migratory cells in each picture (e)

In Vitro Migration and Invasion Assays in Prostate Cancer

71

inserts are placed in a well of a culture plate containing medium supplemented with a growth factor or chemoattractant (e.g., serum). Subsequently, cells are allowed to migrate for a short period by incubation in a humidified incubator at 37 °C. 2. Stopping of the Migration by Fixation and Staining of the Cells on the Filter Inserts After incubation, migratory prostate cancer cells on the filter inserts can be fixed for histological evaluation. Commonly used fixative reagents for filter inserts include 4% paraformaldehyde and methanol [28]. Subsequently, the migrated cells (i.e., the cells on the outer/lower side of the filter) are stained and quantified. Different reagents are available to stain the filter inserts; the most commonly used dye to stain the filter inserts is crystal violet, a histological dye staining nuclei purple/blue [29, 30]. Other possibilities for staining the filter inserts include the Giemsa stain [22] or hematoxylin [31]. 3. Quantification of the Migration Different methods are used for quantifying the migration of the Transwell assay. An easy method is counting the number of cells on the filter. Subsequently, the percentage of migratory cells can be determined by calculating the ratio of migratory cells against the total number of seeded cells. Another possibility includes automatically measuring the percentage of the area of cells present in each insert by using image analysis software such as free available ImageJ software. Besides the histological quantification of the migration, it is possible to fluorescently label the cells (for instance with Calcein-AM). Subsequently, the migratory cells on the filter can be dissociated from the membrane by using Trypsin or commercially available dissociation solutions (e.g., from R&D Systems), by adding the solution to the lower chamber of the Transwell. Subsequently, the fluorescent signal of the collected cells can be quantified by using a fluorescence plate reader. There are several systems available based on quantifying the migration of fluorescent cells including the ThinCert™ system (Greiner Bio-one), the Cultrex Migration Assay (R&D Systems), and the BD FluoroBlok Insert System (BD Biosciences). 1.3.2  Scratch Assay

Another assay to study the migration of prostate cells is the scratch assay, also called the wound-healing assay. In this assay, a scratch is made in a confluent layer of cells using a pipette tip [32, 33]. Subsequently, migration is measured by monitoring the closure of the scratch by taking pictures at different time points or by using time-lapse microscopy. Several adaptations exist to the scratch assay, including the use of fluorescently labeled cells [34] and the use of a high-throughput automated screening [35].

72

Arjanneke F. van de Merbel et al.

Advantages of the scratch assay include its low costs and its simplicity; no special lab equipment or reagents are required. In contrast to the Transwell migration assay, the migration can be examined in real time with the scratch assay. A major disadvantage of the scratch assay includes its inability to discriminate between cellular proliferation and migration. 1.4  Transwell Invasion Assays

The Transwell invasion assay is an adaptation of the Transwell migration assay, and is often used to study the invasive ability of prostate cancer cells in vitro [36]. During the Transwell invasion assay, the filter inserts are coated with an extracellular matrix prior to seeding the cells (Fig. 1b). The filter inserts can be coated with several matrices, including collagen [37], fibronectin [24, 37], laminin [37], or Matrigel [38]. Subsequently, the experimental setup is similar to the Transwell migration assay, i.e., seeding serum-starved cells in the upper chamber, fixing, staining, and quantification. A common read-out of the Transwell invasion assay is the invasive index [39]; this is calculated by taking the ratio of invaded (passed through the ECM-coated filters) against the migrated cells (non-coated filters) [40]. Advantages of the Transwell invasion assay are the easy experimental setup and its versatility. One of the disadvantages of the Transwell invasion assay is that the invasion cannot be monitored in real time.

2  Materials 2.1  Cell Maintenance

1. Prostate cancer cells. For this experiment, LNCaP sublines C4, C4-2, and C4-2B4 were being used. See Note 1. 2. Cell culture medium supplemented with serum. For C4, C4-2, and C4-2B4 cells, we use T-medium supplemented with 10% Fetal Calf Serum (FCS). See Note 2. 3. Starvation medium, i.e., T-medium supplemented with 0.3% FCS. See Note 3.

2.2  Transwell Migration/Invasion Assay

1. 6.5 mm Transwell® 8.0 μm Pore Polycarbonate Membrane Inserts (Corning). See Note 4. 2. 24-well culture plates. 3. For Transwell invasion assay: Growth factor-depleted Matrigel (BD Biosciences). See Note 5. 4. Phosphate Buffered Saline (PBS). 5. Methanol. 6. 3.7% paraformaldehyde (PFA) in PBS. The 3.7% PFA solution is prepared by diluting 37% PFA stock solution in demi water.

In Vitro Migration and Invasion Assays in Prostate Cancer

73

7. Crystal Violet stock solution: 0.4 g of Crystal Violet dissolved 10 ml 100% ethanol. 8. Cotton swaps. 9. A pair of tweezers. 10. Light microscope connected with a microscope camera. 11. Computer with ImageJ software (National Institutes of Health).

3  Methods 3.1  Serum Starvation of the Prostate Cancer Cells 3.2  Preparation of the Transwell Chambers and Seeding of the Cells

1. Refresh the medium of the prostate cancer cells by adding starvation medium to the prostate cells. See Note 3. 1. Place the Transwell chambers in a 24-well plate. See Note 4. 2. For Transwell invasion assay: Thaw the Matrigel on ice and dilute the Matrigel in serum-free cell culture medium to a concentration of 2 mg/ml. See Note 5. 3. For Transwell invasion assay: Carefully add 20 μl of diluted Matrigel to the Transwell insert and spread the Matrigel across the filter by using a 100 μl pipette tip. 4. For the Transwell invasion assay: Incubate the Matrigel-coated Transwell inserts for 1 h at 37 °C. 5. Optional: For better attachment of the prostate cancer cells to the Transwell inserts, equilibrate the Transwell filter inserts prior to seeding the cells. Add 300 μl of T-medium to the lower compartment and 200 μl of medium to the upper compartment. See Notes 6 and 7. Incubate the plate for at least 1 h at 37 °C. Subsequently remove the medium from the upper compartment. 6. Trypsinize the cells, resuspend and seed 60,000 cells in 300 μl starvation medium in the upper compartment. See Note 8. 7. Allow the cells to migrate for 20 h. See Note 9.

3.3  Fixation and Staining of Migratory Cells on the Transwell Inserts

1. Pick up the insert by using a pair of tweezers and carefully decant the medium from the upper compartment. 2. Place the insert in a new well. 3. Add 500 μl 4% PFA solution to the lower compartment and incubate for 10 min at room temperature. 4. Pick up the insert by using a pair tweezers, decant the PFA from the upper compartment, and place the insert in a new well.

74

Arjanneke F. van de Merbel et al.

5. Add 500 μl methanol to the lower compartment and incubate for 10 min at room temperature. 6. Wash the inserts twice with PBS by dipping the inserts in a 50 ml tube filled with PBS. 7. Place the insert in a new well. 8. Dilute the Crystal Violet stock solution 1:20 in demi water. Subsequently, add 500 μl diluted Crystal Violet solution to the lower compartment. Let the Crystal Violet incubate for 30 min at room temperature. 9. Wash the inserts with tap water by dipping the inserts in a 50 ml tube filled with tap water. 10. Clear the upper side of the membrane with a cotton swap. See Note 10. 11. Let the filters dry overnight at room temperature before taking pictures. 3.4  Quantification of the Migration/ Invasion by ImageJ

1. Place the filter insert on a microscope slide. 2. Take five pictures of each membrane at a 5× magnification by using a light microscope. 3. Convert the pictures to black and white (B&W) images. 4. Quantify the B&W images with ImageJ by using the threshold tool. Adjust the threshold in a way that only the cells are highlighted. 5. Quantify by manually counting the number of migratory cells in each picture (Fig. 1d). Measure the percentage area of cells by using the “limit to threshold” and “area fraction” options in the ImageJ software (Fig. 1e) (see Note 11).

4  Notes 1. The C4, C4-2, and C4-2B4 cell lines were established from LNCaP cells [41–43]. For C4 cells, androgen-responsive LNCaP cells were subcutaneously co-inoculated with human bone fibroblast cell line (MS) in male athymic BALB/c mice. After 8 weeks, the mice were castrated. After maintenance of the tumor for 4 more weeks, the tumor was harvested and epithelial cells were enriched in vitro. Compared to the parental LNCaP cells, C4 cells are capable of forming tumors in vivo when co-inoculated with MS cells [41]. The C4-2 cell line was obtained by subcutaneous co-­ inoculation of C4 cells together with MS cells in castrated BALB/c mice. After 12 weeks, the tumor was harvested and epithelial cells were obtained by enrichment in vitro. Compared

In Vitro Migration and Invasion Assays in Prostate Cancer

75

to LNCaP cells, C4-2 cells are androgen-irresponsive and have an increased tumorigenic and metastatic capability in vivo [41–43]. The C4-2B4 cell line was derived by orthotopic injection of C4-2 cells in castrated BALB/c mice. Metastases were harvested from the lumbar spine. Compared to LNCaP and C4-2 cells, C4-2B4 cells are more aggressive and have an enhanced metastatic potential [42]. 2. The culture medium depends on the cell line being used. For culturing C4, C4-2, and C4-2B4 cells, our research group uses T-medium. T-medium is prepared by mixing 80% Dulbecco’s Modified Eagle Medium (DMEM) with 20% F-12K nutrient mixture Kaighn’s modification. This is supplemented with 10% Fetal Calf Serum (FCS), 1% Insulin-Transferrin-­ Selenium, 0.125 mg/ml Biotin, 12.5 mg/ml Adenine, 6.825 ng/ml T3, and 1% Penicillin-Streptomycin (P/S). 3. Prior to seeding the cells in the Transwell chambers, prostate cancer cells are serum-starved. Serum contains several growth factors and cytokines that could affect the migration of prostate cancer cells. An important principle in the Transwell migration/invasion assay is the establishment of a concentration gradient in growth factors across the Transwell chamber; while the serum-starved cells are resuspended and seeded in starvation medium in the upper chamber, the lower chamber contains medium supplemented with 10% serum. This concentration gradient across the Transwell chamber is a stimulus for the prostate cancer to migrate and invade. By starving the prostate cancer cells overnight and resuspending the cells in starvation medium, the prostate cancer cells respond better to the growth factor-gradient across the chamber (see also Note 7). An alternative for starvation medium is charcoal-stripped serum. Charcoal-stripped serum is depleted of lipophilic compounds (i.e., certain growth factors, hormones, and cytokines) that could affect the migration and invasion. 4. During this experiment, filter inserts with a pore size of 8.0 μm were used to assess the migration and invasion of prostate cancer cells. Inserts with different pore sizes are available, ranging from 0.4 to 8.0 μm. The choice of pore size depends on the cell type used and therefore on the research question. For migration and invasion assays, a pore size of 5.0 or 8.0 μm is often used, whereas a smaller pore size is used for other applications (e.g., co-culture experiments and drug transport studies). Besides the different pore sizes, filter inserts are available in different materials and a variety of sizes. The different materials of the filter inserts include polycarbonate, polyester, and collagen-coated polytetrafluoroethylene, and all have different

