. ..—J . ”2:1... . .. . , 2... h . L... ‘ .2 in: . i» . if . 93.. .uvam. ,. i All! 2 . .1 x32... 27 .3. 53-4-2. I A . . a it axlnhx.fl.\: . v... ‘ «Sign 5.» . .Hh. . . . DAD-i- 15”..“ h1uu.rl . 53:..Junn. [3.21: 65: . III... II . .1: :t? 3.;1‘11.39|$a$!|H-Kh5l .. 5:83.: I: veil ALL.“ 1...: LIBRARY Michigan State University This is to certify that the dissertation entitled UNDERSTANDING TRANSFORMATION AND TUMOR INITIATION IN THE CONTEXT OF MESENCHYMAL STEM CELLS AND OSTEOSARCOMA: COMPARATIVE STUDIES IN DOGS AND HUMANS presented by MANISH NEUPANE has been accepted towards fulfillment of the requirements for the PhD. degree in Comparative Medicine and Integrative BiologL 4/“ I 7’/// /\ // ,, Major Prol’essfs/Sign’ature /2' /7'2M2 Date MSU is an Affirmative Action/Equal Opportunity Employer — — -.-.-—.-.----.-.—.-.-.—.—-.-.- PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5I08 K:IProj/Aoc&PrelelRC/DatoDue.indd UNDERSTANDING TRANSFORMATION AND TUMOR INITIATION IN THE CONTEXT OF MESENCHY MAL STEM CELLS AND OSTEOSARCOMA: COMPARATIVE STUDIES IN DOGS AND HUMANS By Manish Neupane A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Comparative Medicine and Integrative Biology 2009 ABSTRACT UNDERSTANDING TRANSFORMATION AND TUMOR INITIATION IN THE CONTEXT OF MESENCHY MAL STEM CELLS AND OSTEOSARCOMA: COMPARATIVE STUDIES IN DOGS AND HUMANS By Manish Neupane Osteosarcoma (OS) is the most common primary bone tumor of human and dogs. This tumor is highly resistant to conventional chemo-and radiotherapies, and is composed of a varying proportion of undifferentiated and differentiated cell types of mesenchymal lineages. These attributes make OS as a plausible candidate for being a cancer driven by stem cells. Increasing evidence suggests that OS is a disease of blocked differentiation, caused by genetic and epigenetic changes that interrupt the process of osteoblastic differentiation. However, the origin of this tumor is still unknown. The issue that whether adult stem cells such as mesenchymal stem cells (MSCs) may require fewer or different steps than more differentiated cells to acquire a transformed phenotype is still unresolved. The current set of studies were undertaken to gain insights into OS-relevant tumorigenic events in the appropriate cellular context as well as to study the functionally relevant markers for identification of tumor initiating fraction of this disease. Canine OS has been shown to be an appropriate and informative model for human OS. We first established the experimental system of canine MSCs by using a novel approach of modulation of cellular redox state, and extensively characterized these cells. We then incorporated defined and regulatable genetic elements into both canine and human MSCs by means of retroviral vectors. These incorporated genes were turned on at different time points of osteogenic differentiation, and evaluated for tumorigenic phenotypes in virro. On the other hand, canine OS cells were sorted on the basis of expression of pluripotency-associated transcription factor-OCT4, and efflux of Hoechst dye (side population assay), which were then evaluated for tumorigenic potential in vitro and/or in vivo. Our results indicate that alterations of multiple pathways seem to be necessary for tumorigenic transformation of MSCs or their differentiated descendants. We found that overexpression of exogenous MET and/or BMIl was not sufficient to transform MSCs at any stage of differentiation. Undifferentiated MSCs underwent senescence after modest expansion of life-span. Although MSCs in early stage of differentiation were prevented from further differentiation, these cells also succumbed to senescence. On the other hand, MSCs at late stage of differentiation were completely refractory to oncogenic transformation, and underwent temtinal differentiation. We found that although both OCT4-positive and OCT4-negative cell populations of a canine OS cell line were capable of giving rise to tumors, only OCT4-positive OS cells recapitulated the original tumor phenotype. Furthermore, cellular heterogeneity seemed to accelerate tumor development. We were also successful in isolating side population (SP) cells from canine OS, and we found that SP did not appear to be a major contributor of putative cancer stem cell phenotype in this tumor. Our results underscore the complexity of OS biology, and these findings will be usefill to understand the disease process and design more effective strategies to treat OS in future. Copyright by MANISH NEUPANE 2009 DEDICATION This dissertation is dedicated to my parents for their inspiration, encouragement, and support. ACKNOWLEDGEMENTS I would like to express my sincere gratitude to my mentor, Dr. Vilma Yuzbasiyan- Gurkan, fdr providing continuous support, opportunities, and resources throughout the course of my PhD, and for her guidance and advices. I am thankful to my committee member, Dr. Chia-Cheng Chang for training me in the laboratory and for providing constant academic inputs. I further acknowledge the contributions of my other committee members, Dr. Jose Cibelli and Dr. Matti Kiupel, for their advices and sharing of knowledge and resources. My labmates have always maintained collegial and supportive atmosphere in the lab, and helped me on technical matters and critical discussions. I thank Joshua Webster, Nikolaos Dervisis, Tuddow Thaiwong, Elizabeth Bartlett, Emmalena Gregory-Bryson, Megan Goodall, Annet Wenker, and Te-Chuan Chen. I would also like to thank Dr. Steven Suhr from Cibelli lab for sharing his expertise on molecular cloning and viral vectors. I acknowledge the support of Companion Animal Fund at College of Veterinary Medicine for funding the research projects related to the dissertation work. I thank Dr. Murari Suvedi for his help, encouragement and support. I am indebted to the contribution of my parents, Jhanka Prasad Neupane and Shannila Neupane, throughout my entire career. I am thankful to my brother Mahesh for his help on several aspects of this dissertation. I appreciate the love and support of my wife Neeru, and acknowledge her help on the current thesis work. vi TABLE OF CONTENTS LIST OF TABLES ........................................................................................................... xi LIST OF FIGURES ........................................................................................................ xii LIST OF ABBREVIATIONS ......................................................................................... xv CHAPTER I INTRODUCTION ............................................................................................................ 1 CHAPTER 2 ISOLATION AND CHARACTERIZATION OF CANINE ADIPOSE-DERIVED MESENCHYMAL STEM CELLS ................................................................................. 30 SUMMARY ............................................................................................................. 31 INTRODUCTION ................................................................................................... 32 MATERIALS AND METHODS ............................................................................ 33 Isolation of canine adipose tissue derived mesenchymal stem cells (cAD-MSCs) .......................................................................................................................... 33 Differentiation of cAD-MSCs (Osteogenesis, Chondrogenesis, and Adipogenesis) ................................................................................................... 34 RT-PCR ............................................................................................................ 35 Immunocytochemistry ...................................................................................... 36 Proliferation potential of cAD-MSCs ............................................................... 37 RESULTS ................................................................................................................ 38 Isolation of putative cAD-MSCs ...................................................................... 38 Expression of stemness markers ....................................................................... 39 Lifespan of cAD-MSCs in different media ...................................................... 39 Differentiation of cAD-MSCs .......................................................................... 39 DISCUSSION .......................................................................................................... 42 APPENDIX .............................................................................................................. 46 REFERENCES ........................................................................................................ 55 CHAPTER 3 EVALUATION OF GENE EXPRESSION IN ADIPOGENIC INDUCTION OF CANINE MESENCHYMAL STEM CELLS ................................................................. 59 SUMMARY ............................................................................................................. 60 INTRODUCTION ................................................................................................... 62 MATERIALS AND METHODS ............................................................................ 64 Isolation of canine adipose tissue derived mesenchymal stem cells (cAD-MSCs) .......................................................................................................................... 64 Verification of stemness phenotypes and multi-lineage differentiation potential .......................................................................................................................... 64 Differentiation studies ...................................................................................... 64 vii RT-PCR ............................................................................................................ 66 Statistical analysis ............................................................................................ 67 RESULTS ................................................................................................................ 69 Effect of different induction regimens on adipogenic differentiation of AD- MSCs ................................................................................................................ 69 Gene expression analysis following adipogenic induction .............................. 70 DISCUSSION .......................................................................................................... 72 APPENDIX .............................................................................................................. 75 REFERENCES ........................................................................................................ 81 CHAPTER 4 TRANS-DIFFERENTIATION OF CANINE ADIPOSE-DERIVED MESENCHYMAL STEM CELLS INTO CELLS WITH NEURAL PHENOTYPE .................................... 85 SUMMARY ............................................................................................................. 86 INTRODUCTION ................................................................................................... 87 MATERIALS AND METHODS ............................................................................ 89 Isolation of cAD-MSCs .................................................................................... 89 Verification of stemness phenotypes and multi-lineage differentiation potential .......................................................................................................................... 89 Induction of neural differentiation .................................................................... 89 RT-PCR ............................................................................................................ 91 Immunocytochemistry ...................................................................................... 92 RESULTS ................................................................................................................ 94 Induction of neural differentiation .................................................................... 94 Expression of neural markers at mRN A level .................................................. 95 Expression of neuronal markers at protein level .............................................. 95 DISCUSSION .......................................................................................................... 96 APPENDIX ............................................................................................................ 101 REFERENCES ...................................................................................................... 106 CHAPTER 5 ASSESSMENT OF RELATIVE PERMISSIBILITY OF MESENCHYMAL STEM CELLS AT VARIOUS STAGES OF OSTEOGENIC DIFFERENTIATION TO TUMORIGENIC TRANSFORMATION BY DEFINED GENETIC INTERVENTION ....................................................................................................................................... 113 SUMMARY ........................................................................................................... 114 INTRODUCTION ................................................................................................. 116 MATERIALS AND METHODS .......................................................................... 120 Source of mesenchymal stem cells (MSCs) ................................................... 120 Construct details ............................................................................................. 121 Production of retrovirus .................................................................................. 123 Verification of ligand-regulatable expression of vectors ............................... 123 Transduction of MSCs .................................................................................... 124 Induction of osteogenesis and regulation of transgene expression ................ 125 SV40 transfection and retroviral transduction of MSCs ................................ 125 Characterization for tumorigenic phenotypes in vitro .................................... 126 viii Evaluation of senescence-associated beta-galactosidase activity ................... 127 Validation of transgene expression ................................................................ 128 RESULTS .............................................................................................................. 129 Expression of MET and BMII mRNA in OS cell lines .................................. 129 Validation of ligand-regulatable expression systems ..................................... 129 Stable transduction of BM]! and MET in MSCs ............................................ 130 Ectopic expression of BMI] and MET at different time points of osteogenic differentiation ................................................................................................. 13 1 Consequences of expression of SV40 ER in MSCs ....................................... 133 Evaluation of tumorigenic phenotypes in vitro .............................................. 134 DISCUSSION ........................................................................................................ 135 APPENDIX ............................................................................................................ 140 REFERENCES ...................................................................................................... 155 CHAPTER 6 EVALUATION OF OCT4 EXPRESSION AS A STEM CELL MARKER 1N CANINE OSTEOSARCOMA ...................................................................................................... 162 SUMMARY ........................................................................................................... 163 INTRODUCTION ................................................................................................. 164 MATERIALS AND METHODS .......................................................................... 168 Construct details ............................................................................................. I68 Establishment of OCT4 reporter cell lines ..................................................... 168 Validation of endogenous OCT4 activity of reporter cell lines ..................... 169 Maintenance of cell lines ................................................................................ 170 Gene expression studies ................................................................................. 170 Immunostaining .............................................................................................. 172 Characterization for tumorigenic phenotypes in vitro .................................... 172 Hoechst 33342 staining and side population (SP) phenotype ........................ 175 Purity following clonal expansion and flow cytometry ................................. 175 Xeno-transplantation and limiting dilution assay ........................................... 176 Histological studies ........................................................................................ 176 Re-establishment of canine osteosarcoma cell lines ...................................... 176 Statistical analysis .......................................................................................... 177 RESULTS .............................................................................................................. 178 Establishment of OCT4 reporter cell lines ..................................................... 178 Expression of endogenous OCT4 mRNA in reporter cell lines ..................... 178 Evaluation of tumorigenic phenotypes in vitro .............................................. 179 SP phenotype .................................................................................................. 181 Xenotransplantation assay .............................................................................. 181 Gene expression studies ................................................................................. 183 DISCUSSION ........................................................................................................ 1 86 APPENDIX ............................................................................................................ 193 REFERENCES ...................................................................................................... 209 CHAPTER 7 SIDE POPULATION PHENOTYPE IN CANINE OSTEOSARCOMA .................... 216 ix SUMMARY ........................................................................................................... 217 INTRODUCTION ................................................................................................. 219 MATERIALS AND METHODS .......................................................................... 222 Establishment of canine OS cell lines ............................................................ 222 SP analysis ...................................................................................................... 222 RT-PCR .......................................................................................................... 224 Characterization for tumorigenic phenotypes in vitro .................................... 225 Statistical Analysis ......................................................................................... 228 RESULTS .............................................................................................................. 229 Establishment of primary culture of canine OS ............................................. 229 Expression of ABC transporters in canine OS cells ....................................... 229 Evaluation of SP phenotype in canine OS cells ............................................. 229 Evaluation of tumorigenic phenotypes in vitro .............................................. 231 Gene expression studies ................................................................................. 234 DISCUSSION ........................................................................................................ 237 APPENDIX ............................................................................................................ 243 REFERENCES ...................................................................................................... 262 CHAPTER 8 CONCLUSIONS AND FUTURE DIRECTIONS ........................................................ 269 LIST OF TABLES Table 2.1. Primers used in RT-PCR of stemness and mesenchymal lineage-associated markers ............................................................................................................................ 46 Table 2.2. Yield of plastic adherent cells in primary culture under conditions favoring MSCs enrichment ............................................................................................................ 47 Table 3.1. Primers used in quantitative RT-PCR of adipogenic markers ...................... 75 Table 4.1. Canine primers used in quantitative RT-PCR of neural markers ............... 101 Table 4.2. Human primers used in quantitative RT-PCR of neural markers ............... 102 Table 5.1 Primers used for sequencing of ME T open reading frame ........................... 140 Table 5.2. Primers used for sequencing of BMII open reading frame ....................... 140 Table 5. 3. Primers used to validate the expression of transgenes ............................... 141 Table 6.1. Primer sets used in qRT-PCR of stemness-related markers ....................... 193 Table 6.2. In vitro growth properties of D-17 canine OS cell line, and its subpopulations sorted on the basis of OCT4 expression ....................................................................... 194 Table 6.3. Tumorigenicity of various number of D-17 cells sorted on the basis of OCT4 expression ..................................................................................................................... 194 Table 7.1. Primer sets used in qRT-PCR of stemness-related markers ....................... 243 Table 7.2. Growth properties of cells sorted on the basis of Hoechst dye efflux ........ 244 xi LIST OF FIGURES Images in this dissertation are presented in color Figure 2.1. Phenotype of cAD—MSCs ............................................................................ 49 Figure 2.2. Osteogenic differentiation of cAD-MSCs ................................................... 51 Figure 2.3. Chondrogenic differentiation of cAD-MSCs ............................................... 52 Figure 2.4. Adipogenic differentiation of cAD-MSCs ................................................... 54 Figure 3.1. Induction of adipogenic differentiation of canine and human AD-MSCs under different treatment conditions. .............................................................................. 77 Figure 3.2. Quantification of intracellular triglyceride content by Oil Red 0 staining ”7.3 Figure 3.3. Expression of adipogenic markers after induction of canine AD-MSCs 79 Figure 3.4. Expression of adipogenic markers after induction of human AD-MSCs.... 80 Figure 4.1. Neural induction of canine and human AD-MSCs .................................... 103 Figure 4.2. Immunostaining for neuronal markers following neural induction ........... 104 Figure 4.3. RT-PCR analysis for expression of neural markers at mRNA level ......... 105 Figure 5.1. Expression of MET and BMII in OS cell lines .......................................... 143 Figure 5.2. Validation of ligand-regulatable expression system in HEK 293-FT cellsig‘.1 Figure 5.3. Verification of ligand-regulatable expression system in MSCs ................ 145 Figure 5.4. Forced expression of MET and/or BMII in undifferentiated MSCs ......... 146 Figure 5.5. Over-expression of MET and BMII at different time points of osteogenic differentiation ................................................................................................................ 147 Figure 5.6. Consequences of SV40 ER incorporation ................................................. 148 Figure 5.7. Effect of ectopic expression of exogenous transgenes on life span of MSCs. ....................................................................................................................................... 150 xii Figure 5.8. In vitro transformation phenotypes ........................................................... 152 Figure 5.9. Verification of transgene expression at mRN A level ................................ 153 Figure 5.10. Verification of transgene expression at the protein level ........................ 154 Figure 6.1. Establishment of D-17 reporter cell lines .................................................. 196 Figure 6.2. mRNA expression of OCT4 in different D-17 OCT4 reporter cell lines .. 197 Figure 6.3. Anchorage independent growth of D-17 OCT4 reporter cell lines ........... 198 Figure 6.4. Matrigel invasion assay of D-17 OCT4 reporter cell lines ........................ 199 Figure 6.5. Chemosensitivity assay of D-l7 OCT4 reporter cell lines ........................ 199 Figure 6.6. Radiosensitivity assay of D-1 7 OCT4 reporter cell lines .......................... 200 Figure 6.7. SP phenotype in D-17 OCT4 reporter cell lines ........ 202 Figure 6.8. Tumorigenicity of D-17 cells sorted on the basis of OCT4 expression and the resultant tumor phenotypes ........................................................................................... 203 Figure 6.9. Expression of stemness-associated genes in D-17 OCT4 reporter cell lines .. ....................................................................................................................................... 205 Figure 6.10. Expression of ABC transporters and stemness-related genes in D-17 OCT4 reporter cell lines ........................................................................................................... 206 Figure 6.11. Re—establishment of OS cell lines from mouse tumor xenografis ............ 207 Figure 6.12. Comparison of expression of stemness markers in original versus re- established cell lines ..................................................................................................... 208 Figure 7.1. Representative morphologies of canine OS cells ...................................... 245 Figure 7.2. Expression of ABCGZ, MDR-I, and MRP-I in canine OS cell lines ........ 245 Figure 7.3. SP profile of canine OS cell lines (Groupl tumors) .................................. 247 Figure 7.4. SP profile of canine OS cell lines (Group 2 tumors) .................................. 248 Figure 7.5. Anchorage independent growth of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux ................................................................................ 249 xiii Figure 7.6. Matrigel invasion assay of subpopulations of canine 08 cells sorted on the basis of Hoechst dye efqux ........................................................................................... 250 Figure 7.7. Chemosensitivity assay of subpopulations of canine OS cells sorted on the basis of Hoechst dye efqux ........................................................................................... 251 Figure 7.8. Radiosensitivity assay of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux ........................................................................................... 253 Figure 7.9. Relative mRN A expression of ABC transporter genes in of subpopulations of canine 08 cells sorted on the basis of Hoechst dye efflux ........................................... 255 Figure 7.10. Expression of stemness-associated genes in subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. ............................................................ 257 Figure 7.11. Expression of Notch family genes in subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux ..................................................................... 259 Figure 7.12. Expression of stemness-related genes-TERI”, MCM7, and BCL2 in subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux .............. 261 xiv ABC AIG ANOVA cAD-MSCs cAMP CPDL CSCs DMSO DOS DTT EGF FABP FACS FBS FTC HRP IBMX LPL MSCs NAC NSP ABBREVIATIONS ATP-binding cassette Anchorage independent growth Analysis of variance Canine adipose-derived mesenchymal stem cells Cyclic adenosine monophosphate Cumulative population doubling level Cancer stem cells Dimethyl sulfoxide Dog osteosarcoma Dithiothreitol Epidermal growth factor Fatty acid binding protein Fluorescence-activated cell sorter Fetal bovine serum Fumitremorgin C Horse radish peroxidase Isobutylrnethlyxanthine Lipoprotein lipase Mesenchymal stem cells N-acetyl-L-Cysteine Non-side population XV OS PBS PPAR qRT-PCR SCID SP TICs YFP a-MEM Osteosarcoma Phosphate buffered saline Peroxisome proliferator-activated receptor Quantitative real time PCR Severe combined immunodeficiency Side population Tumor initiating cells Yellow fluorescent protein Minimum essential medium-alpha xvi CHAPTER 1 Introduction 1. Cancer stem cells (CSCs): Various in vitro and in vivo assays have demonstrated that tumors are composed of heterogeneous population of cells with differences in proliferation, differentiation, and tumor initiation potential despite their monoclonal origin [1-3]. These observations led to the idea that tumors are initiated and maintained by a distinct subpopulation of tumor cells, the so—called cancer stem cells (CSCs) [4-10]. The idea that tumor starts fi'om stem cells was propounded in 1875 when Cohnheim noticed extensive similarities between embryonic tissues and cancer in terms of their proliferation and differentiation capacity, and proposed that stem cells misplaced during embryonic development were source of tumors in later life [11]. Later studies in murine teratocarcinoma [12] and leukemia [2,13] demonstrated the ability of a single cell to generate tumors with heterogeneous progeny. Cancer was regarded as ‘maturation arrest of tissue determined stem cells’ or ‘blocked ontogeny’ in early 705 [14,15]. However, the idea of cancer stem cell was formally validated in 1994 when John Dick and colleagues showed functional differences between biologically distinct leukemic cells in NOD/SCID mouse model [16]. Stem cells have long life span, allowing them to accumulate multiple genetic and epigenetic changes necessary for cellular transformation [4,5,17-19]. Moreover, stem cells share several common phenotypes with cancer cells [20], such as high proliferation potential, increased susceptibility to telomerase activation [21], and higher expression of anti-apoptotic molecules [22]. Since stem cells have extensive proliferation capacity, fewer mutations are required to constitutively activate prom-oncogenic pathways (e. g. WNT, SHH, NOTCH) and inactivate tumor suppressor pathways (e.g. PTEN, p16), increasing their likelihood of becoming tumorigenic [23,24]. These concepts led to the idea of cancer stem cells which are defined as the “cells within a tumor that possesses the capacity to self-renew and to cause the heterogeneous lineages of cancer cells that comprise the tumor” [25]. These distinct sub-population of self-renewing cancer cells drive tumor growth, and are responsible for relapse of tumor after completion of conventional therapy [7-9,25,26]. After initial identification in acute myeloid leukemia [16], cancer stem cells have been recently documented in several solid cancers including brain, breast, prostate, melanoma, lung, colon, head and neck, and pancreatic cancers (reviewed in [27,28]). The frequency of CSCs in a tumor may reflect the grade of trunor or stage of progression, and this has been shown to correlate with poor clinical outcome and patient prognosis [29,30]. Frequency of CSCs has been reported to vary from 0.03% to nearly 100% (reviewed in [31]). Tumors with high frequency of CSCs might indicate highly undifferentiated and aggressive tumors, more genetically unstable tumors, or their selective enrichment following treatment [31]. Carcinogenesis is a multi-stage, multi-mechanism process which proceeds through “initiation”, “promotion”, and “progression” phases [32,33] leading to immortalization of mortal cells (or prevention of mortalization of otherwise immortal cells) followed by neoplastic transformation events, and acquisition of characteristic hallmarks such as: (i) self-sufficiency in growth signals, (ii) insensitivity to growth- inhibitory signals, (iii) limitless replication potential, (iv) evasion of apoptosis, (v) sustained angiogenesis, and (vi) tissue invasion and metastasis [34]. Two major questions underlying the cancer stem cell models are: (i) what is the origin of tumors, and (ii) what drives tumors? The stochastic model has been traditionally 3 put forward to propose the idea that all tumor cells are created equal and any tumor cell is capable of generating new tumor given the right microenvironment [4,28,31]. However, substantial experimental evidence is available to support the hierarchical model which follows the notion that tumor transmitting ability is restricted to a distinct subpopulation of cancer cells, the so-called ‘cancer stem cells’ [4,9,10,17,27]. But, origin of these cancer stem cells is an area of ongoing debate [28], since cancer stem cells and stem cell cancer may not be the same thing. Whether a cell can acquire ability to become cancer stem cell at different stages of differentiation pathway is still highly controversial. In the absence of definitive evidence, there exist several possibilities for the origin of cancer stem cell: (i) multiple alterations in stem cell, (ii) initial alteration in stem cell and additional alterations in committed progenitor cells, (iii) initial alteration in committed progenitor cell, converting it into self-renewing stem cell, (iv) multiple alterations in differentiated cell, changing it to dedifferentiated cancer stem cell. Some of the earlier studies have shown the possibility of occurrence of final hit in progenitor cells [35-37] as well as dedifferentiation of differentiated cells prior to oncogenic events [38-40]. Cell killing, as judged by tumor shrinkage, is regarded as an important indicator of drug efficacy according to the current therapeutic standard. But, cancer stem cells usually constitute less than 1% of the total cancer cells, and are relatively resistant to the conventional chemotherapy [41-43] and radiotherapy regimens [44-46]. Their ability to express drug transporters, robust DNA repair mechanisms, quiescent cell cycle kinetics, asymmetric segregation of DNA strands, and expression of anti-apoptotic molecules are all thought to result in the continued presence of these “cancer stem cells” after conventional therapy and consequent recurrence of tumor following removal of tumor bulk [9,25,47,48]. Based on these premises, targeted therapy against cancer stem cells which spares normal stem cells, has been claimed to prevent tumor recurrence without significant side effects [8,17,41,49]. Differential signal transduction pathways (aberrant self renewal pathway, or cancer stem cell specific survival pathway) [50-55], differentially expressed cell surface molecules [56-60], cancer stem cell niche [61], or differentiation-induced selective depletion of cancer stem cells [62-64] can be exploited for this purpose [17]. In fact, recent studies involving CD44 targeting in human AML in xenograft model [65], mTOR targeting in PTEN deleted leukemic stem cells in mice [54], PML protein targeting in CML [66], and B—catenin removal in skin tumors [67] have demonstrated the possibility of selective targeting. But, if cancer stem cells arise from dedifferentiation of differentiated cells or committed progenitor cells, removing those cancer stem cells will not prevent the dedifferentiation event(s), and may eventually lead to the cancer recurrence despite the therapeutic targeting. So, the frequency with which non-tumorigenic cancer cells will evolve to become tumorigenic will determine the outcome of targeted therapy against cancer stem cells. Thus, determination of what drives the tumor has critical consequences to our approach to treatment. Identification of mutations in specific oncogenes and tumor suppressor genes in specific cancers have implicated the role of cancer associated mutations in initiation and progression of cancers [68]. MET, a tyrosine kinase receptor for hepatocyte growth factor (HGF), was originally identified as a transforming oncogene in a chemically transformed human osteosarcoma cell line (MNNG-HOS) [69]. Mis-expression of MET has been noticed in several mesenchymal tumors including human and canine osteosarcoma (08) [70-74]. In a recent study, lentiviral vector-mediated over-expression of MET was shown to be enough for converting human primary cultured osteoblasts into OS cells [75]. On the other hand, BMII is a polycomb transcriptional repressor, which has been shown to have a role in maintenance of adult stem cells as well as cancer stem cells [4,5,24,76,77]. It reduces the expression of p1 6INK4a and p19ARF, and has been found to be overexpressed in several human cancers [78]. BMII overexpression in mammary epithelial cells led to their immortalization [79], and its overexpression in lymphocytes led to lymphoma [80]. A recent study showed that lentiviral-mediated overexpression of BMII in human placenta derived MSCs (hPDMC) was able to immortalize these cells [81]. 