76

Arjanneke F. van de Merbel et al.

optical properties. Polyester filters are clear and have the best properties for visualizing the cells. Polycarbonate filters are translucent and have a poor cell visibility, whereas ­collagen-­coated polytetrafluoroethylene filters are clear when they are wet. For this experiment, 6.5 mm Transwell inserts were being used that fit in a 24 well plate. Available filter sizes range from 4.26 to 75 mm and fit in different culture plates ranging from a 96-well plate up to a 100 mm dish. 5. In order to study the invasion of prostate, Transwell filter inserts were coated with Matrigel. For coating, we used growth factordepleted Matrigel instead of normal Matrigel to exclude the effect of growth factors present in the normal formulation of Matrigel. When working with Matrigel, work as much as possible on ice, since Matrigel will immediately solidify at room temperature. Precooling all plates, pipette tips, and other equipment will enable easy coating of the Transwell filter inserts and will prevent Matrigel from solidifying. 6. Easy access to the lower compartment is possible by placing the pipette tip through the windows in the insert wall. Make sure that no air bubbles are present under the filter inserts as this can interfere with the migration. 7. Better results may be achieved after prior equilibration of the filter inserts. This equilibrium period improves the attachment of some cell types to the insert. 8. For the 6.5 mm filter inserts, we seed 60,000 cells. However, the optimal seeding density depends on the cell line and insert size being used. The growth and migration of cells on the filter inserts depends on seeding density. 9. The incubation period/migration time depends on the migratory capacity of the cells and on the setup of the experiments (e.g., use of chemoattractant and other growth factors). An optimal migration time is crucial for a migration assay. Too short incubation will result in a low/absent migration, whereas a too long incubation will result in a biased result caused by the proliferation of cells. For C4, C4-2, and C4-2B4 cells, we use an incubation time of 20 h. 10. Prior to quantification, non-migrated cells are removed by swapping the upper chamber. Be careful not to put too much pressure on the membranes when using the cotton swaps. The membranes are fragile and too much pressure will break the membrane or will cause the membrane to be uneven. In order to make the cotton swaps ready for use, pull and stretch the tip to make softer, wet it with water, and make it even. 11. There are different options to quantify the degree of migration in the Transwell assay. For example semiautomatically measur-

In Vitro Migration and Invasion Assays in Prostate Cancer

77

ing the percentage area of the migratory cells by using the thresholding option of the ImageJ software (Fig. 1d). Another option is to manually count the number of migratory cells in each field (Fig. 1e). Although measuring the ­percentage area of migratory cells is semiautomatic and therefore could reduce observer bias (i.e., the same threshold is set to all pictures of the different conditions), this method could have some serious drawbacks. Setting a correct threshold is often very difficult since selection of the migratory cells often results in selecting the pores of the filter insert as well. This could result in an incorrect quantification and could result in a loss of effect. Therefore, it is recommended to use at least two different methods to quantify the migration during this assay. References 1. Siegel RL, Miller KD, Jemal A (2015) Cancer statistics, 2015. CA Cancer J Clin 65(1):5–29. https://doi.org/10.3322/caac.21254 2. Bastian PJ, Carter BH, Bjartell A, Seitz M, Stanislaus P, Montorsi F, Stief CG, Schroder F (2009) Insignificant prostate cancer and active surveillance: from definition to clinical implications. Eur Urol 55(6):1321–1330. https:// doi.org/10.1016/j.eururo.2009.02.028 3. Pienta KJ, Bradley D (2006) Mechanisms underlying the development of androgen-­ independent prostate cancer. Clin Cancer Res 12(6):1665–1671. https://doi. org/10.1158/1078-0432.ccr-06-0067 4. Harris WP, Mostaghel EA, Nelson PS, Montgomery B (2009) Androgen deprivation therapy: progress in understanding mechanisms of resistance and optimizing androgen depletion. Nat Clin Pract Urol 6(2):76–85. https://doi.org/10.1038/ncpuro1296 5. Friedl P, Gilmour D (2009) Collective cell migration in morphogenesis, regeneration and cancer. Nat Rev Mol Cell Biol 10(7):445–457. https://doi.org/10.1038/nrm2720. nrm2720 [pii] 6. Larue L, Bellacosa A (2005) Epithelial-­ mesenchymal transition in development and cancer: role of phosphatidylinositol 3′ kinase/ AKT pathways. Oncogene 24(50):7443–7454. https://doi.org/10.1038/sj.onc.1209091 7. Nakaya Y, Sheng G (2013) EMT in developmental morphogenesis. Cancer Lett 341(1):9– 15. https://doi.org/10.1016/j. canlet.2013.02.037 8. Kalluri R, Neilson EG (2003) Epithelial-­ mesenchymal transition and its implications for

fibrosis. J Clin Invest 112(12):1776–1784. https://doi.org/10.1172/jci20530 9. van der Pluijm G (2011) Epithelial plasticity, cancer stem cells and bone metastasis formation. Bone 48(1):37–43. https://doi. org/10.1016/j.bone.2010.07.023. S87563282(10)01367-0 [pii] 10. Hugo H, Ackland ML, Blick T, Lawrence MG, Clements JA, Williams ED, Thompson EW (2007) Epithelial–mesenchymal and mesenchymal–epithelial transitions in carcinoma progression. J Cell Physiol 213(2):374–383. https://doi.org/10.1002/jcp.21223 11. Williams LV, Veliceasa D, Vinokour E, Volpert OV (2013) miR-200b inhibits prostate cancer EMT, growth and metastasis. PLoS One 8(12):e83991. https://doi.org/10.1371/ journal.pone.0083991 12. Kong D, Li Y, Wang Z, Banerjee S, Ahmad A, Kim HR, Sarkar FH (2009) miR-200 regulates PDGF-D-mediated epithelial-­mesenchymal transition, adhesion, and invasion of prostate cancer cells. Stem Cells (Dayton, Ohio) 27(8):1712–1721. https://doi.org/10.1002/ stem.101 13. Li P, Yang R, Gao WQ (2014) Contributions of epithelial-mesenchymal transition and cancer stem cells to the development of castration resistance of prostate cancer. Mol Cancer 13:55. https://doi. org/10.1186/1476-4598-13-55 14. Nauseef JT, Henry MD (2011) Epithelial-to-­ mesenchymal transition in prostate cancer: paradigm or puzzle? Nat Rev Urol 8(8):428– 439. https://doi.org/10.1038/ nrurol.2011.85

78

Arjanneke F. van de Merbel et al.

15. Haeger A, Wolf K, Zegers MM, Friedl P (2015) Collective cell migration: guidance principles and hierarchies. Trends Cell Biol 25(9):556– 566. https://doi.org/10.1016/j. tcb.2015.06.003 16. Das T, Safferling K, Rausch S, Grabe N, Boehm H, Spatz JP (2015) A molecular mechanotransduction pathway regulates collective migration of epithelial cells. Nat Cell Biol 17(3):276–287. https://doi.org/10.1038/ ncb3115 17. Ilina O, Friedl P (2009) Mechanisms of collective cell migration at a glance. J Cell Sci 122(Pt 18):203–3208. https://doi.org/10.1242/ jcs.036525 18. Kramer N, Walzl A, Unger C, Rosner M, Krupitza G, Hengstschlager M, Dolznig H (2013) In vitro cell migration and invasion assays. Mutat Res 752(1):10–24. https://doi. org/10.1016/j.mrrev.2012.08.001 19. Keller R (2005) Cell migration during gastrulation. Curr Opin Cell Biol 17(5):533–541. https://doi.org/10.1016/j.ceb.2005.08.006 20. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144(5):646–674. https://doi.org/10.1016/j. cell.2011.02.013 21. Justus CR, Leffler N, Ruiz-Echevarria M, Yang LV (2014) In vitro cell migration and invasion assays. J Vis Exp (88). https://doi. org/10.3791/51046 22. Yu D, Zhong Y, Li X et al (2015) ILs-3, 6 and 11 increase, but ILs-10 and 24 decrease stemness of human prostate cancer cells in vitro. Oncotarget 6(40):42687–42703. https://doi. org/10.18632/oncotarget.5883 23. Wang Z, Gao Y, Liu Y, Chen J, Wang J, Gan S, Xu D, Cui X (2015) Tectonic1 contributes to the growth and migration of prostate cancer cells in vitro. Int J Mol Med 36(4):931–938. https://doi.org/10.3892/ijmm.2015.2313 24. Banerjee S, Li G, Li Y, Gaughan C, Baskar D, Parker Y, Lindner DJ, Weiss SR, Silverman RH (2015) RNase L is a negative regulator of cell migration. Oncotarget 6(42):44360–44372. https://doi.org/10.18632/oncotarget.6246 25. Guo W, Keener AL, Jing Y, Cai L, Ai J, Zhang J, Fisher AL, Fu G, Wang Z (2015) FOXA1 modulates EAF2 regulation of AR transcriptional activity, cell proliferation, and migration in prostate cancer cells. Prostate 75(9):976– 987. https://doi.org/10.1002/pros.22982 26. Zoni E, van der Horst G, van de Merbel AF et al (2015) miR-25 modulates invasiveness and dissemination of human prostate cancer cells via regulation of alphav- and alpha6-­ integrin expression. Cancer Res 75(11):2326–

2336. https://doi.org/10.1158/0008-5472. can-14-2155 27. Kroon J, in’t Veld LS, Buijs JT, Cheung H, van der Horst G, van der Pluijm G (2014) Glycogen synthase kinase-3beta inhibition depletes the population of prostate cancer stem/progenitor-like cells and attenuates metastatic growth. Oncotarget 5(19):8986–8994. https://doi. org/10.18632/oncotarget.1510 28. van den Hoogen C, van der Horst G, Cheung H et al (2010) High aldehyde dehydrogenase activity identifies tumor-initiating and metastasis-­ initiating cells in human prostate cancer. Cancer Res 70(12):5163–5173. https://doi.org/10.1158/0008-5472. can-09-3806 29. Gillies RJ, Didier N, Denton M (1986) Determination of cell number in monolayer cultures. Anal Biochem 159(1):109–113 30. Chiba K, Kawakami K, Tohyama K (1998) Simultaneous evaluation of cell viability by neutral red, MTT and crystal violet staining assays of the same cells. Toxicol In Vitro 12(3):251–258 31. Sun C, Zhao X, Xu K, Gong J, Liu W, Ding W, Gou Y, Xia G, Ding Q (2011) Periostin: a promising target of therapeutical intervention for prostate cancer. J Transl Med 9:99. https:// doi.org/10.1186/1479-5876-9-99 32. Rodriguez LG, Wu X, Guan JL (2005) Wound-­ healing assay. Methods Mol Biol (Clifton, NJ) 294:23–29 33. Liang CC, Park AY, Guan JL (2007) In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro. Nat Protoc 2(2):329–333. https://doi. org/10.1038/nprot.2007.30 34. Menon MB, Ronkina N, Schwermann J, Kotlyarov A, Gaestel M (2009) Fluorescence-­ based quantitative scratch wound healing assay demonstrating the role of MAPKAPK-2/3 in fibroblast migration. Cell Motil Cytoskeleton 66(12):1041–1047. https://doi. org/10.1002/cm.20418 35. Yarrow JC, Perlman ZE, Westwood NJ, Mitchison TJ (2004) A high-throughput cell migration assay using scratch wound healing, a comparison of image-based readout methods. BMC Biotechnol 4:21. https://doi. org/10.1186/1472-6750-4-21 36. Marshall J (2011) Transwell((R)) invasion assays. Methods Mol Biol (Clifton, NJ) 769:97–110. https://doi. org/10.1007/978-1-61779-207-6_8 37. Pawar SC, Demetriou MC, Nagle RB, Bowden GT, Cress AE (2007) Integrin alpha6 cleavage: a novel modification to modulate cell

In Vitro Migration and Invasion Assays in Prostate Cancer

79

migration. Exp Cell Res 313(6):1080–1089. 41. Wu HC, Hsieh JT, Gleave ME, Brown NM, Pathak S, Chung LW (1994) Derivation of https://doi.org/10.1016/j.yexcr.2007. androgen-independent human LNCaP pros01.006 tatic cancer cell sublines: role of bone stromal 38. Gang X, Wang G, Huang H (2015) Androgens cells. Int J Cancer J 57(3):406–412 regulate SMAD ubiquitination regulatory fac42. Wu TT, Sikes RA, Cui Q et al (1998) tor-­1 expression and prostate cancer cell inva- Establishing human prostate cancer cell xenosion. Prostate 75(6):561–572. https://doi. grafts in bone: induction of osteoblastic reacorg/10.1002/pros.22935 tion by prostate-specific antigen-producing 39. Albini A, Benelli R (2007) The chemoinvasion tumors in athymic and SCID/bg mice using assay: a method to assess tumor and endothelial LNCaP and lineage-derived metastatic subcell invasion and its modulation. Nat Protoc lines. Int J Cancer 77(6):887–894 2(3):504–511. https://doi.org/10.1038/ 43. Thalmann GN, Anezinis PE, Chang SM, nprot.2006.466 Zhau HE, Kim EE, Hopwood VL, Pathak S, 40. Albini A, Benelli R, Noonan DM, Brigati C von Eschenbach AC, Chung LW (1994) (2004) The “chemoinvasion assay”: a tool to Androgen-independent cancer progression study tumor and endothelial cell invasion of and bone metastasis in the LNCaP model of basement membranes. Int J Dev Biol 48(5-­ human prostate cancer. Cancer Res 6):563–571. https://doi.org/10.1387/ 54(10):2577–2581 ijdb.041822aa

Chapter 5 Transplantable Animal Studies and Whole-Body Optical Imaging in Prostate Carcinoma Geertje van der Horst, Maaike van der Mark, Henry Cheung, and Gabri van der Pluijm Abstract Current treatments of advanced prostate cancer only marginally increase overall survival and can be regarded as predominantly palliative. Hence, there is an urgent need for novel therapeutic strategies for the treatment of primary tumors and, more importantly perhaps, for the prevention of tumor progression and metastasis formation. Clinically relevant preclinical models are therefore urgently needed. An ideal, clinically relevant preclinical model would mimic the genetic and phenotypic changes that occur at the different stages of human prostate cancer progression and subsequent metastasis. In this chapter, transplantable xenograft prostate cancer models are described, in which human prostate cancer cells are transplanted into host animals (e.g., immune-deficient mice). Cancer cells can be administered to the small laboratory animals in various ways, including inoculation of the prostate tumor cells subcutaneously, at the anatomical site of origin (orthotopically), or at the metastatic site. In addition, we describe imaging methods suitable for small laboratory animals with emphasis on optical imaging (bioluminescence and fluorescence). Key words Prostate carcinoma, Preclinical model, Xenograft model, Optical imaging, Bioluminescence

1  Introduction 1.1  Clinical problem of Prostate Carcinoma

Prostate carcinoma is the most common cancer in males and the second leading cause of cancer death in the Western world. Due to an expanding population at risk, the socioeconomic and medical impact of prostate carcinoma is increasing. Current treatments of primary prostate tumors are initially very effective. However, in 20–30% of newly diagnosed patients with apparent organ-confined prostate cancer, beneficial responses are followed by tumor recurrence at distant sites leading to incurable, devastating metastatic disease. Current treatments of advanced prostate cancer only marginally increase overall survival and can be regarded as predominantly palliative [1, 2]. Hence, there is an urgent need for novel ­therapeutic strategies for the treatment of primary tumors and, more importantly perhaps, for the prevention of tumor progression and metastasis

Zoran Culig (ed.), Prostate Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 1786, https://doi.org/10.1007/978-1-4939-7845-8_5, © Springer Science+Business Media, LLC, part of Springer Nature 2018

81

82

Geertje van der Horst et al.

formation. Clinically relevant preclinical models are therefore urgently needed. 1.2  Preclinical Models

To establish clinically relevant models to predict therapy response and metastasis formation, these models ought to mimic both the genetic and epigenetic changes that occur at the various stages of human prostate cancer progression and subsequent metastasis. Unfortunately, no such models yet exist for small laboratory animals, due to the lack of spontaneous metastasis, the very long latency of metastasis, or the presence of only intravascular metastases [3, 4]. As a result, numerous preclinical animal studies have been developed, in which prostate cancer cells are transplanted into a host animal (i.e., transplantable tumor models). In syngeneic models, the inoculated cells are of the same species and genetic background [5], whereas in xenograft models, the cancer cells are of human origin inoculated into immune-deficient mice, including BALB/c nu/nu nude and severe combined immune-deficient (SCID) mice. A disadvantage of transplantable animal models is that only specific stages of the metastatic cascade are characterized (e.g., no spontaneous tumor formation), and some crucial features of the interaction between tumor cells and the surrounding tumor microenvironment might be altered in these models. In addition, the syngeneic inbred mouse models lack genetic complexity, whereas the xenograft preclinical models have an incomplete immune system.

1.3  Transplantable Animal Models

In the transplantable animal models, various methods exist to inoculate the prostate cancer cells, including subcutaneous inoculation, orthotopic inoculation (i.e., at the anatomical site of origin—the prostate), or inoculation at the metastatic site (the bone/bone microenvironment). Moreover, cells can be administered via the left cardiac ventricle to mimic metastatic spread of the tumor cells throughout the body (systemic inoculation).

1.4  Subcutaneous Transplantation

Subcutaneous administration is a straightforward method of inoculation and remains a valuable approach for high-throughput drug screening. However, to more accurately studying the processes of tumor progression and metastasis, it is important to administer prostate cancer cells to a more biologically relevant environment such as the prostate or to the metastatic site.

1.5  Orthotopic Transplantation

Administration of prostate cancer cells into the prostate, the tissue from which the tumor cells were originally derived, is called orthotopic inoculation. Orthotopic inoculation of an established prostate cancer cell line, PC-3M-Pro4 results in tumor growth in the prostate, as well as metastasis towards the locoregional lymph nodes. However, no reliable bone metastases have been formed in

Animal Studies and Imaging in Prostate Cancer

83

this model [6–8]. Current drawbacks of orthotopic transplantation include the lack of distant metastasis formation and the possibility of tumor cells leaking into the peritoneum following surgery as well as the trauma of opening the mouse peritoneum itself. 1.6  Intraosseous Transplantation

Another transplantable animal model consists of inoculation of prostate tumor cells into one of the metastatic sites, the bone/ bone marrow microenvironment (intraosseous or intra-bone inoculation). This technique can be used to study the interaction between the cancer cells and the metastatic environment (stromal compartment), recapitulating the interactions in established prostate cancer bone metastasis. Intraosseous inoculation results in either osteolytic or osteoblastic lesions or a mixture of those, depending on the prostate cancer cell line used. For example, osteoblastic lesions are formed after inoculation of C4-2B, MDA-­ PCa-­2b, LAPC-9, and LuCaP 23.1 prostate cancer cell lines [9], whereas the human prostate cancer cell lines PC-3, Du-145, and RM-1 cells predominantly cause osteolytic lesions [7]. Key factors in the process of tumor-induced changes in the bone such as MMP-7 and MMP-13 have been identified using inoculation of mice with these established cell lines [10, 11]. Drawbacks of these models are the initial bone damage and resulting repair mechanisms due to the procedure as well as potential inflammation at the injection site. In addition, since the tumor cells have not entered the bone microenvironment via the endothelium (one of the early steps of the metastatic cascade), the precise location of the tumor cells might be different (i.e., the tumor cells potentially do not reside at their specific niche) [12, 13].