2.1 Cancers in dogs: Dog is emerging as a promising biomedical model system, and it enjoys medical surveillance and clinical literature second only to human [82]. Canine models have remained instrumental in demonstrating the potential of gene therapy and stem cell-based therapies [83,84] as well as in evaluating the efficacy of anti-cancer drugs [85]. Dogs live in the same environment as human, and naturally occurring canine tumors reflect the genetic heterogeneity noticed in human patients [86]. There is a greater genetic homology between dogs and humans than between either species and mouse [87]. In addition, dogs represent relatively more outbred population than inbred laboratory animals [86]. It has been estimated that more than 1 million pet dogs in United States are diagnosed and managed with cancers [88]. Naturally occurring canine tumors develop in the context of an intact immune system and syngeneic tumor microenvironment. These tumors are comparable to human tumors in terms of time course of development, clinical and molecular features, and response to conventional therapies [88]. Canine cancers such 6 as osteosarcoma, lymphoma, soft tissue sarcomas, melanoma, and mammary tumors share histopathological features and biological complexities with human cancers [86,88]. Canine genome has now been sequenced, and successful efforts have been made to develop high-throughput technologies and to validate the reagents for canine system [82,88]. We are now in a position to apply molecular tools for interrogating canine biology with the same level of sophistication to its human counterpart. These all attributes make canine tumors uniquely suitable for translational studies on biology of cancer as well as for therapeutic targeting [89-93]. 2.2. Fundamental features of osteosarcoma: Osteosarcoma is an aggressive, malignant mesenchymal tumor that arises within the bone, and is the most common primary bone tumor of human and dogs [94,95]. It has been estimated that over 8,000 new cases are diagnosed in veterinary medicine every year in the United States [96]. Similarly, approximately 900 new cases occur annually in people in the United States (American Cancer Society, http://www.cancer.org/docroot/home/index.asp). The worldwide incidence of OS is 1.7 per million in individuals younger than 10 years of age, and 8.2 per million in individuals between 10 to 19 years age [97,98]. OS accounts for 4% of all malignancies, 85% of all skeletal malignancies, and 98% of appendicular primary canine bone tumors [94]. Canine OS closely resembles human OS in terms of histopathological appearance and biological behavior [96], and is an excellent model to study tumor biology and therapeutic intervention [86,87,92]. Canine OS is commonly seen in large and giant breed dogs, and metaphyseal region of the long bones is the most common primary site [99]. In human, OS mostly 7 develops in second decade of life in the metaphyses of the long bones which are then undergoing rapid growth in length [97,98]. When it occurs in humans older than 40 years old, it occurs as a consequence of previous bone irradiation or due to pre-existing disease such as Paget’s disease [100]. Less than 20% of affected dogs survive more than 2 years, and about 20% of affected children survive more than 5 years [101,102]. The standard- of-care is amputation or limb-sparing surgery followed by adjuvant chemotherapy. Death from OS usually results from progressive pulmonary metastasis with resultant respiratory failure. 1ntra-medullary high grade sarcoma is the most common type of OS, which accounts for about 75 % of all bone tumors. This conventional OS is divided into 3 subclasses (viz. osteoblastic, chondroblastic, and fibroblastic), based on the predominant type of matrix and abundance of osteoid deposition [103,104]. The non-conventional OS (25%) is further sub-classified into telangiectatic OS, low grade OS, small cell 08, parosteal OS, and periosteal OS [103,104]. This is true for both human and canine OS. 2.3. Etiology of osteosarcoma: The precise etiology of both human and canine OS is yet unknown. Exposure to beryllium oxide, orthopedic prostheses, and the PB] virus are known to induce OS in small animal models (reviewed in [86,105]). SV40 viral DNA has been detected in up to 50% of OS tumors (reviewed in [105]). Similarly, radiation exposure and trauma have been regarded as other contributing factors of OS pathogenesis (reviewed in [105]). Most OS are sporadic, and inherited predisposition accounts for a limited number of cases. These predisposing disease conditions include Li-Fraumeni syndrome, hereditary retinoblastoma, Rothmund-Thomson syndrome, and Werner syndrome [86,106]. In 8 contrast to other sarcomas such as Ewing’s sarcoma and alveolar rhabdomyosarcoma, OS do not have specific molecular genetic abnormalities that can serve as useful diagnostic or prognostic markers of the disease [107]. OS are highly heterogeneous tumors with complex structural and numerical chromosomal abnormalities [107]. The commonly involved genetic alterations include inactivation of tumor suppressor genes such as RBI, 7’ P53, CDKNZA, PT EN (in canine), and overexpression of oncogenes such as MET, MYC, HERZ, PDGFR, and [GP 1R [86,106,107]. Similarly, proteins such as ezrin, annexin 2 (AnxA2), CXCR4 and Fas ligand have been shown to play an important role in metastasis of the diseases [86,106,107]. Canine OS closely resembles human OS in terms of histopathological appearance, molecular markers, biological behavior, and responsiveness to conventional therapies [86,96], and is an excellent model to study tumor biology and therapeutic intervention [89-93]. In fact, canine OS has been accepted as a suitable spontaneous tumor model for the human counterpart disease, as well as being a potential system in which to evaluate cancer progression [108,109]. Furthermore, normalized cluster analysis of microarray data from canine and human OS samples did not segregate into distinct clusters based on species, providing a strong genomic evidence of the similarity and relevance of the canine OS as a model for the human disease [88]. 2.4. Osteosarcoma as a disease of differentiation: Osteosarcomas have been viewed as a clinically and molecularly heterogeneous group of malignancies characterized by varying degrees of mesenchymal differentiation [105]. Increasing evidence suggests that OS is a disease of differentiation caused by genetic and epigenetic changes that interrupt the process of osteoblastic differentiation 9 [110,11 1]. These tumors contain varying proportion of cells of mesenchymal lineage including undifferentiated cells, osteoblasts, chondroblasts, and fibroblasts [112]. Moreover, OS are quite resistant to conventional chemotherapy and radiotherapy‘[112]. All of these attributes make canine OS as a plausible candidate for cancer derived from and/or driven by mesenchymal stem cells (MSCs). 3. MSCs as an experimental system to understand the origin of osteosarcoma: MSCs are located in several connective tissue compartments [113], and have the ability to give rise to several cell types of mesenchymal lineages including osteoblasts, chondrocytes, myocytes, and fibroblasts [114]. MSCs are thought to be the cell of origin for various types of sarcoma [115]. A recent study has shown that Ewing’s tumors (ET) originate from MSCs which accounts for the predominant localization of ET in bones and soft tissues, two major sources for these stem cells [116]. Although < 1% in occurrence, OS can arise from soft tissues and visceral organs as well (extraskeletal OS), without any involvement of bone/periosteum. These tumors are highly metastatic in terms of their biological behavior [117-119]. However, the issue that whether normal adult stem cells may require fewer or different steps than differentiated cells to acquire a transformed phenotype is still unresolved [120,121]. The use of MSCs from both adipose and bone marrow origin in our experimental approach will allow us to assess the potential for osteosarcomagenesis from MSCs residing in different niches and/or their descendents. 4. Approaches to identify CSCszl Although a few cell surface markers have been used to enrich and prospectively isolate cancer stem cells/tumor-initiating cells in humans (leukemia-CD34’7CD38', breast cancer-CD44’/CD24'"°“’, brain- nestin and CD133+, colon cancer-Lgr5 and CD133+, 10 multiple myeloma-CD138', prostate cancer-CD442 pancreatic cancer-CD44’7CD24", liver cancer- a-fetoprotein (reviewed in [122]), these are tissue-specific but not universal markers. Moreover, most of the currently used markers to enrich and prospectively isolate tumor stem cells are cell surface-based [28], and they do not have functional relevance to initiation and propagation of tumor. In addition, only a limited number of validated antibodies are available against these cell surface markers for model species such as dog. Thus, there is an acute need for markers that have a mechanistic basis and causal relationship to the process of carcinogenesis. Isolation of tumor stem cells on the basis of state of stemness (such as expression of pluripotency-associated marker OCT4) and its functional phenotype (such as ability to efqux Hoechst dye) might represent a better approach to enrich the tumor-initiating cells. 4.1. OCT4 expression as a candidate marker for identification of tumor-initiating cells: Several studies have suggested OCT4 (also known as OCT 3 /4, POUSFI) as a candidate marker for CSCs. This transcription factor is implicated in the pluripotency of stem cells. It is expressed in embryonic stem cells [123,124], several adult stem cells [125-128], and many cancers {125,129-139], including bone sarcomas in human [140]. Canine OS cell lines have been shown to expresses the message for OCT 4 [141]. In addition, our research group has also identified OCT4 positive cells in various canine tumors, including OS using immunohistochemistry [134]. Genomic fusion between OCT4 and EWSRI (Ewing Sarcoma Region) has been associated with a case of a human bone tumor, indicating OCT4 reactivation [142]. In another study, forced expression of II OCT4 was sufficient to transform non-tumorigenic Swiss 3T3 fibroblasts into high grade fibrosarcoma-producing cells [143]. However, direct evidence for the role and usefulness of OCT4 in cancer stem cells has not yet been generated-leading to much controversy in this area [137,144-149]. This issue can only be resolved through rigorous study of pure population of OCT4 expressing cells. But, OCT4 is a transcription factor and unlike other cell surface markers, OCT4 expressing tumor cells cannot be fractionated by conventional antibody based flow- sorting approach. This difficulty in prospective isolation of OCT4 expressing cells will prevent the purification of these cells from heterogeneous tumor cell population, which would limit the usefulness of this marker to a restricted subset of studies. One of the feasible approaches would be to establish reporter cell lines whereby expression of reporter marker(s) reflects the activity of endogenous OCT4 promoter. Introduction of fluorescent reporter genes under the control of cell type-specific promoters is considered as one of the best methods to identify and enrich particular progenitor or specialized cell types [150]. Reporter gene-based approach can be extrapolated to other functional markers as well as several other tumor types to isolate and evaluate tumor stem cell fraction. There are two main strategies to generate reporter cell lines: (i) random integration of plasmid containing reporter gene driven by the promoter of gene of interest into the particular cell line(s); (ii) homologous recombination to ‘knock-in’ the reporter gene into the endogenous locus of gene of interest into the particular cell line(s). Randomly integrated transgenic reporters have been effectively used to tag human embryonic stem cells 12 [151,152] as well as their differentiated derivatives [153-155]. Using this approach, Levings et al [156] have recently identified tumor-initiating cells in human OS. 4.2. Side population as a functional phenotype of tumor-initiating cells: Several studies have suggested that a subpopulation of cells, called the side population (SP) which are defined by their characteristics in flow cytometry in the presence of Hoechst 33342 vital dye, to harbor and be greatly enriched for the stem cells in a given tumor cell population [157]. In the absence of well-defined surface markers for tumor-initiating cells, SP analysis represents a feasible and clinically useful approach to isolate tumor-initiating cells from a heterogeneous population of cells. SP assay has been demonstrated to enrich primitive and undifferentiated cells (reviewed in [158]), including mouse embryonic stem cells [159]. In recent studies on human malignancies, SPs have been identified in several tumors and cancer cell lines (reviewed in [160]), including human mesenchymal tumors such as OS [161,162] and Ewing’s sarcoma [163]. Originally described by Goodell et a1 [164,165], SP cells are typically identified on lower left quadrant of FACS dot plot as a population of cells displaying low blue and low red fluorescence, after incubation of target cells with the fluorescent dye Hoechst 33342 and subsequent analysis of dual-wavelength fluorescence [166]. These low- staining SP cells disappear from the FACS profile after treatment with ATP-binding cassette (ABC) transporter inhibitors such as verapamil, reserpine or fumitremorgin C (FTC). Chemotherapeutic resistance has remained a major cause of tumor relapse and therapeutic failure. One mechanism by which cancer cells become resistant to chemotherapy is the expression of ABC transporter proteins which use the energy of ATP hydrolysis to actively pump out a wide variety of substrates, including drugs as well as 13 fluorescent dyes such as Hoechst, across the cell membrane [167]. Thus, Hoechst efflux- based SP analysis seems to be an attractive surrogate system to increase our understanding of therapeutic failure and tumor recurrence. Chemotherapeutic resistance has remained a major cause of tumor relapse and therapeutic failure. One mechanism by which cancer cells become resistant to chemotherapy is the expression of ABC transporter proteins which use the energy of ATP hydrolysis to actively pump out a wide variety of substrates, including drugs as well as fluorescent dyes such as Hoechst, across the cell membrane [167]. Thus, Hoechst efflux- based SP analysis seems to be an attractive surrogate system to increase our understanding of therapeutic failure and tumor recurrence. In the present study, a comparative approach was undertaken by carrying out parallel experiments in canine and human derived MSCs. Canine OS has been shown to be a relevant and informative model for human OS. Canine OS is a frequently observed neoplasm in veterinary oncology, and we have ready access to a large number of such cases though Comparative Oncology Center on-campus. In this regard, further understanding of the biology of OS may yield a relevant spontaneously occurring tumor model which can then provide further opportunities for initiating follow-up intervention studies that may serve as a critical step for improving the therapeutic outcome of canine OS as well as in translating these therapeutic approaches to human subjects. 14 10. ll. 12. REFERENCES . Fialkow PJ. Clonal origin of human tumors. Annu Rev Med 1979;30:135-143. Bruce WR, Van Der Gaag A. A quantitative assay for the number of murine lymphoma cells capable of proliferation in vivo. Nature 1963;199:79-80. F idler IJ, Hart IR. Biological diversity in metastatic neoplasms: origins and implications. Science 1 982;2 1 7:998- 1003. Reya T, Morrison SJ, Clarke MF et al. Stem cells, cancer, and cancer stem cells. Nature 2001;414:105-111. Pardal R, Clarke MF, Morrison SJ. Applying the principles of stem-cell biology to cancer. Nat Rev Cancer 2003;32895-902. Jordan CT. Cancer stem cell biology: from leukemia to solid tumors. Curr Opin Cell Biol 2004;16:708-712. Clarke MF, Fuller M. Stem cells and cancer: two faces of eve. Cell 2006;124:1111-1115. Polyak K, Hahn WC. Roots and stems: stem cells in cancer. Nat Med 2006;12:296-300. Wicha MS, Liu S, Dontu G. Cancer stem cells: an old idea--a paradigm shift. Cancer Res 2006;66:1883-1890. Dick JE. Stem cell concepts renew cancer research. Blood 2008;112:4793-4807. Cohnheim V. Congenitales, quergestreiftes muskelsarkom der nieren. Virchows Arch Pathol Anat Physiol Klin Med 1875;65:64-69. Kleinsmith LJ, Pierce GB, Jr. Multipotentiality of single embryonal carcinoma cells. Cancer Res 1964;24:1544-1551. 15 13. I4. 15. l6. I7. 18. 19. 20. 21. 22. 23. 24. Park CH, Bergsagel DE, McCulloch EA. Mouse myeloma tumor stem cells: a primary cell culture assay. J Natl Cancer Inst 1971;46:411-422. Pierce GB, Johnson LD. Differentiation and cancer. In Vitro 197 l ;7: 140-145. Potter VR. Phenotypic diversity in experimental hepatomas: the concept of partially blocked ontogeny. The 10th Walter Hubert Lecture. Br J Cancer 1978;38:1-23. Lapidot T, Sirard C, Vonnoor J et al. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice. Nature 1994;367:645-648. Jordan CT, Guzman ML, Noble M. Cancer stem cells. N Engl J Med 2006;355:1253-1261. Beachy PA, Karhadkar SS, Berman DM. Tissue repair and stem cell renewal in carcinogenesis. Nature 2004;432:324-331. Marx J. Mutant stem cells may seed cancer. Science 2003;301:1308-1310. Chang CC. Recent translational research: stem cells as the roots of breast cancer 1. Breast Cancer Res 2006;8:103. Sun W, Kang KS, Morita I et a]. High susceptibility of a human breast epithelial cell type with stem cell characteristics to telomerase activation and inunortalization. Cancer Res 1999;59:61 18-6123. Merritt AJ, Potten CS, Watson AJ et a1. Differential expression of bcl-2 in intestinal epithelia. Correlation with attenuation of apoptosis in colonic crypts and the incidence of colonic neoplasia l. J Cell Sci 1995;108 (Pt 6):2261-227l. Cheng T. Cell cycle inhibitors in normal and tumor stem cells. Oncogene 2004;23:7256-7266. Pardal R, Molofsky AV, He S et al. Stem cell self-renewal and cancer cell proliferation are regulated by common networks that balance the activation of 16 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. prom-oncogenes and tumor suppressors. Cold Spring Harb Symp Quant Biol 2005;70:177-185. ’ Clarke MF, Dick JE, Dirks PB et al. Cancer Stem Cells-Perspectives on Current Status and Future Directions: AACR Workshop on Cancer Stem Cells. Cancer Res 2006;66:9339-9344. Zhang M, Rosen JM. Stem cells in the etiology and treatment of cancer. Curr Opin Genet Dev 2006;16:60-64. Ward R], Dirks PB. Cancer stem cells: at the headwaters of tumor development. Annu Rev Pathol 2007 ;2: 1 75-189. Visvader JE, Lindeman GJ. Cancer stem cells in solid tumours: accumulating evidence and unresolved questions. Nat Rev Cancer 2008;8:755-768. van RA, Feller N, Kelder A et al. High stem cell frequency in acute myeloid leukemia at diagnosis predicts high minimal residual disease and poor survival. Clin Cancer Res 2005;11:6520-6527. Zeppernick F, Ahmadi R, Campos B et al. Stem cell marker CD133 affects clinical outcome in glioma patients. Clin Cancer Res 2008;14:123-129. Marotta LL, Polyak K. Cancer stem cells: a model in the making. Curr Opin Genet Dev 2009;19:44-50. Pitot HC, Goldsworthy TL, Moran S. The natural history of carcinogenesis: implications of experimental carcinogenesis in the genesis of human cancer. J Supramol Struct Cell Biochem 1981;17:133-146. Moolgavkar SH, Luebeck EG. Multistage carcinogenesis and the incidence of human cancer. Genes Chromosomes Cancer 2003;38:302-306. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57-70. Jarnieson CH, Ailles LE, Dylla SJ et al. Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML. N Engl J Med 2004;351:657- 667. I7 36. 37. 38. 39. 40. 4]. 42. 43. 44. 45. Passegue E, Jamieson CH, Ailles LE et al. Normal and leukemic hematopoiesis: are leukemias a stem cell disorder or a reacquisition of stem cell characteristics? Proc Natl Acad Sci U S A 2003;100 Suppl 1:11842-11849. Cozzio A, Passegue E, Ayton PM et al. Similar MLL-associated leukemias arising from self-renewing stem cells and short-lived myeloid progenitors. Genes Dev 2003; 1 7:3029-3035. Dai C, Celestino JC, Okada Y et al. PDGF autocrine stimulation dedifferentiates cultured astrocytes and induces oligodendrogliomas and oligoastrocytomas from neural progenitors and astrocytes in vivo. Genes Dev 2001;15: 1913-1925. Uhrbom L, Dai C, Celestino JC et a]. Ink4a-Arf loss cooperates with KRas activation in astrocytes and neural progenitors to generate glioblastomas of various morphologies depending on activated Akt. Cancer Res 2002;62:5551- 5558. Bachoo RM, Maher EA, Ligon KL et a1. Epidermal growth factor receptor and Ink4a/Arf: convergent mechanisms governing terminal differentiation and transformation along the neural stem cell to astrocyte axis. Cancer Cell 2002;] :269-277. Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer 2005;52275-284. Li X, Lewis MT, Huang J et a]. Intrinsic resistance of tumorigenic breast cancer cells to chemotherapy. J Natl Cancer Inst 2008;100:672-679. Oravecz-Wilson K1, Philips ST, Yilmaz OH et al. Persistence of leukemia- initiating cells in a conditional knockin model of an imatinib-responsive myeloproliferative disorder. Cancer Cell 2009; 16: 137-148. Baumann M, Krause M, Hill R. Exploring the role of cancer stem cells in radioresistance. Nat Rev Cancer 2008;8:545-554. Bao S, Wu Q, McLendon RE et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 2006;444:756-760. 18 46. 47. 48. 49. 50. 5]. 52. 53. 54. 55. 56. Diehn M, Cho RW, Lobo NA et al. Association of reactive oxygen species levels and radioresistance in cancer stem cells. Nature 2009;458:780-783. Cairns J. Somatic stem cells and the kinetics of mutagenesis and carcinogenesis. Proc Natl Acad Sci U S A 2002;99: 10567-10570. Donnenberg VS, Donnenberg AD. Multiple drug resistance in cancer revisited: the cancer stem cell hypothesis. J Clin Pharmacol 2005 ;45:872-877. Al-Hajj M, Becker MW, Wicha M et al. Therapeutic implications of cancer stem cells. Curr Opin Genet Dev 2004; 14:43-47. Guzman ML, Swiderski CF, Howard DS et al. Preferential induction of apoptosis for primary human leukemic stem cells. Proc Nat] Acad Sci U S A 2002;99: 16220-16225. Xu Q, Simpson SE, Scialla T] et a]. Survival of acute myeloid leukemia cells requires PI3 kinase activation. Blood 2003;102:972-980. Guzman ML, Rossi RM, Kamischky L et al. The sesquiterpene lactone parthenolide induces apoptosis of human acute myelogenous leukemia stem and progenitor cells. Blood 2005; 105:4163-4169. Xu Q, Thompson JE, Carroll M. mTOR regulates cell survival after etoposide treatment in primary AML cells. Blood 2005;106:4261-4268. Yilmaz OH, Valdez R, Theisen BK et al. Pten dependence distinguishes haematopoietic stem cells from leukaemia-initiating cells. Nature 2006;441:475- 482. Todaro M, Alea MP, Di Stefano AB et al. Colon cancer stem cells dictate tumor growth and resist cell death by production of interleukin-4. Cell Stem Cell 2007 ;1:389-402. Jordan CT, Upchurch D, Szilvassy SJ et al. The interleukin-3 receptor alpha chain is a unique marker for human acute myelogenous leukemia stem cells. Leukemia 2000;14:1777-1784. l9 57. 58. 59. 60. 61. 62. 63. 65. 66. 67. F euring-Buske M, Frankel AE, Alexander RL et a]. A diphtheria toxin-interleukin 3 fusion protein is cytotoxic to primitive acute myeloid leukemia progenitors but spares normal progenitors. Cancer Res 2002;62: 1730-1736. Bonnet D, Warren EH, Greenberg PD et al. CD8(+) minor histocompatibility antigen-specific cytotoxic T lymphocyte clones eliminate human acute myeloid leukemia stem cells. Proc Natl Acad Sci U S A 1999;96:8639-8644. Hamann PR, Hinman LM, Hollander I et a]. Gemtuzumab ozogamicin, a potent and selective anti-CD33 antibody-calicheamicin conjugate for treatment of acute myeloid leukemia. Bioconjug Chem 2002;13:47-58. Schatton T, Murphy GF, Frank NY et al. Identification of cells initiating human melanomas. Nature 2008;451:345-349. Calabrese C, Poppleton H, Kocak M et al. A perivascular niche for brain tumor stem cells. Cancer Cell 2007 ;1 1:69-82. Benedetti L, Grignani F, Scicchitano BM et a]. Retinoid-induced differentiation of acute promyelocytic leukemia involves PML-RARalpha-mediated increase of type II transglutaminase. Blood 1996;87:1939-1950. Jain M, Arvanitis C, Chu K et al. Sustained loss of a neoplastic phenotype by brief inactivation of MYC. Science 2002;297:102-104. Ohno R, Asou N, Ohnishi K. Treatment of acute promyelocytic leukemia: strategy toward further increase of cure rate. Leukemia 2003;] 7: 1454-1463. Jin L, Hope KJ, Zhai Q et al. Targeting of CD44 eradicates human acute myeloid leukemic stem cells. Nat Med 2006; 12: 1 167-1 174. Ito K, Bemardi R, Morotti A et al. PML targeting eradicates quiescent leukaemia- initiating cells. Nature 2008;453:1072-1078. Malanchi I, Peinado H, Kassen D et al. Cutaneous cancer stem cell maintenance is dependent on beta-catenin signalling. Nature 2008;452:650-653. 20 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004;10:789-799. Cooper CS, Park M, Blair DG et al. Molecular cloning of a new transforming gene from a chemically transformed human cell line. Nature 1984;311:29-33. Ferracini R, Di Renzo MF, Scotlandi K et al. The Met/HGF receptor is over- expressed in human osteosarcomas and is activated by either a paracrine or an autocrine circuit. Oncogene 1995;10:739-749. Wallenius V, Hisaoka M, Helou K et al. Overexpression of the hepatocyte growth factor (HGF) receptor (Met) and presence of a truncated and activated intracellular HGF receptor fragment in locally aggressive/malignant human musculoskeleta] tumors. Am J Pathol 2000;156:821-829. Ferracini R, Angelini P, Cagliero E et a]. MET oncogene aberrant expression in canine osteosarcoma. J Orthop Res 2000;18:253-256. MacEwen EG, Kutzke J, Carew J et al. c-Met tyrosine kinase receptor expression and function in human and canine osteosarcoma cells. Clin Exp Metastasis 2003;20:421-430. De MR, Miretti S, Iussich S et al. met oncogene activation qualifies spontaneous canine osteosarcoma as a suitable pre-clinical model of human osteosarcoma. J Pathol 2009;218:399-408. ‘ Patane S, Avnet S, Coltella N et a]. MET overexpression turns human primary osteoblasts into osteosarcomas. Cancer Res 2006;66:4750-4757. Lessard J, Sauvageau G. Bmi-I determines the proliferative capacity of normal and leukaemic stem cells. Nature 2003;423:255-260. Leung C, Lingbeek M, Shakhova O et al. Bmil is essential for cerebellar development and is overexpressed in human medulloblastomas. Nature 2004;428:337-341. Valk-Lingbeek ME, Bruggeman SW, Van LM. Stem cells and cancer; the polycomb connection. Cell 2004;1 18:409-418. 2] 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. Dimri GP, Martinez JL, Jacobs JJ et al. The Bmi-l oncogene induces telomerase activity and immortalizes human mammary epithelial cells. Cancer Res 2002;62:4736-4745. Haupt Y, Bath ML, Harris AW et al. bmi-l transgene induces lymphomas and collaborates with myc in tumorigenesis. Oncogene 1993;8:3161-3164. Zhang X, Soda Y, Takahashi K et al. Successful immortalization of mesenchymal progenitor cells derived from human placenta and the differentiation abilities of immortalized cells. Biochem Biophys Res Commun 2006;351:853-859. O'Brien SJ, Murphy WJ. Genomics. A dog's breakfast? Science 2003;301:1854- 1855. Acland GM, Aguirre GD, Ray J et al. Gene therapy restores vision in a canine model of childhood blindness. Nat Genet 2001;28:92-95. Ponder KP, Melniczek JR, Xu L et al. Therapeutic neonatal hepatic gene therapy in mucopolysaccharidosis VII dogs. Proc Natl Acad Sci U S A 2002;99:13102- 13 107. Guzman ML, Rossi RM, Neelakantan S et al. An orally bioavailable parthenolide analog selectively eradicates acute myelogenous leukemia stem and progenitor cells. Blood 2007;110:4427-4435. Mueller F, Fuchs B, Kaser-Hotz B. Comparative biology of human and canine osteosarcoma. Anticancer Res 2007;27:155-164. Hansen K, Khanna C. Spontaneous and genetically engineered animal models; use in preclinical cancer drug development. Eur J Cancer 2004;40:858-880. Paoloni M, Khanna C. Translation of new cancer treatments from pet dogs to humans. Nat Rev Cancer 2008;8: 147-156. Khanna C, Lindblad-Toh K, Vail D et a]. The dog as a cancer model. Nat Biotechnol 2006;24:1065-1066. 22 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. Hahn KA, Bravo L, Adams WH et al. Naturally occurring tumors in dogs as comparative models for cancer therapy research. In Vivo l994;8:133-l43. Knapp DW, Waters DJ. Naturally occurring cancer in pet dogs: important models for developing improved cancer therapy for humans. Mol Med Today 1997;328- 1]. Vail DM, MacEwen EG. Spontaneously occurring tumors of companion animals as models for human cancer. Cancer Invest 2000; 1 82781-792. Mack GS. Clinical trials going to the dogs: canine program to study tumor treatment, biology. J Natl Cancer Inst 2006;98:161-162. Thompson KG, Pool RR. Tumors of bone. In: Meuten DJ, ed. Tumors in Domestic Animals. Iowa State Press, Ames, IA 2002:245-83. Barroga EF, Kadosawa T, Okumura M et al. Establishment and characterization of the growth and pulmonary metastasis of a highly lung metastasizing cell line from canine osteosarcoma in nude mice. J Vet Med Sci 1999;61:361-367. Withrow SJ, Powers BE, Straw RC et al. Comparative aspects of osteosarcoma. Dog versus man. Clin Orthop Relat Res 1991;159-168. Nagarajan R, Weigel BJ, Thompson RC et a]. Osteosarcoma in the first decade of life. Med Pediatr Oncol 2003;41:480-483. Ragland BD, Bell WC, Lopez R et a]. Cytogenetics and molecular biology of osteosarcoma. Lab Invest 2002;82:365-373. Knecht CD, Priester WA. Musculoskeletal tumors in dogs. J Am Vet Med Assoc 1978;172:72-74. Malawer MM, Link MP, Donaldson SS. Sarcomas of bone. In: DeVita DTJr, Rosenberg SA, Hellman S, eds. Cancer: principles and practice of oncology. Philadelphia, PA: Lippincott Williams and Wilkins 2001:1891-935. Demell ws, Straw RC, Withrow SJ. In: Withrow s, MacEwen EG, eds. Small Animal Clinical Oncology. Philadelphia: WB Saunders Co 20011378418. 23 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. Bielack SS, Kempf-Bielack B, Delling G et a]. Prognostic factors in high-grade osteosarcoma of the extremities or trunk: an analysis of 1,702 patients treated on neoadjuvant cooperative osteosarcoma study group protocols. J Clin Oncol 2002;20:776-790. Unni KK. Osteosarcoma of bone. J Orthop Sci 1998;3z287-294. Marina N, Gebhardt M, Teot L et al. Biology and therapeutic advances for pediatric osteosarcoma Oncologist 2004;92422-441. Tang N, Song WX, Luo J et al. Osteosarcoma development and stem cell differentiation. Clin Orthop Relat Res 2008;466:2114-2130. Kansara M, Thomas DM. Molecular pathogenesis of osteosarcoma. DNA Cell Biol 2007;26:1-18. Wang LL. Biology of osteogenic sarcoma. Cancer J 2005;11:294-305. Khanna C, Wan X, Bose S et a]. The membrane-cytoskeleton linker ezrin is necessary for osteosarcoma metastasis. Nat Med 2004;10:182-186. Khanna C, Hunter K. Modeling metastasis in vivo. Carcinogenesis 2005;26:513- 523. Haydon RC, Luu HH, He TC. Osteosarcoma and osteoblastic differentiation: a new perspective on oncogenesis. Clin Orthop Relat Res 2007;454:237-246. Thomas D, Kansara M. Epigenetic modifications in osteogenic differentiation and transformation. J Cell Biochem 2006;98:757-769. Straw RC. Tumors of the skeletal system. In: Withrow SJ, MacEwen GE, eds. Small Animal Clinical Oncology. W.B. Saunders Co. Philadelphia 1996:287-305. Young HE, Mancini ML, Wright RP et al. Mesenchymal stem cells reside within the connective tissues of many organs. Dev Dyn 1995;202:137-144. Caplan AI, Bruder SP. Mesenchymal stem cells: building blocks for molecular medicine in the let century. Trends Mol Med 2001;72259-264. 