1.7  Systemic Inoculation of Cancer Cells

To study the process of homing to distant target organs and subsequent growth at the metastatic site, systemic inoculation of tumor cells either in the lateral tail vasculature or in the left cardiac ventricle are widely used experimental metastasis models [6, 14–18]. Using this technique, the later stages of prostate cancer metastasis can be examined, mainly the ability of the cancer cells to home to the metastatic site, survive in the stroma of the metastatic target organ, proliferate and finally form an overt metastasis. Moreover, these models are valuable in screening the response to compounds targeting formation and/or growth of prostate cancer metastases. Inoculation in the tail vasculature predominantly results in pulmonary metastasis, whereas bone metastasis is more prevalent when tumor cells are inoculated in the left cardiac ventricle. A potential disadvantage of these models is that early steps in the metastatic cascade are bypassed. Advantages of these models include control on both the amount of and characteristics of the cells that are inoculated. Using intracardiac inoculation, the tumorigenic and ­metastatic potential of different subpopulations of cancer cells can

84

Geertje van der Horst et al.

be investigated, such as cells with a stem/progenitor phenotype [19] or with altered expression levels of a gene of interest [20, 21]. 1.8  Imaging of Small Laboratory Animals

Although it is essential to improve the preclinical models to more closely mimic the course of human prostate cancer, accurate imaging and data acquisition of the tumor progression and metastasis as well as the biological processes involved is also crucial for the strength of the experimental approach. To study molecules, pathogenic processes, drug delivery and response, several novel imaging technologies have been developed allowing real-time imaging in a preclinical and occasionally clinical setting [6, 22, 23]. Several imaging technologies have been developed for small animals, such as microcomputed tomography (μCT), micropositron emission tomography (μPET), single photon emission computed tomography (SPECT), magnetic resonance imaging (MRI), ultrasound imaging, and optical imaging (reviewed in [3, 4]). It is important to note that some imaging techniques are better suited for certain applications than others. For example, to monitor tumor cell biology, tumor burden, progression, and metastasis, highly sensitive approaches such as PET/SPECT and optical imaging (photo-­ acoustic imaging, bioluminescence and fluorescence imaging) are more appropriate, while to obtain anatomical detail CT and MRI are more suitable [24]. Combining several imaging techniques, called multimodality imaging, may provide a better solution to overcome the limitations of the independent techniques. Multimodality imaging will expand as well as improve the available information for the preclinical models. In this chapter, we will focus on noninvasive whole-body optical imaging (encompassing bioluminescence and fluorescence), since optical imaging is sensitive, permits longitudinal and quantitative real-time gene expression, cellular localization, and drug response studies in small laboratory animals [3, 4]. In addition, optical imaging is the least costly, requires shortest imaging times of 5 min or less, is easy to use, and can image several animals at once.

1.9  Noninvasive Whole-Body Optical Imaging

Optical imaging is based on emission of light from labeled cells or probes (the source of the light being either fluorescent or bioluminescent). Noninvasive, whole-body optical imaging enables examination of longitudinal and quantitative real-time gene expression, cellular localization, as well as drug response studies in small laboratory animals.

1.10  Bioluminescence

Bioluminescence imaging detects photons emitted by an enzymatic reaction in which luciferase catalyzes the conversion of d-luciferin into oxyluciferin in an ATP-dependent process. Bioluminescence imaging is a widely used imaging technique because of the low background, high signal-to-noise ratio, and the

Animal Studies and Imaging in Prostate Cancer

85

short acquisition time. Several luciferases have been detected, of which firefly luciferase derived from the firefly Photinus pyralis is the most widespread used (see Note 4). It has to be noted that bioluminescence imaging has limited depth information and limited spatial resolution. 1.11  Fluorescence

Using fluorescence imaging, reporter proteins (either fluorescent proteins or fluorochromes targeted to specific cell compartments or molecules) need to be excited by an external excitation source. Subsequently, fluorescence imaging detects the emission of light at specific wavelength from the reporter proteins. Fluorescence imaging can be divided into Reflectance Imaging (FRI) and Molecular Tomography (FMT), which can provide 3D information [25]. Like bioluminescence imaging, fluorescence imaging is restricted due to the limited penetration depth and spatial resolution. Moreover, autofluorescence of mammalian tissues is the most important source of background impairing the sensitivity of the fluorescence signal (see Note 7).

2  Materials 2.1  Cell Maintenance

For PC-3 cells and their derivatives (e.g., PC-3M-Pro4), Dulbecco’s modified eagle’s medium DMEM with 4.5 g/L glucose, GlutaMAX™ Supplement and pyruvate supplemented with 10% FCII (Fetal Clone II, Hyclone) and penicillin/streptomycin and 0.8 mg/mL geneticin (G418, neomycin) is used (see Note 1). For cell maintenance, cells are passaged at 80–90% confluence.

2.2  Constructs

For generation of stable cell lines, we used a modified pGL4 reporter construct (Promega) containing the mammalian optimized firefly luciferase 2 under control of the CAGGS promoter (CMV early enhancer/chicken beta actin CAG) with the neomycin selection marker (geneticin, G418) (see Note 2). For the generation of stable reporter cell lines, we generated modified pGL4 reporter constructs with destabilized FFLuc2 under control of the appropriate promoter sequence (e.g., BRE4 or CAGA12 elements for respectively the BMP and TGFβ pathway).

2.3  Imaging Systems

For bioluminescence and fluorescence imaging of small laboratory animals, the Xenogen IVIS Lumina Series III Pre-clinical In Vivo Imaging System (Perkin Elmer) was used. The sensitive range of the CCD camera sets the wavelength range of the IVIS Lumina for fluorescence applications from 400 to 950 nm. For fluorescence imaging, CRi Maestro in vivo multispectral imaging system for fluorescence imaging can be used (see Note 3). The wavelength range of the Maestro for fluorescence applications ranges from 500

86

Geertje van der Horst et al.

to 900 nm. For anesthetization of the mice, the Isoflurane Tec-3 Anesthesia Vaporizer System is utilized. For analyzing the obtained results, Living Image software and Maestro software are used for respectively the IVIS Lumina and the Cri Maestro. 2.4  Luciferase Substrates

2.5  Materials for Intraosseous Inoculation

The substrate for Photinus pyralis (firefly) luciferase is luciferin (also known as d-luciferin, (d-(−)-(2-(6′-hydroxy-2′benzothiazolyl) thiazone-4-carboxylic acid)), a monopotassium salt (Promega). The substrate for Renilla reniformis (Renilla) luciferase and Gaussia princeps (Gaussia) luciferase is coelenterazine (ViviRen™ Live Cell Substrate (Promega)). The substrate for NanoLuc is furimazine (Promega) (see Note 4). 1. Isoflurane Tec-3 Anesthesia Vaporizer System. 2. 70% EtOH. 3. Pair of bended tweezers. 4. Pair of tweezers. 5. Needles (25GA 5/8 0.5 × 16). 6. Insulin syringes. 7. PBS. 8. 1 mL syringes. 9. Surgical scissors. 10. Tissue paper. 11. Gauze pads. 12. Stainless steel wound clips 9 mm.

2.6  Materials for Orthotopic Inoculation

1. Isoflurane Tec-3 Anesthesia Vaporizer System. 2. 70% EtOH. 3. Pair of tweezers. 4. Needles (30G 1/2, BDMicro-Fine). 5. 1 mL syringes. 6. Surgical scissors. 7. Gauze pads. 8. Stainless steel wound clips 9 mm. 9. Growth Factor Reduced Matrigel (BD Biosciences).

Animal Studies and Imaging in Prostate Cancer

2.7  Materials for Anesthesia

87

1. Isoflurane Tec-3 Anesthesia Vaporizer System. 2. 70% EtOH. 3. 1 mL syringes. 4. Needles (30G 1/2, BDMicro-Fine). 5. Gauze pads.

3  Methods 3.1  Cell Lines (See Note 5)

For optical imaging using bioluminescence, tumor cell lines stably expressing luciferase can be used (see Notes 4–6). For optical imaging using fluorescence, preferably a red-shifted fluorochrome (see Note 7) is used. Fluorescence imaging can be performed with either stable expressing cell lines or labeling of the cells with a fluorochrome depending on the experimental questions (see Note 8). Many bone metastatic subclones of established prostate cancer cell line exist that have variable propensity for bone colonization and growth. For our experiments we use a subclone of PC-3 cells, PC-3M-Pro4 that readily forms bone metastasis after inoculation in vivo. Before inoculation, the tumor cells are maintained in their appropriate medium containing G418 (or appropriate antibiotic selection for the luciferase or fluorochrome) and passaged approximately every 72 h (1:10 dilution). Ideally, cells should be maintained in culture only briefly before inoculation (preferably less than 2 weeks). One day before inoculation, refresh the medium with medium without antibiotic selection. Tumor cells were harvested at 70–80% confluence with 0.15% trypsin/EDTA solution at 37 °C (keep the trypsinization time 10% of normal

92

Geertje van der Horst et al.

body weight or 2 cm3). (8) Tumors that interfere with eating or impair ambulation. In case of human endpoints or the beforehand set experimental endpoints, mice were anesthetized and sacrificed by cervical ­dislocation. Tumors were dissected and processed for either further histological analysis or RNA isolation for gene expression profiling. Evaluation of the metastasis should be performed by immunohistochemistry (see Note 16): For example, bone metastases in a subset of mice were also examined by Goldner staining after mice were sacrificed using decalcified bone.