24 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. Helrnan LJ, Meltzer P. Mechanisms of sarcoma development. Nat Rev Cancer 2003;3:685-694. Tirode F, Laud-Duval K, Prieur A et al. Mesenchymal stem cell features of Ewing tumors. Cancer Cell 2007;] 1:421-429. Bane BL, Evans HL, Ro .IY et al. Extraskeletal osteosarcoma. A clinicopathologic review of 26 cases. Cancer 1990 ,:65 2762-2770. Patnaik AK. Canine extraskeletal osteosarcoma and chondrosarcoma: a clinicopathologic study of 14 cases. Vet Pathol 1990;27:46-55. Chung EB, Enzinger FM. Extraskeletal osteosarcoma. Cancer 1987;60:1 132- 1142. Vitale-Cross L, Amomphimoltham P, Fisher G et a]. Conditional expression of K- ras in an epithelial compartment that includes the stem cells is sufficient to promote squamous cell carcinogenesis. Cancer Res 2004;64:8804-8807. Funes JM, Quintero M, Henderson S et al. Transformation of human mesenchymal stem cells increases their dependency on oxidative phosphorylation for energy production. Proc Natl Acad Sci U S A 2007; 104:6223-6228. Park CY, Tseng D, Weissman IL. Cancer stern cell-directed therapies: recent data from the laboratory and clinic. Mo] Ther 2009;17:219-230. Rosner MH, Vigano MA, Ozato K et a]. A POU-domain transcription factor in early stem cells and germ cells of the mammalian embryo. Nature 1990;345:686- 692. Loh YH, Wu Q, Chew JL et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat Genet 2006, 38: 43 1- 440. Tai MH, Chang CC, Kiupel M et a]. Oct4 expression in adult human stem cells: evidence in support of the stem cell theory of carcinogenesis. Carcinogenesis 2005;26:495-502. 25 126. 127. 128. 129. 130. 131. 132. I33. 134. 135. 136. Jiang Y, Jahagirdar BN, Reinhardt RL et al. Pluripotency of mesenchymal stem cells derived from adult marrow. Nature 2002;418:41-49. Kucia M, Reca R, Campbell FR et al. A population of very srnal] embryonic-like (V SEL) CXCR4(+)SSEA-1(+)Oct-4+ stem cells identified in adult bone marrow. Leukemia 2006;20:857-869. Matthai C, Horvat R, Noe M et al. Oct-4 expression in human endometrium. Mol Hum Reprod 2006;12:7-10. Monk M, Holding C. Human embryonic genes re-expressed in cancer cells. Oncogene 2001 ;20:8085-8091 . Hochedlinger K, Yamada Y, Beard C et al. Ectopic expression of Oct-4 blocks progenitor-cell differentiation and causes dysplasia in epithelial tissues. Cell 2005;121:465-477. Looijenga LH, Stoop H, de Leeuw HP et a]. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003;63:2244- 2250. Ezeh UT, Turek PJ, Reijo RA et a]. Human embryonic stem cell genes OCT4, NAN 0G, STELLAR, and GDF3 are expressed in both seminoma and breast carcinoma. Cancer 2005;104:2255-2265. Atlasi Y, Mowla SJ, Ziaee SA et al. OCT-4, an embryonic stem cell marker, is highly expressed in bladder cancer. Int J Cancer 2007;120:1598-1602. Webster JD, Yuzbasiyan-Gurkan V, Trosko JE et a]. Expression of the embryonic transcription factor Oct4 in canine neoplasms: a potential marker for stem cell subpopulations in neoplasia. Vet Pathol 2007;44:893-900. Chen YC, Hsu HS, Chen YW et al. Oct-4 expression maintained cancer stem-like properties in lung cancer-derived CD133-positive cells. PLoS ONE 2008;32e2637. Chiou SH, Yu CC, Huang CY et al. Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and high-grade oral squamous cell carcinoma. Clin Cancer Res 2008; 14:4085-4095. 26 137. I38. 139. 140. 14]. 142. 143. 144. I45. 146. 147. 148. Hu T, Liu S, Breiter DR et al. Octamer 4 small interfering RNA results in cancer stem cell-like cell apoptosis. Cancer Res 2008;68:6533-6540. Iki K, Pour PM. Expression of Oct4, a stem cell marker, in the hamster pancreatic cancer model. Pancreatology 2006;6z406-413. Karoubi G, Gugger M, Schmid R et al. OCT4 expression in human non-small cell lung cancer: implications for therapeutic intervention. Interact Cardiovasc Thorac Surg 2009;82393-397. Gibbs CP, Kukekov VG, Reith JD et a]. Stem-like cells in bone sarcomas: implications for tumorigenesis. Neoplasia 2005;72967-976. Wilson H, Huelsmeyer M, Chun R et a]. Isolation and characterisation of cancer stem cells from canine osteosarcoma. Vet J 2008;175:69-75. Yamaguchi S, Yamazaki Y, Ishikawa Y et a]. EWSRI is fused to POUSF] in a bone tumor with translocation t(6;22)(p21;q12). Genes Chromosomes Cancer 2005;43:217-222. Gidekel S, Pizov G, Bergman Y et al. Oct-3/4 is a dose-dependent oncogenic fate determinant. Cancer Cell 2003;4z361-370. Cantz T, Key G, Bleidissel M et al. Absence of OCT4 expression in somatic tumor cell lines. Stem Cells 2008;26:692-697. Lengner CJ, Camargo FD, Hochedlinger K et al. Oct4 expression is not required for mouse somatic stem cell self-renewal. Cell Stem Cell 2007;] :403-415. Liedtke S, Stephan M, Kogler G. Oct4 expression revisited: potential pitfalls for data misinterpretation in stem cell research. Biol Chem 2008;389:845-850. Lin H, Shabbir A, Molnar M et al. Stem cell regulatory function mediated by expression of a novel mouse Oct4 pseudogene. Biochem Biophys Res Commun 2007;355:111-116. Suo G, Han J, Wang X et a1. Oct4 pseudogenes are transcribed in cancers. Biochem Biophys Res Commun 2005;337:1047-1051. 27 I49. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. Zangrossi S, Marabese M, Broggini M et al. Oct-4 expression in adult human differentiated cells challenges its role as a pure stem cell marker. Stem Cells 2007;25: 1675-1680. Giudice A, Trounson A. Genetic modification of human embryonic stem cells for derivation of target cells. Cell Stem Cell 2008;2z422-433. Gerrard L, Zhao D, Clark A] et al. Stably transfected human embryonic stem cell clones express OCT4-specific green fluorescent protein and maintain self-renewal and pluripotency. Stem Cells 2005;23:124-133. Eiges R, Schuldiner M, Drukker M et al. Establishment of human embryonic stem cell-transfected clones carrying a marker for undifferentiated cells. Curr Biol 2001;11:514-518. Singh RN, Nakano T, Xuing L et al. Enhancer-specified GFP-based FACS purification of human spinal motor neurons from embryonic stem cells. Exp Neurol 2005;196:224-234. Huber I, Itzhaki I, Caspi O et al. Identification and selection of cardiomyocytes during human embryonic stem cell differentiation. FASEB J 2007;21:2551-2563. Lavon N, Yanuka O, Benvenisty N. Differentiation and isolation of hepatic-like cells from human embryonic stem cells. Differentiation 2004;72:230-238. Levings PP, McGarry SV, Currie TP et al. Expression of an exogenous human Oct-4 promoter identifies tumor-initiating cells in osteosarcoma. Cancer Res 2009;69:5648-5655. Hadnagy A, Gaboury L, Beaulieu R et a]. SP analysis may be used to identify cancer stem cell populations. Exp Cell Res 2006;312:3701-3710. Challen GA, Little MH. A side order of stem cells: the SP phenotype. Stem Cells 2006;24:3-12. Zhou S, Schuetz JD, Bunting KD et al. The ABC transporter Bcrpl/ABCG2 is expressed in a wide variety of stem cells and is a molecular determinant of the side-population phenotype. Nat Med 2001;7: 1028- I 034. 28 160. 161. I62. I63. 164. 165. 166. 167. Wu C, Alman BA. Side population cells in human cancers. Cancer Lett 2008;268:1-9. Wu C, Wei Q, Utomo V et a]. Side population cells isolated from mesenchymal neoplasms have tumor initiating potential. Cancer Res 2007;67:8216-8222. Tsuchida R, Das B, Yeger H et al. Cisplatin treatment increases survival and expansion of a highly tumorigenic side-population fraction by upregulating VEGF/Fltl autocrine signaling. Oncogene 2008;27:3923-3934. Hirschmann-Jax C, Foster AE, Wulf GG et al. A distinct "side population" of cells with high drug efflux capacity in human tumor cells. Proc Natl Acad Sci U S A 2004;101:14228-14233. Goodell MA, Brose K, Paradis G et al. Isolation and functional properties of murine hematopoietic stem cells that are replicating in vivo. J Exp Med 1996;183:1797-1806. Goodell MA, Rosenzweig M, Kim H et a1. Dye efflux studies suggest that hematopoietic stem cells expressing low or undetectable levels of CD34 antigen exist in multiple species. Nat Med 1997;3:1337-1345. Scharenberg CW, Harkey MA, Torok-Storb B. The ABCG2 transporter is an efficient Hoechst 33342 efflux pump and is preferentially expressed by immature human hematopoietic progenitors. Blood 2002;99:507-512. Borst P, Elferink RO. Mammalian ABC transporters in health and disease. Annu Rev Biochem 2002;71:537-592. 29 CHAPTER 2 Isolation and characterization of canine adipose-derived mesenchymal stem cells 30 SUMMARY This study is the first documentation of the isolation and extensive characterization of mesenchymal stem cells from canine adipose tissue. Methods previously used by our group to isolate and differentiate human adipose-derived mesenchymal stem cells (hAD-MSCs) have been modified and optimized for derivation of similar cells from canine adipose tissues. The canine adipose-derived mesenchymal stem cells (CAD-MSCs) showed lower proliferation ability and were refractory to osteogenic and adipogenic differentiation under conditions employed to differentiate hAD-MSCs. The differentiation of cAD-MSCs into osteoblasts and adipocytes was effectively achieved under modified conditions, by using laminin-coated plates and PPAR-y ligands, respectively. The formation of micromass was sufficient to induce Chondrogenesis, unlike hAD-MSCs, which require TGF-B. These cells displayed anchorage independent growth in soft agar, and their colony forming efficiency in plastic was comparable with human counterparts. The CAD-MSCs expressed genes associated with pluripotency while their differentiated progeny expressed appropriate lineage specific genes. The optimization of growth and differentiation of CAD-MSCs should facilitate future stem cell-based reparative and regenerative studies in dogs. The dog is a promising biomedical model which is suitable for evaluation of novel therapies such as those employing stem cells in experimental and in spontaneous disease settings. 31 INTRODUCTION Tissue engineering holds much promise for the regenerative treatment of various tissue disorders as well as for the delivery of therapeutic genes. One major area of great potential is the use of mesenchymal stem cells (MSCs) for the restoration of bone defects. By virtue of their cellular differentiation potential [1] and trophic effects [2], MSCs are promising candidates for regenerative and reparative medicine. The MSCs exist in several connective tissue compartments [3], and have been isolated from various tissues [4]. Adipose tissue is an attractive source of mesenchymal stem cells because of its abundance and ease of access with minimal donor site morbidity. The current study was undertaken with the goal of isolation, expansion and characterization of MSCs from adult canine adipose tissue (CAD-MSCs). Repair of major fractures, better outcomes of major reconstruction surgeries and implants such as hip replacement are of equal concern in both human and veterinary medicine. The dog is a promising biomedical model for evaluation of novel therapies such as those employing stem cells in experimental and in spontaneous disease settings [5]. Recent evidence on use of mesoangioblast stem cells for the treatment of muscular dystrophy in dogs demonstrated the potential of this model system [6]. 32 MATERIALS AND METHODS Materials: All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise stated. Isolation of canine adipose tissue derived mesenchymal stem cells (cAD-MSCs): The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee. Methods used to isolate mesenchymal stem cells from human adipose tissue [7], were adapted to isolate mesenchymal stem cells from adult canine adipose tissue. The dogs ranged in age fi'om 1 to 3 years, and were clinically healthy dogs undergoing experimental surgeries unrelated to this study. Briefly, adipose tissue was collected from subcutaneous, omental, and inguinal fat depots of dogs, using standard surgical procedures. Each adipose tissue sample was weighed, and digested overnight at 37°C with Collagenase type IA (1mg/ml) in D medium, a modified Eagle’s MEM medium (5 ml medium/ gm of tissue) supplemented with N-Acetyl-L-Cysteine (NAC) (2 mM), L-ascorbic acid 2-phosphate (Asc 2P) (0.2 mM), penicillin, streptomycin & amphotericin. Following centrifugation and washing of the pellet, cells were incubated (about 8 gm of tissue/25 cm2 flask) in D medium with 10% fetal bovine serum (F BS), NAC (2 mM) and Asc 2P (0.2 mM) in incubator supplied with humidified air and 5% C02. Unattached cells were removed the next day by washing with phosphate buffer saline. Adherent cells were cultured in the K-NAC medium with 5% FBS. The K-NAC medium is a modified MCDB 153 medium (Keratinocyte-SFM, GIBCO—Invitrogen Corporation, Carlsbad, CA) supplemented with NAC (2 mM) and Asc 2P (0.2 mM). This low calcium (0.09 mM) medium contains recombinant epidermal grth factor (rEGF, 5 33 ng/ml), bovine pituitary extract (BPE, 50 rig/ml), insulin (5 jig/ml), hydrocortisone (74 ng/ml) and 3, 3’, 5-triiodo-D.L.-thyronine (T3) (6.7 ng/ml) [8]. The medium was renewed every 3 days. The cells, grown to near confluence, were quantified, and subcultured or cryopreserved for further studies in K-NAC medium with 10% FBS and 10% DMSO. Differentiation of cAD—MSCs (Osteogenesis, Chondrogenesis, and Adipogenesis): Cells derived from subcutaneous and omental fat and expanded in the K-NAC medium with 5% FBS were used for differentiation studies. The cells were treated by different induction cocktails in D medium with 10% FBS. All studies were carried out with same number of controls. Osteogenesis: Cells were plated at the seeding density of 1000 cells/cm2 in 6 well plates (regular or laminin-coated) and treated with dexamethasone (0.1 11M) or 1, 25- dihydroxyvitarrrin D; (0.0] pM), Asc 2P (50 uM) and B-glycerophosphate disodium (10 mM) (DAG or VAG cocktail) in D medium containing 10% FBS for 6-8 weeks with medium change once in every 3 days. Alizarin Red staining and von Kossa staining were carried out to detect calcified extracellular matrix deposits. Quantitative Assay for Calcium: Calcium concentrations in treated and control plates were assayed with an inductively coupled plasma-atomic emission spectrophotometer (Vista AX), (V arian, Australia) using Yttrium internal standard and Cesium ionization suppressant at 370.602 nm. Chondrogenesis: For Chondrogenesis, micromass cultures of cells (1 x 105 cells/10 uL) were incubated and formed in 24-well plates for 2.5 hours and then treated by TGF-Bl (10 ng/ml), Asc 2P (50 uM) and insulin (6.25 rig/ml) (TAI cocktail) for 14 days, with 34 medium change once in every 3 days. The micromasses was stained both grossly and histologically (5 micron thick paraffin embedded sections), with Alcian blue for the presence of sulfated proteoglycan—rich matrix. Adipogenesis: For adipogenesis, cells were plated at the seeding density of 10,000 cells/cm2 in 6 well plates, and treated with the conventional protocol of IBMX (500 pM), dexamethasone (1 uM), indomethacin (100 pM), and insulin (10 jig/ml) (IDII cocktail) for 21 days with medium change once in every 3 days. We optimized the adipogenic induction cocktail for cAD-MSCs. Cells were treated with rosiglitazone (5 11M) (BioVision, Mountain View, CA), dexamethasone (1 uM), and insulin (5 jig/ml) (RDI cocktail) for 2 weeks with medium changed every 3 days. Induction efficacy of rabbit serum (5% or 10%), and high glucose (4.5 g/L) was also examined. Oil Red 0 staining was done to examine the lipid droplet formation. RT—PCR: Total RNAs were extracted from cells using Versagene RNA Purification Kit (5 Prime Inc, Gaithersburg, MD) and treated with DNase I (Ambion, Austin, TX) to remove contaminating DNA. cDNAs were synthesized from 1 pg total RNA using random hexamer primers and Superscript III reverse transcriptase(1nvitrogen). Primers derived from coding regions of respective genes in canine genome were used to amplify the target sites (Table 2.1). To ensure that primers would uniquely amplify the target transcripts, primers to some of the genes (OCT 4, BSP, OSTEOCALCIN) were designed to flank an intron, allowing us to rule out genomic contamination by simple inspection of product size. 25 pl PCR reactions were prepared with 2p] cDNA, 5 pmol of each primer, 0.5 units of Taq polymerase (Invitrogen), and final concentrations of 40 M dNTPs, 2 mM 35 MgClz, 20 mM Tris-HCI, and 50 pl KC]. Cycling conditions were as follows: 94°C for 4 min; 30-35 cycles at 94°C for 1 min, optimal annealing temperature (Table 2.1) for 1 min, 72°C for 1 min; followed by 72°C for 5 min. The PCR products were separated on 2% agarose gel by electr0phoresis, stained with ethidium bromide, visualized under UV light, and digital images captured with Alphalmager software (San Leandro, CA). Respective tissue samples (bone, cartilage, and fat from dog) were used as positive controls to validate the primers, and no template controls (water instead of cDNA) were used as negative controls. For OCT 4, NANOG, & SOX2 genes, PCR products were gel-purified using Qiaex H Gel Extraction Kit (Qiagen, Valencia, CA), sequenced with automated sequencer, and verified after sequence alignment with canine genome. Immunocytochemistry: After osteogenic induction, cells were trypsinized and cytospin preparations were made. The cytospin slides were fixed in cold acetone (-20°C) for 3 minutes and washed twice with tris buffered saline. Endogenous peroxidase was neutralized with 3% hydrogen peroxide for 6 minutes, and slides were incubated with 1:2 dilution of ready-to- use monoclonal mouse anti-human OSTEOCALCIN (BioGenex, San Ramon, CA) for 30 minutes, using an autostainer (BondMax) (Vision BioSystem, MA). Polymer reagent (Vision BioSystem, MA) was used for immunolabeling, and the immunoreaction was visualized with 3, 3’-diaminobenzidine chromogen (Dako, CO) under a microscope followed by counterstaining with Hematoxylin (SurgiPath, IL). After chondrogenic induction, the micromasses were fixed in 10% neutral buffered formalin, embedded in paraffin block, and similar immunostaining protocol (as 36 mentioned above) was followed. 1: 100 dilution of rabbit polyclonal anti-collagen 2 (DakoCytomation, Carpenteria, CA) was used as a source of primary antibody after validating the antibody with canine tissues, and a streptavidin—immunoperoxidase staining procedure (Dako) was used for immunolabeling. Proliferation potential of cAD-MSCs: Cumulative pepulation doubling level (cpdl): For determination of the cumulative population doubling level (cpdl), 100,000 cells were plated in 75 cm2 flask and grown in K-NAC medium containing 5% F BS until near confluence, to quantify the final cell yield and subcultured. The population doubling (pd) at each subculture was calculated by using the following equation, pd = ln(N/N,)/ln2, where Ni and Nf are initial and final cell ° numbers, respectively, and In is the natural log. The pds of continuous subculture were added to obtain cpd]. Colony formation efficiency in soft agar and on plastic: As described [9], 50,000 cells in 3 ml of 0.33% agarose medium were plated on top of 3 ml pre-hardened 0.5% agarose medium in each triplicate 60 mm dishes with grids to aid colony counting. 2.5 ml medium (K-NAC with 5% FBS) was then added and renewed every 3 days. The numbers of anchorage independent colonies were scored after 3 weeks. Colony forming efficiency on plastic was assayed by plating 200 cells in each of triplicate 100 mm plates in K-NAC medium with 5% FBS as well as 10% FBS for 3 weeks. The colonies were then stained with 1% crystal violet, and scored. 37 RESULTS Isolation of putative cAD-MSCs: Canine AD-MSCs were successfully isolated from subcutaneous and omental fat of 3 different dogs. As shown in Figure 2.1A, both serpiginous and fibroblast-like cells were observed in primary culture. These cells could maintain this phenotype and be expanded in culture (see confluent cells in Figure 2.1B). Similar to human adipose- derived MSCs, symmetric division of serpiginous (Figure 2.1C, black arrow) and fibroblast-like cells (Figure 2.1D, white arrow) as well as asymmetric division of serpiginous cells (Figure 2.1D, black arrow) were evident at low density (Figure 2.1D). These cells also displayed anchorage independent growth in soft agar (Figure 2.1E) at a frequency of 14.2 d: 3.7%. On plastic surface, the colony forming efficiency in K-NAC medium with 10% FBS was 17.75 :1: 3.2%. It was considerably low (4.67 i 2.3%) with 5% FBS. Some variability in proliferation rate was observed among initial cell cultures derived from different tissue sites, with cells derived from subcutaneous fat becoming confluent in 25 cm2 flasks in 5-6 days as compared to those derived from omental fat which took 11-12 days to reach confluency. This variation in proliferation capacity was also paralleled with cell yield studies in primary culture, where subcutaneous fat yielded the greatest number of cells per gm of fat (Table 2.2). Although cells from inguinal fat were attached to the plastic surface, they showed little proliferation, and cell yield was not calculated for them. However, cells from subcutaneous and omental fats were apparently similar in terms of their efficiency to form colonies in soft agar and plastic (difference of less than 2 standard deviations from the above-mentioned mean values). 38 Expression of stemness markers: RT-PCR analysis revealed that cAD-MSCs express pluripotency-associated transcription factors, OCT4, NANOG, and SOX2 (Figure 2.1G). Life span of cAD-MSCs in different media: The lifespan of CAD-MSCs derived from subcutaneous fat was 25 i 1.2 cpdl in 83 days after 8 passages, when cells were grown in the K-NAC medium with 5% F BS. The proliferation potential was significantly enhanced when these cells were grown in the 1:1 mixture of K-NAC and D medium supplemented with 5% F BS (cpdl = 40) in 66 days after 9 passages) at the cell seeding density of 400 cells/cm2 (Figure 2.1F). The proliferation potential of CAD-MSCs isolated from omental fat was quite similar (difference of about 1.5 standard deviation from the above-mentioned mean value). Differentiation of cAD-MSCs: The cAD-MSCs isolated from subcutaneous fat as well as those from omental fat were differentiated into 3 mesodermal lineages (osteogenic, chondrogenic and adipogenic). Osteogenesis: Upon induction with DAG cocktail, the cells became more cuboidal-like in phenotype, and continued to proliferate actively and formed cell aggregates that would roll in a sheet and easily detach (Figure 2.2 A, B and C). Deposition of calcified extracellular matrix was evident in treated cells that formed cell aggregates (Figure 2.2D), but not in monolayer cells as revealed by Alizarin Red staining (Figure 2.2E) as well as von Kossa (Figure 2.2F). Cell phenotype did not change in untreated cells. 39 Laminin coating of plates not only prevented the clumping of cells during differentiation, but also induced extensive deposition of calcified extracellular matrix on monolayer cell culture following 6 weeks of treatment (Figure 2.2 G, H and I, displaying unstained, Alizarin Red, and von Kossa stained cells, respectively). This mineralization was more pronounced when 1, 25-dihydroxyvitamin D3 was used in the in duction regimen (as shown in Figure 2.2) instead of dexamethasone (not shown). RUNXZ expression was stronger in osteo-induced cells, although basal expression was noticed in non-induced cells as well (Figure 2.20). COLIAI was strongly expressed in both induced cells and control CAD-MSCs. 0n the other hand, expression of OSTERIX, BSP, and OST E OCALCIN was seen only in induced cells, with no basal expression in control cells (Figure 2.20). After 20 days of osteogenic differentiation, the cells had clear diffuse cytoplasmic staining for OSTEOCALCIN, predominantly produced by osteoblasts (Figure 2.2M and 2.2N). Quantitative measurement of calcium showed that treated cells had about 16 fold higher level of calcium than in control cells (27.5 :t 8.6 pg per well (6-well plate) for treatment versus 1.7 i 0.4 pg per well in control). Chondrogenesis: Within a day after seeding the CAD-MSCs in micromass culture, three- dimensional aggregates were observed in both treated (Figure 2.3A) and control wells (Figure 2.3D). Both treated (Figure 2.3B) and control micromasses (Figure 2.3B) stained positive for Alcian Blue indicating presence of sulfated proteoglycans. This was confirmed by staining paraffin embedded sections corresponding to each group (Figure 2.3C- treated and Figure 2.3F, control). Putting the cells in micromasses (both treated and 40 control) resulted in specific expression of C 0L2A, AGGRECAN, COMP, COLIOA, and SOX 9, all of which were undetected in monolayer of cAD-MSCs expanded in KNAC medium (Figure 2.31). Moreover, immunostaining of sections of both treated and control micromasses confirmed the expression of COL2A (Figure 2.3G,H). Adipogenesis: Following the induction regimen (IDII) that was effective for adipogenic differentiation of hMSCs, adipogenic differentiation of CAD-MSCs seemed to be limited (Figure 2.4A- unstained and Figure 2.4B, Oil Red 0 stained). The use of rabbit serum (5%) in combination with rosiglitazone (5 pM), dexamethasone (1 pM), insulin (5 pg/ml) and high glucose (4.5 gm/L) considerably increased the adipogenic differentiation (in terms of both numbers and size of fat globules). Fat globules were noted within 4-5 days of treatment (Figure 2.4C), which continued to increase in size, and stained positive with Oil Red 0 (Figure 2.4D). In contrast, no differentiation was observed in the untreated group (Figure 2.4B, unstained and Figure 2.4F, Oil Red 0 stained). Expression of PPARyZ, CEBPa, FABP4, and LPL were specific to adipo-indcued cells (Figure 2.4G). Basal level of LEPTIN mRNA was observed in non-induced control cells, and expression level was increased following adipogenic induction (Figure 2.4G). Cells from both subcutaneous and omental fat from all the three donors were apparently equal in terms of their differentiation potential across all the three lineages tested. An estimation of randomly selected microscopic fields of treated plates showed that approximately > 80% of the total cells underwent differentiation (judged by the pattern of staining). 4] DISCUSSION This study has demonstrated for the first time the isolation of canine adult MSCs with extensive proliferation potential and multi-lineage differentiation ability from adipose tissue (CAD-MSCs). Although some studies have been done to demonstrate the in viva potential of canine bone marrow derived MSCs for tissue regeneration, irnmunogenecity and gene delivery [IO-14], most of them lacked in Vitro characterization of the cells used for these purposes. A recent study has demonstrated the in viva potential of BMP-2 modified cells isolated from canine adipose tissue [15], but those cells were not characterized for their proliferation potential, expression of stemness markers, and lineage specific markers upon induction. This study used the low calcium medium supplemented with antioxidants for the expansion of MSCs, which had been shown to extend the life span of hAD-MSCs [16]. Serpiginous shaped cells, which divide symmetrically as well as asymmetrically, were interspersed in the primary culture and subsequent passages of CAD-MSCs. The appearance of this stem cell morphology for CAD-MSCs paralleled hAD-MSCs [17], and other stem cell types [18,19]. The proliferation potential of CAD-MSCs in K-NAC medium with 5% FBS (cpdl = 25 in 82 days) was higher than hAD-MSCs reported by Zuk et al [20] (21 cpdl in 165 days after 13 passage), but lower than hAD-MSCs reported by Lin et a] [21] (35 cpdl in 62 days after 7 passages) using the same medium. This may reflect the species difference in growth requirement or proliferation potential of AD-MSCs. However, this lower proliferation potential of CAD-MSCs could be greatly improved by using a medium with 1:1 mixture of K-NAC and D medium supplemented with 5% FBS (40 cpdl in 66 days after 9 passages). Thus, these cells can be adequately expanded for a variety of 42 therapeutic applications. Our study showed that canine adipose tissues from different sites showed considerable variation in terms of cell yield, with subcutaneous fat yielding the greatest number of MSCs/gm of fat compared to omental and inguinal fat. Similar regional differences have been documented in various species [22,23]. OCT 4, NANOG, and SOX2 are pluripotency markers, which are usually ascribed to embryonic stem cells. However, their expression has been docrunented in some of the somatic stem cells [24,25]. CAD-MSCS expressed all of these three markers at mRNA level, suggesting the stemness of these cells isolated from canine fat. cAD-MSCs were able to undergo anchorage independent growth, a property which has been documented for a number of human adult stem cells including human mesenchymal [26] and liver [27] stem cells. Following osteogenic induction of cAD-MSCs with conventional DAG cocktail, mineralization occurred only in discrete foci where cell aggregates formed. This pattern is different from human AD-MSCs, [28,29] but is similar to observations in canine bone marrow derived MSCs [30]. Consistent with the recently documented role of laminin in osteogenesis of human MSCs [3 I], inclusion of laminin in the induction regimen allowed differentiation and mineralization in the monolayer. As had been demonstrated earlier [32,33], the use of 1, 25 dihydroxyvitarnin D3 allowed more extensive deposition of calcified ECM after induction, compared to dexamethasone. Consistent with earlier studies in other systems [34], CAD-MSCs also expressed basal level of early stage transcription factor RUNX2 and extracellular matrix protein COL IA, which were upregulated following osteogenic induction. OSTERIX, a 43 late stage transcription factor, and two other osteoblast specific markers, BSP and OSTEOCALCIN, were expressed only following osteogenic induction. For human AD-MSCs, treatment with TGF [31 is necessary for the formation of three-dimensional aggregates and chondrogenic differentiation [35,36]. In contrast, cell- cell contact in micromass was sufficient for chondrogenic differentiation of CAD-MSCS. This finding is similar to that of bovine bone marrow derived MSCs [37]. Chondrogenic transcription factor, SOX9 and chondrocyte markers COL2A, AGGRE CAN, COMP, and C 0L1 0A were expressed after micromass culture (both TGF til-induced and uninduced) of CAD-MSCs, whereas cells expanded in monolayer did not express any of these markers. Canine MSCs were refractory to the commonly used induction condition for adipogenic differentiation of human MSCs, and needed significant optimization. Replacement of fetal bovine serum with rabbit serum and inclusion of rosiglitazone and higher glucose concentration in the medium, enhanced the adipogenic differentiation of canine MSCs. Adipogenic transcription factors PPARyZ and CEBPa, and adipocyte markers FABP4 and LPL were specifically expressed following adipogenic induction, whereas LEPTIN was expressed at basal level, even in undifferentiated MSCs. In conclusion, this study has clearly demonstrated that canine adipose tissue is a promising source of MSCs. Immediate and clinically relevant uses of CAD-MSCs include repair of torn tendons, ligaments and cartilage as well as broken bones, which parallel needs in human medicine. Employment of stem cells both in experimental as well as spontaneously occurring disease states in the dog will provide rigorous model systems to assess the translation of stem cell based therapy into clinics under both autologous and allogenic settings. Therefore, we propose that the methods demonstrated in this paper will allow pertinent issues of stem cell biology to be addressed in the canine model system and facilitate the realization of therapeutic potential of stem cells. However, there are species differences in the properties of adipose-derived MSCs. As with any study in a model organism, studies in dogs should be cautiously interpreted, even though they provide important insights into the biology and therapeutic potential of these cells. The exact nature or mechanism(s) underlying these differences should be explored in future studies. 45 APPENDIX Table 2.1. Primers used in RT-PCR of stemness and mesenchymal lineage-associated markers Prod uct Ann. Marker Gene Primer sequence (5' - 3') Size Tern s p. (C) OCT4 Forward GAGTGAGAGGCAACCTGGAG 274 60 Reverse GTGAAGTGAGGGCTCCCATA NANOG Forward GAATAACCCGAATTGGAGCAG 141 60 Stem Reverse AGCGATTCCTCTTCACAGTTG “855 SOX2 Forward AGTCTCCAAGCGACGAAAAA 142 58 Reverse GCAAGAAGCCTCTCCTTGAA RUNX2 Forward GTCTCCTTCCAGAATGCTTCC 100 62 Reverse GGAACTGAGGATGAGGAGAC COLIAI Forward GTAGACACCACCCTCAAGAGC 119 62 Reverse TTCCAGTCGGAGTGGCACATC OSX Forward ACGACACTGGGCAAAGCAG 285 60 Osteo Reverse CATGTCCAGGGAGGTGTAGAC blasts BSP Forward TTGCTCAGCATTTTGGGAATGG 295 60 Reverse AACGTGGCCGATACT'TAAAGAC OC Forward GAGGGCAGCGAGGTGGTGAG 134 62 Reverse TCAGCCAGCTCGTCACAGT'I’GG COL2A Forward GAAACTCTGCCACCCTGAATG 156 64 Reverse GCTCCACCAGT'TCTTCTTGG AGC Forward ATCAACAGTGCTTACCAAGACA 122 58 Reverse ATAACCTCACAGCGATAGATCC COMP Forward GTGGTGGACAAGATTGATGTG 120 54a Chondr Reverse CACCCAGTTGGGATCTATCTG mytes COLIOA Forward AGTAACAGGAATGCCGATGTC 121 52 Reverse TCTI‘GGGTCATAATGCTGTI‘GC SOX9 Forward GCTCGCAGTACGACTACACTGAC 101 60 Reverse GTTCATGTAGGTGAAGGTGGAG Reverse TGGCTCCATGAAGTCACCAAAGG CEBPa Forward AGTCAAGAAGTCGGTGGACAAG 151 601) Reverse GCGGTCATTGTCACTGGTGAG FABP4 Forward ATCAGTGTAAACGGGGATGTG 117 60 a Reaction contained 10% DMSO b Reaction contained 10% DMSO 46 Table 2.1 (contd’). Reverse GACT'ITTCTGTCATCCGCAGTA LPL Forward ACACATTCACAAGAGGGTCACC 134 60 Reverse CTCTGCAATCACACGGATGGC LEP Forward CTATCTGTCCTGTGTTGAAGCTG 102 60 Reverse GTGTGTGAAATGTCATTGATCCTG BMG Forward TCTACATTGGGCACTGTGTCAC 136 60 House Reverse TGAAGAGTTCAGGTCTGACCAAG keeping Table 2.2. Yield of plastic adherent cells in primary culture under conditions favoring MSCs enrichment Site Number of Cells/gm of fat dogs (flags—d.) Subcutaneous ‘ 3 528,000 :t 236,000 Omental 3 182,000 in 95,000 Inguinal 3 Much lower (not calculated) 47 Figure 2.1. Phenotype of CAD-MSCs. Morphology of CAD-MSCs at low (A) and high cell density (B). The cultures contain serpiginous-shaped and cuboid or fibroblast-like cells. The serpiginous cells showed symmetrical (C) or asymmetrical (D) division. (E) Anchorage independent colonies of CAD-MSCs developed in soft agar. Scale bar = 100 pm. (F) Cumulative population doubling of CAD-MSCs in K-NAC medium with 5% FBS or K-NAC : D (1:1) medium with 5% FBS. (G) mRN A expression of stemness markers OCT4, NANOG, and SOX2 in cAD-MSCs; [32 MICROGLOBULIN (BMG) is a housekeeping gene; M represents 100 bp DNA ladders. 48 -o- K-NAC:D(1 :1 )+ 5% FBS -o- K-NAC+ 5% FBS I If I f I I I I I I 0 10 20 30 40 50 60 70 80 90 100 G Days in culture M OCT4 NANOG SOX2 BMG Figure 2.1. 49 Figure 2.2. Osteogenic differentiation of cAD-MSCs. After differentiation induction with DAG supplementation, the cells continued to grow and started to cluster together leaving the vacant areas in tissue culture plate (A). They then rolled and formed aggregates (B, C) which eventually detached from the surface. Cell aggregates (D) showed positive red Alizarin Red staining (E) and black von Kossa staining (F). On laminin-coated plates (with VAG supplementation), the cells stayed in monolayer and were in time covered by calcified extracellular matrix (G) which was confirmed by Alizarin Red staining (H) as well as von Kossa staining (1). In contrast, uninduced cells did not show any morphological changes (J) and were negative for Alizarin Red (K) or von Kossa staining (L). Scale bars: A-C = 400 pm; D, F, G = 100 pm; E, H-L = 200 pm. Osteocalcin was abundantly distributed in the induced cells that underwent osteogenic differentiation (M), and was absent in uninduced cells (N). Scale bars = 20 run. (0) mRNA expression of osteogenenic markers; RUNX2 (RXZ), COLIAI (CNI), OSTERIX (OSX), 0ST EOCALCIN (0C), BONE SIALOPROTEIN (BSP). 