4  Notes 1. Each cell line has its own special medium supplemented with various antibiotics and serum, also depending on the laboratory. In our laboratory, FCII is used since it is a fetal bovine serum (FBS) alternative, containing additional growth factors and supplements. The benefit of using FCII is the lower variability in batches compared to different fetal bovine serum batches. PC-3M-Pro4 cells are PC-3M cells passaged four times in the prostate of Balb/c nu/nu mice to generate highly metastatic clones that are capable of readily forming bone metastases. 2. Several constructs are available for bioluminescence measurements, of which some are optimized for use in mammalian cells to reduce background signals such as the pGl4 vectors (Promega). In addition, the intensity of the luciferase signal critically depends on the promoter in front of the luciferase gene. For example, the CMV promoter tends to be hypermethylated in vivo [26]. In addition, some promoters react to certain signaling pathways or therapeutic compounds, in that case it should be taken into consideration that the luciferase signal could be different compared to the tumor burden. For in vivo experiments with established prostate cancer cell lines, we prefer to use EF1α or CAGGS promoters, since these promoters appear to be stable in vivo and not responsive to major signaling pathways expressed in the prostate and bone microenvironment (such as the TGFβ superfamily). 3. Other possible equipment include the IVIS100 Imaging System (Perkin Elmer) capable of bioluminescence imaging and the IVIS® Spectrum in vivo imaging systems (Perkin Elmer). The latter can, like the IVIS Lumina, be used for multimodal imaging (both fluorescence and bioluminescence). The IVIS spectrum is, depending on the spectrum version, suited for multimodal 3D tomography combined with integrated X-ray and μCT. This provides the possibility to project

Animal Studies and Imaging in Prostate Cancer

93

the optical data sets on data acquired with other imaging techniques such as MRI or μCT. However, for fluorescence imaging, CRi Maestro multispectral imaging system is preferred due to the inbuilt multispectral imaging software algorithms which enable rapid imaging and quantification of more than one fluorescent s­ ignal in real time. The employed spectral unmixing enables imaging of multiple fluorescent signals simultaneously together with the removal of the majority of the autofluorescence from the image (depending on the fluorescent proteins used, see also Note 7). The increase in sensitivity by using multispectral imaging is approximately 300-fold, and separating the autofluorescence from the label fluorescence improves quantitation and lowers the detection limit noticeably. All machines include heated platforms to keep the mice at normal body temperature since there will be less light emission if anesthesia has decreased the animal’s body temperature. 4. Bioluminescent imaging is compromised by low signal intensity, impairing imaging of cells in deeper tissues and thereby restricting the sensitivity of the imaging method. Choosing the optimal luciferase for the designed experiments is therefore essential. For cell tracking, stable and high expression of the luciferase is essential. Firefly luciferase, derived from the firefly Photinus pyralis, which catalyzes the substrate luciferin, is the most extensively used luciferase enzyme in cell-based bioluminescent imaging [27]. The luminescent reaction is ATPdependent. A major improvement has been made with firefly luciferase 2, a mammalian codon-optimized firefly (see Note 5). NanoLuc® (Nluc) luciferase is a small enzyme (19.1 kDa vs. 61 kDa-sized firefly luciferase) about 100-fold brighter than either firefly or Renilla reniformis luciferase using a novel substrate, furimazine, to produce high intensity luminescence [28]. The luminescent reaction is ATP-­ independent and designed to suppress background luminescence for maximal assay sensitivity (Invitrogen). Still, the detection of NanoLuc in deep tissues is limited due to the emission of predominantly blue light by this enzyme. Renilla luciferase, derived from the anthozoan sea pansy Renilla reniformis, catalyzes coelenterazine, a substrate distinct from luciferin. However, the absorption properties of tissues limit its use in vivo [27]. The humanized Gaussia luciferase enzyme, derived from the copepod Gaussia princeps, likewise uses coelenterazine as a substrate and does not require ATP, but emits a markedly more intense signal compared to Renilla luciferase and may therefore overcome the limitations associated with Renilla luciferase. It has to be taken into account that the native Gaussia luciferase enzyme is secreted, markedly attenuating the in vivo bioluminescent signal of cells that express the enzyme.

94

Geertje van der Horst et al.

Therefore, Gaussia luciferase can be used as a biomarker for monitoring tumor progression and treatment response of systemic metastases rather than for cell tracking. A blood Gaussia luciferase assay revealed early growth of metastatic tumors before BLI could visualize their presence [29]. Several versions of the Gaussia luciferase exist, including a transmembrane version, that can be used for cell tracking purposes. Other, naturally occurring, variants of luciferase (Chromo-Luc™, Promega) were obtained from click beetles (Pyrophorus plagiophthalamus), and produce light at several different wavelengths of emission (from blue to red). Based on the proposed studies, one or several luciferases can be chosen for cell tracking experiments. For reporter studies, e.g., examining expression levels of certain signaling pathways, luciferase reporters have been generated that have a low intrinsic stability to quickly reflect transcriptional dynamics by adding a rapid degradation signal. As shown before, dual imaging using different luciferases is possible [28, 30, 31]. Moreover, bioluminescence imaging can be supplemented with other optical imaging (fluorescence) or other imaging techniques (multimodal imaging, see also [32]). 5. To compare the mammalian codon-optimized firefly luciferase 2 with the conventional firefly luciferase, stable luciferase 2 positive PC-3M-Pro4 prostate cancer cell lines were generated using a vector containing the enhanced luciferase gene under control of the CAGGS promotor (as described in [33]). Several clones were selected for further analysis and their luciferase activities were measured. Although variations occurred between the different clones, on average the PC-3M-Pro4 luciferase 2 (PC-3M-Pro4luc2) clones emitted 95 RLU/cell; this is on average a 202-fold (202 ± 95) increase compared to PC-3M-Pro4 cells with the previous luciferase generation (PC-3M-Pro4luc, 0.7 RLU/cell). We assessed the levels of bioluminescent signal of a titration series of PC-3M-Pro4luc vs. PC-3M-Pro4luc2 cells (10–50,000 cells). As shown in Fig. 2, lower amounts of PC-3M-Pro4luc2 cells compared to PC-3M-Pro4luc could be measured using an in vitro luciferase assay (Dual Luciferase Reporter Assay System, Promega). Even for the lower range of cells (10–1000 cells), an R2 was calculated of approximately 1, indicating that luciferase 2 activity can predict the amount of cells quite accurately (Fig. 2a, b for respectively high and low range of cells). In contrast, PC-3MPro4luc cells were not accurately measurable using a luciferase assay in these low amounts (R2 = 0.1). We subsequently compared the PC-3M-Pro4luc with the PC-­3M-­Pro4luc2 in a titration series implanted subcutaneously (Fig. 2c, d). The 10,000 and 1000 cell conditions were accurately and reproducibly measurable

95

Animal Studies and Imaging in Prostate Cancer

B in vitro

1200 1000 800 600 400 200 0

R2 = 1

BLI (10^5 RLU)

B L I (10^5 R L U)

A

R = 0,9979 10000

20000

30000

40000

CC

in vitro

25

R2 = 0,9908

20 15 10 5

2

0

30

0

50000

R2 = 0,1021 0

200

BLI (10^5 RLU)

D 35

Pro4luc

Pro4luc2

400

600

800

1000

1200

subcutaneous

30 25 20 15 10 5 0

R 2= 1

R 2= 0,9522 0

2000

4000

6000

8000

10000

Fig. 2 Comparison between Pro4luc and Pro4luc2 cells growing in vitro and in a subcutaneous in vivo model. Quantification of biochemically measured luciferase activity and its correlation with cell number with a cell number range of respectively 10–50,000 (a) and 10–1000 (b). The R2 coefficient was calculated with SPSS 17.0. Representative images of mice with subcutaneously inoculated PC-3M-Pro4luc or PC-3M-Pro4luc2 cells 3 h after inoculation (c) with the quantification in relative light units in (d). (open circles: Pro4luc; closed circles: PC-3M-Pro4luc2). Data are representative of three independent experiments (n = 4)

in all mice inoculated. In all mice (10/10), the 100 PC-3MPro4luc2 cell condition could be measured, whereas no measurable signal was detected in the 100 PC-3M-Pro4luc cells condition (0/10). In 1/10 even 20 PC-3M-Pro4luc2 cells could be detected ­subcutaneously using the IVIS Lumina. We next inoculated prostate cancer cells in the most frequent metastatic site in prostate cancer, the bone. As shown in Fig. 3, PC-3M-Pro4luc2 displayed significantly enhanced bioluminescent signal compared to PC-3M-Pro4luc (on average 130fold increase) although tumor volume was similar (Fig. 3c, d). To compare the metastatic properties, the PC-3M-Pro4luc and PC-3M-Pro4luc2 cells were inoculated into the left cardiac ventricle. As shown in Fig. 4, BLI signal could already be detected after 24 h in the mice inoculated with PC-3M-­ Pro4luc2 cells. In contrast, BLI signal appeared 14 days after inoculation of PC-3M-Pro4luc cells. Total tumor burden of mice inoculated with PC-3M-Pro4luc2 cells increased in the first week, then decreased until day 10 followed by exponential growth thereafter. The presence of tumor cells has been validated by immunohistochemical staining (Goldner staining) showing a metastasis in the bone marrow (Fig. 4d, e). An example of a micrometastasis (measured after 48 h) in a tibia (Fig. 4f) displays possible arrest in the bone marrow vasculature.

96

Geertje van der Horst et al.

A

Pro4luc

total tumor burden (10^5 R L U)

B

Pro4luc2

600000 500000 400000 300000 200000 100000 0

0

35000

**

*

30000 25000 20000 15000

*

10000 5000 0 0

7

14

*

* 7

14

21

28

Fig. 3 Comparison between Pro4luc and Pro4luc2 cells growing intraosseous. Representative images at day 3 after inoculation of either PC-3M-Pro4luc or PC-3M-Pro4luc2 cells in the bone (a). Total tumor burden for the mice inoculated with either PC-3M-Pro4luc2 (closed circles) or PC-3M-Pro4lucvehicle (open circles). In the insert, the first 14 days of the experiment are shown (b). (open circles: PC-3M-Pro4luc; closed circles: PC-3MPro4luc2). Representative images of a Goldner staining showing respectively PC-3M-Pro4luc (c) and PC-3MPro4luc2 (d) tumor cells in the tibia. GP growth plate, T tumor, B bone. Scale bars represent 100 μm

The number of metastases in the mice inoculated with PC-3MPro4luc2 cells was initially significantly higher, but after 14 days no significant differences were found. Most likely, these metastases are also present in the mice inoculated with the PC-3M-Pro4 luc cells, but remain below the detection

97

Animal Studies and Imaging in Prostate Cancer

A Pro4 luc

Pro4 luc2

day1

day2

day7

day10

C 50000 45000 40000 35000 30000 25000 20000 15000 10000 5000 0

1000 800

*

*

600 400 200 0

0

0

D

*

7

day1

14

*

* 7

14

*

*

No. of metastases/mouse

Total tumor burden BLI (10^5 RLU)