50 Treatment Control Figure 2.2. 51 Figure 2.3. Chondrogenic differentiation of cAD-MSCs. Micromass formed under both TGF-B induced (A) and uninduced (D) condition. Treated (B) as well as control (B) micromasses stained positive for Alcian blue staining which were confirmed by staining of paraffin embedded sections of these micromasses (C, treatment and F, control). Scale bars: A, B, D, E = 200 pm; C, F e 100 pm. Immunostaining of micromasses with chondrocytes specific marker COL2A (G, treatment; H, control). Scale bars: G = 50 pm; H = 50 pm. (1) mRNA expression of chondrogenic markers: COL2 A ((‘N2), SOX9, AGGRECAN (AGC), COMP, COLIOA (CNIO). 52 Figure 2.4. Adipogenic differentiation of CAD-MSCs. Following adipogenic induction with IDII protocol, only few cells with fat droplets were noticed (A) and accordingly, Oil Red 0 stain was confined to a few cell aggregates but not in cell monolayer (B). When the induction regimen was optimized by supplementation with rabbit serum, rosiglitazone, and high glucose, numerous large fat droplets appeared within 4-5 days of induction (C) which showed positive Oil Red 0 staining (D).Untreated cells displayed fibroblast-like morphology (E), and were completely negative for Oil Red 0 staining (F). Scale bars: A, C-F = 100 pm; B = 200 pm. (G) mRNA expression of adipogenic markers; PPARyZ, C EBPa, FABP4, LPL, LEPT IN (LEP). 53 . '1 .r. I ' II v.11. felt. I! ‘II v II .bn. . LEP BM G M PPARGZ CEBPAZFABP4 LPL t n e m t a e r. T. Control Figure 2.4. 54 10. 11. REFERENCES . Caplan AI, Bruder SP. Mesenchymal stem cells: building blocks for molecular medicine in the 2 lst century. Trends Mol Med 2001;72259-264. . Caplan Al, Dennis JE. Mesenchymal stem cells as trophic mediators. J Cell Biochem 2006;98: 1076-1084. Young HE, Mancini ML, Wright RP et a]. Mesenchymal stem cells reside within the connective tissues of many organs. Dev Dyn 1995;202:137-144. Kolf CM, Cho E, Tuan RS. Mesenchymal stromal cells. Biology of adult mesenchymal stem cells: regulation of niche, self-renewal and differentiation. Arthritis Res Ther 2007;9z204. Neff MW, Rine I. A fetching model organism. Cell 2006;124:229-231. Sampaolesi M, Blot S, D'Antona G et al. Mesoangioblast stem cells ameliorate muscle function in dystrophic dogs. Nature 2006;444:574-579. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived hmnan mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Bruder SP, Kraus KH, Goldberg VM et al. The effect of implants loaded with autologous mesenchymal stem cells on the healing of canine segmental bone defects. J Bone Joint Surg Am 1998;80:985-996. Arinzeh TL, Peter SJ, Archarnbault MP et al. Allogeneic mesenchymal stem cells regenerate bone in a critical-sized canine segmental defect. J Bone Joint Surg Am 2003;85-A: 1927-1935. 55 12. 13. I4. 15. 16. l7. l8. 19. 20. 21. 22. Yamada Y, Ueda M, Naiki T et al. Autogenous injectable bone for regeneration with mesenchymal stem cells and platelet-rich plasma: tissue-engineered bone regeneration. Tissue Eng 2004;10:955-964. De Kok I, Drapeau SJ, Young R et al. Evaluation of mesenchymal stem cells following implantation in alveolar sockets: a canine safety study. Int J Oral Maxillofac Implants 2005;20:51 1-518. Mosca JD, Hendricks JK, Buyaner D et a]. Mesenchymal stem cells as vehicles for gene delivery. Clin Orthop Relat Res 2000;S71-S90. Li H, Dai K, Tang T et al. Bone regeneration by implantation of adipose-derived stromal cells expressing BMP—2. Biochem Biophys Res Commun 2007 ;356:836- 842. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Tang DG, Tokumoto YM, Apperly JA et al. Lack of replicative senescence in cultured rat oligodendrocyte precursor cells. Science 2001 ;291 :868-87 1. Zulewski H, Abraham EJ, Gerlach MJ et a]. Multipotentia] nestin-positive stem cells isolated from adult pancreatic islets differentiate ex vivo into pancreatic endocrine, exocrine, and hepatic phenotypes. Diabetes 2001;50:521-533. Zuk PA, Zhu M, Mizuno H et al. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 2001;7z211-228. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Tchkonia T, Tchoukalova YD, Giorgadze N et al. Abundance of two human preadipocyte subtypes with distinct capacities for replication, adipogenesis, and 56 23. 24. 25. 26. 27. 28. 29. 30. 31. apoptosis varies among fat depots. Am J Physiol Endocrinol Metab 2005;2882E267- E277. Lacasa D, Agli B, Moynard D et a]. Evidence for a regional-specific control of rat preadipocyte proliferation and differentiation by the androgenic status. Endocrine 1995;32789-793. Kucia M, Reca R, Campbell FR et al. A population of very small embryonic-like (VSEL) CXCR4(+)SSEA-1(+)Oct-4+ stem cells identified in adult bone marrow. Leukemia 2006;20:857-869. Izadpanah R, Joswig T, Tsien F et al. Characterization of multipotent mesenchymal stem cells from the bone marrow of rhesus macaques. Stem Cells Dev 2005;14:440- 451. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Chang CC, Tsai JL, Kuo K et al. Expression of Oct-4, alpha fetoprotein and vimentin and lack of gap-junctional intercellular communication (GJIC) as common phenotypes for human adult liver stem cells and hepatoma cells. Proc Am Assoc Cancer Res 2004;45:642. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Zuk PA, Zhu M, Mizuno H et al. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 2001;72211-228. Kadiyala S, Young RG, Thiede MA et al. Culture expanded canine mesenchymal stem cells possess osteochondrogenic potential in vivo and in vitro. Cell Transplant 1997;62125-134. Klees RF, Salasznyk RM, Kingsley K et al. Laminin-S induces osteogenic gene expression in human mesenchymal stem cells through an ERK-dependent pathway. Mol Biol Cell 2005;16:881-890. 57 32. 33. 34. 35. 36. 37. Zuk PA, Zhu M, Ashjian P et al. Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 2002;13:4279-4295. Volk SW, Diefenderfer DL, Christopher SA et al. Effects of osteogenic inducers on cultures of canine mesenchymal stem cells. Am J Vet Res 2005;66:1729-1737. Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143-147. Zuk PA, Zhu M, Mizuno H et a]. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 2001;72211-228. Zuk PA, Zhu M, Ashjian P et al. Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 2002;13:4279-4295. Bosnakovski D, Mizuno M, Kim G et al. Chondrogenic differentiation of bovine bone marrow mesenchymal stem cells in pellet cultural system. Exp Hematol 2004;32:502-509. 58 CHAPTER 3 Evaluation of gene expression in adipogenic induction of canine mesenchymal stem cells 59 SUMMARY Understanding the process of adipogenesis will provide critical insights into the pathophysiology of energy metabolism and obesity. Mesenchymal stem cells (MSCs) offer a suitable experimental model system to study the biology of adipose tissue development and disorders. Several attributes make the dog as an excellent animal model for obesity research, and we propose that use of the canine model for such studies will provide valuable insights into both basic biology and translational medicine. Comparative studies provide a special window to the mechanistic basis of adipogenesis. Our lab has successfully isolated and extensively characterized MSCs from both human and dog adipose tissues. Our earlier studies had revealed the differences in the effectiveness of different induction regimens between these two species. This study was undertaken to assess the species-specific differences in terms of expression of a panel of genes important in the process of adipogenesis, including adipogenic transcription factors, and early as well as late stage markers of adipogenesis. Both human and canine MSCs were exposed to following different potentially adipogenic media with different chemical supplements: 1) Control in high glucose DMEM + 5% FBS, 2) RD] (Rosiglitazone, Dexamethasone, and Insulin) + 5% F BS, 3) RD] + Linoleic acid + 5% FBS, 4) RD] + 5% Rabbit Serum, 5) Linoleic acid + 5% FBS, and 6) IDII (Indomethacin, Dexamethasone, Insulin, and IBMX) + 5% FBS. Accumulation of intracellular lipid was determined by histochemical staining. Quantitative RT-PCR was carried out to evaluate the pattern of gene expression across each treatment group between the two species. This study established a model experimental system for in vitro adipogenesis of MSCs. We found differences between human and canine AD-MSCs, 60 both in terms of stimuli and pattern of gene expression following induction. The molecular and cellular events underlying these differences warrant more detailed study. Directed differentiation of MSCs into adipocyte will provide an opportunity for better understanding of diseases of energy metabolisms such as obesity and obesity-related disorders. 61 INTRODUCTION It has been estimated that about 66% of adults in the United States are either overweight or obese, and in dogs, the incidence is as high as 40% [1]. Since obesity results from an increase in adipocyte size and differentiation of preadipocytes into mature adipocytes [2], understanding the process of adipocyte differentiation will provide critical insights into pathophysiology of obesity and obesity-related disorders [3,4]. Moreover, the balance between osteogenesis and adipogenesis is a target research area in the study of osteoporosis [5,6]. The process of adipogenesis has been extensively studied in established rodent cell lines such as 3T3-L1/F442A and Ob17, or in primary preadipoctyes isolated fiom stromal vascular fraction of adipose tissues [2,7-10]. Undifferentiated and multipotent mesenchymal stem cells (MSCs) [1 1-13] represent an excellent experimental system [3] to study the biology of adipose tissue development and disorders. Although the origin and development of adipocytes is not completely understood, a recent study has identified a subpopulation of CD34+ cells with the potential of MSCs as in viva precursors of white adipose tissue [14]. Multipotent MSCs undergo lineage-commitrnent to become preadipocytes, which then terminally differentiate into mature, lipid-filled, insulin- sensitive adipocytes [2,15]. Moreover, cells that reside outside the adipose tissue have been shown to influence and contribute to its fate [16,17]. Dogs share a common environment with humans, and provide a natural model for several spontaneously occurring diseases that have human counterparts [18]. Humans and dogs have similar nutritive requirements, obesity-induced disorders (such as diabetes), food consumption patterns (small number of large meals), and metabolic characteristics. 62 All of these attributes make the dog as an excellent animal model for obesity research, and we propose that use of the canine model for such studies will provide valuable insights into both basic biology and translational medicine. The process of adipogenesis involves the sequential expression of multiple transcription factors, acquisition of metabolic competence, and synthesis and secretion of fat-specific proteins [2,10,15,19]. Various growth factors, hormones and cytokines as well as the components of cell-cell and cell-matrix interactions are involved in this process [2]. This physiological process is mediated in large part by up-regulation of two major adipogenic transcription factors-PPARy and CEBPa [7, 15,19]. PPARy is a central regulator of fat cell differentiation, and its expression reaches at peak level during terminal differentiation [8,9]. CEBPa is turned on during early stage of adipocytic differentiation prior to the expression of most adipogenic genes [20]. Differentiated adipocytes express lipid-metabolizing genes such as FABP4 (aP2) and lipoprotein lipase (LPL), and adipocytokine-LEPTIN. F ABP4 is involved in fatty acid transport and metabolism in adipocytes, and has been implicated in obesity and obesity-associated disorders such as type II diabetes and atherosclerosis [21]. LPL functions as a triglyceride hydrolase and is involved in receptor-mediated lipoprotein uptake [22]. LEPTIN is primarily secreted by white adipocytes, and acts as a lipostatic factor and contribute to the regulation of body weight [23]. To obtain a better understanding of effects of various induction regimens, we exposed human and canine AD-MSCs to various adipogenic media with different chemical supplements, and then examined the accumulation of intracellular lipid and marker genes associated with early and late stages of adipogenesis. 63 MATERIALS AND METHODS Materials: All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise stated. Isolation of canine adipose tissue derived mesenchymal stem cells (cAD-MSCs): The MSCs previously isolated from human and canine adipose tissues by our research group [24,25] were used in this study. Briefly, stromal-vascular fiaction was extracted after digestion of minced subcutaneous and omental adipose tissues (dog) or lipoaspirates (human) with collagenase type IA (1 mg/ml), and plastic adherent cells were expanded in a modified MCDB 153 medium (Keratinocyte—SFM) supplemented with N- acetyl-L-cysteine (NAC) (2 mM) and L-ascorbic acid 2-phosphate (Asc 2P) (0.2 mM) (referred to as K-NAC medium), and 5% FBS . Verification of stemness phenotypes and multilineage differentiation potential: Before using in this study, culture expanded AD-MSCs were validated for their expression of pltuipotency-associated transcription factors (0C T 4, NANOG, and SOX2), ability to form anchorage independent colonies in semisolid media, and potential to differentiate into mesenchymal lineages (osteoblasts, chondrocytes, and adipocytes) as described earlier [24,25]. Early passage cells (passage 3 to 5) were used throughout the course of this study. Differentiation studies: Standard Induction Regimen: For cAD-MSCs, cells were plated at the seeding density of 10,000 cells/cm2 in 6-welled plates, and treated with rosiglitazone (SpM) (BioVision, Mountain View, CA), dexamethasone(1pM), and insulin (5 pg/ml) OtDI cocktail) 64 supplemented with 5% rabbit serum in high glucose DMEM (4.5 mg glucose per ml of medium) for 15 days with medium changed every 3 days. For hAD-MSCs, cells were plated at the seeding density of 10,000 cells/cm2 in 6 well plates, and treated with IBMX (500 pM), dexamethasone (1 pM), indomethacin (100 pM), and insulin (10 pg/ml) (IDII cocktail) for 15 days with medium change once in every 3 days. Experimental induction Groups: Cells were plated at the seeding density of 10,000 cells/cm2 in 6-welled plates with high glucose DMEM medium with 5% F BS and induced for 15 days as following: i. Control group: no supplements (only DMEM with 5% FBS) ii. RD] group: RD] cocktail iii. RD] + L group: RDI cocktail supplemented with linoleic acid (30 pM) iv. RD] cocktail supplemented with 5% rabbit serum (and FBS removed) v. L group: Linoleic acid (30 pM) only vi. IDII group: IDI] cocktail. The concentration of linoleic acid (30 pM) used in this study was determined after preliminary dose-response and toxicity studies (data not shown). Morphological differentiation was judged by examination of randomly selected microsc0pic fields of treated plates (percentage of treated cells that underwent differentiation as judged by the size and number of fat droplets-containing cells). Accumulation of cytoplasmic lipid was confirmed by Oil Red 0 staining. Oil Red 0 staining: Cells were stained with Oil Red 0 as described previously [26]. Briefly, cells were fixed in 4% paraformaldehyde for 30 min, washed with 60% isopropanol, and stained with Oi] Red. 0 solution (0.21% dye in 60% isopropanol) for 30 65 l min. Cells were then destained with 60% isopropanol followed by repeated washing under running water. The triglyceride accumulation was assessed microscopically with an inverted microscope. For quantification of the triglyceride content, incorporated dye was extracted with 100% isopropanol and optical density was measured spectrophotometrically at 500 nm. RT-PCR: RNA extraction and cDNA synthesis: Total RNA was extracted from undifferentiated MSCs and adipo-induced MSCs (15 days post-treatment) using Versagene RNA Purification Kit (Gentra), and treated with DNase I (Ambion, Austin, TX) to remove any contaminating genomic DNA. RNA from canine adipose tissue was extracted in-house, and human adipose tissue RNA was purchased from a commercial source (Ambion, Austin, TX). One pg total RNA was reverse transcribed in 30 pl reaction volume, using 10 units of Superscript 1]] reverse transcriptase (Invitrogen, Carlsbad,CA), and final concentrations of 5 pM anchored oligo-dT primers, 0.5 mM dNTPs, 50 mM Tris-HCl, 75 mM KC], 3 mM MgC12, and 5 mM DTT, and 1x first strand buffer supplied by the manufacturer (Invitrogen, Carlsbad,CA). The reverse transcription reaction was performed at 50°C for 1 hr, and then inactivated at 70°C for 15 min. The cDNAs were purified by Qiaquick PCR purification kit (Qiagen) before using them for real time PCR. Validation of primers: All primers were designed using Primer 3 software [27]. Primers derived from coding regions of respective genes in canine and human genome were used to amplify the target sites (Table 3.1). All of the primers were fust validated with cDNAs synthesized from canine and human adipose tissue RNAs. Presence of single arnplicon of expected product size was verified by agarose gel electrophoresis following conventional 66 PCR. Before using AACT method, reaction efficiencies were determined by standard curves generated from total RNAs. A ten-fold serial dilution of 100 ng total RNA at five different points confirmed that both the slope and R2 values were close to the theoretical values (slope: -3.8 to -3.3; R220.99). A validation experiment was done to demonstrate the similarities in efficiency of target amplification and reference amplification (slope difference <0.1). Dissociation curve analysis was performed to confirm the specificity of each product and absence of primer-dimer complexes. Real time PCR: Cells from one of the dog AD-MSCs isolated from subcutaneous adipose tissue (designated as BR7-I) and one of the human AD-MSCs isolated from lipoaspirate (designated as PLA-3) were used for quantitative RT-PCR studies. Quantitative RT-PCR was carried out using AB] 7700 sequence detection system (Applied Biosystems, Foster city, CA) on the thermal profile of 50°C for 2 min, 95°C for 10 min, 40 cycles at 95°C for 15 sec, 60°C for 1 min. Each 15 pl reaction contained 1.5 pl diluted cDNA (10 ng starting total RNA) and 200 nM primers mixed with 2X SYBR Green master mix (Applied Biosystems, Foster city, CA). All reactions were performed in triplicate, and reaction mixture with RNA instead of cDNA (no reverse transcriptase control) was used in each run to ensure the absence of genomic DNA contamination. Expression of each target gene was normalized against ,62-MICROGLOBULIN (endogenous reference gene), and relative quantification was carried out between induced-MSCs and undifferentiated MSCs by 2M" method [28,29]. Statistical analysis: Data were presented as mean i s.d. for three individual determinations of the same cell line. Untreated control values were set as 1. The Kolmogorov Smimov’s test 67 was used to verify the normal distribution of the data. For real time PCR data, one way analysis of variance (ANOVA) followed by Tukey’s post hoc test for multiple comparisons was carried out using SigmaStat version 2.03 (SPSS Inc., Chicago, IL). For quantification of triglyceride content, one way AN OVA was performed followed by Dunnett’s method of post hoc comparison. A value of p <0.05 was considered statistically significant. 68 RESULTS The induction regimen (IDII) that was effective for adipogenic differentiation of hMSCs, seemed to only have a week adipogenic effect on CAD-MSCs. The use of rabbit serum (in combination with rosiglitazone considerably increased the adipogenic differentiation (in terms of both numbers and size of fat globules). Effect of different induction regimens on adipogenic differentiation of AD-MSCs: As described earlier [24], the IDII induction regimen was very effective in adipogenic differentiation of hAD-MSCs (Figure 3.1A2). But, adipogenic differentiation of CAD-MSCs seems to be limited with this regimen (Figure 3.1B2). The use of another induction cocktail (RDI) considerably improved the adipogenic differentiation of cAD- MSCs (Figure 3.183) and was effective in differentiation of hAD-MSCs as well (Figure 3. 1 A3). When rabbit serum (5%) was included in the RD] induction cocktail, differentiation of CAD-MSCs (Figure 3.1B4), but not hAD-MSCs (Figure 3.1A4), was further improved. Fat globules were noted within 5-6 days of treatment, which continued to increase in size, and stained positive with Oil Red 0. Linoleic acid alone was able to induce only tiny droplets which did not increase in size throughout the course of differentiation in both canine (Figure 3.1B5) and human (Figure 3.1A5) AD-MSCs, unless it was supplemented with RD] cocktail (RDI+L) (human = Figure 3.]A6; dog = Figure 3.1B6). Oil Red 0 content retained by adipo-induced cells from each experimental group was extracted and quantified. As shown in Figure 3.2, insignificant amount of dye was retained by the cells from linoleic acid- and IDH- treated group for CAD-MSCs, and linoleic acid treated group for hAD-MSCs (p>0.05 when compared to untreated control). 69 Cells from all other treatment groups had retained significantly higher amount of dye when compared to untreated control (p<0.05). Gene expression analysis following adipogenic induction: We evaluated the transcriptional phenotype of adipo-induced cells by assessing the expression of a panel of adipocyte-specific genes under different induction conditions by quantitative RT-PCR (Figure 3.3 and 3.4). Following IDII induction of CAD-MSCs, the induced cells barely increased the expression of PPARyZ, LPL and LEPTIN (< 1.5 fold) and modestly increased the expression of CEBPa and FABP4 when compared to untreated control. On the other hand, IDII induction of hAD-MSCs led to the considerable up-regulation of all of these genes. Expression of all of these adipogenic markers were highly increased after induction of both canine and human AD-MSCs with RD] cocktail. When rabbit serum was included in the RD] cocktail, there was a remarkable up-regulation of all adipogenic markers in both canine and human AD-MSCs. Linoleic acid alone induced only modest increase in expression of these genes in both CAD-MSCs and hAD-MSCs. When linoleic acid was combined with RD] cocktail for induction of canine and human AD-MSCs, expression of these genes were comparable with RD] alone treatment group in both species. All these changes in expression of marker genes were congruent with phenotypic responses. Out of the all adipogenic markers examined, LPL expression consistently showed the highest fold change, from about 40 to more than 160 fold after induction, in all canine treatment groups that displayed grossly adipocytic morphology (RDI, RDI+RS, RDI+L). In contrast, FABP4 expression showed the highest fold change, (from about 40 to 100 70 fold after induction, in human cells that effectively underwent adipogenic differentiation (IDII, RDI, RDI+RS, RDI+L). 7] DISCUSSION Adipogenesis of human MSCs is commonly induced with IBMX (phosphodiesterase inhibitor that increases intracellular CAMP), dexamethasone (glucocorticoid agonist), insulin (stimulator of IGF-I receptor), and indomethacin (PPARy activator) (IDII treatment) [2]. We had found earlier [25] that effective adipogenic differentiation of cAD-MSCs required substantial optimization of the standard induction regimen used for human and rodent MSCs, and specifically, this involved the use of rabbit serum in conjunction with rosiglitazone, dexamethasone, insulin, and high glucose supplementation [25]. There are variations among well- characterized cell lines in terms of induction requirements. For instance, adipogenesis of rodent cell lines 3T3-L1 and 3T3-F 442A requires supplementation of dexamethasone, insulin, and cyclic AMP whereas another rodent cell line Ob] 7 requires incubation with fatty acids [30]. Similarly, a preadipocyte cell line which was refractory to rosiglitazone induction, undergoes adipogenesis in response to IBMX, indomethacin, and hydrocortisone [3 1]. A previous microSAGE analysis had revealed that single cell-derived human MSCs simultaneously expressed mRNAs of various mesenchymal lineages [32]. Consistent with this, we found that undifferentiated M SCs express basal levels of adipogenic markers (detectable by quantitative RT-PCR), and these genes were variably up-regulated after induction with different adipogenic signals. Changes in expression level of these adipogenic markers were congruent with phenotypic responses after induction. 72 Effect of IDII treatment was dramatically different between canine and human AD-MSCs. The phenotypic differentiation was not evident in induced CAD-MSCs unlike in induced hAD-MSCs, and accordingly, expression of adipogenic genes were barely upregulated after induction. Indomethacin is one of the critical components of IDII mixture. Although it acts as a PPARy agonist at higher concentrations, it also blocks cyclooxygenase activity and inhibits the formation of adipogenic prostaglandins- including PPAR 7 activator 15-deoxy-A'2’14-PGJZ [33]. In one of the earlier reports, indomethacin was even found to block the terminal differentiation of Ob1771 murine preadipocyte cell line [34]. On the other hand, RD] treated cells displayed upregulation of all adipogenic markers (transcription factors PPARyZ and CEBPa as well as late adipogenic genes FABP4, LPL and LEPTIN) in both canine and human cells. Rosiglitazone is a highly selective ligand that binds to PPAR'y with nanomolar affinities [33]. This compound was found to stimulate adipogenesis in human MSCs [35], and our study corroborated this finding in the context of both canine and human MSCs. Similarly, in agreement with its reported adipogenic activity [3], rabbit serum was found to have dramatic effect on adipogenic induction of both canine and human AD-MSCs. Genes of adipocyte lineage were expressed even higher when RD] treated cells were supplemented with 5% rabbit serum. Diascro et a] [36] had reported that rabbit serum is highly enriched in fiee fatty acids compared to FBS, and these fatty acids were able to induce osteoblasts into adipocytes. They reported that linoleic acid was one of the most abundant free fatty acids in rabbit serum, and its ability to activate PPARy- mediated transcription was comparable with rabbit serum and rosiglitazone [36]. Earlier studies had implicated the role of fatty 73 acids in the differentiation of preadipose cells to adipose cells [37,3 8], although they did not affect the commitment of adipoblasts to preadipose cells [37]. In the present study, we compared the effect of linoleic acid on adipogenic induction of undifferentiated MSCs isolated from canine and human adipose tissues. Linoleic acid alone was unable to induce mature adipocyte-like phenotype, although tiny droplets were noticed inside the treated cells. Cells treated with only linoleic acid showed modest up-regulation of adipogenic markers in both canine and human cells, consistent with the lack of maturation in adipocytes. It was found earlier that, unlike rosigltizaone, linoleic acid was relatively ineffective to induce biochemical and functional differentiation of human preadipocytes [39]. This endogenous ligand of PPARy is a weaker agonist with rrricromolar affinity [9], and additional factors seem to be required for complete adipogenic differentiation of MSCs. As such, transcriptional activity of PPARy is mediated by complex interaction with co-activators and co-repressor proteins [30]. As expected, when linoleic acid was supplemented with RDI mixture, we found increased expression of all adipogenic markers, including the late-stage markers. In summary, we established a model experimental system for in vitro adipogenesis of MSCs. There are differences between human and canine AD-MSCs, both in terms of stimuli and pattern of gene expression following induction. These might be the result of epi genetic context or differences in the state of cellular permissiveness. The molecular and cellular events underlying these differences warrant more detailed study. Studies of differentiation of MSCs into adipocyte will provide an opportunity for better understanding of diseases of energy metabolisms such as obesity and obesity-related disorders. 74 APPENDIX Table 3.1. Primers used in quantitative RT-PCR of adipogenic markers Amp Ann licon Tem Markers Gene Primer sequence (5' - 3') Size p(°C) PPAR Forward ACACGATGCTGGCGTCCTTGATG y2 119 60 Reverse TGGCTCCATGAAGTCACCAAAGG CEBP Forward AGTCAAGAAGTCGGTGGACAAG 603 a 151 Reverse GCGGTCATTGTCACTGGTGAG FABP Forward ATCAGTGTAAACGGGGATGTG 4 117 60 Adipocyte Reverse GACT'I'I‘TCTGTCATCCGCAGTA S LPL Forward ACACATTCACAAGAGGGTCACC 134 60 (Canine) Reverse CTCTGCAATCACACGGATGGC LEPT Forward CTATCTGTCCTGTGTTGAAGCTG IN Reverse GTGTGTGAAATGTCAT'I’GATCCTG 102 60 Forward TCTACATTGGGCACTGTGTCAC Housekee ping BZM (Canine) Reverse TGAAGAGTTCAGGTCTGACCAAG 136 60 PPAR Forward ACACAATGCTGGCCTCCTTGATG y2 Reverse GCTCCATAAAGTCACCAAAAGG 117 60 CEBP Forward TGAGGATGTATACCCCTGGTG a Reverse TAGGCAATGCTGAAGGCATAC 123 60 FABP Forward TGCAGATGACAGGAAAGTCAAG Adipocyte 4 5 Reverse CCACCACCAGTTTATCATCCTC 105 60 (Human) LPL Forward CGACTCGTCTTTCTCCTGATG Reverse ACCTCCATTCGGGTAAATGTC 128 60 LEPT Forward AAGAAGAGACAGGAGGGCAAG IN Reverse TCTGGGTCACATTTTCCAGAG 147 60 Forward AAT‘TCCAAATTCTGCTTGCTI‘G Housekee ping BZM (Human) Reverse ACATCAAACATGGAGACAGCAC 146 60 a Reaction contained 10% DMSO 75 Figure 3.]. Induction of adipogenic differentiation of canine and human AD-MSCs under different treatment conditions. Third passage AD-MSCs were induced to differentiate with various induction cocktails. Undifferentiated human (A1) and canine (B1)AD-MSCs are shown retained typical MSC-like morphologies. Numerous fat droplets were noticed within one week of induction of hAD-MSCs by IDII regimen (A2), whereas only few cells with fat droplets were noted after induction of cAD-MSCs under similar condition (B2). Use of RD] cocktail increased both number and sizes of fat droplets after induction of cAD-MSCs (B3), and was equally effective in adipogenic differentiation of hAD-MSCs (A3). Supplementation of RD] cocktail with rabbit serum further augmented the number and size of fat droplets in both canine (B4) and human (A4) cells. Linoleic acid was unable to induce distinct droplets in both canine (BS) and human (A5) AD-MSCs whereas addition of RD] was very effective in both cases (A6, B6). Magnification = 200x. 76 F'AI'IFI {I I ’1 ' ,4341’1’; ‘ 6"?” ;‘4 I Figure 3.1. 77 1.2 - A l ' E S 0.8 - '0 0.6 - E 8 0.4 - lDog 0'2 1 IHuman 0 \ ’\ S (P s9 s9 Treatment groups Figure 3.2. Quantification of intracellular triglyceride content by Oil Red 0 staining. After induction with different supplementations, cells were stained with Oil Red 0. Accumulated intro-cytoplasmic dye was extracted with isopropanol and the dye-content was measured spectrophotometrically at 500 nm. Data were represented as mean i s.d. of triplicates. 78 Canine AD—MSCs 180 - - PPARr - CEBPa E FABP4 0160 - - LPL g — LEPTIN '5 / 2 / ‘3 40 - 20 - 0 . RD] L RDI+L RD1+RS IDII Treatment groups Figure 3.3. Expression of adipogenic markers after induction of canine AD-MSCs. mRNA expression of adipogenic markers PPARy2, CEBPa, FABP4, LPL, and LEPTIN were measured by qRT-PC R, following adipogenic induction of canine AD-MSCs with different induction cocktails. mRNA level of each gene was normalized to that of 192- MICROGLOBULIN, and expressed as fold change relative to the respective control group. 79 Human AD-MSCs 125 100 IPPARy Eh .5 DFABP4 % 50 ILPL m 25 ILeplin 0 RD] L RDI+L RD1+RS IDII Treatment groups Figure 3.4. Expression of adipogenic markers after induction of human AD-MSCs. mRNA expression of adipogenic markers PPARyZ, CEBPa, FABP4, LPL, and LEPTIN were measured by qRT—PCR, following adipogenic induction of human AD-MSCs with different induction cocktails. mRNA level of each gene was normalized to that of [92- MICROGLOBULIN, and expressed as fold change relative to the respective control group. 80 10. 11. REFERENCES . German AJ. The growing problem of obesity rn dogs and cats. J Nutr 2006;136:19408-19468. Gregoire FM, Smas CM, Sul HS. Understanding adipocyte differentiation. Physiol Rev 1998;78:783-809. . Janderova L, McNeil M, Murrell AN et a]. Human mesenchymal stem cells as an in vitro model for human adipogenesis. Obes Res 2003;11:65-74. Camp HS, Ren D, Leff T. Adipogenesis and fat-cell function in obesity and diabetes. Trends Mol Med 2002;8:442-447. Nuttall ME, Gimble JM. Controlling the balance between osteoblastogenesis and adipogenesis and the consequent therapeutic implications. Curr Opin Pharmacol 2004;4:290-294. Gimble JM, Zvonic S, Floyd ZE et al. Playing with bone and fat. J Cell Biochem 2006;98:251-266. Rosen ED, Walkey CJ, Puigserver P et a]. Transcriptional regulation of adipogenesis. Genes Dev 2000;14: 1293-1307. Rosen ED, Spiegelman BM. Molecular regulation of adipogenesis. Annu Rev Cell Dev Bi012000;16:145-171. Lowell BB. PPARgamma: an essential regulator of adipogenesis and modulator of fat cell function. Cell 1999;99:239-242. Cowherd RM, Lyle RE, McGehee RE, Jr. Molecular regulation of adipocyte differentiation. Semin Cell Dev Biol 1999;10:3-10. Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143-147. 8] 12. 13. 14. 15. 16. 17. 18. I9. 20. 21. 22. 23. Zuk PA, Zhu M, Mizuno H et al. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 2001;71211-228. Schaffler A, Buchler C. Concise review: adipose tissue-derived stromal cells-~basic and clinical implications for novel cell-based therapies. Stem Cells 2007;25:818- 827. Rodeheffer MS, Birsoy K, Friedman JM. Identification of white adipocyte progenitor cells in vivo. Cell 2008;135:240-249. MacDougald OA, Mandrup S. Adipogenesis: forces that tip the scales. Trends Endocrinol Metab 2002; 1 3:5-1 1. Crossno JT, Jr., Majka SM, Grazia T et a]. Rosiglitazone promotes development of a novel adipocyte population from bone marrow-derived circulating progenitor cells. J Clin Invest 2006;116:3220-3228. Hausman GJ, Hausman DB. Search for the preadipocyte progenitor cell. J Clin Invest 2006;] 16:3103-3106. Neff MW, Rine J. A fetching model organism. Cell 2006;124:229-231. Brun RP, Kim JB, Hu E et al. Adipocyte differentiation: a transcriptional regulatory cascade. Curr Opin Cell Biol 1996;8z826-832. Mandrup S, Lane MD. Regulating adipogenesis. J Biol Chem 1997;272:5367-5370. Maeda K, Cao H, Kono K et al. Adipocyte/macrophage fatty acid binding proteins control integrated metabolic responses in obesity and diabetes. Cell Metab 2005;]:107-119. Fried SK, Russell CD, Grauso NL et a]. Lipoprotein lipase regulation by insulin and glucocorticoid in subcutaneous and omental adipose tissues of obese women and men. J Clin Invest 1993;92:2191-2198. Friedman JM, Halaas IL. Leptin and the regulation of body weight in mammals. Nature 1998;395:763-770. 82 24. 25. 26. 27. _ 28. 29. 30. 31. 32. 33. 34. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Neupane M, Chang CC, Kiupel M et al. Isolation and characterization of canine adipose-derived mesenchymal stem cells. Tissue Eng Part A 2008;14: 1007-1015. Ramirez-Zacarias J L, Castro-Munozledo F, Kuri-Harcuch W. Quantitation of adipose conversion and triglycerides by staining intracytoplasmic lipids with Oil red O. Histochemistry 1992;97:493-497. Rozen S, Skaletsky HJ. Primer3 on the W for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics Methods and Protocols: Methods in Molecular Biology. Totowa, NJ: Humana Press 2000:365-86. Livak KJ, Schrrrittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402-408. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 2008;3z1101-1108. Grimaldi PA. The roles of PPARs in adipocyte differentiation. Prog Lipid Res 2001;40:269-281. Torii I, Morikawa S, Nakano A et al. Establishment of a human preadipose cell line, HPB-AML-I: refractory to PPARgamma-mediated adipogenic stimulation. J Cell Physiol 2003;197:42-52. Tremain N, Korkko J, Ibberson D et a]. MicroSAGE analysis of 2,353 expressed genes in a single cell-derived colony of undifferentiated human mesenchymal stem cells reveals mRNAs of multiple cell lineages. Stem Cells 2001;19:408-418. Lehmann JM, Lenhard JM, Oliver BB et al. Peroxisome proliferator-activated receptors alpha and gamma are activated by indomethacin and other non-steroidal anti-inflammatory drugs. J Biol Chem 1997;272:3406-3410. Negrel R, Gaillard D, Ailhaud G. Prostacyclin as a potent effector of adipose-cell differentiation. Biochem J 1989;257:399-405. 83 35. 36. 37. 38. 39. Benvenuti S, Cellai I, Luciani P et a1. Rosiglitazone stimulates adipogenesis and decreases osteoblastogenesis in human mesenchymal stem cells. J Endocrinol Invest 2007;302RC26-RC30. Diascro DD, Jr., Vogel RL, Johnson TE et al. High fatty acid content in rabbit serum is responsible for the differentiation of osteoblasts into adipocyte-like cells. J Bone Miner Res 1998;13:96-106. Amri EZ, Ailhaud G, Grimaldi PA. Fatty acids as signal transducing molecules: involvement in the differentiation of preadipose to adipose cells. J Lipid Res 1994;35:930-937. Ibrahimi A, Teboul L, Gaillard D et al. Evidence for a common mechanism of action for fatty acids and thiazolidinedione antidiabetic agents on gene expression in preadipose cells. Mol Pharmacol 1994;46: 1070-1076. Hutley LJ, Newell FM, Joyner JM et al. Effects of rosiglitazone and linoleic acid on human preadipocyte differentiation. Eur J Clin Invest 2003;33:574-581. 84 CHAPTER 4 Trans-differentiation of canine adipose-derived mesenchymal stem cells into cells with neural phenotype 85 SUMMARY The potential of mesenchymal stem cells (MSCs) to differentiate into mesodermal cells is well established. There is growing evidence that MSCs also may trans- differentiate into cell types of other germ layer lineages. Further investigation of such phenotypic plasticity will provide important insights into their suitability for cell-based regenerative and reparative medicine. The purpose of this study was to evaluate the expression of neural markers after treatment of human and canine adipose-derived mesenchymal stem cells (AD-MSCs) with neural induction regimen. AD-MSCs were isolated and expanded in culture in a low calcium medium supplemented with antioxidants. Differentiation was induced in serum free media with different induction agents. Panels of neural markers were used to evaluate the change in mRNA and protein expression by conventional and quantitative reverse transcription (RT)-PCR, and immunocytochemistry. RT-PCR data revealed that induced cells strongly expressed mRNAs for neural genes. In addition, immunocytochemistry showed that only induced cells with neural morphology, but not the undifferentiated MSCs, expressed neural markers. Thus, these findings reveal that MSCs can be induced to become neural-like cells under suitable conditions, which not only assume neural morphologies but also express neural markers at RNA and protein levels. This study demonstrates that MSCs from an abundant and accessible source have the potential to transdifferentiate into neural cells. While further functional studies are warranted, such cells hold great potential for treatment of neurodegenerative diseases. Evaluation of these cells in the canine system and their use in interventional studies in the canine will facilitate translational studies in the human. 86 INTRODUCTION Developmental potential of resident adult stem cells had been thought to be restricted to the cell lineages of the tissue where they resided or, at the best, to the germ layer lineages of the parental cell. However, a growing body of evidence suggests that differentiation potential of mesenchymal stem cells (MSCs) [1,2] is not restricted to mesodermal lineages [3]. They have been shown to transdifferentiate into developmentally unrelated cell types such as epithelial cells [4], hepatocytes [5,6], and neural cells [7]. MSCs can be induced to express the properties of neural cells in vitro [7-11], and they have been shown to facilitate functional recovery in vivo after transplantation into brain and spinal cord [12-16]. Since neural tissues have limited self-renewal or repair potential [17-19], stem cell-based therapies represent a promising approach for the treatment of neurological disorders. MSCs can be isolated from clinically accessible sources and adequately expanded for autologous transplantation [2,20]. They secrete growth factors and cytokines, which have angiogenic, antifibrotic, and mitotic effects, and are capable of intrinsic tissue repair [21]. Moreover, their immunoprivileged and immunomodulatory properties [22,23] make them uniquely suitable for allogenic transplantation and off-the-shelf therapies. Thus, MSCs are promising for cell-based therapy of neurodegenerative diseases and traumatic injuries, and may circumvent immunologic concerns and ethical issues in their applications. Adipose tissue is emerging as a promising source of MSCs because of its abundance as well as case of access with minimal donor site morbidity [24-26]. We propose that the study of therapeutic potential in canine model will facilitate the 87 translation of stem cell based therapy in clinical practice and will be of benefit to both human and canine species. The dog is an excellent biomedical model [27], and the level of sophistication of canine veterinary medicine will allow us to assess the promise of stem cells for novel therapeutic approaches [28]. Dogs develop many of the same neurological diseases as that of human, and there are canine equivalents of diseases such as Alzheimer’s disease, parkinsonism, epilepsy, stroke, spinal cord injury, muscular dystrophy, and neurometabolic disorders (reviewed in [29]). Therefore, establishment of canine-based experimental system to assess the neural differentiation potential of clinically accessible canine cells will allow evaluating the usefulness of this translational model for cell-based therapies. 88 MATERIALS AND METHODS Materials: All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise stated. Isolation of cAD-MSCs: The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee. The MSCs previously isolated from human and canine adipose tissues by our research group [30,31] were used to investigate their neural differentiation potential. Briefly, stromal-vascular fraction was extracted after digestion of minced subcutaneous adipose tissues (dog) or lipoaspirates (human) with collagenase type IA (1 mg/ml), and plastic adherent cells were expanded in a modified MCDB 153 medium (Keratinocyte-SFM) supplemented with N-acetyl—L-cysteine (NAC) (2 mM) and L- ascorbic acid 2-phosphate (Asc 2P) (0.2 mM) (referred to as K-NAC medium), and 5% FBS . Verification of stemness phenotypes and multilineage differentiation potential: Before using in this study, culture expanded AD-MSCs were validated for their expression of pluripotency-associated transcription factors (0C 7’ 4, NANOG, and SOX2), ability to form anchorage independent colonies in semisolid media, and potential to differentiate into mesenchymal lineages (osteoblasts, chondrocytes, and adipocytes) as described earlier [30,31]. Induction of neural differentiation: Three different treatment regimens were tested in three different cell culture media for their neural induction potential. Early passage cells (passage 3 to 5) were used 89 for all experiments. Morphological changes were examined throughout the course of induction by phase contrast microscopy, followed by RT-PCR, and immunocytochemistry for examination of neural markers. Induction media: The following three induction media were used for neural differentiation studies: (a) D medium: a modified Eagle’s MEM [32] (b) MSU-1 medium: a 1:1 mixture of D medium and MCDB 153 supplemented with growth factors and hormones [33] (c) Neurobasal medium with 1% N2 supplement (NB-N2) medium: a commercial medium obtained from Invitrogen corporation (Carlsbad,CA). Protocol 1. Cells were plated at the density of 2000 cells/cm2 in 35 mm plates in K-NAC medium with 5% F BS. After keeping the cells in K-NAC medium with 5% FBS for 1 day, they were switched to and maintained in induction medium under serum fi'ee condition for 2 days. Then, pre-induction was done by treating the MSCs with B- mercaptoethanol (lmM) for 1 day in induction medium. Differentiation was induced by treating the pre-induced cells with nicotinamide (10 mM), forskolin (5 pM), IBMX (200 pM), and retinoic acid (0.5 pM) (NF IR cocktail) for 5 days. Protocol 2. Cells were plated at the density of 2000 cells/cm2 in 35 mm plates in K-NAC medium with 5% F BS. On the next day, they were switched to induction medium (without serum), and treated with butylated hydroxyanisole (BHA) (200 pm), potassium chloride (5 mM), valproic acid (2 mM), forskolin (10 pM), hydrocortisone (1 pM), and insulin (5 pg/ml) for 7 days. This protocol is a modification of previously published protocol [9]. 90 Protocol 3. Cells were plated at the density of 2000 cells/cm2 in 35 mm plates in K-NAC medium with 5% FBS. On the next day, the cells were treated with basic fibroblast growth factor (bFGF) (10 pg/ml) and murine nerve growth factor (NGF) (10 ng/ml) in induction medium for 10 days. RT-PCR: RNA extraction and cDNA synthesis: Total RNA was extracted from undifferentiated MSCs and neural-induced MSCs (5 days post-induction) using Versagene RNA Purification Kit (Gentra), and treated with DNase I (Ambion, Austin, TX) to remove any contaminating genomic DNA. RNA from canine brain was extracted in-house, and human brain tissue RNA was purchased from a commercial source (Ambion, Austin, TX). One pg total RNA was reverse transcribed in 30 pl reaction volume, using 10 units of Superscript 1]] reverse transcriptase (Invitrogen, Carlsbad,CA), and final concentrations of 5 pM anchored oligo-dT primers, 0.5 mM dNTPs, 5 mM DTT, and 1x first strand buffer supplied by the manufacturer (Invitrogen, Carlsbad,CA). The reverse transcription reaction was performed at 50°C for 1 hr, and then inactivated at 70°C for 15 min. The cDNAs were purified by Qiaquick PCR purification kit (Qiagen) before using them for real time PCR. Validation of primers: All primers were designed using Primer 3 software [34]. Primers derived fi'om coding regions of respective genes in canine and human genome were used to amplify the target sites (Tables 4.1 and 4.2). All of the primers were first validated with cDNAs synthesized from canine and human brain tissue RNAs. Presence of single amplicon of expected product size was verified by agarose gel electrophoresis following conventional PCR. Before using AACT method, reaction efficiencies were determined by 91 standard curves generated from total RNAs. A ten-fold serial dilution of 100 ng total RNA at five different points confirmed that both the slope and R2 values were close to the theoretical values (slope= -3.8 to -3.3; R220.99). A validation experiment was done to demonstrate the similarities in efficiency of target amplification and reference amplification (slope difference <0.1). Dissociation curve analysis was performed to confirm the specificity of each product and absence of primer-dimer complexes. Real time PCR: Cells from two of the dog AD-MSCs isolated from subcutaneous adipose tissue (designated as BR7-I and BR20-A) and one of the human AD-MSCs isolated from lipoaspirate (designated as FLA-3) were used for quantitative RT-PCR studies. Quantitative RT-PCR was carried out with 2X SYBR Green master mix (Applied Biosystems, Foster city, CA) and 100 nM primers in 15 p1 reaction using AB] 7700 sequence detection system (Applied Biosystems, Foster city, CA) on the thermal profile of 50°C for 2 min, 95°C for 10 min, 40 cycles at 95°C for 15 sec, 60°C for 1 min. All reactions were performed in triplicate, and reaction mixture with RNA instead of cDNA (no reverse transcriptase control) was used in each run to ensure the absence of genomic DNA contamination. Expression of each target gene was normalized against [32- MICROGLOBULIN (endogenous reference gene), and relative quantification was carried out between induced-MSCs and undifferentiated MSCs by 2M” method [35,36]. Immunocytochemistry: Cells were fixed in cold acetone (-20°C) for 3 minutes, and washed twice with tris-buffered saline. Non-specific binding of primary antibody was blocked by serum-free protein block (DakoCytomation, CA), and cells were incubated overnight at 4°C with primary antibody. Primary antibodies used were: polyclonal anti-NF 200 (rabbit, 1:200, 92 Sigma) and polyclonal anti-NSE (rabbit, 1:1000, Dako). HRP-labeled polymer (EnVision, K1491; DakoCytomation) was used for immunolabeling, and the immunoreaction was visualized with 3, 3’-diaminobenzidine chromogen (Dako, CA) under a microscope followed by counterstaining with hematoxylin (SurgiPath, IL). 93 RESULTS Induction of neural differentiation: Out of the 3 different neural induction regimens, we found protocol 1 as the most . efficient regimen in terms of differentiation into neural-like phenotypes. As judged by morphological phenotypes, differentiation was found to be very robust with protocol 1, which involved treatment of B-mercaptoethanol pre-induced cells with NFIR cocktail. Among the various media tested, MSU-1 medium without EGF (designated as MSU-1(- E)) was the most optimum for neural induction (see next paragraph). With protocol 2, some multipolar refractile cells with neurite-like extensions were seen but cell viability was low following induction. The differentiated cells could not remain attached to the substratum for long period, and this resulted in high rates of cell death. Supplementation of induction cocktail with murine NGF (10 ng/ml) could not prevent this phenomenon. With protocol 3, cells remained spindle-shaped, and did not assume typical neural morphology even after 10 days of exposure to protocol 3. Based on these observations, we chose protocol 1 and MSU-1(-E) medium as our standard induction regimen for firrther studies. Following 24 hr of induction with NFIR cocktail (preceded by 24 hr of pre- induction, and 2 days of maintenance in serum free, MSU-1(-E) medium), cells started to show the appearance of bipolar to multipolar refractile bodies with neurite-like extensions. They formed extensive branching which were organized to form network with increase in time (Figure 4.1A2, A3, A6, A7). In contrast, the untreated cells kept their fibroblast-like appearance (Figure 4. 1 A4, A8). 94 Expression of neural markers at mRNA level: We found that even undifferentiated MSCs express basal level of neural markers, most of which were up-regulated after treatment with induction cocktail. Quantitative RT-PCR revealed that neuronal markers microtubule-associated protein 2 (MAPZ), B- tubulin3 (B-TUBB3), neuron specific enolase- y (y-NSE), and neurofilament-200 (NF - 200) were considerably up-regulated whereas neural precursor marker nestin (NES) was down-regulated after induction (Figure 4.3). On the other hand, expression of oligodendrocyte marker galactosylceramidase (GALC) was barely changed even after 1 week of induction, whereas astrocyte marker- glial fibrillary acidic protein (GFAP) could not be detected in either before or after induction (Figure 4.3). Although there were variations between human and canine MSCs regarding level of expression following induction, they followed the similar trend of gene expression. Expression of neuronal markers at protein level: To characterize the identity of induced cells with neural morphology, we examined for the expression of neuronal markers in these cells (Figure 4.2). We stained these cells for a neuron-specific intermediate filament, NF -200. Cells displaying neural morphologies were found to express NF-200 (Figure 4.2B1, B5), whereas this marker could not be detected in un-induced cells (Figure 4.2B2, B6). To investigate neuronal characteristics further, we examined these cells for the expression of neuronal marker-y- NSE. Cells exhibiting neural morphologies stained positive for y-NSE (Figure 4.2B3, B7), whereas untreated cells were y-NSE negative (Figure 4.2B4, B8). In contrast to the expression of neuronal proteins, we could not detect the expression of standard astrocytic marker-GFAP (data not shown) in induced cells from both dog and human CAD-MSCs. 95 DISCUSSION Our research group has isolated MSCs from adipose tissues of different mammalian species by modulating the cellular redox state [30,31,37]. In the current study, we have identified optimum induction regimen that can effectively trans- differentiated AD-MSCs into cells with neural-like phenotypes. Our findings indicate that canine and human AD-MSCs can be induced to differentiate into non-mesenchymal cell types, particularly cells with neuronal characteristics. Neurogenesis in adult is limited to selected regions of the brain, and neuronal repair after injury is very limited [17,18]. Cell-based therapies hold a promise for effective treatments of central nervous system disorders, including degenerative diseases as well as traumatic and ischemic injuries. It was reported earlier that transplantation of fetal doparrrinergic neurons into Parkinson’s disease patients showed clinical success [38]. One of the earlier reports on ability of MSCs to differentiate into astrocytes and neurofilament-containing cells after injection into neonatal mouse brain had indicated the plasticity of these cells [39]. We have shown that AD-MSCs can be isolated from easily accessible and abundant source such as fat, and can be readily induced to differentiate into cells with neuron-like phenotypes. Thus, these cells may provide a feasible alternative to less accessible neural stem cells or more controversial embryonic stem cells [40]. In fact, it was recently reported that transplantation of fetal neural stem cells led to the development of brain tumor in a human patient [41]. Such findings strongly suggest the need of using well-characterized cells, and assessing their safety in a more relevant animal model before using them in human. We argue that establishment of dog as a preclinical model for cell-based therapies will allow us to evaluate both safety and 96 efficacy of such therapies, which can then be translated into human setting with decreased risk and increased relevance. We used 3 different treatment regimens for the induction of neural phenotypes from AD-MSCs, based on our preliminary studies and published literature [7,9]. Although AD-MSCs exhibited neural-like phenotype after treatment with BHA- containing cocktail (protocol 2), these cells showed signs of cell death as reported previously [42-44]. Although bFGF could contribute to the neural commitment of undifferentiated stem cells [45,46], treatment of MSCs with growth factors (bFGF and NGF in the presence or absence of EGF) (protocol 3) was not very effective in our hands, as the cells remained fibroblast-like even after 10 days of induction. We did not pursue any further characterization of these cells. We found that treatment with NFIR cocktail (protocol 1) was most effective for both induction and maintenance of neurOnal differentiation. To determine the neural phenotypic characteristics of AD-MSCs after exposure to induction media, we evaluated the gene expression pattern of these cells with a panel of neural markers. Undifferentiated AD-MSCs expressed mRNAs of many of neural markers at low level. This is consistent with earlier reports on expression of neural mRNAs and proteins in undifferentiated MSCs [47-51]. We found that induced cells up- regulated the expression of neuron specific rrrarkers NSE, B TUBB3, NF -200, and MAP2, and down-regulated the expression of neuronal precursor marker NES. Our RT-PCR data showed significant reduction in the expression of NES mRN A after 5 days of induction. NES is an intermediate filament protein, and its expression 97 decreases with neuronal maturation [52]. Woodbury et a] [7] had reported that NES expression was seen in differentiating cells at 5 hours after treatment, but it became undetectable at 6 days. NSE- y is normally found in mature neurons and in cells of neuronal origin [53]. B-III tubulin is a neuronal cytoskeletal dimer [54]. Similarly, NF - 200 is an intermediate filament protein found in mature neurons [55], and microtubule- associated protein 2 (MAP2) is a cytoskeletal protein required for development and maintenance of neurons [56]. Upon immunostaining, we detected the expression of both NF—200 and NSE-y proteins only in differentiated cells, unlike their expression at mRNA level in control cells as well. Like Woodbury et a1 [7] and Deng et a] [57], but unlike Sanchez-Ramos et a1 [8] and Safford et al [9], we could not notice the expression of fibrillary astrocyte marker- GFAP in differentiated cells at either mRNA or protein level. GFAP is an intermediate filament protein that is normally expressed by astrocytes and schwann cells [58]. Furthermore, expression of oligodendrocyte marker-GALC [59] did not change significantly after induction. These findings suggest that induced cells followed neuronal- like rather than glial-like fate after treatment with our regimen. EGF has been reported to induce glial but not neuronal differentiation of neural precursors (discussed in [60]), and deletion of EGF from the MSU-1 induction medium in the present study might have contributed to the neuronal-like features of the induced cells. Although some studies have raised concerns that morphological changes and increase in immunoreactivity for neural markers after chemical induction of MSCs might reflect the artifacts of cellular toxicity and cytoskeletal changes [42,43,61-63], other 98 studies have provided more robust validation of neural differentiation potential of MSCs [64-66]. However, molecular mechanisms implied in this differentiation process have remained largely unknown. Treatment with forskolin (adenylate cyclase activator) and IBMX (phosphodiesterase inhibitor) increases the level of intracellular second messenger cAMP. cAMP increases Ca2+ influx and excitability, and has been shown to activate protein kinases such as canonical protein kinase A (PKA) pathway [42,46]. These changes could have been responsible for neural-like changes in phenotypes and protein expression [42,46]. In fact, elevated intracellular cAMP level was able to induce immature neuronal phenotype in human MSCs [57], and cAMP up-regulation has been implicated in neuroendocrine differentiation of human prostate cancer cells [67,68]. A recent study has provided evidence for the role of insulin growth factor (IGF)-1 signaling in IBMX-induced neural differentiation of MSCs [69]. On the other hand, Spinella et a] [70] had shown that retinoic acid was able to differentiate embryonic carcinoma cells into neurons through regulation of retinoic acid-responsive genes. Similarly, retinoic acid in combination with Sonic hedge hog induced the expression of sensory neuron markers in MSCs [71], and a subset of MSCs underwent neural differentiation after treatment with retinoic acid containing-cocktail [8]. Further investigations are required to elucidate the molecular events mediated by the combination of chemicals that regulates the expression of neural markers and the emergence of neural-like phenotypes. In summary, we have been able to develop the suitable in vitro conditions by which canine MSCs can be trans-differentiated into cells with neural phenotypes, as shown by the expression of neural markers at mRN A and protein level. The results of our 99 in vitro studies are very promising. The ability of expanded MSCs culture isolated from canine adipose tissue to trans-differentiate into neural phenotypes offers the feasibility of harnessing MSCs for neurologic diseases and damages. This study establishes a large animal model system for functional studies and therapeutic intervention. However, the rigorous demonstration of trans-differentiation ability awaits further validation by functional characterizations and studies on engraftment, cell survival, and in vivo differentiation in animal models. 100 APPENDIX Table 4.1. Canine primers used in quantitative RT-PCR of neural markers Amp licon Ann Size Temp Markers Gene Primer sequence (5' - 3') (bp) (°C) Forward CTCAGCCCAGGAGGAGATAAC NF -200 Reverse AGCTGCCTCTCTAGTGAGTCC 102 60 B- Forward CATGAACACCTTCAGTGTGG T UBB3 Reverse TGTCGTACAAAGCCTCATTGTC I40 62 Forward CATCCGCCTGAGATTAAGGATC Neurons AMP-2 Reverse TGCTCTGTGACCCATGTTCTC 119 58 Forward CATCGTGATGGCAAATATGAC NSE Reverse TC C CTGAC AAAGTCCTGGTAG 104 62 Forward ACTCAGTGGGTCTGGAATGG NESTYN Reverse TC CTGCTGCAAACTGTTCAC 117 62 Astrocytes Forward AGGGACAGAACCTCAAAGATG GFAP Reverse TC C AGCAGTTTCCTGTAGGTG 112 64 Oligodendr Forward AC C AC TCAGTT'TACCCAACCAG ocytes GALC Reverse TCAACGATGATGGTGAGGTTAC 116 62 Housekeepi Forward TCTACATTGGGCACTGTGTCAC 136 60 ng B2M Reverse TGAAGAGTTCAGGTCTGACCAAG a Reaction contained 5% DMSO 10] Table 4.2. Human primers used in quantitative RT-PCR of neural markers Ampl icon Ann Size Tem Markers Gene Primer sequence (5' - 3') (bp) pf C Forward ACCTGCTCAATGTCAAGATGG NF -200 Reverse GAATTTTGGGGAGTCCTICTG 129 60 B- Forward GCCTGACAATTTCATCTTTGG T UBB3 Reverse TCGCAGT’I’I‘TCACACTCCTTC 127 60 Forward CTGCCAGACCTGAAGAATGTC Neurons MAP-2 Reverse AGAGCCACATTTGGATGTCAC 132 60 Forward CATCGTGATGGCAAATATGAC NSE Reverse TCCCTGACAAAGTCCTGGTAG 104 62 Forward TGATTCTGTGAGTGTCAGTGTC NES T IN Reverse AGGAAACAGGGTCAGACTCTTC 14] 60 Astrocytes Forward AGAGGAACATCGTGGTGAAGAC GFAP Reverse C TGCTTGGACTCCTTAATGACC 65 60 Oligodendr F orWard TACGAGTGGTGGT'I'GATGAAAG ocytes GALC Reverse CACGACATAATAGGCAGTCAG 144 60 Housekeepi Forward AATTCCAAATTCTGCTI‘GCTTG ng BZM Reverse ACATCAAACATGGAGACAGCAC 146 60 102 Figure 4.1. Neural induction of canine and human AD-MSCs. Morphology of un- differentiated canine (Al) and human (A5) AD-MSCs. After induction with NFIR cocktail, undifferentiated cells aSSumed neural-like morphologies featuring bipolar to multipolar refractile bodies with long neurite-like extensions as seen in (A2, A3 = cAD- MSCs; A6, A7 = hAD-MSCs) (A2, A6 = day 5 after induction) and (A3, A7 = day 7 after induction). Cells maintained in the control medium retained fibroblast-like morphology (A4 = CAD-MSCs; A8 = hAD-MSCs). Magnification: A1, A5 = 40X; others =200x. 103 . “Q". ~ < u. .f‘) w“ (I 1' °-°.M"I°‘ ' ‘ o .D a a. .- . !: ... - '.: s " '1: A! I . n ' I »' u c ‘ ° _. -- -, -. j. w .’°~.p '5' o I.“ : ...'- .‘I. ' ’ a I. ’. .‘. a g. . " . lv'- -'.r '5": ‘ '.' .' '. .17". 0"°. . 0’23 Figure 4.2. Immunostaining for neuronal markers following neural induction. Induced cells (canine CAD-MSCs) with neural morphology expressed neuronal markers NF-200 (BI) and y-NSE (B3). In contrast, un-induced cells with fibroblast-like morphology did not stain for any of those proteins (BZ and B4). Magnification = 200x. Human cAD- MSCs with immunostaining for NF-200 (B5) and y—NSE (B7). are shown in the lower panel. Un-induced cells did not stain for either of those proteins (B6 and B8). Note: Since immunostaining of human cells were carried out in cytospin preparation, these cells lost the neural-like architecture during the procedure. A 50 ‘ cAD-MSCs Fold change MAPZ BT UB3 y-NSE NF200 NESTIN GALC Neural markers B '0 7 hAD-MSCs Fold change o w 4» O\ 00 MAP2 BT UB3 y-NSE NF200 NESTYN GALC Neural markers Figure 4.3. RT-PCR analysis for expression of neural markers at mRNA level. Real time qRT-PCR revealed that mature neuronal markers MAPZ, ,B- TUBULIN 111, y-NSE, and NF - 200 were considerably up-regulated following induction of CAD-MSCs (A) and hAD- MSCs (B), whereas a neural precursor marker — NEST IN was down-regulated, and oligodendrocyte maker -GALC was barely changed. 105 10. ll. REFERENCES . Caplan AI. Mesenchymal stem cells. J Orthop Res 1991;92641-650. Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143-147. . Kuo TK, Ho JH, Lee OK. Mesenchymal Stem Cell Therapy for Non- musculoskeletal Diseases: Emerging Applications. Cell Transplant 2009. Brzoska M, Geiger H, Gauer S et al. Epithelial differentiation of human adipose tissue-derived adult stem cells. Biochem Biophys Res Commun 2005;330:142-150. Lee KD, Kuo TK, Whang-Peng J et al. In vitro hepatic differentiation of human mesenchymal stem cells. Hepatology 2004;40:1275-1284. Seo MJ, Suh SY, Bae YC et al. Differentiation of human adipose stromal cells into hepatic lineage in vitro and in vivo. Biochem Biophys Res Commun 2005;328:258- 264. Woodbury D, Schwarz EJ, Prockop DJ et al. Adult rat and human bone marrow stromal cells differentiate into neurons. J Neurosci Res 2000;61:364-370. Sanchez-Ramos J, Song S, Cardozo-Pelaez F et al. Adult bone marrow stromal cells differentiate into neural cells in vitro. Exp Neurol 2000;164:247-256. Safi‘ord KM, Safford SD, Gimble JM et al. Characterization of neuronal/glial difl‘erentiation of murine adipose-derived adult stromal cells. Exp Neurol 2004;187:319-328. J iang Y, Henderson D, Blackstad M et al. Neuroectodermal differentiation from mouse multipotent adult progenitor cells. Proc Natl Acad Sci U S A 2003; 100 Suppl 12] 1854-11860. Lee OK, [(0 YC, Kuo TK et al. Fluvastatin and lovastatin but not pravastatin induce neuroglial differentiation in human mesenchymal stem cells. J Cell Biochem 2004;93:917-928. 106 12. l3. 14. 15. l6. 17. 18. 19. 20. 21. 22. Himes BT, Neuhuber B, Coleman C et al. Recovery of function following grafting of human bone marrow-derived stromal cells into the injured spinal cord. Neurorehabil Neural Repair 2006;20:278-296. Zhao LR, Duan WM, Reyes M et al. Human bone marrow stem cells exhibit neural phenotypes and ameliorate neurological deficits after grafting into the ischemic brain of rats. Exp Neurol 2002;174:11-20. Hofstetter CP, Schwarz EJ, Hess D et al. Marrow stromal cells form guiding strands in the injured spinal cord and promote recovery. Proc Natl Acad Sci U S A 2002;99:2199-2204. Chopp M, Li Y. Treatment of neural injury with marrow stromal cells. Lancet Neurol 2002;] :92- 100. Akiyama Y, Radtke C, Kocsis JD. Remyelination of the rat spinal cord by transplantation of identified bone marrow stromal cells. J Neurosci 2002;22:6623- 6630. . Gage FH. Mammalian neural stem cells. Science 2000;287:1433-1438. Magavi SS, Leavitt BR, Macklis JD. Induction of neurogenesis in the neocortex of adult mice. Nature 2000;405:951-955. Temple S, varez-Buylla A. Stem cells in the adult mammalian central nervous system Curr Opin Neurobiol 1999;92135-141. Zuk PA, Zhu M, Mizuno H et al. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 2001;72211-228. Caplan Al, Dennis JE. Mesenchymal stem cells as trophic mediators. J Cell Biochem 2006;98: 1076- 1084. Bartholomew A, Sturgeon C, Siatskas M et al. Mesenchymal stem cells suppress lymphocyte proliferation in vitro and prolong skin graft survival in vivo. Exp Hematol 2002;30:42-48. 107 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Le BK, Tammik C, Rosendahl K et a1. HLA expression and immunologic properties of differentiated and undifferentiated mesenchymal stem cells. Exp Hematol 2003;31:890-896. Schafi‘ler A, Buchler C. Concise review: adipose tissue-derived stromal cellsnbasic and clinical implications for novel cell-based therapies. Stem Cells 2007;25:818- 827. Gimble JM, Katz AJ, Bunnell BA. Adipose-derived stem cells for regenerative medicine. Circ Res 2007;100:1249-1260. Fraser JK, Wulur I, Alfonso Z et al. Fat tissue: an underappreciated source of stem cells for biotechnology. Trends Biotechnol 2006;24:150-154. Neff MW, Rine J. A fetching model organism. Cell 2006;124:229-231. Schneider MR, Adler H, Braun J et al. Canine embryo.derived stem cells--toward clinically relevant animal models for evaluating efficacy and safety of cell therapies. Stem Cells 2007;25: 1850-1851. Valenzuela M], Dean SK, Sachdev P et al. Neural precursors from canine skin: a new direction for testing autologous cell replacement in the brain 1. Stem Cells Dev 2008;17:1087-1094. Lin TM, Tsai JL, Lin SD et a1. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Neupane M, Chang CC, Kiupel M et al. Isolation and characterization of canine adipose-derived mesenchymal stem cells. Tissue Eng Part A 2008; 14: 1007-1015. Chang CC, Boezi JA, Warren ST et al. Isolation and characterization of a UV- sensitive hypermutable aphidicolin-resistant Chinese hamster cell line. Somatic Cell Genet 1981;7z235-253. Kao CY, Nomata K, Oakley CS et al. Two types of normal human breast epithelial cells derived from reduction mammoplasty: phenotypic characterization and response to SV40 transfection. Carcinogenesis 1995;16:531-538. 108 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. Rozen S, Skaletsky HJ. Primer3 on the W for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics Methods and Protocols: Methods in Molecular Biology. Totowa, NJ: Humana Press 2000:365-86. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402-408. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 2008;321101-1108. Wang KH, Kao AP, Wangchen H et al. Optimizing proliferation and characterization of multipotent stem cells from porcine adipose tissue. Biotechnol Appl Biochem 2008;51:159-166. Freed CR, Greene PE, Breeze RE et al. Transplantation of embryonic dopamine neurons for severe Parkinson's disease. N Engl J Med 2001;344:710-719. Kopen GC, Prockop DJ, Phinney DG. Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 1999;96:10711-10716. McKay R. Stem cells in the central nervous system. Science 1997;276:66-71. Amariglio N, Hirshberg A, Scheithauer BW et al. Donor-derived brain tumor following neural stem cell transplantation in an ataxia telangiectasia patient. PLoS Med 2009;6ze1000029. Neuhuber B, Gallo G, Howard L et al. Reevaluation of in vitro differentiation protocols for bone marrow stromal cells: disruption of actin cytoskeleton induces rapid morphological changes and mimics neuronal phenotype. J Neurosci Res 2004;77:192-204. Choi CB, Cho YK, Prakash KV et a1. Analysis of neuron-like differentiation of human bone marrow mesenchymal stem cells. Biochem Biophys Res Commun 2006;350:138-146. Egusa H, Schweizer FE, Wang CC et al. Neuronal differentiation of bone marrow- derived stromal stem cells involves suppression of discorth phenotypes through gene silencing. J Biol Chem 2005;280:23691-23697. 