B

day14 18

day17

day21

day24

******

16 14 12

**

10 8

*

6 4 2 0

21

28

0

7

14

21

28

F

E

Fig. 4 Comparison of Pro4luc and Pro4luc2 cells in an experimental metastasis model. Representative images of mice at day 1, 2, 7, 10, 14, 17, 21, and 24 after inoculation of either PC-3M-Pro4luc or PC-3M-Pro4luc2 cells (a). Total tumor burden for the mice inoculated with the respective cells. In the insert, the first 17 days of the experiment are shown (b). The number of metastases/mouse (c). (open circles: PC-3M-Pro4luc; closed circles: PC-3M-Pro4luc2). Bioluminescent images of mice at day 24 with RLU of the left tibia and corresponding images of a Goldner staining showing respectively mice inoculated with PC-3M-Pro4luc (d) and PC-3MPro4luc2 (e) cells. GP growth plate, T tumor, B bone. Scale bars represent 80 μm. Representative image of a mouse 48 h after inoculation of PC-3M-Pro4luc2 cells (f). Goldner staining of the left tibia of this mouse (J) with corresponding magnification. B trabecular bone, BM bone marrow, S sinusoids, T tumor, OC osteoclast, Ocy osteocyte, Ob lining osteoblast

98

Geertje van der Horst et al.

limit of the IVIS Lumina. We further examined the tumor homing and early growth using the PC-3M-Pro4luc2 cell line. As shown in Fig. 5a, some metastases form but do not grow into overt metastases. For example, the metastasis in the left tibia grows into an overt metastasis, whereas the metastasis in the right tibia disappears and becomes undetectable after 10 days (with both IVIS Lumina measurements and immunohistochemistry). Moreover, the first few days many micrometastases can be detected in the distal spine of which none grow into overt metastasis, whereas some metastases in the lumbar spine do develop into overt metastases (Fig. 5b, c). This indicates that the prostate cancer cells home to these sites, but do not grow into overt metastasis. This model can therefore be used to study the kinetics of homing and metastasis of prostate cancer cells in real time. 6. Which cell line to use depends on the research question. Cell lines (stably) expressing reporter genes (either fluorescent or bioluminescent) are needed for the optical imaging procedures. Otherwise caliper measurements should be performed supplemented with species-specific qPCR on several tissues that might contain metastases. 7. A significant problem for optical imaging using fluorescence is the limited transmission of the light through the tissues due to scattering and absorption of the light in the tissues. Autofluorescence from endogenous molecules, such as hemoglobin and cytochromes, causes significant background signals. Since autofluorescence of the tissue is less prominent and tissue penetration is improved in the far-red or near-infrared (NIR), fluorescent proteins in these regions are more suitable for deeper, noninvasive imaging of small animals (reviewed in [34]). For example, living colors© fruit fluorescent proteins (RFP variants) such as dTomato and mCherry are more suitable for in vivo experiments in small laboratory animals. In addition to autofluorescence, the limited depth of the fluorescence signaling impairs the feasible experiments. Combinations of high-speed intravital imaging using a multiphoton laser scanner microscope with the dorsal skinfold window model has increased the depth of the tissue that can be interrogated [35]. However, experiments are still restricted to the dorsal skinfold chamber or subcutaneous regions. It remains difficult to measure sensitively in bones. Compared to bioluminescence, relatively large tumors (~1 mm3) are needed to measure fluorescent signals in bone (reviewed in [24]). 8. The simplest approach to labeling cells for cell tracking is to add a fluorescent dye that is taken up by viable cells. However, because these dyes are diluted during each cell division, the

Animal Studies and Imaging in Prostate Cancer

A

99

B

1

2

7

10

14

days 1

C

7

10

14 days

paws tail mandible liver hip tibia neck/shoulder scrotum spine

5 4,5 amount of metastases/mouse

2

4 3,5 3 2,5 2 1,5 1 0,5 0 0

7

14

21

28

Fig. 5 Kinetics of tumor growth. Representative images at respectively day 1, 2, 7, 10, and 14 after inoculation of PC-3M-Pro4luc2 cells (a). Representative images of mice 24 h after inoculation of PC-3M-Pro4luc2 cells with a lower threshold setting (b). Amount of metastases/mouse at different sites (c)

signal per cell reduces over time, thus limiting some long-term experiments. However, this procedure can be useful in determining a cell population with limited cell division (e.g., stem cells) (reviewed in [36]). 9. In case of PC-3M-Pro4 cells and other established prostate cancer cell lines, Balb/c nu/nu are a suitable animal model. Other immune-deficient mice can also be utilized. For example, for experiments with the established prostate cancer cells VCaP or C4-2B cells, NOD SCID mice are preferred since these cells did not show tumor growth in Balb/c nu/nu mice. NSG mice are usually adequate for experimental metastasis procedures. Regarding the choice of the appropriate mouse model, it should be taken into consideration that light is transmitted more efficiently through the tissues of white or hairless mice because melanin absorbs substantial amounts of light and light is scattered more by dark fur. 10. Fat depositions surrounding the prostate and other organs are rarely found in young Balb/c nu/nu mice but more regularly in other mouse strains and older Balb/c nu/nu mice.

100

Geertje van der Horst et al.

11. 4.5–5-week-old mice preferably since the amount of bone metastasis is highest since these mice have higher bone turnover compared to older mice. 12. The amount of cells inoculated depends on the experiment. When using drug efficacy studies, 105 cells/100 μL can be inoculated. However, determining the effect of different subpopulations of cells can be performed with less cells. For example, inoculation of 10,000 cells of the ALDHhigh stem/ progenitor subpopulation resulted in tumor progression and metastasis [19]. 13. Occasionally, the luciferin is misinoculated intraperitoneally. If no or less bioluminescence signal appears than expected because of previous measurements or a palpable tumor, you should reinject d-luciferin. 14. A problem of bioluminescence imaging can be overexposure of smaller foci by large signals. It is possible to cover the large focus with a piece of black carton or plexiglass to measure smaller foci. Measure mice separately when one mouse has a significantly higher tumor burden. At the end of the experiment, quickly resect the large tumors and subsequently remeasure the mouse to visualize potential additional (micrometastatic) tumors previously not visible. Another possibility is to isolate the organs presumably containing tumor cells and measure the organs ex vivo. 15. If the image is saturated, first the F-stop (diaphragm) can be adjusted. It is not necessary to remeasure the other animals, since changing the F-stop does not change the measured bioluminescent signal. The software corrects for the F-stop. 16. Histological analysis of the BLI signal remains pivotal, since environmental factors and therapeutic interferences may cause discrepancies between tumor burden and bioluminescence intensity. The intensity of the signal measured by in vivo imaging may depend on various factors, such as d-luciferin absorption through the peritoneum, cell membrane permeability, availability of co-factors, intracellular pH, and transparency of overlying tissue. In addition, because luciferases are oxygenases, the oxygen requirement may limit the use of luciferases as reporters in anaerobic environments, such as necrotic cores of large tumors. Similarly, the blood flow may not be sufficient to deliver the substrate to the tumor cells expressing luciferase.

Animal Studies and Imaging in Prostate Cancer

101

References 1. Bastian PJ, Carter BH, Bjartell A, Seitz M, Stanislaus P, Montorsi F, Stief CG, Schroder F (2009) Insignificant prostate cancer and active surveillance: from definition to clinical implications. Eur Urol 55:1321–1330 2. Heidenreich A, Bastian PJ, Bellmunt J et al (2014) EAU guidelines on prostate cancer. Part II: treatment of advanced, relapsing, and castration-resistant prostate cancer. Eur Urol 65:467–479 3. van der Horst G, van der Pluijm G (2012) Preclinical models that illuminate the bone metastasis cascade. Recent Results Cancer Res 192:1–31 4. van der Horst G, van der Pluijm G (2012) Preclinical imaging of the cellular and molecular events in the multistep process of bone metastasis. Future Oncol 8:415–430 5. Khanna C, Hunter K (2005) Modeling metastasis in vivo. Carcinogenesis 26:513–523 6. Buijs JT, Rentsch CA, van der Horst G et al (2007) BMP7, a putative regulator of epithelial homeostasis in the human prostate, is a potent inhibitor of prostate cancer bone metastasis in vivo. Am J Pathol 171:1047–1057 7. Buijs JT, van der Pluijm G (2009) Osteotropic cancers: from primary tumor to bone. Cancer Lett 273:177–193 8. An Z, Wang X, Geller J, Moossa AR, Hoffman RM (1998) Surgical orthotopic implantation allows high lung and lymph node metastatic expression of human prostate carcinoma cell line PC-3 in nude mice. Prostate 34:169–174 9. Schwaninger R, Rentsch CA, Wetterwald A et al (2007) Lack of noggin expression by cancer cells is a determinant of the osteoblast response in bone metastases. Am J Pathol 170:160–175 10. Lynch CC, Hikosaka A, Acuff HB et al (2005) MMP-7 promotes prostate cancer-­ induced osteolysis via the solubilization of RANKL. Cancer Cell 7:485–496 11. Nannuru KC, Futakuchi M, Varney ML, Vincent TM, Marcusson EG, Singh RK (2010) Matrix metalloproteinase (MMP)-13 regulates mammary tumor-induced osteolysis by activating MMP9 and transforming growth factor-­ beta signaling at the tumor-bone interface. Cancer Res 70:3494–3504 12. Wang N, Docherty FE, Brown HK et al (2014) Prostate cancer cells preferentially home to osteoblast-rich areas in the early stages of bone metastasis: evidence from in vivo models. J Bone Miner Res 29:2688–2696