109 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. Palmer TD, Markakis EA, Willhoite AR et al. Fibroblast grth factor-2 activates a latent neurogenic program in neural stem cells from diverse regions of the adult CNS. J Neurosci 1999;19:8487-8497. Jori F P, Napolitano MA, Melone MA et al. Molecular pathways involved in neural in vitro differentiation of marrow stromal stem cells. J Cell Biochem 2005;94:645- 655. Tondreau T, Lagneaux L, Dejeneffe M et al. Bone marrow-derived mesenchymal stem cells already express specific neural proteins before any differentiation. Differentiation 2004:7223 19-326. Mareschi K, Novara M, Rustichelli D et al. Neural differentiation of human mesenchymal stem cells: Evidence for expression of neural markers and eag K+ channel types. Exp Hematol 2006;34:1563-1572. Deng J, Petersen BE, Steindler DA et al. Mesenchymal stem cells spontaneously express neural proteins in culture and are neurogenic after transplantation. Stem Cells 2006;24:1054-1064. Blondheim NR, Levy YS, Ben-Zur T et al. Human mesenchymal stem cells express neural genes, suggesting a neural predisposition. Stem Cells Dev 2006;15:141-164. Woodbury D, Reynolds K, Black IB. Adult bone marrow stromal stem cells express gemtline, ectoderrnal, endodermal, and mesodermal genes prior to neurogenesis. J Neurosci Res 2002;69:908-917. Lendahl U, Zimmerman LB, McKay RD. CNS stem cells express a new class of intermediate filament protein. Cell 1990;60:585-595. Marangos PJ, Zis AP, Clark RL et al. Neuronal, non-neuronal and hybrid forms of enolase in brain: structural, immunological and fimctional comparisons. Brain Res 1978;150:117-133. Roskams AJ, Cai X, Ronnett GV. Expression of neuron-specific beta-III tubulin during olfactory neurogenesis in the embryonic and adult rat. Neuroscience 1998;83:191-200. 110 55. 56. 57. 58. 59. 60. 61. 62. 63. 65. Julien JP. Neurofilament functions in health and disease. Curr Opin Neurobiol 1999;91554-560. Murakami K, Fellous A, Baulieu EE et a1. Pregnenolone binds to microtubule- associated protein 2 and stimulates microtubule assembly. Proc Natl Acad Sci U S A 2000;97:3579-3584. Deng W, Obrocka M, Fischer I et al. In vitro differentiation of human marrow stromal cells into early progenitors of neural cells by conditions that increase intracellular cyclic AMP. Biochem Biophys Res Commun 2001;282:148-152. Bignami A, Dahl D. Specificity of the glial fibrillary acidic protein for astroglia. J Histochem Cytochem 1977;25:466-469. Satoh JI, Tai T, Kim SU. Differential expression of gangliosides and galactolipids in fetal human oligodendrocytes and astrocytes in culture. Brain Res Dev Brain Res 1996;93:172-181. Takagi Y, Takahashi J, Saiki H et al. Dopaminergic neurons generated from monkey embryonic stem cells function in a Parkinson primate model 1. J Clin Invest 2005;115:102-109. Lu P, Blesch A, Tuszynski MH. Induction of bone marrow stromal cells to neurons: differentiation, transdifl‘erentiation, or artifact? J Neurosci Res 2004;77:174-191. Krabbe C, Zimmer J, Meyer M. Neural transdifferentiation of mesenchymal stem cells--a critical review. APMIS 2005;] 13:831-844. Bertani N, Malatesta P, Volpi G et al. Neurogenic potential of human mesenchymal stem cells revisited: analysis by immunostaining, time-lapse video and microarray. J Cell Sci 2005;118:3925-3936. Tondreau T, Dejenefi‘e M, Meuleman N et a1. Gene expression pattern of functional neuronal cells derived from human bone marrow mesenchymal stromal cells. BMC Genomics 2008;9: 166. Wislet-Gendebien S, Hans G, Leprince P et al. Plasticity of cultured mesenchymal stem cells: switch from nestin-positive to excitable neuron-like phenotype. Stem Cells 2005;23:392-402. 11] 66. 67. 68. 69. 70. 71. Dezawa M, Kanno H, Hoshino M et al. Specific induction of neuronal cells from bone marrow stromal cells and application for autologous transplantation. J Clin Invest 2004;113:1701-1710. Bang YJ, Pimia F, Fang WG et a1. Terminal neuroendocrine differentiation of human prostate carcinoma cells in response to increased intracellular cyclic AMP. Proc Natl Acad Sci U S A 1994;91:5330-5334. Cox ME, Deeble PD, Lakhani S et a1. Acquisition of neuroendocrine characteristics by prostate tumor cells is reversible: implications for prostate cancer progression. Cancer Res 1999;59:3821-3830. Ning H, Lin G, Fandel T et al. Insulin grth factor signaling mediates neuron-like differentiation of adipose-tissue-derived stem cells. Differentiation 2008;76:488- 494. Spinella MJ, Kerley J S, White KA et al. Retinoid target gene activation during induced tumor cell differentiation: human embryonal carcinoma as a model. J Nutr 2003;133:273S-276S. Kondo T, Johnson SA, Yoder MC et al. Sonic hedgehog and retinoic acid synergistically promote sensory fate specification from bone marrow-derived pluripotent stem cells. Proc Natl Acad Sci U S A 2005;102:4789-4794. 112 CHAPTER 5 Assessment of relative permissibility of mesenchymal stem cells at various stages of osteogenic differentiation to tumorigenic transformation by defined genetic intervention 113 SUMMARY Osteosarcomas (OS) have been viewed as a clinically and molecularly heterogeneous group of malignancies characterized by varying degrees of mesenchymal differentiation. Increasing evidence suggests that OS is indeed a disease of blocked differentiation, caused by genetic and epigenetic changes that interrupt the process of osteoblastic differentiation. However, the origin of this tumor is still unknown. The issue that whether normal adult stem cells such as MSCs may require fewer or different steps than more differentiated cells to acquire a transformed phenotype is still unresolved. The current study was undertaken to gain insights into this issue of origin in OS model system. We evaluated the consequences of forced expression of OS relevant oncogenes- MET and BM“ in canine and human MSCs at different time points of osteogenic differentiation, using inducible expression system. For the purpose of this study, open reading frame (ORF) of BMII and MET were cloned into retrovirus-based conditional expression vectors, transduced (singly or in combination) into MSCs, and then expressed at three different time points of osteogenic differentiation (undifferentiated MSCs, early stage differentiation and late stage differentiation). Our results demonstrated that ectopic expression of MET (alone or in combination with BMII) could not immortalize or transform the MSCs isolated from bone marrow and adipose tissues of human and dog. We found that expression of these genes, alone or in combination, were not sufficient to completely transform undifferentiated M SCs or their osteogenic descendants. Following ectopic overexpression, undifferentiated M SCs underwent senescence after modest extension of 114 life-span. Although the overexpression of these genes in MSCs at early stage of osteogenic differentiation blocked the further differentiation of these cells down the osteogenic lineages despite the maintenance of osteogenic stimuli, these cells also eventually succumbed to senescence. In contrast, MSCs at late stage of osteogenic differentiation were unaffected by overexpression of these genes and displayed osteogenic phenotype in the presence of suitable stimuli. These results suggest that additional pathways need to be perturbed for tumorigenic transformation of MSCs at different stages of osteogenic differentiation. We verified this by incorporation of SV40 early region (SV40 ER) which immortalized canine M SCs (but not human MSCs), and made them permissive to transformation by MET. On the other hand, failure to transform human MSCs with any combination of SV40 ER, MET and BMII indicates the existence of species differences, and possible requirement for the perturbation of additional or different pathways. Definitive evidence for tumorigenic transformation of these cells will be confirmed in future studies by xenotransplantation experiments. Extension of our observations to larger sample size in fiiture will provide better insights into the mechanistic bases for tumorigenic transformation process. We conclude that identification of origin of OS may lead to better therapeutic targeting, identification of new markers for disease progression, and earlier detection of tumors. 115 INTRODUCTION There is a growing body of evidence indicating that stem cells are involved not only in the generation of multicellular organisms and regeneration of their tissues, but that cells with stem cell attributes are also integral to the initiation and maintenance of several cancers [1-5]. Substantial experimental data is now available to support the notion that tumor transmitting ability is restricted to a small subpopulation of cancer cells, the ‘cancer stem cells’ (C SCs) [1-3]. These distinct sub-populations of self-renewing cancer cells drive tumor growth, and are responsible for relapse of tumor after completion of conventional therapy [4,5]. Cancer stem cells have been documented in several solid cancers including brain, breast, prostate, melanoma, lung, colon, head and neck, and pancreatic cancers (reviewed in [2,3]). However, the origin of these CSCs is an area of ongoing debate [3]. It has been argued that cancer stem cells (C SCs) may not necessarily implicate “the stem cell’s cancer”. Whether a cell at any stage of differentiation pathway can acquire an ability to become a CSC is still highly controversial. In the absence of definitive evidence, several possibilities exist for the origin of CSCs, including (i) undifferentiated stem cells, (ii) lingeage committed progenitor cells, or (iii) differentiated cells. Stem cells have long life span, allowing them to accumulate multiple genetic and epi genetic changes necessary for cellular transformation [6-8]. Moreover, since they have extensive proliferation capacity, fewer mutations are required to constitutively activate proto-oncogenic pathways (e. g. WNT, SHH, NOTCH) and inactivate tumor suppressor pathways (e.g. PTEN, p16), increasing their likelihood of becoming tumorigenic [9]. In fact, two recent studies have convincingly demonstrated that intestinal stem cells represent the cells-of-origin for 116 malignant transformation in colon cancer [10,11]. But, some other studies have shown the possibility of occurrence of final hit in progenitor cells [12-14] as well as dedifferentiation of differentiated cells prior to oncogenic events [1 5-1 7]. Many of these studies have shown that these CSCs are generally resistant to conventional chemotherapy and radiotherapy regimens, and are thus responsible for relapse of tumor after completion of conventional therapy [4,5]. Based on these premises, targeted therapy against C SCs which spares normal stem cells, has been claimed to prevent tumor recurrence without significant side effects (reviewed in [18]). But, if CSCs arise from dedifferentiation of differentiated cells or committed progenitor cells, removing those CSCs will not prevent the dedifferentiation event(s), and may eventually lead to the cancer recurrence despite the therapeutic targeting. So, the frequency with which non-tumorigenic cancer cells will evolve to become tumorigenic will determine the outcome of targeted therapy against CSCs. Thus, determination of what drives the tumor has critical consequences to our approach to treatment. MET, a tyrosine kinase receptor for hepatocyte growth factor (HGF), was originally identified as a transforming oncogene in a chemically transformed human osteosarcoma (OS) cell line [19]. This single receptor can transduce multiple biological activities including proliferation, survival, motility, and morphogenesis. Misexpression of MET has been noticed in several mesenchymal tumors including human and canine OS [20-23]. In a recent study, lentiviral vector-mediated overexpression of MET was shown to be enough for converting human primary cultured osteoblasts into OS cells [24]. 117 On the other hand, BMII is a polycomb transcriptional repressor, which has been shown to have a role in maintenance of adult stem cells as well as cancer stem cells [6,7,9,25,26]. It reduces the expression of pl6INK4a and p19ARF, and has been found to be overexpressed in several human cancers [27]. BMII overexpression in mammary epithelial cells led to their immortalization [28], and its overexpression in lymphocytes led to lymphoma [29]. A recent study showed that lentiviral-mediated overexpression of BM] 1 in human placenta derived MSCs (hPDMC) was able to immortalize these cells [30]. Mesenchymal Stem Cells (MSCs) are located in several connective tissue compartments [3]], and have the ability to give rise to several cell types of mesenchymal lineages including osteoblasts, chondrocytes, myocytes, and fibroblasts [32]. MSCs are thought to be the cell of origin for various types of sarcoma [33]. A recent study has shown that Ewing’s tumors (ET) originate from MSCs which accounts for the predominant localization of ET in bones and soft tissues, two major sources for these stem cells [34]. Although < 1% in occurrence, OS can also arise from soft tissues and visceral organs without any involvement of bone/periosteum (extraskeletal OS) [35-37]. The use of MSCs from both bone marrow and adipose origin in our experimental approach will allow us to assess the potential for OS arising from MSCs residing in different niches and/or their descendents. Osteosarcoma (OS) is the most common primary bone tumor of human and dogs [38,39]. It is a highly malignant tumor with poor clinical prognosis. 1t accounts for 4% of all malignancies, 85% of all skeletal malignancies, and 98% of appendicular primary bone tumors [38]. Canine OSAclosely resembles human OS in terms of histopathological 118 appearance and biological behavior [40], and is an excellent model system to study tumor biology and therapeutic intervention [41-43]. Osteosarcoma is quite resistant to conventional chemotherapy and radiotherapy, and contains varying proportion of undifferentiated cells, osteoblasts, chondroblasts, and fibroblasts [44]. Increasing evidence suggests that OS is indeed a disease of blocked differentiation, caused by genetic and epi genetic changes that interrupt the process of osteoblastic differentiation [45,46].These attributes make canine OS as a promising candidate for cancer derived from and/or driven by stem cells [47]. In the current study, we evaluated the consequences of ectopic expression of OS relevant oncogenes-MET and BMII in canine and human MSCs at different time points of osteogenic differentiation, using inducible expression system. Our study revealed that MSCs and their osteogenic descendants are refiactory to these tumorigenic insults, and perturbation of additional pathways, including those altered by SV40 early region genes might be necessary for tumorigenic transformation of MSCs at different stages of osteogenic differentiation. 119 MATERIALS AND METHODS Materials: All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise stated. Source of mesenchymal stem cells (MSCs): Mesenchymal stem cells (MSCs), isolated and characterized from canine and human bone marrow and subcutaneous adipose tissue, were used in this study. Bone marrow-derived MSCs (BM-MSCs) (designated as BM66-D for canine BM-MSCs, and BM—8 for human BM-MSCs) were isolated from plastic adherent fraction of mononuclear cells (separated by density gradient centrifugation) of bone marrow aspirate following a modification of publishedprotocols [48,49], and were expanded in K-NAC medium containing 5% FBS. Adipose tissue-derived MSCs (AD-MSCs) (designated as BR7I for canine AD-MSCs, and PLA-3 for human AD-MSCs) were isolated from collagenase- extracted stromal vascular fraction of adipose tissue, after expansion of plastic adherent cells in low calcium KNAC medium containing 5% FBS. These AD-MSCs were previously isolated and characterized by our research group [49,50]. K-NAC medium is a modified MCDB 153 medium (Keratinocyte—SFM) (Invitrogen Corporation, Carlsbad, CA) supplemented with N-acetyl-L-cysteine (NAC) (2 mM) and L-ascorbic acid 2- phosphate (Asc 2P) (0.2 mM), and has been found to be optimum for maintaining the proliferation and differentiation potential of MSCs from different species in our experimental system [49,50]. Before using in this study, culture expanded MSCs were validated for their expression of transcripts for pluripotency-associated transcription factors (0C T 4, 120 NANOG, and SOX2), and potential to differentiate into mesenchymal lineages (osteoblasts, chondrocytes, and adipocytes) as described earlier [49,50]. Early passage cells (passage 3 to 5) were used in this study and they were maintained in K-NAC medium with 5% F BS, unless otherwise indicated. Construct details: Expression of transgene(s) were regulated by tetracycline-regulatable (pSCNIT and pHIT) or ecdysone-inducible (pBORIS-I and pBORIS-III) vectors (gifts from Dr. Steven Suhr, Michigan State University). The pSCNIT and leT vectors are moloney murine leukemia virus (MMLV)-based, replication-defective retroviral vectors (Tet-off system) with neomycin (for pSCNIT vector) or hygromycin (for pI-IIT vector) resistance gene, internal ribosome entry site (IRES) and tetracycline transactivator protein (tTA) located upstream of a minimal human cytomegalovirus (hCMV) promoter fused to the tetracycline operator sequence (tetO-CMV) which directs transgene expression [51]. The pBORIS-I and pBORIS-III vectors are also MMLV-based retroviral vectors with neomycin resistance gene, Bombyx mori ecdysone receptor (EcR) transactivator, and an ecdysone responsive promoter that directs transgene expression [52]. Subcloning of MET: Primers were designed to introduce Sfil and Pmel restriction sites at the 5’ and 3’ ends of the amplicon respectively (see below), and 4.2 kb open reading frame of human MET was amplified from pRRL.sin.ppt.hCMV.METpre vector [24] (gifi from Dr. Maria Flavia Di Renzo, University of Torino, Turin, Italy ) with iproof high fidelity DNA polymerase (Bio-Rad, Hercules, CA). The amplified fragment was subcloned into Sfil and Pmel sites of pSCNIT and leT vectors, and the MET sequence 121 insert was verified by bi-directional sequencing. Sequencing primers are listed in Table 5.1. Forward primer: 5'- GAGAGAGCGGCCGCCTGGGCCGAAA T GAA GGCCCCC GCT GT GCTT G—3' (Sfil site underlined, and 5’ end of MET italicized); Reverse primer: 5 '- GAGAGAGTTTAAACCT A T GA T GT CTCCCA GAA GGA G-3 '(Pmel site underlined, and 3’ end of MET italicized) Subcloning of BMII: Primers were designed to introduce Sfil and Pmel restriction sites at the 5’ and 3’ ends of the amplicon respectively (see below), and 1 kb open reading frame of human BMII was amplified from pOTB7 vector (Open Biosystems, Huntsville, AL) with iproof high fidelity DNA polymerase (Bio-Rad). The amplified fi'agment was subcloned into SfiI and Pmel sites of pI-IIT vector as well as pBORIS-l and pBORIS-III vectors, and the BMII sequence insert was verified by bi-directional sequencing. Sequencing primers are listed in Table 5.2. Forward primer: 5 '- GAGAGAGGCCGCCTGGGCCACGCGTGGTA T GCA T C GAA CAA CGA GAA T C-3 ' (Sfil site underlined, and 5’ end of BMII italicized); Reverse primer: 5 '- GAGAGAATCGATGTTTAAAC T CAA CCA GAA GAA G TT GCT GA T G-3 ' (Pmel site underlined, and 3’ end of BMII italicized) Subcloning of YFP: YFP (yellow fluorescence protein) fragment was removed from pSCN1T3-YFP-HMG vector (obtained from Dr. Steven Suhr, Michigan State University), 122 and inserted into multiple cloning sites of pHIT, pBORIS-I, and pBORIS-III vectors using Mia] and Pmel. Production of retrovirus: For production of retrovirus [53], 150 mm tissue culture plates were seeded with 1><107 HEK 293-F T cells in a-MEM one day prior to transfection. For 6 plates, 75 ug pVSV-G, 75 ug pGag-pol (gifts from Dr. Steven Suhr, Michigan State University), and 150 pg transgene (either YFP, MET or BMII transgene) containing plasmid (in the backbone of either pHIT, pSCNIT, pBORIS-I, or pBORIS-IH vector) were combined together in 12.5 ml water. DNA-calcium phosphate complex was prepared by drop-wise addition of 2.5 ml of calcium chloride (2 M), followed by slow drop-wise addition of 15 ml of 2x HBS solution. Then, 5 ml of the final mixture was added drop-wise to each plate. The DNA-precipitate-containing medium was replaced with fresh medium (25 ml u—MEM with 10% FBS per plate) 8 hrs post-transfection, and the conditioned medium with viral particles were collected 48 hrs post-transfection. Viral supematants were filtered through 0.45 uM low protein binding filters (Nalgene, Rochester, NY) following low speed centrifugation (2000 rpm for 5 min), and concentrated by ultracentrifugation (24,000><10s cells per well, and transfected in triplicates with 10, 25, 100, and 800 ng of pHIT, pBORIS-I, and pBORIS- III vectors containing YFP insert, using Lipofectarnine 2000 (Invitrogen) according to the manufacturer’s instructions. After 8 hrs of transfection, lipofectamine-containing culture medium was replaced with fresh media with-or without lug/ml doxycycline or 1 uM tebufenozide, and YFP expression was monitored for 72 hrs. Transduction of MSCs: MSCs were seeded in 24-well plate at 1><104 cells per well in K-NAC medium containing 5% F BS. After overnight attachment, culture medium was removed and viral supernatant diluted with 300 [.11 medium and supplemented with 8 ug/ml polybrene, were added at different multiplicity of infection (MOI) (MOIs of 50, 25, 10, 5, 1, 0.1, and 0.01) to each culture well. MSCs were transduced with either ME T -or BMII-containing virus individually or with both viruses (co-transduced simultaneously or transduced in step- wise manner). Plates were centrifuged at 1000>< g for 1 hr at 30°C. Then, 300 pl of fresh culture medium was added to each well and plates were kept at 37°C in C02 incubator. On the next day, virus containing culture media was replaced with fresh culture media with or without ligands (1-2 ug/ml doxycycline or 1 uM tebufenozide). In some cases, cells were subjected to second round of transduction using similar method. Regulation of vector expression in MSCs was assessed by monitoring YFP expression in the presence or absence of ligands after 48 hrs of transduction. For stable selection, transduced cells were passaged at 1:10, cultured in the presence of doxycycline, and selected with antibiotics (200-400 ug/ml G418 or 50-100 124 rig/ml hygromycin) after 48 hrs for 10 days. Individual resistant clones were picked up and expanded in the presence of selection antibiotics and doxycycline. All transduction experiments were carried out with YFP-containing virus in parallel in order to rule out any possible artifact unrelated to the oncogene of interest. Same constructs were used in both human and canine cells, as the human and canine proteins are identical at 92% for MET and 98.7% for BMll. Induction of osteogenesis and regulation of transgene expression: Transgene(s)-inserted MSCs clones were expanded in the presence of doxycycline (l ug/ml) and plated at the seeding density of 1000 cells/cm2 in 24-well laminin-coated plates, and then treated with 1, 25-dihydroxyvitamin D3 (0.01 pM), ascorbate-Z-phosphate (50 pM) and B-glycerophosphate disodium (10 mM) (V AG cocktail) in D medium containing 10% F BS with medium change once in every 3 days [66]. (For human MSCs, l, 25-dihydroxyvitamin D3 was replaced by 0.1 uM dexamethasone). Expression of the transduced genes (MET and/or BMII, or YF P) were induced (by removing doxycycline in the culture media) at two different time points after osteogenic induction of MSCs (week 2 and week 4) as well as in undifferentiated’MSCs (maintained in K-NAC medium with 5% FBS). Although arbitrary, these time points were chosen on the basis of available evidences to reflect lineage commitment (2 weeks) and late stage of osteogenesis (4 weeks) from MSCs in our system [49,50]. SV40 transfection and retroviral transduction of MSCs: Early passage (passage 3-5) MSCs were plated in 24-well plate at 2><10S cells per well and transfected by pRNS-l plasmid (containing an origin-defective SV40 genome) 125 [54], using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. To get stable transfectants, transfected cells were passaged at 1:10 and selected with 200 jig/ml G418 after 48 hrs for 10 days. Individual resistant clones were picked up and expanded in the presence of 100 jig/ml G418 in order to determine their life span in vitro. SV40-ER containing clones with extended life span (>50 cpdl) were transduced with ME T - or BMII -containing retrovirus (with pI-IIT vector backbone) as described above and selected with 100 ug/ml hygromycin for stable expression. Characterization for tumorigenic phenotypes in vitro: Cells which continued to proliferate beyond the standard life span of normal MSCs (~35 cpdl) after expression of transgenes were chosen for characterization of following tumorigenic phenotypes: (i) Determination of cumulative population doubling level (cpdl): 1 x 105 cells were plated in 75 cm2 flask, grown in original culture medium (K-NAC medium with 5% FBS or differentiation medimn with 10% FBS) until near confluence for quantification of final cell yield, and the cells were continuously subcultured to determine cumulative population doubling [49]. (ii) Anchorage independent growth: 5 x 10" cells in 3 ml of 0.33% agarose medium were plated on top of 3 ml pre-hardened 0.5% agarose medium in each triplicate 60 mm dishes with grids to aid colony counting. 2.5 ml liquid medium (K-NAC medium with 5% F BS) was then added and renewed once every 3 days. At the end of 3 weeks, the numbers of anchorage independent colonies developed were scored under a microscope [49]. 126 (iii) Matrigel invasion assay: 5 x 104 cells were seeded in 500 pl of serum free culture medium (K-NAC medium) on the upper side of Matrigel Matrix (with 8 pm pore size PET membrane) of each well of 24-well cell culture insert (BD Biosciences, San Jose, CA). The lower chamber of the transwell insert were filled with 750 u] K-NAC medium containing 10% FBS. After 24 hr of incubation, the filters were removed and cells that invaded the Matrigel and attached to the lower chamber of the transwell were fixed with 100% methanol for 10 min, and stained with 0.2% cresyl violet for 20 min [55]. Cresyl violet stain on the membranes was eluted with 100% ethanol/0.2 M sodium citrate (1:1), and absorbance was read at 570 nm using SpectraMax M5 microplate reader (Molecular Devices, Sunnyvale, CA). The percentage of invading cells was calculated by comparison of absorbance in cells attached to the underside of the membrane (topside wiped with cotton wool) against absorbance of total cells in the membrane (topside not wiped). (iv) Determination of saturation density: To determine saturation density, 1 x 104 cells were seeded in each of multiple 60-mm dishes on day 0, and grown in K-NAC medium with 5% FBS. The cultures were re-fed every day for 10 days, and cells were counted from triplicate dishes on days 4-10. The saturation density of control vector was normalized to 1.0, and the relative values were reported for each experimental group. Evaluation of senescence-associated beta—galactosidase activity: Cells were plated in 6—well plate at 2>~<105 cells per well, and beta-galactosidase activity was measured at pH 6 [56], using senescence beta-galactosidase staining kit (Cell Signaling Technology, Danvers, MA) according to the manufacturer’s instructions. 127 Validation of transgene expression: To make sure that the transgenes are getting expressed in the context of MSCs, we performed RT-PCR to look at the expression of exogenously introduced genes (MET, BMII, or SV40). Total RNA was isolated from different experimental groups, using Versagene RNA isolation kit (5 Prime Inc, Gaithersburg, MD). RNAs were treated with TURBO DNA-free DNase I (Ambion, Austin, TX) to remove any contaminating genomic DNA. One ug total RNA was reverse transcribed in 30 ul reaction volume, using 10 units of Superscript III reverse transcriptase (Invitrogen, Carlsbad,CA), and final concentrations of 5 uM anchored oligo-dT primers, 0.5 mM dNTPs, 5 mM DTT, and 1x first strand buffer supplied by the manufacturer (Invitrogen, Carlsbad,CA) . The reverse transcription reaction was performed at 50°C for 1 hr, and then inactivated at 70°C for 15 min. Twenty-five ul PCR reactions were prepared with 2111 cDNA, 5 pmol of each primer, 0.5 units of Taq polymerase (Invitrogen, CA), and final concentrations of 40 uM dNTPs, 2 mM MgC12, 20 mM Tris-HCl, and 50 [.11 KC]. Cycling conditions were as follows: 94°C for 4 min; 35 cycles at 94°C for l min, 60°C for 1 min, 72°C for 1 min; followed by 72°C for 5 min. The PCR products were separated on 2% agarose gel by electrophoresis, stained with ethidium bromide, visualized under UV light, and digital images captured with Alphalmager software (San Leandro, CA). Primer sequences used for PCR are listed in Table 5.3. Transgene expression was also verified at protein level by immunostaining with antibodies for BMII and MET. 128 RESULTS Expression of MET and BMII mRNA in OS cell lines: We first evaluated the expression of MET and BMII transcripts in a panel of OS cell lines (both canine and human) by quantitative RT-PCR. The expression levels were compared to that of mRNAs from normal osteoblasts (induced to differentiate from MSCs). Our results showed that many of the canine and human OS cell lines expressed remarkably higher level of mRNAs for both MET (Figure 5.1A) and BMII (Figure 5.1B) than those expressed by normal osteoblasts, although there was a considerable variation among cell lines regarding the level of expression. Validation of ligand-regulatable expression systems: In order to examine the role of MET and BMII overexpression in MSCs at different time points of osteogenic differentiation, we established ligand regulatable expression system by cloning the open reading frame of MET and BMII as well as that of YFP into doxycycline-regulatable vectors (leT or pNIT) and ecdysone-regulatable vectors (pBORIS-l and pBORlS-I‘II). We first sought to determine the inducibility of the expression system by introducing the various amounts of YFP insert-containing pHIT, pBORIS-I, and pBORIS-III vectors into easily transfectable system of HEK 293-FT cells. In case of tetracycline-regulatable pHIT vector, very robust expression of YFP was seen with 100 and 800 ng of plasmid (transfected into 2 X 105 cells per well of 24-well plate), which was efficiently turned off in the presence of 1 ug/ml of doxycycline (Figure 5.2A- D). In contrast, YF P expression was very weak from pBORIS-III vector (Figure 5.2E-H) 129 and even weaker from pBORlS-I vector (Figure 5.21-L) (both are ecdysone-regulatable vectors), regardless of amount of the plasmid. We then packaged the transgene-containing vectors into replication-incompetent retroviruses and transduced them into MSCs. Similar to the context of HEK 293-FT, YFP expression from leT vector was very robust in MSCs (Figure 5.3A, B), whereas the fluorescence was weak from pBORIS-III (Figure 5.3C, D) and virtually undetectable from pBORI S-l (Figure 5.3E, P). So, for the purpose of present study, we focused on tetracycline-regulatable vectors for expression of all our transgenes of interest, and did not pursue these BORIS vectors any further. Stable transduction of BMII and MET in MSCs: In an attempt to analyze the transformation of MSCs from two different species (canine sand human) and their osteogenic descendants, we transduced early passage cells (passage 3 to 5) of canine and human MSCs (one cell line each of both bone marrow- derived and adipose tissue-derived MSCs; 4 MSCs cell lines in total) with tetracycline- regulatable pI-IlT-BM11(with hygromycin resistance gene) and pSCNIT-MET (with neomycin resistance gene) vectors at several different MOls (see materials and methods). Turning on the transgene expression at this stage revealed that transduction above MOI of 5 had somewhat toxic effects on the cells (true for both BMII and MET containing vectors), but cell morphologies and grth rates of the surviving cells were similar to those transduced at lower MOI. So, we carried out all subsequent transductions at MOI of 1. We transduced all four MSCs (dog and human) (MOI = 1), picked up 24 clones from 130 each cell line (for each of BMIl and MET transduction, alone or in combination) and expanded them in the presence of selection antibiotic and doxycycline. Ectopic expression of BMII and MET at different time points of osteogenic differentiation: We then induced osteogenic differentiation of transgene(s)-containing MSC clones, and turned on the expression of transgene(s) at two different time points of osteogenic differentiation (as well as in undifferentiated MSCs). Our results demonstrated that forced expression of BMII and/or MET could not prevent the eventual senescence of undifferentiated MSCs (Figure 5.4). We examined 24 clones each of BMII and MET (alone or in combination)-expressing cells from 4 different MSCs. The majority of MSC clones expressing BMII or MET (Figure 5.10) alone underwent senescence within 30-35 cpdl, which is close to the end of normal life- span of human and canine MSCs in our system (Figure 5.7A) [49,50]. These senescent cells adopted flattened and enlarged morphology, did not proliferate further despite the presence of mitogenic stimuli, and showed positive staining for SA-B-Gal (Figure 5.4C, D, G, H). A few of the clones expressing BMIl and MET together (2 clones each of canine BM-MSCs and AD-MSCs, and 4 clones of human AD-MSCs) were able to proliferate beyond 30 cpdl, but all of them succumbed to senescence within the next 10- 15 cpdl when sub-cultured at low cell density (1800 cells/cmz) (Figure 5.4E-H) (Figure 5.7 B, C). And, none of these cells were able to form large-sized colonies (>30 urn) even after 2 months of incubation in sofi agar with regular medium change (data not shown). We noted that several clones of BM11 and MET expressing human AD-MSCs could 131 grow at very high density, and when maintained at high cell density from early stage (after ~20 cpdl), they remained very confluent and quiescent rather than becoming senescent (Figure 5.41). We verified the expression of mRNAs for BMII and MET in these cells (Figure 5.8A). After 4 months in culture with regular media change, some cells in one of the clone (designated as PLA3-clone 7) outgrew as distinct sub- populations of rapidly growing spindle-shaped cells with criss-crossed morphology. We sub-cloned two of these sub-populations (designated as PLA3-clone 7.1 and PLA3-clone 7.2), and expanded them in culture (Figure 5.4], K). Intriguingly, both of these subclones continued to grow for next 36 cpdl (Figure 5.7D), and then succumbed to senescence (Figure 5.4L). When we over-expressed the transgene(s) after 2 weeks of osteogenic differentiation (Figure 5.5A, D), these cells could not undergo fruther differentiation despite the presence of osteogenic stimuli in the culture media (Figure 5.5B, E). Out of the 24 clones examined for each transgene overexpression, almost all of them senesced within next 15 cpdl (Figure 5.5C, F). Although one of the BMII over-expressing human AD-MSCs clones kept dividing for 25 cpdl, it eventually underwent senescence. In contrast, turning on the transgene(s) expression after 4 weeks of osteogenic differentiation (Figure 5.5G) could not prevent the terminal differentiation of any of the clonal cells we had examined, and these cells were eventually covered by mineralized deposits (Figure 5.5H, I). 132 We subjected early-stage clones from all of these MSCs to second round of transduction with the same combination of BMII and/or MET-containing retroviruses, but this could not rescue the eventual senescence of these cells. Consequences of expression of SV40 ER in MSCs: Based on our data, immortalization seemed to be a major limiting step in the transformation process. SV40 ER has been shown to play a crucial role in immortalization of normal diploid human cells [57]. We incorporated large T antigen containing-SV40 early region (SV40 ER) intoone each of canine MSCs (BM 66-D) and human MSCs (FLA-3) used in this study, and verified the expression by RT-PCR (Figure 5.9). Expression of SV40 ER was sufficient to immortalize canine MSCs (Figure 5.6A). These cells continued to proliferate beyond 200 cpdl (Figure 5.7E), and we did not notice any crisis period during their continuous growth in culture. On the other hand, SV40 ER could not prevent the senescence of human MSCs (Figure 5.6E), although it extended the life span of these cells modestly (from 35 cpdl to 74 cpdl) (Figure 5.6D). We then asked whether the expression of MET or BMII in SV40 ER-expressing cells can induce transformation phenotype in these cells or not. For this, we transduced SV40 ER- transfected cells (at ~50 cpdl) with retroviral vectors carrying MET or BMll (pSCNIT- MET or pSCNIT-BMII). Since human MSCs could not be immortalized with SV40-ER expression alone, ect0pic expression of either of MET or BMIl also could not rescue these cells from eventual senescence (Figure 5.6F). On the other hand, SV40-ER immortalized canine MSCs exhibited in vitro transformation phenotypes (described below) in response to MET overexpression (Figure 5.68). 133 Evaluation of tumorigenic phenotypes in vitro: SV40-ER immortalized canine BM-MSCs (BM66-D) were subjected to a variety of in vitro surrogate assays for transformation phenotypes, along with other experimental groups (Figure 5.7). Although BMII overexpression showed some changes in phenotypes, these changes were relatively less prominent when compared to the effects of MET. SV40-ER immortalized and MET over-expressing canine MSCs exhibited increased ability to form large sized colonies (>50 pm) in soft agar (Figure 5.7A), which was significantly higher than those formed by immortal cells (SV40 only), and BMII overexpressing immortal cells (p <0.001). Control cells (transfected with YFP control vector) formed only small-sized abortive colonies (<30 um). Moreover, MET over- expressing, immortalized cells had significantly higher invasive ability (p <0.001) (Figure 5.7B) and saturation density (p <0.001) (Figure 5.7C) when compared to other experimental groups (including BMII over-expressing, SV40-immortalized only, or vector control). 134 DISCUSSION Multiple lines of evidence suggest that genetic and epigenetic alterations in stem cells might represent the path of least resistance in the tumorigenic transformation process [10,11]. It has been argued that these cells may require less than the estimated 4 to 7 distinct genetic changes needed for malignant transformation of more differentiated cells [58]. However, other studies indicated that CSCs phenotype can be acquired, and more differentiated cells do not necessarily behave as evolutionary dead ends [13-16]. It is now well-known that even a normal differentiated cell (from multiple species and multiple sources) can be reprogrammed back to a functional pluripotent embryonic stem- like cell by expressing a right combination of transcription factors (reviewed in [59]). These findings raise the possibility that cells at different stages of differentiation can be responsive to tumorigenic event(s), but the frequency and resultant tumorigenic phenotype will be different depending upon the respondent cell types and the particular constellation of genetic changes. This may be the reason why OS are so heterogeneous in terms of their phenotypes, behavior, and clinical outcomes. Using liposarcoma as a model, Matushansky et a1 [60] suggested that transformation of mesenchymal tumors can occur at distinct states of cell maturation, and this might be responsible for a spectrum of histological subtypes seen in tumors originating from any particular organ. In support of this notion, it was recently demonstrated that depending on whether glioblastomas were initiated in neural precursor cells or their differentiated derivatives, distinct tumor phenotypes were seen [61]. Similarly, deletion of a single gene Pax-S in differentiated B lymphocytes was able to dedifferentiate these cells, culminating into aggressive lymphomas [62]. 135 Earlier studies had shown that only when MSCs were immortalized by forced expression of BMII and hTERT [63] or by inactivation of p53 and Rb together with hTERT activation [64], they were permissive to tumorigenic transformation by H-Ras oncogene. However, activated mutations of Ras genes are rarely found in OS [65]. So, we sought to evaluate the role of OS relevant oncogene(s). MET has been found to be over- expressed in many cases of both human and canine OS [20,23]. Importantly, a recent study demonstrated that overexpression of MET was sufficient to transform normal human osteoblasts into tumorigenic OS cells [24], which is in marked contrast to the usually postulated requirement of multiple defined genetic changes for neoplastic transformation of human primary cells [66]. However, the so—called osteoblasts used in this study [24] were not rigorously characterized for osteoblastic phenotype, and were more likely to be the mesenchymal stem cells (MSCs), which have the potential to become osteoblasts under appropriate microenvironment. On the other hand, BMII is involved in the self-renewal of many somatic stem cells [7], and this polycomb protein has been reported to immortalize some human epithelial cells [28] as well as human placenta-derived MSCs [30], although it failed to immortalize human fibroblasts [67]. Based upon these evidences, we attempted to evaluate the consequences of forced overexpression of MET and/or BMII in MSCs at different stages of osteogenic differentiation. Our results demonstrated that forced expression of MET (alone or in combination with BMll) could not immortalize or transform the MSCs isolated from bone marrow and adipose tissues of human and dog. We found that expression of these genes, alone or in combination, were not sufficient to completely transform undifferentiated MSCs or 136 their osteogenic descendants. Following forced overexpression, undifferentiated MSCs underwent senescence afier modest extension of life-span. Although the overexpression of these genes in MSCs at early stage of osteogenic differentiation blocked the further differentiation of these cells down the osteogenic lineages despite the maintenance of osteogenic stimuli, these cells also eventually succumbed to senescence. In contrast, MSCs at late stage of osteogenic differentiation were unaffected by overexpression of these genes and displayed osteogenic phenotype in the presence of suitable stimuli. These results suggest that additional pathways need to be perturbed for tumorigenic transformation of M SCs at different stages of osteogenic differentiation. We verified this by incorporation of SV40 ER which immortalized canine MSCs (but not human MSCs), and made them permissive to transformation by MET. On the other hand, failure to transform human MSCs with any combination of SV40 ER, MET and BMII indicates the existence of species differences, and possible requirement for the perturbation of additional or different pathways. Definitive evidence for tumorigenic transformation of these cells will be confirmed in future studies by xenotransplantation experiments. SV40 ER contributes to the immortalization process by large T antigen-mediated inactivation of tumor suppressors-TP53 and retinoblastoma protein (RBI), and further facilitates the transformation process by small T antigen-mediated inactivation of protein phosphatase 2A (PP2A) pathway [57]. However, it has been argued that SV40-mediated immortalization is accompanied by widespread genomic alterations, and this might confound the exact genetic elements necessary for neoplastic transformation [57]. For instance, it was shown earlier that SV40 large T antigen induces a CCAAT box binding factor which transactivates cyclin A, cdc2, DNA polymerase a, thymidine kinase etc 137 [68]. We propose that modulation of SV40 ER—regulated pathways in inducible manner in the context of canine MSCs along the hierarchy of osteogenic differentiation will provide better insights in the biology of osteosarcomagenesis. Unlike the report of Patane et al [24] on successful transformation of human osteoblasts by MET overexpression, neither human nor canine MSCs could be transformed by overexpression of MET (singly or in combination with BMII) in our experimental settings. In that study, it had taken almost 2 months to see the foci of transformation after transduction with lentiviral vector, suggesting the occurrence of other secondary transforming events subsequent to MET overexpression. It is possible that heterogeneity in the cellular context might play important role in the tumorigenic transformation events. We argue that polymorphism(s) in the coding or non-coding regulatory regions of the genes which modify tumor susceptibility might have considerable effects on transformation phenotype. In support of this notion, Akagi et a1 [69] had shown that two out of four different human fibroblasts could not be transformed by the same combination of SV40ER, catalytic subunit of telomerase, and activated H- RAS oncogene which was able to transform the other two fibroblasts, including BJ fibroblasts. These human B] fibroblasts, which have been successfully used for neoplastic transformation assays, were shown to exhibit very high antioxidant capacity and slow telomere shortening unlike several other fibroblasts (reviewed in [69]), and this might have contributed to their permissibility to tumorigenic transformation. Therefore, we cannot rule out the possibility that overexpression of the same combination of BMII and MET may be able to immortalize and/or transform MSCs obtained from other donors. Extension of our observations to larger sample size in firture will provide better 138 insights into the mechanistic bases for tumorigenic transformation process. We propose the use MSCs from OS patients (both human and canine) and OS-prone canine breeds (e.g. Irish Wolfhound, Rottweiler, Greyhound) and incorporation of various mutant versions of MET in the background of TP53 and/or RB knockdown for future studies. In conclusion, identification of origin of OS may lead to better therapeutic targeting, identification of new markers for disease progression, and earlier detection of tumors. 139 APPENDIX Table 5.1. Primers used for sequencing of MET open reading frame Primer ID 1 Sequence (5’-3’) Forward Primers hMETl stF (1-600) ATGAAGGCCCCCGCTGTG hMET600-1200F TITCCCAGATCATCCATTGC hMET1200-1800F GCCGTGATGAATATCGAACAG hMET1800-2400F GAAATGCACAGTTGGTCCTG hMED400-3000F TACCACTCCTTCCCTGCAAC hMET3000-3600F ATGCCGACAAGTGCAGTATC hMET3600-4200F TCAAGGTI‘GCTGATTTTGGTC SCNTHa GCTCGTTTAGTGAACCGTCAG Reverse Primers hMET1-6OOR CGAATGCAATGGATGATCTG hME T 600-1200R GCAAAGCTGTGGTAAACTCTG hMET1200-1800R AATGTATTCATCGTGCTCTCAC hMET1800-24OOR TGCAGGGAAGGAGTGGTAC hMET2400-3000R GCATGAACCGTTCTGAGATG hMETBOOO-3600R ACCAAAATCAGCAACCTTGAC hMEIlastR (3600-4200) CTATGATGTCTCCCAGAAGG SCNT3_Rb TTATGTATTTI‘CCATGCCTTGC Table 5.2. Primers used for sequencing of BMII open reading frame Primer ID | Segmence (5’-3’) Forward Primers hBMII l-SOOF TGCATCGAACAACGAGAATCA hBMlI SOO-IOOOF TGTCATGTATGAGGAGGAACC BORIS-I-Fc CGCTAGAACAAGGG'ITCAATG Reverse Primers hBMII 1-500 R TGCTGGGCATCGTAAGTATC hBMII 500-1000R TCAACCAGAAGAAGTTGCTG d BORIS-I-R TCTG'I'I‘CCTGACCTTGATCTG a Primers specific to pSCNT3 vector backbone b . . aners specrfic to pSCN T3 vector backbone c Primers specific to pBORIS-I vector backbone d . . aners specrfic to pBORIS-I vector backbone 140 Table 5. 3. Primers used to validate the expression of transgenes Ann Amplicon Temp Gene Primer sequence (5' - 3') Size (bp) (°C) BMII Forward TGTCATGTATGAGGAGGAACC (Human) Reverse TC AAC C AGAAGAAGTTGCTG 490 60 MET Forward TATCAGCTTCCCAACTTCACC (Human) Reverse GTGAAC CTC C GACTGTATGTC 358 60 Forward GTGGCTATGGGAACTGGAG SV40 Reverse CTCTACAGATGTGATATGGCTG 227 60 GA PDH Forward CTC TGC TC C TCCTGTTC GAC (Human) Reverse ACGACCAAATCCGTTGACTC 112 60 32M Forward TCTACATTGGGCACTGTGTCAC Canine) Reverse TGAAGAGTTCAGGTCTGACCAAG 136 60 141 Figure 5.1. Expression of MET and BMII in OS cell lines. Relative expression level of MET (A) and BMII (B) genes were determined by quantitative RT-PCR. mRN A level of each gene was normalized to that of ,BZ-MICROGLOBULIN, and expressed as fold change relative to the osteoblasts differentiated fiom mesenchymal stem cells. D-17, ABRAMS, and GRACIE are established cell lines of canine OS, whereas MSU-DOS-3 to MSU-DOS-IOM are primary culture isolates from canine OS samples established in our laboratory. UZOS and SaOS-2 are human OS cell lines, and MSCs represent mesenchymal stem cells. 142 . 0 3 . . _ 5 O 5 2 2 II. 10- owuaau 2o...— whom—2 N-m0m_2 TwUmE wON-D N-m050 um were counted 3 wks afier seeding. Data were represented as meanisd of triplicates. For matrigel invasion assay (B), a total of 5 x 104 cells were seeded in on the upper side of Matrigel Matrix and incubated for 24 hrs using 10% F BS as a chemoattractant. Cells that had invaded the Matrigel were quantified colorimetrically after cresyl violet staining. Data were represented as meanisd of triplicates. For determination of saturation density (C), a total of 1 x 104 cells were seeded in each of multiple 60-mm dishes, and cells were counted cultures from triplicate dishes on days 4-10. The saturation density of control vector was normalized to 1.0, and the relative values were reported for each experimental group. Data were represented as meanisd of triplicates. 151 Number of large colonies) (per 60 mm plate) > —NW#MO\\IW\O OOOOOOOOOO L 1 r 1 r l I J BM66D + control vector BM66D + BM66D + BM66D + SV40 ER SV40 ER + SV40 ER + BMI 1 Cell lines MET % Invasion 50 4O 30 20 BM66D + BM66D + BM66D + control vector SV40 ER SV40 ER + BM] 1 Cell lines BM66D + SV40 ER + MET Relative saturation density BM66D + BM66D + BM66D + control vector SV40 ER SV40 ER + BM] 1 Cell lines BM66D + SV40 ER + MET Figure 5.8. 152 S V40 32M M 2§§ 83 C Q D Q ‘1' [T E V- N E 2 53 § 83 2 2 3. E 332 Figure 5.9. Verification of transgene expression at mRNA level. RT-PCR was carried out to verify the expression of transgenes. PLA3-clone 7 (~20 cpdl) showed the expression of both BMII and MET (A), but these cells eventually underwent senescence. Expression of SV-40ER was retained in immortalized canine BM-MSCs (BM66-D), before (B) and after introduction of MET (C) or BMII (D). Expression of exogenously introduced MET (C) or BMII (D) transgenes were verified after their incorporation into SV-40 immortalized cells. M = 100 bp DNA ladder; B2M (flZ-MICROGLOBULIN) and GAPDH are house-keeping genes. M represents 100 bp DNA ladder, with the lowest band representing 100 bps. 153 ’h .. “;‘g‘fi 3M5? ”u, r .‘ki flg“. *a C 0“ ’3 9 ' 9 9 .43? 1.. 9' ti D. ’ m ' V 9. o .6 . s o... O S .{.. . O Q. Figure 5.10. Verification of transgene expression at the protein level. Exogenous expression of MET and BMII in MSCs was verified by immunostaining with human MET and BM] 1 antibodies. Canine MSCs expressed human MET (A) and BMII (B) proteins after transduction with pSCNIT-MET and pHIT-BMI 1 vectors. Expression of human MET and BMII proteins in SV40-immortalized and human MET or BMII transduced canine MSCs are shown in Figure 10C and 10D respectively. Magnification = 200x 154 10. ll. REFERENCES . Clarke MF, Dick JE, Dirks PB et al. Cancer Stem Cells-—Perspectives on Current Status and Future Directions: AACR Workshop on Cancer Stem Cells. Cancer Res 2006;66:9339-9344. Ward RJ, Dirks PB. Cancer stem cells: at the headwaters of tumor development. Annu Rev Pathol 2007;2:175-189. . Visvader JE, Lindeman GJ. Cancer stem cells in solid tumours: accumulating evidence and unresolved questions. Nat Rev Cancer 2008. Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer 2005;5z275-284. . Baumann M, Krause M, Hill R. Exploring the role of cancer stem cells in radioresistance. Nat Rev Cancer 2008;8:545-554. Reya T, Morrison SJ, Clarke MF et al. Stem cells, cancer, and cancer stem cells. Nature 2001;414:105-111. . Pardal R, Clarke MF, Morrison SJ. Applying the principles of stem-cell biology to cancer. Nat Rev Cancer 2003;3:895-902. . Jordan CT, Guzman ML, Noble M. Cancer stem cells. N Engl J Med 2006;355:1253-1261. Pardal R, Molofsky AV, He S et al. Stem cell self-renewal and cancer cell proliferation are regulated by common networks that balance the activation of proto-oncogenes and tumor suppressors. Cold Spring Harb Symp Quant Biol 2005;70:177-185. Barker N, Ridgway RA, van Es JH et al. Crypt stem cells as the cells-of-origin of intestinal cancer. Nature 2009;457:608-611. Zhu L, Gibson P, Currle DS et al. Prominin 1 marks intestinal stem cells that are susceptible to neoplastic transformation. Nature 2009;457:603-607. 155 12. l3. I4. 15. l6. 17. 18. 19. 20. 21. Jamieson CH, Ailles LE, Dylla SJ et al. Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML. N Engl J Med 2004;351:657- 667. Passegue E, Jamieson CH, Ailles LE et al. Normal and leukemic hematopoiesis: are leukemias a stem cell disorder or a reacquisition of stem cell characteristics? Proc Natl Acad Sci U S A 2003;100 Suppl 1:11842-11849. Cozzio A, Passegue E, Ayton PM et al. Similar MLL-associated leukemias arising from self-renewing stem cells and short-lived myeloid progenitors. Genes Dev 2003;17:3029-3035. Dai C, Celestino JC, Okada Y et al. PDGF autocrine stimulation dedifferentiates cultured astrocytes and induces oligodendrogliomas and oligoastrocytomas from neural progenitors and astrocytes in vivo. Genes Dev 2001;15:1913-1925. Uhrbom L, Dai C, Celestino JC et al. Ink4a-Arf loss cooperates with KRas activation in astrocytes and neural progenitors to generate glioblastomas of various morphologies depending on activated Akt. Cancer Res 2002;62:5551- 5558. Bachoo RM, Maher EA, Ligon KL et al. Epidermal growth factor receptor and Ink4a/Arf: convergent mechanisms governing terminal differentiation and transformation along the neural stem cell to astrocyte axis. Cancer Cell 2002; 1 :269-277. Park CY, Tseng D, Weissman IL. Cancer stem cell-directed therapies: recent data from the laboratory and clinic. Mol Ther 2009;17:219-230. Cooper CS, Park M, Blair DG et al. Molecular cloning of a new transforming gene from a chemically transformed human cell line. Nature 1984;311:29-33. F erracini R, Di Renzo MF, Scotlandi K et al. The Met/HGF receptor is over- expressed in human osteosarcomas and is activated by either a paracrine or an autocrine circuit. Oncogene 1995;10:739-749. F erracini R, Angelini P, Cagliero E et al. MET oncogene abenant expression in canine osteosarcoma. J Orthop Res 2000;18:253-256. 156 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. MacEwen EG, Kutzke J, Carew J et al. c-Met tyrosine kinase receptor expression and function in human and canine osteosarcoma cells. Clin Exp Metastasis 2003;20:421-430. De MR, Miretti S, Iussich S et al. met oncogene activation qualifies spontaneous canine osteosarcoma as a suitable pre-clinical model of human osteosarcoma. J Pathol 2009;218:399-408. Patane S, Avnet S, Coltella N et al. MET overexpression turns human primary osteoblasts into osteosarcomas. Cancer Res 2006;66:4750-4757. Lessard J, Sauvageau G. Bmi-l determines the proliferative capacity of normal and leukaemic stem cells. Nature 2003;423:255-260. Leung C, Lingbeek M, Shakhova O et al. Bmil is essential for cerebellar development and is overexpressed in human medulloblastomas. Nature 2004;428:337-341. Valk-Lingbeek ME, Bruggeman SW, Van LM. Stem cells and cancer; the polycomb connection. Cell 2004;1 18:409-418. Dimri GP, Martinez JL, Jacobs 1] et al. The Bmi-l oncogene induces telomerase activity and immortalizes human mammary epithelial cells. Cancer Res 2002;62:4736-4745. Haupt Y, Bath ML, Harris AW et al. bmi-I transgene induces lymphomas and collaborates with myc in tumorigenesis. Oncogene 1993;8:3161-3164. Zhang X, Soda Y, Takahashi K et al. Successful immortalization of mesenchymal progenitor cells derived from human placenta and the differentiation abilities of immortalized cells. Biochem Biophys Res Commun 2006;351:853-859. Young HE, Mancini ML, Wright RP et al. Mesenchymal stem cells reside within the connective tissues of many organs. Dev Dyn 1995;202:137-144. Caplan AI, Bruder SP. Mesenchymal stem cells: building blocks for molecular medicine in the let century. Trends Mol Med 2001;7z259-264. 157 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. Helman LJ, Meltzer P. Mechanisms of sarcoma development. Nat Rev Cancer 2003;3:685-694. Tirode F, Laud-Duval K, Prieur A et al. Mesenchymal stem cell features of Ewing tumors. Cancer Cell 2007;] 12421-429. Chung EB, Enzinger FM. Extraskeletal osteosarcoma Cancer 1987;60:1132- 1142. Bane BL, Evans HL, Ro JY et a1. Extraskeletal osteosarcoma. A clinicopathologic review of 26 cases. Cancer 1990;65:2762-2770. Patnaik AK. Canine extraskeletal osteosarcoma and chondrosarcoma: a clinicopathologic study of 14 cases. Vet Pathol 1990;27246-55. Thompson KG, Pool RR. Tumors of bone. In: Meuten DJ, ed. Tumors in Domestic Animals. Iowa State Press, Ames, IA 20022245-83. Barroga EF, Kadosawa T, Okumura M et a1. Establishment and characterization of the growth and pulmonary metastasis of a highly lung metastasizing cell line from canine osteosarcoma in nude mice. J Vet Med Sci 1999;61:361-367. Withrow SJ, Powers BE, Straw RC et al. Comparative aspects of osteosarcoma. Dog versus man. Clin Orthop Relat Res 1991;159—168. Vail DM, MacEwen EG. Spontaneously occurring tumors of companion animals as models for human cancer. Cancer Invest 2000;18:781-792. Hansen K, Khanna C. Spontaneous and genetically engineered animal models; use in preclinical cancer drug development. Eur J Cancer 2004;40:858-880. Mueller F, Fuchs B, Kaser-Hotz B. Comparative biology of human and canine osteosarcoma. Anticancer Res 2007;27:155-164. Straw RC. Tumors of the skeletal system. In: Withrow SJ, MacEwen GE, eds. Small Animal Clinical Oncology. W.B. Saunders Co. Philadelphia 1996:287-305. 158 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. Haydon RC, Luu HH, He TC. Osteosarcoma and osteoblastic differentiation: a new perspective on oncogenesis. Clin Orthop Relat Res 2007;454:237-246. Thomas D, Kansara M. Epi genetic modifications in osteogenic differentiation and transformation. J Cell Biochem 2006;98:757-769. Tang N, Song WX, Luo J et al. Osteosarcoma development and stem cell differentiation. Clin Orthop Relat Res 2008 ;466221 14-2 1 30. Zeng L, Rahrmann E, Hu O et al. Multipotent adult progenitor cells from swine bone marrow. Stem Cells 2006;24:2355-2366. Lin TM, Tsai JL, Lin SD et al. Accelerated growth and prolonged lifespan of adipose tissue-derived human mesenchymal stem cells in a medium using reduced calcium and antioxidants. Stem Cells Dev 2005;14:92-102. Neupane M, Chang CC, Kiupel M et al. Isolation and characterization of canine adipose-derived mesenchymal stem cells. Tissue Eng Part A 2008;14:1007-1015. Suhr ST, Senut MC, Whitelegge JP et a1. Identities of sequestered proteins in aggregates from cells with induced polyglutarnine expression. J Cell Biol 2001;153:283-294. Suhr ST, Gil EB, Senut MC et al. High level transactivation by a modified Bombyx ecdysone receptor in mammalian cells without exogenous retinoid X receptor. Proc Natl Acad Sci U S A 1998;95:7999-8004. Suhr ST, Gage FH. Gene therapy in the central nervous system: the use of recombinant retroviruses. Arch Neurol 1999;56:287-292. Litzkas P, Jha KK, Ozer HL. Efficient transfer of cloned DNA into human diploid cells: protoplast fusion in suspension. Mol Cell Biol 1984;422549-2552. Lin SL, Chang D, Ying SY. Hyaluronan stimulates transformation of androgen- independent prostate cancer. Carcinogenesis 2007;28:310-320. 159 56. 57. 59. 60. 61. 62. 63. 65. 66. Dimri GP, Lee X, Basile G et al. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci U S A 1995;92:9363-9367. Pipas JM. SV40: Cell transformation and tumorigenesis. Virology 2009;384:294- 303. . Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57-70. Yamanaka S. A fresh look at iPS cells. Cell 2009;137:13-17. Matushansky I, Hernando E, Socci ND et al. A developmental model of sarcomagenesis defines a differentiation-based classification for liposarcomas. Am J Pathol 2008;172: 1069- 1 080. Bruggeman SW, Hulsman D, Tanger E et al. Bmil controls tumor development in an Ink4a/Arf-independent manner in a mouse model for glioma. Cancer Cell 2007;12:328-341. Cobaleda C, Jochum W, Busslinger M. Conversion of mature B cells into T cells by dedifferentiation to uncommitted progenitors. Nature 2007;449:473-477. Shima Y, Okamoto T, Aoyama T et al. In vitro transformation of mesenchymal stem cells by oncogenic H-rasVal 12. Biochem Biophys Res Commun 2007;353:60-66. . Funes JM, Quintero M, Henderson S et al. Transformation of human mesenchymal stem cells increases their dependency on oxidative phosphorylation for energy production. Proc Natl Acad Sci U S A.2007;104:6223-6228. ntillon-Klussmann F, Garcia-Delgado M, Villa-Elizaga I et al. Mutational activation of ras genes is absent in pediatric osteosarcoma. Cancer Genet Cytogenet 1995;79:49-53. Hahn WC, Weinberg RA. Rules for making human tumor cells. N Engl J Med 2002;347:1593-1603. 160 67. 68. 69. Itahana K, Zou Y, Itahana Y et al. Control of the replicative life span of human fibroblasts by p16 and the polycomb protein Bmi-l. Mol Cell Biol 2003;23:389- 40 l . Tanimoto A, Kao CY, Chang CC et al. Deregulation of cdc2 gene expression correlates with overexpression of a 110 kDa CCAAT box binding factor in transformed cells. Carcinogenesis 1998;19:1735-1741 . Akagi T, Sasai K, Hanafusa H. Refractory nature of normal human diploid fibroblasts with respect to oncogene-mediated transformation. Proc Natl Acad Sci U S A 2003;100:13567-13572. 161 CHAPTER 6 Evaluation of OCT4 expression as a stem cell marker in canine osteosarcoma 162 SUMMARY Osteosarcoma (OS) is the most common primary bone tumor of hmnan and dogs. The tumor is highly resistant to conventional chemo-and radiotherapies, and is composed of a varying proportion of undifferentiated and differentiated cell types of mesenchymal lineages. These attributes make OS as a plausible candidate for being a cancer driven by stem cells. Cancer stem cells (CSCs), cells comprising the tumor-initiating fraction of a tumor, have been proposed to express pluripotency-associated transcription factor, OCT4. The purpose of this study was to test the validity of OCT4 expression as a marker for tumor initiating subset of canine OS, using the OS cell line D-17. OCT4 reporter clonal cell lines were established from D-17 canine OS cells, by introducing a vector with an OCT4 promoter fused to a yellow fluorescent protein (YFP) gene. Expression of YFP was used to sort the cells into OCT4 positive and negative fractions. These cells were assessed for endogenous OCT4 activity, and extensively characterized for their in vitro phenotypes and gene expression profile. Tumorigenic potential of these cell fractions were assessed by injecting various number of OCT4 positive and OCT4 negative as well as parental cells into highly immunodeficient mice. Our data show that heterogeneous parental cell population was more tumorigenic than discrete populations of cancer cells sorted on the basis of OCT4 expression. However, only the cells that were positive for OCT4 promoter activity recapitulated the OS tumor phenotype. These findings suggest that heterogeneity enhances the ability of these cells to initiate tumors in the xenotransplantation model, and use of OCT4 expression in combination with other CSCs markers might enable to identify enriched population of C SC 5 in OS. 163 INTRODUCTION Despite being clonal in origin [1], tumor cells within a single tumor are functionally heterogeneous in terms of their ability to proliferate, differentiate, and propagate a new tumor [2]. Two models have been proposed to explain this tumor heterogeneity and differences in tumorigenic capacity among the tumor cells: clonal evolution model and hierarchical model. Clonal evolution model [3,4] posits that tumor heterogeneity arises from continuous selection and expansion of tumor cells with growth advantage (dominant clones), and any tumor cell is capable of generating a new tumor given the right rrricroenvironment. Hierarchical model [5,6] follows that tumor heterogeneity arises from a specific subset of cells which maintains the cellular hierarchy within a tumor, and tumor-transmitting ability is restricted to this defined subpopulation of cancer cells, the so-called “cancer stem cells” (CSCs). Cancer stem cells are defined as the “cells within a tumor that possesses the capacity to self-renew and to cause the heterogeneous lineages of cancer cells that comprise the tumor” [7]. CSCs have been documented in several solid cancers including brain, breast, prostate, melanoma, lung, colon, head and neck, and pancreatic cancers (reviewed in [8,9]). Many of these studies have shown that these CSCs are generally resistant to conventional chemotherapy and radiotherapy regimens and are responsible for relapse of tumor after completion of conventional therapy [10-13]. Osteosarcoma is the most common primary bone tumor of humans and dogs [14,15]. It is a highly malignant tumor with poor clinical prognosis. OS accounts for 4% of all malignancies, 85% of all skeletal malignancies, and 98% of appendicular primary. 164 canine bone tumors [14]. Canine OS closely resembles human OS in terms of histopathological appearance and biological behavior [16], and is an excellent model to study tumor biology and therapeutic intervention [17-19]. There is a greater genetic homology between dogs and humans than between either species and mouse [20]. In addition, dogs live in the same environment as human, and naturally occurring canine tumors reflect the genetic heterogeneity noticed in human patients (reviewed in [19]). Osteosarcoma is quite resistant to conventional chemotherapy and radiotherapy, and contains varying proportion of undifferentiated cells, osteoblasts, chondroblasts, and fibroblasts [21]. Increasing evidence suggests that OS is a disease of differentiation caused by genetic and epigenetic changes that interrupt the process of osteoblastic differentiation [22,23]. All of these attributes make canine OS as a likely candidate for cancer derived from and/or driven by stem cells. The current study was undertaken to examine the validity of a candidate C SC marker, OCT4, as a marker of the stem cell fraction of canine OS, a highly malignant canine tumor that is a major cause of mortality and morbidity. Traditional cell sorting strategies rely on the expression of cell-specific marker on the cell surface and availability of suitable antibody for its detection. Most of the currently used markers to enrich and prospectively isolate tumor stem cells are cell surface-based [9], and they do not have functional relevance to initiation and propagation of tumor. Thus, there is an acute need for markers that have a mechanistic basis and causal relationship to the process of carcinogenesis. Furthermore, only a limited number of validated antibodies are available against these cell surface markers for the dog. Several studies have suggested OCT4 (also known as OCT 3 /4, POU5F1) as a candidate marker for CSCs. This 165 transcription factor is implicated in the pluripotency of stem cells. It is expressed in embryonic stem cells [24,25], several adult stem cells [26-29], and many cancers [26,30- 40], including bone sarcomas in human [41]. Canine OS cell lines have been shown to expresses the message for OCT 4 [42]. In addition, our research group has also identified OCT4 positive cells in various canine tumors, including OS using immunohistochemistry [35]. Genomic fusion between OCT4 and EWSRI (Ewing Sarcoma Region) has been associated with a case of a human bone tumor, indicating OCT4 reactivation [43]. In another study, forced expression of OC T4 was sufficient to transform non-tumorigenic Swiss 3T3 fibroblasts into high grade fibrosarcoma—producing cells [44]. However, direct evidence for the role and usefulness of OCT4 in cancer stem cells has not yet been generated-leading to much controversy in this area [38,45-50]. We think that this issue can only be resolved through rigorous study of pure population of OCT4 expressing cells. But, OCT4 is a transcription factor and unlike other cell surface markers, OCT4 expressing tumor cells cannot be fractionated by conventional antibody based flow-sorting approach. This difficulty in prospective isolation of OCT4 expressing cells will prevent the purification of these cells from heterogeneous tumor cell population, which would limit the usefulness of this marker to a restricted subset of studies. One of the feasible approaches would be to establish reporter cell lines whereby expression of reporter marker(s) reflects the activity of endogenous OCT4 promoter. Introduction of fluorescent reporter genes under the control of cell type-specific promoters is considered as one of the best methods to identify and enrich particular progenitor or specialized cell types [51]. Therefore, we used fluorescent reporter of 166 OCT4 activity in order to isolate tumor stem cell fraction from canine OS. Reporter gene- based approach can be extrapolated to other functional markers as well as several other trunor types to isolate and evaluate tumor stem cell fraction. There are two main strategies to generate reporter cell lines: (i) random integration of plasmid containing reporter gene driven by the promoter of gene of interest into the particular cell line(s); (ii) homologous recombination to ‘knock-in’ the reporter gene into the endogenous locus of gene of interest into the particular cell line(s). We have Opted for the first strategy, given the fact that randomly integrated transgenic reporters have been effectively used to tag human embryonic stem cells [52,53] as well as their differentiated derivatives [54—56]. Using this approach, Levings et al [57] have recently identified tumor initiating cells in human OS. 167 MATERIALS AND METHODS Materials: All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise stated. Construct details: PhOCT4-Venus plasmid (obtained from Dr. Jose Cibelli, Michigan State University) was constructed by replacing CMV promoter and Far-Red fluorescent protein (HcRedl) gene with human 2.8 Kb 0C T 4 promoter (-2800 to +44, relative to the transcription start site) and ‘Venus’ (an improved version of YFP) in the backbone of pHcRedl-Nuc plasmid (Clontech). This construct has neomycin resistance gene under the control of SV40 early enhancer, which allows us to establish both OCT 4 promoter positive (if any) and OCT 4 promoter negative stable cell lines after G418 (Invitrogen Corporation, Carlsbad, CA) selection. 