13. Yu C, Shiozawa Y, Taichman RS, McCauley LK, Pienta K, Keller E (2012) Prostate cancer and parasitism of the bone hematopoietic stem cell niche. Crit Rev Eukaryot Gene Expr 22:131–148 14. Buijs JT, Henriquez NV, van Overveld PG et al (2007) Bone morphogenetic protein 7 in the development and treatment of bone metastases from breast cancer. Cancer Res 67:8742–8751 15. van der Horst G, van den Hoogen C, Buijs JT et al (2011) Targeting of alpha(v)-integrins in stem/progenitor cells and supportive microenvironment impairs bone metastasis in human prostate cancer. Neoplasia 13:516–525 16. van der Pluijm G, Sijmons B, Vloedgraven H, Deckers M, Papapoulos S, Lowik C (2001) Monitoring metastatic behavior of human tumor cells in mice with species-specific polymerase chain reaction: elevated expression of angiogenesis and bone resorption stimulators by breast cancer in bone metastases. J Bone Miner Res 16:1077–1091 17. van der Pluijm G, Que I, Sijmons B, Buijs JT, Lowik CW, Wetterwald A, Thalmann GN, Papapoulos SE, Cecchini MG (2005) Interference with the microenvironmental support impairs the de novo formation of bone metastases in vivo. Cancer Res 65:7682–7690 18. Peyruchaud O, Serre CM, NicAmhlaoibh R, Fournier P, Clezardin P (2003) Angiostatin inhibits bone metastasis formation in nude mice through a direct anti-osteoclastic activity. J Biol Chem 278:45826–45832 19. van den Hoogen C, van der Horst G, Cheung H et al (2010) High aldehyde dehydrogenase activity identifies tumor-initiating and metastasis-­ initiating cells in human prostate cancer. Cancer Res 70:5163–5173 20. van den Hoogen C, van der Horst G, Cheung H, Buijs JT, Pelger RC, van der Pluijm G (2011) The aldehyde dehydrogenase enzyme 7A1 is functionally involved in prostate cancer bone metastasis. Clin Exp Metastasis 28:615–625 21. van den Hoogen C, van der Horst G, Cheung H, Buijs JT, Pelger RC, van der Pluijm G (2011) Integrin alphav expression is required for the acquisition of a metastatic stem/progenitor cell phenotype in human prostate cancer. Am J Pathol 179:2559–2568 22. Kaijzel EL, van der Pluijm G, Lowik CW (2007) Whole-body optical imaging in animal models to assess cancer development and progression. Clin Cancer Res 13:3490–3497

102

Geertje van der Horst et al.

23. Black PC, Shetty A, Brown GA, Esparza-Coss E, Metwalli AR, Agarwal PK, McConkey DJ, Hazle JD, Dinney CP (2010) Validating bladder cancer xenograft bioluminescence with magnetic resonance imaging: the significance of hypoxia and necrosis. BJU Int 106:1799–1804 24. Conway JR, Carragher NO, Timpson P (2014) Developments in preclinical cancer imaging: innovating the discovery of therapeutics. Nat Rev Cancer 14:314–328 25. Graves EE, Weissleder R, Ntziachristos V (2004) Fluorescence molecular imaging of small animal tumor models. Curr Mol Med 4:419–430 26. Teschendorf C, Warrington KH Jr, Siemann DW, Muzyczka N (2002) Comparison of the EF-1 alpha and the CMV promoter for engineering stable tumor cell lines using recombinant adeno-associated virus. Anticancer Res 22:3325–3330 2 7. Klerk CP, Overmeer RM, Niers TM, Versteeg HH, Richel DJ, Buckle T, Van Noorden CJ, van Tellingen O (2007) Validity of b ­ ioluminescence measurements for noninvasive in vivo imaging of tumor load in small animals. BioTechniques 43(7–13):30 28. Stacer AC, Nyati S, Moudgil P, Iyengar R, Luker KE, Rehemtulla A, Luker GD (2013) NanoLuc reporter for dual luciferase imaging in living animals. Mol Imaging 12:1–13 29. Chung E, Yamashita H, Au P, Tannous BA, Fukumura D, Jain RK (2009) Secreted Gaussia luciferase as a biomarker for monitoring tumor progression and treatment response of systemic metastases. PLoS One 4:e8316

30. Henriquez NV, van Overveld PG, Que I, Buijs JT, Bachelier R, Kaijzel EL, Lowik CW, Clezardin P, van der Pluijm G (2007) Advances in optical imaging and novel model systems for cancer metastasis research. Clin Exp Metastasis 24:699–705 31. Mezzanotte L, Que I, Kaijzel E, Branchini B, Roda A, Lowik C (2011) Sensitive dual color in vivo bioluminescence imaging using a new red codon optimized firefly luciferase and a green click beetle luciferase. PLoS One 6:e19277 32. Deroose CM, De A, Loening AM, Chow PL, Ray P, Chatziioannou AF, Gambhir SS (2007) Multimodality imaging of tumor xenografts and metastases in mice with combined small-­ animal PET, small-animal CT, and bioluminescence imaging. J Nucl Med 48:295–303 33. van der Horst G, van Asten JJ, Figdor A, van den Hoogen C, Cheung H, Bevers RF, Pelger RC, van der Pluijm G (2011) Real-time cancer cell tracking by bioluminescence in a preclinical model of human bladder cancer growth and metastasis. Eur Urol 60:337–343 34. Hilderbrand SA, Weissleder R (2010) Near-­ infrared fluorescence: application to in vivo molecular imaging. Curr Opin Chem Biol 14:71–79 35. Reeves KJ, Hurrell JE, Cecchini M, van der Pluijm G, Down JM, Eaton CL, Hamdy F, Clement-Lacroix P, Brown NJ (2015) Prostate cancer cells home to bone using a novel in vivo model: modulation by the integrin antagonist GLPG0187. Int J Cancer 136:1731–1740 36. Parish CR (1999) Fluorescent dyes for lymphocyte migration and proliferation studies. Immunol Cell Biol 77:499–508

Chapter 6 Protocols for Tissue Microarrays in Prostate Cancer Studies Tatjana Vlajnic, Serenella Eppenberger-Castori, and Lukas Bubendorf Abstract Tissue microarray (TMA) technology is a method for high-throughput analysis of tissue biomarkers, commonly used in translational cancer research. TMAs allow performing a variety of in situ applications on hundreds of tissue samples simultaneously using the same protocols as for conventional slides. Thereby, precious material from patient samples remains largely preserved while costs in resources and time in laboratory processing decrease. Therefore, a TMA is a powerful tool to identify and study biomarkers that may have a potential diagnostic, prognostic, and predictive value. Depending on the research question, there are different types of TMAs, such as progression TMA, outcome TMA, and tumor heterogeneity TMA. Since the first introduction of the TMA method almost 20 years ago, most laboratories used manual tissue arrayers for manufacturing. Nowadays, automatic or semiautomatic devices are commercially available, which largely facilitates the technical construction. However, preparatory work remains the most time-consuming part in preparing TMAs. This chapter focuses on issues involved in design and construction of prostate cancer TMAs. Key words Tissue microarray, TMA, Tissue arrayer, Immunohistochemistry, In situ hybridization, Prostate cancer, Tumor heterogeneity

1  Introduction Tissue microarray (TMA) technology is widely used for high-­ throughput analysis of a large number of samples. The principal method was originally described by Battifora in 1986 as the concept of a multi-tumor “sausage” tissue block for testing novel immunohistochemical antibodies [1]. A decade later, Kononen et al. modified the current technique using a novel punching instrument that allows precise arraying of tissue cores of regular size and shape [2]. Briefly, they punched out cylindrical cores from representative regions of numerous paraffin-embedded tissue blocks, so-called “donor” blocks, and inserted them into a “recipient” paraffin block. This method enables parallel analysis of multiple samples simultaneously with identical experimental conditions

Zoran Culig (ed.), Prostate Cancer: Methods and Protocols, Methods in Molecular Biology, vol. 1786, https://doi.org/10.1007/978-1-4939-7845-8_6, © Springer Science+Business Media, LLC, part of Springer Nature 2018

103

104

Tatjana Vlajnic et al.

for all samples. Further advantages are preservation and minimal destruction of valuable material from patient samples while saving time and costs in laboratory processing. Consecutive sections from a TMA block are suitable for a number of analyses such as immunohistochemistry, immunofluorescence, and in situ hybridization methods for detection of DNA and RNA, using the same protocols as for conventional histological sections. With recent progress in technology, DNA array-based and large-­scale sequencing studies of tumor and non-tumor tissue have led to the discovery of countless disease-related genes that need validation with regard to their potential clinical impact. In prostate cancer (PC), genome-profiling studies have identified recurrent somatic mutations in important oncogenes and tumor suppressor genes [3, 4]. However, further functional analyses are needed to understand the clinical importance of these mutations. Downstream protein expression analysis might reveal potential prognostic and predictive biomarkers as well as novel candidates for targeted therapies. Such analyses require a large dataset of patient samples with long-term clinical follow-up information. TMAs are very well suited for this purpose. Furthermore, prostate carcinoma is known to be a heterogeneous tumor on both morphological and molecular levels. It is important to be aware of this fact when constructing a TMA and when analyzing and interpreting the results. On the other hand, it is possible to address specifically the issue of tumor heterogeneity using a TMA specifically designed to address and measure tumor heterogeneity. Finally, it is important to keep in mind that a TMA is a tool to discover general associations of biomarkers with histological, molecular, or clinical endpoints rather than to examine individual tumors.

2  Materials 1. Formalin-fixed paraffin-embedded (FFPE) blocks and corresponding hematoxylin and eosin (H&E) slides. 2. Ultrafine waterproof point pen. 3. All donor blocks must be included in standard tissue-­embedding cassettes (Unisette Tissue in Stacks; 3 × 2.5 × h 0.5 cm; Biosystems AG, Switzerland). 4. P250 slide scanner (3D-Histech Ltd., Sysmex Suisse AG, Horgen, Switzerland). 5. Pannoramic Viewer Program (3D-Histech Ltd., Hungary). 6. Paraffin block (Sysmex Suisse AG, Horgen, Switzerland). 7. Scalpel.

Tissue Microarrays in Prostate Cancer Studies

105

8. Tissue arrayer: TMA GrandMaster® (TMA-GM; 3D-Histech Ltd., Sysmex Suisse AG, Horgen, Switzerland). 9. Adhesive glass slides SuperFrost™ Plus (Thermo Fisher Scientific AG, Switzerland) 10. HM 325 Rotary Microtome (Thermo Scientific™ AG Switzerland). 11. Microtome blades (Type S35, Feather, Japan). 12. Floatation water bath. 13. Slide drying hotplate. 14. Laboratory oven.

3  Methods 3.1  TMA Design

Before the actual technical TMA construction, a considerable amount of time must be dedicated in planning of the TMA design and preparation of the material. This comprises determining the type of TMA, selecting suitable cases from the pathology database, reviewing histological slides, and collecting tissue blocks (see Notes 1 and 2). It is of great importance to collect the complete medical history of the patients, with special emphasis on previous treatments of PC (androgen deprivation therapy (ADT), chemotherapy, radiotherapy). Some types of TMAs require long-term clinical follow-up information. Another important issue in the design of a TMA is the choice of an appropriate population size to obtain significant results. For that purpose, it is helpful to involve a biostatistician to calculate the statistical power.

3.1.1  Type of TMA

There are different types of TMAs depending on the research question to be addressed. We describe here those most commonly used in PC studies. 1. Outcome/Prognosis TMA: This type of TMA is used to compare tissue biomarkers and clinical outcome in order to identify and evaluate potential prognostic markers. Most important issue in this context is a sufficiently large number of cases with available long-term clinical follow-up data (ideally at least 5 years). In case of PC, this is usually represented by serial serum PSA measurements. An increase in serum PSA level indicates biochemical recurrence in patients treated with radical prostatectomy (rPE) or radiotherapy, and tumor progression in patients under active surveillance. Complete medical history should include information about the hormonal status of PC, which is hormone-­sensitive PC or castration resistant PC (CRPC), in patients who received ADT and history of other previous treatments (chemotherapy, radiotherapy incl. brachytherapy).

106

Tatjana Vlajnic et al.

2. Progression TMA: Progression TMAs are used to study molecular alterations or the expression profile of a biomarker across different stages of disease during progression and in comparison to normal tissue. This type of TMA contains normal prostate tissue or benign prostatic hyperplasia (BPH), premalignant lesions (prostatic intraepithelial neoplasia, PIN), and tumors at various stages, from untreated localized PC to local recurrences and metastases (Fig. 1). In cases with recurrent tumors, it is important to differentiate between hormone-sensitive PC and CRPC. Thereby, several components can be obtained from the same patient (e.g., PIN and invasive carcinoma) (see Note 3). This TMA can also contain different Gleason grades. An alternative is a TMA tailored to compare different Gleason categories (i.e., grades or scores) (see Note 4). We recommend using at least 50–100 samples for each tumor stage/category in order to achieve sufficient statistical power.

Fig. 1 Detail of a progression TMA section. (a, b) Staining with hematoxylin and eosin (H&E) (a magnification ×20). (c) Immunohistochemical analysis showing focal positivity for chromogranin (b, c magnification ×100)

Tissue Microarrays in Prostate Cancer Studies

107

3. Prevalence TMA: These TMAs allow for a general estimate of the prevalence of a molecular alteration of interest or expression of a given protein in PC. Most practical way to get together the patient population is to enrol consecutive cases of prostate carcinoma because detailed clinicopathological information is not required for this type of TMA. The number of tumors included in such a TMA can be kept low (100–200), since they are meant to provide a rough estimate of the prevalence of biomarkers to be preselected and prioritized for further studies using progression TMAs or outcome TMAs. Prevalence TMAs are ideal tools to screen a large number of PC for different candidate proteins. 4. Tumor heterogeneity TMA: This type of TMA is designed to address intratumoral genotypic or phenotypic variation of a biomarker, referred to as intratumoral heterogeneity. Thereby, intratumoral heterogeneity can be reflected at different levels, such as different Gleason patterns or different histological subtypes (e.g., ductal, acinar, mucinous). Heterogeneity can also exist at a cellular/nuclear level if some tumor cells show small nuclei with inconspicuous nucleoli and others large vesicular nuclei with prominent nucleoli. However, genotypic heterogeneity does not necessarily go along with changes in protein expression. For example, some prostate carcinomas harbor a TMPRSS2-ERG gene fusion without detectable ERG protein expression [5]. Moreover, intratumoral heterogeneity in ERG expression of two adjacent tumor areas can exist without a difference in morphology between these areas [6]. Likewise, in some “usual” prostate adenocarcinomas, areas of neuroendocrine differentiation may be demonstrated only by immunohistochemical positivity for neuroendocrine markers (synaptophysin, chromogranin, CD56), without accompanying characteristic neuroendocrine morphology [7]. Therefore, when making TMAs with emphasis on intratumoral heterogeneity, we take cores from different areas of the tumor, even if they cannot be distinguished morphologically. Such a tumor “mapping” may for example comprise cores from the center and periphery of each quadrant. If morphologically different tumor areas are present (e.g., different Gleason patterns or different histological subtypes), we core them separately and designate the samples accordingly (see Note 5). 5. CRPC TMA: The genomic profile of the primary tumor and metastatic sites at the stage of CRPC might be different than at the time of the initial diagnosis of PC. This difference is caused by continuous acquisition of genomic alterations during disease progression.

108

Tatjana Vlajnic et al.

Additionally, new resistance mutations to antiandrogen drugs can occur under therapy selection pressure. Nevertheless, treatment of patients at the stage of CRPC remains most challenging and therapeutic options are still limited. Thus, there is an urgent need to identify novel therapeutic targets and predictive markers. For example, Antonarakis et al. showed recently that a splice variant of androgen receptor (AR) called AR-V7 can serve as a negative predictive marker for response to treatment with second generation antiandrogen drugs [8]. In regard to the abovementioned importance, it seems reasonable to generate a TMA consisting solely of CRPC. Alternatively, CRPC cases may be included in a progression TMA. However, at this stage of disease, primary tumors or metastatic lesions are only rarely biopsied. The major source of readily available tissue is palliative transurethral resection specimens (TURP) from patients with local obstruction. There is also a growing demand of matched specimens from prior to and after castration resistance to explore the molecular mechanisms on a per patient basis. Such rare and biologically previous matched specimens can be incorporated into a CRPC TMA or in a separate matched CRPC TMA. 6. Test TMA: In a test TMA, we include a small set of cases with prostate carcinomas (e.g., ten cases) across different stages and Gleason grades, as well as a few cores of benign tissue from other organs. This type of TMA serves to test new antibodies and to optimize technical steps in staining protocols, such as pretreatment conditions or antibody concentration (see Note 6). It is also suitable to compare reactions between different types of tissues (TURP vs. rPE) or between samples that have been stored for different periods to rule out a potential technical bias [9]. 7. TMA from experimental tissues (cell line TMA and xenograft TMA): Common PC cell lines are particularly suitable as (positive or negative) control cores in a tissue-based PC TMA because their genotype and phenotype have been studied extensively. Additionally, cells from cell lines or from primary tissue cultures can be used after treatment with specific regimens. Such functional arrays can also be obtained from PC xenografts [10]. Several groups have described protocols for preparing TMAs using cell lines. In our opinion, the most convenient method is to embed cells in a paraffin block and process them like regular FFPE blocks [11, 12]. 3.1.2  Donor Tissue Selection

In general, we use FFPE tissue for TMA construction (see Note 7). Thereby, PC tissue can be obtained from different sources: rPE, TURP, or core needle biopsies. Most commonly, we use tissue

Tissue Microarrays in Prostate Cancer Studies

109

derived from rPE or TURP specimens. Patients who get diagnosed with advanced stage PC usually do not undergo surgery. In these cases, a small needle biopsy might be the only tumor tissue available. Potential problems that might be associated with different tissue sources are explained below. 1. Radical prostatectomy specimens: RPE specimens are the most valuable tissue source for TMA construction. The main concern regarding rPE specimens is the tissue quality that largely depends on tissue fixation (see Note 8). If the prostate gland was fixed in 10% buffered formalin as a whole, the tissue in the central zone is more subjected to ischemia than the tissue in the periphery due to prolonged penetration of formalin. This inhomogeneous fixation can have an impact on immunohistochemical reactions or in situ hybridization because of poor quality of DNA and RNA [13]. Therefore, the results may differ depending on whether the tissue was taken from central or peripheral parts of the prostate. To overcome this problem, it is advisable to collect tissue from the better fixed peripheral part of the prostate. Injection of formalin before routine processing appears as an elegant method to enable homogeneous fixation of fresh rPE specimens without affecting the integrity of the organ. Interestingly, however, this procedure does not seem to have a major positive effect on immunoreactivity [13, 14]. Recently, we have established a novel approach for processing fresh radical prostatectomy specimens using ceramic foam plates [15]. Before putting them into formalin, we place freshly cut whole mount sections between ceramic foam plates, which allows a homogeneous fixation. Furthermore, necrotic areas should not be sampled to avoid false positive results due to nonspecific immunohistochemical reactions. 2. TURP specimens: TURP material is very well suited for TMAs as long as the individual fragments are sufficiently thick (3–4 mm). However, given the nature of the procedure, the tissue may show extensive cautery artifacts. These can hamper the immunohistochemical reaction leading to false results. Moreover, they show a poorly preserved morphology which makes it difficult to assign the correct Gleason grade. Therefore, tissue with cautery artifacts should be avoided for a TMA. 3. Core needle biopsies: Core needle biopsies may be the only available cancer tissue in patients with advanced disease and in patients who are ­subsequently treated with radiotherapy. Such biopsies may be a very valuable source of tissue for biomarker analysis, provided that the material has not been used up during diagnostic procedures. Several studies

110

Tatjana Vlajnic et al.

have described possible methods how to use core needle biopsies for TMA construction [16–18]. 4. Fresh frozen material: It is possible to create TMAs from non-fixed, i.e., fresh frozen tissue [19]. Better preservation of antigens and nucleic acids has been regarded as an advantage of fresh tissue over fixed tissue. However, most protocols are nowadays optimized to work on FFPE tissue with comparable quality. 3.1.3  Core Number and Size

1. Core number: Several studies have shown that the prognostic utility of TMAs in general is comparable to regular histological sections since both represent only a minor fraction of the whole tumor bulk [20–24]. This is also true for PC, which is considered a highly heterogeneous tumor on both phenotypic and molecular level. If there is a sufficiently large number of patients (ideally >500 patients), a single core per tumor is sufficient to uncover clinically relevant associations [25–29]. For example, studies with the world largest prognostic prostate TMA containing single cores from >10,000 rPE specimens have revealed a number of meaningful data [30, 31]. Using only one core per tumor largely facilitates the scoring and statistical analysis of a TMA and enables to include a larger number of cases in a single (recipient block) TMA. Taking more than one core per sample also results in more TMA blocks to be constructed. Analyzing multiple cores per tumor is recommended for TMAs containing a small number of tumors. A certain loss of interpretable spots is inevitable and ranges from 10% to 30% of spots per TMA depending on the technique of analysis (e.g., IHC or FISH) [32]. This can occur due to floating off during technical processing or replacement of tumor by stroma in deeper block sections. Therefore, if the study comprises only a small set of patients (