0C T 4 promoter is highly conserved across several species, and canine OCT4 has 89% similarity at nucleotide level and 91% similarity at amino acid level with its human counterpart. Establishment of OCT4 reporter ceIlIines: D-17 canine OS cell line was purchased from ATCC (Manassas, VA) and was maintained in Minimum Essential Medium-alpha (Ir-MEM) (Invitrogen) with 10% F BS. D-17 cells were plated in 24-well plate at 2x 105 cells per well and transfected by ApaLI- linearized phOCT4-Venus plasmid, using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. To get stable transfectants, transfected cells were passaged at 1:10 and selected with 400 ug/ml G418 (Invitrogen) afier 48 hrs for 10 days. Individual resistant clones were picked up, monitored for fluorescence expression, and 168 expanded in the presence of 200 ug/ml G418. Stable cell lines were established at clonal level, and re-cloned twice to ensure the clonal origin of these subpopulations. Since the construct would randomly integrated into the host cell genome, several cell clones were analyzed to pick up the right clonal cell line which faithfully reflects the endogenous OCT4 activity. Validation of endogenous OCT4 activity of reporter cell lines: Total RNA was isolated from several sub-clones of YFP-positive and YFP- negative D-17 cells as well as from parental D-17 cell line, using Versagene RNA isolation kit (5 Prime Inc, Gaithersburg, MD). RN As were treated with TURBO DNA- fiee DNase I (Ambion, Austin, TX) to remove any contaminating genomic DNA One ug total RNA was reverse transcribed in 30 ul reaction volume, using 10 units of Superscript III reverse transcriptase (Invitrogen, Carlsbad,CA), and final concentrations of 5 uM anchored oligo-dT primers, 0.5 mM dNTPs, 5 mM DTT, and 1x first strand buffer supplied by the manufacturer (Invitrogen, Carlsbad,CA) . The reverse transcription reaction was performed at 50°C for 1 hr, and then inactivated at 70°C for 15 min. Twenty-five ul PCR reactions were prepared with 2rd cDNA, 5 pmol of each primer, 0.5 units of Taq polymerase (Invitrogen, CA), and final concentrations of 40 uM dNTPs, 2 mM MgC12, 20 mM Tris-HCl, and 50 ul KCl. Cycling conditions were as follows: 94°C for 4 min; 35 cycles at 94°C for 1 min, 60°C for l min, 72°C for 1 min; followed by 72°C for 5 min. The PCR products were separated on 2% agarose gel by electrophoresis, stained with ethidium bromide, visualized under UV light, and digital images captured with Alphalmager software (San Leandro, CA). Endogenous expression 169 of OCT4 mRNA was examined with primers specific to Exon 1 of canine OCT 4, using BZ-microglobulin as a house-keeping gene. One set of intronic primer that would amplify OCT4 genomic DNA, but not OCT4 cDNA was also included to further rule out the presence of any contaminating genomic DNA. cDNA from canine testis was used as a positive control to validate the OCT4 primer, and no template control (water instead of cDNA) was used as a negative control. Primer sequences used to validate endogenous OCT4 expression are listed below: OCT4 Exon-1 primer: Forward = 5 '-GAAGCAGAAGAGGATCACCCTA-3 '; Reverse = 5'-CCGCAGCTTACACATATTCTTG-3' (Product size = 145 bp) QQT4 ingonig primer; Forward = 5'-TTTCCTAAGTGCCTGGCTGT-3 '; Reverse = 5'-ACTTCGGCTCAGTTCATGCT-3' (Product size = 251 bp) Maintenance of cell lines: Parental D-I 7 cell line was maintained in u—MEM with 10% FBS, and YF P positive and negative cell lines (with integrated reporter plasmid) were maintained in the presence of G418 (200 jig/ml) in u-MEM with 10% FBS. Gene expression studies: Total RNA was isolated and reverse transcribed as described above. The cDNAs were purified by Qiaquick PCR purification kit (Qiagen) before using them for real time PCR. Validation of primers: All primers were designed using Primer 3 sofiware [58]. Primers derived from coding regions of respective genes in the canine genome were used to 170 amplify the target sites (Table 6.1). All of the primers were first validated with cDNAs synthesized from canine testis tissue RNAs and/or with canine genomic DNA. Presence of single amplicon of expected product size was verified by agarose gel electrophoresis following conventional PCR. Before using AACT method, reaction efficiencies were determined by standard curves generated from total RNA (or genomic DNA). A ten-fold serial dilution of 100 ng total RNA (or genomic DNA) at five different points confirmed that both the slope and R2 values were close to the theoretical values (slope= -3.8 to -3.3; R220.99). A validation experiment was done to demonstrate the similarities in efficiency of target amplification and reference amplification (slope difference <0.1). Dissociation curve analysis was performed to confirm the specificity of each product and absence of primer-dimer complexes. Real time PCR: Both OCT4 promoter activity-positive and OCT4 promoter activity- negative clonal cell lines as well as parental D-l7 cell line (original cell lines as well as those re-established from xenograft) were used for quantitative RT-PCR studies. Quantitative RT-PCR was carried out with 2X SYBR Green master mix (Applied Biosystems, Foster city, CA) and 100 nM primers in 15 ul reaction using AB] 7700 sequence detection system (Applied Biosystems, Foster city, CA) on the thermal profile of 50°C for 2 min, 95°C for 10 min, 40 cycles at 95°C for 15 sec, 60°C for l min. All reactions were performed in triplicate, and reaction mixture with RNA instead of cDNA (no reverse transcriptase control) was used in each run to ensure the absence of genomic DNA contamination. Expression of each target gene was normalized against ,62- MICROGLOBULIN (endogenous reference gene), and relative quantification was carried (Z'MCT method) by comparing the expression against D17 parental cell line [59,60]. 17] Immunostaining: Cytospin preparations were made from YFP positive and YFP negative clonal cell lines as well as parental D-l7 cell line. Cytospin cells were stained with mouse anti- human monoclonal OCT4 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) as previously described [35]. Characterization for tumorigenic phenotypes in vitro: (i) Determination of population doubling time and saturation density: To determine the growth properties, 1 x 10" cells were seeded in each of multiple 60-mm dishes on day 0, and grown in a-MEM with 10% FBS. For doubling time (DT) calculation, the cultures were re-fed every 3 days, and cells were counted fi'om triplicate dishes on days 2-10. The numbers of cells per 60-mm dish were plotted against the days of culture, and the slope of the linear region of each growth curve was used to calculate the doubling time. For saturation density determination, the cultures were re-fed every day for 10 days, and cells were counted from triplicate dishes on days 4-10. The saturation density was reported as the maximum number of cell per cmz. (ii) Anchorage independent growth: 5 x 104 cells in 3 ml of 0.33% agarose medium were plated on top of 3 ml pre-hardened 0.5% agarose medium in each triplicate 60 mm dishes with grids to aid colony counting. 2.5 ml liquid medium u-MEM with 10% FBS) was then added and renewed once every 3 days. At the end of 2 weeks, the numbers of anchorage independent colonies developed were scored under a microscope. Colonies were stained with 0.005% crystal violet and photographed. 172 (iii) Matrigel invasion assay: For invasion assay, 5 x 10" cells were resuspended in 500 pl of serum free a-MEM medium, and seeded on the upper side of Matrigel Matrix (with 8 pm pore size PET membrane) of each well of 24-well cell culture insert (BD Biosciences, San Jose, CA). The lower chamber of the transwell insert were filled with 750 pl a-MEM containing 10% FBS as a chemoattractant. After 24 hrs of incubation, the filters were removed and cells that invaded the Matrigel and attached to the lower chamber of the transwell were fixed with 100% methanol for 10 min, and stained with 0.2% cresyl violet for 20 min. Cresyl violet stain on the membranes was eluted with 100% ethanol/0.2 M sodium citrate (1:1), and absorbance was read at 570 nm using SpectraMax M5 microplate reader (Molecular Devices, Sunnyvale, CA). The percentage of invading cells was calculated by comparison of absorbance in cells attached to the underside of the membrane (topside wiped with cotton wool) against absorbance of total cells in the membrane (topside not wiped) [61]. (iv) Serum dependence assay: Cells were seeded at 2 x 105 per well in 24-well plates in a-MEM medium with 10% FBS. After overnight attachment, the cells were washed twice with PBS, and cultured in low-serum (0.1% FBS) or serum-free medium. Cell proliferation and viability was monitored for 2 weeks by counting the cells from triplicate wells with a hemocytometer using trypan blue. (v) Chemoresistance and Radioresistance: Chemosensitivity assay: Cells were seeded in 96-well plates at 4000 cells per well in 100 pl a-MEM with 10% FBS, and cultured for 24 hrs. Five replicates of each experimental group were tested with doxorubicin (1.5 pg/ml) and cisplatin (4 rig/ml), 173 using 0.1% DMSO as a vehicle control [62]. Cell viability was determined 48 hrs later by a colorimetric 3-(4,5-dimethylthiazol-Z-yl)-5-(3 carboxymethoxyphenyl)-2-(4- sulfophenyl)-2H-tetrazolium, inner salt (MTS) cell proliferation assay [CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (MTS), Promega, Madison, WI] according to the instructions of the manufacturer (Promega). Briefly, 20 ul of substrate solution was added to each well, and the cells were incubated for 2 hrs at 37°C. Absorbance was measured at 490 nm wavelength for each well using Spectramax M5 microplate reader (Molecular Devices, Sunnyvale, CA). Drug resistance was represented as percentage viability which was calculated as: (mean absorbance of the test well) / (mean absorbance of the control) X 100%. Radiosensitivity assay: Cells were seeded in triplicate in 6-well plates (100 cells per well for 0 and 2 Gy dose, 200 cells per well for 5 Gy dose, and 400 cells per well for 10 Gy dose), and cultured for 24 hrs in a-MEM with 10% FBS. Early passage neonatal fibroblasts from two SCID dogs (with mutation in DNA-PKC) and one of their normal littermate (gift from Dr. Kathy Meek, Michigan State University) were included for assessing the intrinsic radiosensitivity of canine 08 cells. Radiation survival was defined as the ability of cells to maintain clonogenic capacity and form colonies after radiation exposure. Cells from each experimental group were treated with single graded dose (300 cGy/min) of X-ray radiation (0, 2, 5, 10 Gy), using a 6 MV linear accelerator (Clinac 2100 C, Varian Inc., Palo Alto, CA) and incubated for 10 days (until most cell clones reached >50 cells, but colonies were still distinctly apart). Crystal violet (1%) staining was carried out to examine the number of surviving colonies, and survival curve was 174 established for each cell line [63,64]. Results are expressed as mean :1: sd of triplicate wells. Hoechst 33342 staining and side population (SP) phenotype: Parental D-17 cells and three clones each of OCT4P+ and OCT4P- D-l7cells were trypsinized and resuspended at 1 X 106 cells/ml in pre-warrned a—MEM (Invitrogen) with 2% FBS, stained with Hoechst 33342 dye at a final concentration of 3 rig/ml in the presence or absence of 100 uM veraparnil (added 15 min before Hoechst staining), and incubated at 37°C for 70 min in dark with intermittent mixing. To detect SP, cells were analyzed on FACSVantage SE (Becton Dickinson, San Jose, CA) by excitation of Hoechst dye with 350 nm UV laser, and simultaneous detection of blue and red fluorescence (blue, 425-475 nm; red, 660 nm). Dead cells and cellular debris were excluded based on forward and sideward scatter signals. The SP gate was defined as the diminished region on the dot plot in the presence of verapamil. Twenty to fifty thousand live cells were analyzed from each cell population. Purity following clonal expansion and flow cytometry: Both OCT4P+ and OCT4P- stable clonal cell lines were sorted by FACS immediately before use for gene and protein expression studies. For in vivo study, these flow sorted cells were plated and detached after overnight attachment in tissue culture dishes. Cells were trypsinized, resuspended in PBS, and sorted on the basis of YFP expression on a F ACSVantage SE (BD Biosciences, San Jose, CA). Dead cells and cellular debris were excluded on the basis of scatter signals. Bright fluorescence was detected through 530/30 band pass filter after excitation with 488 nm argon laser (100 175 mW), and cells were sorted directly into 6-well plate using 130 um nozzle with a high purity protocol. These cells were detached after 24 hrs, and used for in vivo tumorigenesis study (see below). Keno-transplantation and limiting dilution assay: The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Michigan State University. Cells (parental D-17 and one clone each of FACS purified clonal cell line of OCT4P+ and OCT4P- D-l7) were trypsinized, washed twice with PBS, and resuspended in 1:1 mixture of media and Maui gel (BD biosciences) (in a total volume of 0.2 ml) in aliquots of 5X 10’, 5X105, 5X10", 5X 103 , 5X102 and 1X 102. Five replicates of each dose group were subcutaneously injected into the hind flank of 6-8 week-old female NOD/SCID/ILZRy" mice (Jackson Laboratory, Bar Harbor, ME) under isoflurane anesthesia. The mice were monitored twice a week for palpable tumor formation for 20 wks time period. When tumors reached approximately 1 cm in diameter, these mice were euthanized and tumor samples were further processed for gene and protein expression studies, evaluation by histopathology as well as for re- establishment of cell lines. Histological studies: Tumor samples were fixed in 10% neutral buffered formalin and processed for light microscopy. S-um sections were stained with H&E, Masson’s Trichome and Von Kossa stains, and selected sections of the tumors were stained with OCT4 antibody (Santa Cruz, CA). Re—establishment of canine OS cell lines: 176 Cell lines were re-established from all experimental groups displaying tumors. Briefly, tumor pieces were minced with scalpel blade, and digested at 37°C for 2 hours with collagenase type IA (1 mg/ml) in a—MEM with 10% FBS, 100 U/ml penicillin, 100 rig/ml streptomycin and 0.25 ug/ml amphotericin. Cells were washed twice with a-MEM containing10% FBS, filtered through a 70 um nylon mesh cell strainer (BD biosciences, San Jose, CA), and plated in 100 mm plates. Unattached cells were removed the next day by washing with PBS. Adherent cells were expanded in a-MEM with 10% FBS, subcultured, and cryopreserved for further studies. YFP-positive and YFP-negative cell lines were maintained in the presence of 200 ug/ml G418. Statistical analysis: All quantitative data were presented as the mean :1: standard deviation. Statistical significance was determined by paired t test, or one-way ANOVA followed by Tukey’s multiple comparison test, using SigmaStat version 2.03 (SPSS Inc., Chicago, IL)-as appropriate. A p-value of less than 0.05 was considered significant. For limiting dilution assays, the frequency of tumor-initiating cells was calculated on WEHI web interface (hmz/mioinfwehi.edu.au/software/elda/index.html) using the method described by Hu and Smyth [65]. 177 RESULTS Establishment of OCT4 reporter cell lines: Several clones of OCT4 reporter cell lines were established from D-I7 canine OS cells. We were able to select and expand YFP positive (and therefore, OCT4P+) (Figure 6.1A,B) and YFP negative (and therefore, OCT4P-) (Figure 6.1C,D) clonal lines, both of which remained neomycin resistant and maintained the fluorescence expression pattern after several passages. We have observed that YFP positive cell lines of clonal origin display the heterogeneity in terms of fluorescence. Despite their origin from single YFP positive cell, some of the descendant cells within these clones lose fluorescence expression. We have consistently noticed this phenomenon across several clones and even after repeated subcloning, the frequency of YFP positive cells within any given clone varied from 75% to 96% (see FACS plot) (Figure 6.1E-H). However, we have not seen the converse event i.e. appearance of YFP positive cells within the YFP negative clones. So, it seems that there exists a hierarchy of differentiation within these clones, and OCT4 positive putative CSC is able to divide asymmetrically and give rise to more differentiated descendants that would lose OCT4 (and, therefore YFP) expression. Expression of endogenous OCT4 mRNA in reporter cell lines: RT-PCR results revealed that YFP expression fairly reflected the endogenous OCT 4 expression, although there was some difference among clonal cell lines in terms of levels of expression (Figure 6.2). Based on these results, YFP-positive (therefore, OCT4 178 promoter activity-positive) clone 4.2 and YFP-negative (therefore, OCT4 promoter activity-negative) clone 3 were chosen for further studies. Evaluation of tumorigenic phenotypes in vitro: Cell growth analysis: As shown in Table 6.2, cells from OCT4P+ groups were found to divide slightly faster than OCT4P- cells and parental cells (p<0.001; Tukey’s post hoc test). Similarly, OCT4P+ cells had slightly higher saturation density than OCT4P- cells, but the difference was not statistically significant (p =0.104; one-way ANOVA). Our results indicate that there was no considerable difference in cell growth properties among the experimental groups. Serum dependence assay: When grown in low serum (0.1%) or serum-free medium, majority of the cells from parental group or OCT4P+ groups died by day 12 (Table 6.2). Intriguingly, OCT4P- cells not only remained viable, but also underwent firrther cell division even in the complete absence of serum (plated cell number increased from 2 X 105 to 5.2 X 10:5 per well in average), indicating their significantly lesser dependence on serum (p<0.001; Tukey’s post hoc test). Anchorage independent growth (AIG): The cells were tested for their ability to form anchorage-independent colonies in soft agar. Cells from all three experimental groups were able to grow in soft agar (Parental D-17 =59.08%, OCT4P+ =58.18% OCT4P- =54.08%) at comparable frequencies (Figure 6.3A-G). However, very few colonies from D-17 OCT4P-groups were larger than 100 pm (0.32%), in contrast to the colonies from D-l 7 OCT4P+ (3.97%) 179 and parental D-17 (13.2%) groups (p<0.001; Tukey’s post hoc test) (Figure 6.3G). These data indicate that although OCT4P- cells can grow at high frequency in soft agar, the colony sizes were always significantly smaller than those formed by parental cell line or OCT4P+ cells. Matrigel invasion assay: To investigate the possible differences in invasiveness among parental, OCT4P+, and OCT4P- cells, we performed matrigel invasion assay by adapting the method reported by Lin et al [61]. We first validated the assay using HT-1080 human fibrosarcoma cells as a positive control and NIH-3T3 mouse embryonic fibroblast cells as a negative control (data not shown). As shown in Figure 6.4, we found that both OCT4P+ and OCT4P- cells were significantly more invasive than parental D-17 cells (p<0.001and p<0.05 respectively; Tukey’s post hoc test). Although OCT4P+ cells were slightly more invasive than OCT4P- cells, the difference was not statistically significant (p>0.05). Chemosensitivity assay: We performed MTS assay, and examined the viability of cells from all experimental groups after 48 hrs of treatment with cisplatin and doxorubicin, the two commonly used drugs in the treatment of canine OS. Although there were statistically significant differences in survival advantage among these groups after treatment with both drugs (p<0.001; Tukey’s post hoc test), the difference seemed too little to be biologically significant (Figure 6.5). Radiosensitivity assay: The difference in sensitivity among unfiactionated, OCT4P+, and OCT4P- cells were measured by clonogenic survival assay. We counted the colonies after 10 days of 180 exposure to single graded dose of radiation, and determined the colony forming efficiency of each treatment group. All three experimental subgroups were significantly more radioresistant than normal canine fibroblasts (p<0.001), which were in turn more radioresistant than intrinsically radiosensitive SCID fibroblasts (see Figure 6.6). Parental cell line exhibited somewhat higher level of radioresistance than either OCT4P+ or OCT4P- subgroups at 2 and 5 Gy (p<0.05; Tukey’s post hoc test). On the other hand, difference in OCT4 promoter activity did not show any meaningful differences in terms of their clonogenic survival response following radiation treatment (p>0.05; Tukey’s post hoc test). SP phenotype: Cells capable of actively extruding Hoechst dye (SP cells) have been shown to be enriched for CSC phenotype [66]. We, therefore, asked whether the cells sorted on the basis of OCT4 promoter activity can enrich for SP phenotype or not. Our results (based on 3 clones each of OCT4P+ and OCT4P- cells) showed that the fi'equency of SP cells in OCT4P+ population was enriched by ~5 fold when compared to parental D-17 cells (0.51% i 0.008 % in OCT4P+ cells versus 0.1% in unsorted parental cells) (p<0.001), and ~10 fold when compared to OCT4P- cells (0.51% :h 0.008 % in OCT4P+ cells versus 0.048% i 0.005 % in OCT4P- cells) (p<0.001) (Figure 6.7). Xenotransplantation assay: To determine the frequency of tumorigenic cells for each injection group, we injected the cells in decreasing numbers in immunodeficient mice (Table 6.3), and performed limiting dilution analysis. We provided more permissive environment for 181 potential tumor initiating cells by injecting the cells with Matrigel in NOD/SCID/IL2R7'/' (NOG) mice. Injection of Matrigel-mixed tumor cells into this mouse model has been found to improve the sensitivity of xenotransplantation assay [67]. We found that tumorigenic frequency for parental cell group was 1/713 [95% confidence interval (CI), 1/2101-1/242]. Similarly, the frequency was 1/ 1596 [95% CI, 1/4321-1/590] for OCT4P+ cell group, andl/ 15425 [95% CI, 1/42717-1/5570] for OCT4P- cell group. Based on these data, the estimated tumorigenic cell frequency for parental cell group was about 2.2 times higher than OCT4P+ cells (Chi sq = 1.01; P>chi sq = 0.315) and about 22 times higher than OCT4P- cells (Chi sq = 19.1; P>chi sq = 0.00124). OCT4P+ cells were about 10 times more tumorigenic than OCT4P- cells (Chi sq = 10.5; P>chi sq = 0.00118). Although cells from all experimental groups were tumorigenic in immunodeficient mice (Figure 6.8A-F) tumor behavior was considerably different among these groups. Parental cells produced tumors with very short latency whereas both OCT4P+ and OCT4P— cells produced tumors with relatively longer latency (see Table 6.3). We noticed tumors from all experimental groups at dose as low as 500 cells, but the latency period for tumor development was considerably longer in all cases. Interestingly, injection of both parental cell line (Figure 6.8G) and OCT4P- (Figure 6.8K) cells induced highly anaplastic sarcoma without evidence of characteristic features of OS. On the other hand, OCT4P+ cells were able to induce tumors with histological features of OS (Figure 6.81). Tumors induced by parental cells (Figure 6.8H) and OCT4P+ cells (Figure 6.8J) were highly invasive and locally destructive whereas tumors formed by OCT4P- cells were highly invasive with both gross and histological evidence of distant metastasis (Figure 6.8L,M). 182 Gene expression studies: We examined the gene expression profile of all three experimental subgroups of parental cells as well as OCT4P+ and OCT4P- cells using a panel of stemness-related genes, including pluripotency-associated gene-NANOG, self-renewal-associated gene BMII, NOTCH family genes-NOT CH1, DLL4, HE Y1, and JA GI, sonic hedgehog family gene-GL1], WNT family genes-LRPS and WIFI, telomerase catalytic subunit T ERT , anti- apoptotic gene BCL2, minichromosome maintenance protein-MCM7, and multi-drug resistance ABC transporters-ABCG2,MDR1,MRP1, and ABCA3. As shown in Figure 6.9 and Figure 6.10, there were marked variations among the sub-groups in terms of expression of different genes. When compared to parental cells, notch ligands-NOTCH] and DLL4 were significantly upregulated in OCT4P+ cells (> 5 fold, and >2 fold respectively) and even more upregulated in OCT4P- cells (~35 fold, and >9 fold respectively) (p<0.001 for both cases), whereas another ligand JA G] was significantly downregulated in OCT4P+ cells (~35 fold) (p<0.001). Similarly, expression of HE Y1, a downstream effector of Notch signaling pathway, was reduced in both OCT4P+ (~2.5 fold) and OCT4P- cells (~3 fold) (p<0.001 for both). On the other hand, GL1 I -a downstream component of Shh pathway was expressed at significantly higher level in both of these subgroups (>8 fold in both cases) (p<0.001). Expression of Wnt co-receptor LRP5 was significantly upregulated in OCT4P+ group when compared to both parental (~9.5 fold) and OCT4P- groups (p<0.001). It was interesting that mRNA expression of anti-apoptotic regulator BC L2 was significantly downregulated (> 200 fold) in cells sorted on the basis of OCT4 expression (p<0.001 for both). TERT expression was significantly higher in OCT4P+ group (~4 fold), whereas expression of proliferation 183 marker MCM 7 was somewhat lower in OCT4P+ cells (~1.6 fold) (p<0.05 for both genes) when compared to both OCT4P' cells and parental cells. Out of 4 different ABC transporters we examined, expression of MDRI was significantly higher (~24 fold) in OCT4P- group when compared to other two subgroups (p<0.001). We re-established the cell lines fiom ttunors developed in immunodeficient mice, and monitored for the expression of fluorescence. As shown in Figure 6.11, cells isolated from tumors developed in OCT4P+ experimental group remained YFP positive, whereas those isolated from tumors developed in OCT4P- cells remained YFP negative. Further characterization of these re-established cell lines will be pursued in future studies. We then asked how the expression profile of these genes change in tumor cells re- established fi'om xenograft when compared to the original cells used for xenotransplantation. The results are shown in Figure 6.12. Many of the genes showed variable extent of changes following xenotransplantation. In parental D-l 7 cells, expression of HE Y] was dramatically downregulated (more than 40 fold) following transplantation in mouse (p<0.001), whereas expression of BMI] was considerably upregulated (~10 fold; p<0.001). In OCT4P+ cells, JAG] expression was reduced by about 8 folds. Similarly, WIFI expression was reduced by about 15 fold in case of OCT4P+ cells whereas its expression was undetectable in parental and OCT4P- cells (p<0.001). On the other hand, re-established OCT4P- cells had significantly downregulated the expression of DLL4, HE Y1 , and MDRI (p < 0.001). Interestingly, expression of NANOG gene, which was significantly upregulated in original OCT4P- cells, could not be detected in cells re—established from OCT4P- tumors. Expression of T ERT was lesser in all three subgroups (p<0.001) whereas MC M 7 expression was 184 increased in all of them (p<0.001). Among the other ABC transporters, ABCG2 expression was upregulated and MRP1 was downregulated. Expressions of other genes such as LRP5, GL1], NOTCH], BCL2, and ABCA3 were not remarkably altered. 185 DISCUSSION The identification of CSCs in mesenchymal tumors in general and, OS in particular, has remained elusive in comparison to tumors of hematopoietic, neural, and epithelial origin. This is partly due to the lack of reliable marker to identify and enrich primitive cell types of mesenchymal lineages [68]. Moreover, most of the currently used markers for enriching CSCs lack causal relationship with stemness. This underscores the necessity of firnctional markers for directly validating their roles in CSC function. One recent study has successfully used low 268 proteasome activity to isolate CSCs from human breast and brain cancer cell lines, demonstrating the usefulness of functional marker-based approach [69]. In addition, Levings et al [57] have recently demonstrated the validity of this approach for identification of tumor initiating cells in human OS. OCT4 expression has been proposed as a candidate C SC marker, and its expression has been documented in both human and canine bone tumors [35,41,42]. These findings prompted us to examine whether OCT4 activity is a useful functional marker to identify C SC in canine OS, which has been regarded as an excellent model system to study the biology of spontaneously occurring OS [17-19]. To monitor OCT4 activity in live cells, we established OCT4 reporter cell lines by introducing a vector with an OC T4 promoter fused to a gene for fluorescent protein. This allowed us to separate the original tumor cell population into cells with and without OCT4 activity, based on the expression of fluorescent reporter. We then characterized these cells for their tumorigenic phenotypes (both in vitro and in vivo) and relevant gene expression profile. 186 We first evaluated in vitro tumorigenic phenotypes of one clone each of OCT4P+ and OCT4P- cells in conjunction with parental cell population. We could not find remarkable difference in growth rate or saturation density among any of the canine OS subsets. However, to our surprise, OCT4P- cells were capable of tolerating serum withdrawal for two weeks unlike OCT4P+ and parental cells, indicating that they are able to produce autocrine growth factors. On the other hand, we found that AIG colonies formed by OCT4P- cells were significantly smaller (<100 um), although there was no difference in total number of colonies formed by cells of all 3 experimental groups. AIG is a necessary but not sufficient feature of tumorigenic cells [70]. With this regard, comparing OCT4 activity with the ability to form large-sized colonies in a spectrum of primary culture isolates and established cell lines of OS may provide more informative evidence. Our data further demonstrated that both OCT4P+ and OCT4P- cells were more capable of invading Matrigel matrix than parental cell population, and this difference was also reflected in the invasive phenotype of tumors developed in immunodeficient mice. There is a relatively poor overlap between markers used to identify C SCs [9]. Wu et a1 [68] had shown earlier that Hoechst efflux-based side population (SP) assay can enrich the CSC frequency in human mesenchymal tumors. When we evaluated the cells sorted on the basis of OCT4 promoter activity for their SP phenotype, we found that there was a modest enrichment in SP phenotype in OCT4P+ cells compared to parental cell (0.5% from 0.1%; ~5 fold enrichment) population and OCT4P- cells (0.5% from 0.05%; ~10 fold enrichment). CSCs have been postulated to be more resistant to conventional chemotherapy and radiotherapy than rest of the tumor bulk, and this differential sensitivity has been 187 attributed to tumor relapse following initial shrinkage [10,12]. We examined the sensitivity of OCT4P+ and OCT4P- subsets along with parental canine OS cells to two commonly used drugs in the treatment of canine OS, doxorubicin and cisplatin. However, our findings indicated that OCT4P+ cells did not show considerably different survival advantage over OCT4P- cells or parental cell population upon treatment with standard doses doxorubicin or cisplatin. It is noteworthy that C SCs are not always more resistant than their nontumorigenic progeny [71]. In testicular cancer, for instance, undifferentiated cells are more sensitive to cisplatin therapy than the differentiated cells that they form [72]. Similarly, our data on radiosensitivity assay showed that the differences between OCT4P+ and OCT4P- cells were not biologically significant, and parental cells were rather more resistant than those sorted on OCT4 promoter activity. When we compared their clonogenic survival differences with neonatal fibroblasts from normal and SCID dogs following X-ray irradiation, the OS cells were significantly more radioresistant than both normal fibroblasts and radiosensitive SCID fibroblasts. Thus, canine OS cells seem to be intrinsically more radioresistant as suggested in a recent study [64], but this difference is not related to OCT4 activity. Our data on limiting dilution assay suggests that OCT4 promoter activity does not enrich tumor initiating frequency of parental cell population. Although the frequency of tumor initiating cells in OCT4P+ group were higher than those in OCT4P- group, it was the original D-l7 cell population which was more tumorigenic than either of the groups sorted on the basis of OCT4 promoter activity. Moreover, there was no considerable difference in the growth kinetics of OCT4P+ versus OCT4P- cells. Both experimental groups gave rise to relatively slow growing tumors when compared to the tumors 188 developed from parental cell injection group. High frequency of tumor initiating cells in a tumor cell population suggests that either these tumors are sustained by a high frequency of CSC-like cells or, alternatively, they may conform to the clonal evolution model of tumorigenesis. Recent studies suggest that tumorigenic potential is a common property of tumor cells in some cancers [67,73,74]. Whether a cancer follows a C SC model seems to be determined by cell-of-ori gin, constellations of mutations, and the degree of differentiation block [71]. D-17 cells were originally established from metastatic lesion of canine OS. Nieves et a1 [75] had shown earlier that D-17 cells induced poorly differentiated sarcoma in nude mice without osteoid, cartilage or collagen production. Moreover, there was no gross or histological evidence of metastatic disease in mice receiving intravenous or intra-tibial injection of D-17 cells. Tumor phenotype induced by parental D-17 cell line in our study confirmed those observations. But, it was interesting that unlike parental cell population and OCT4P- subpopulations, OCT4P+ cells were able to faithfully recapitulate the phenotype of OS in mice. Ability to recapitulate original tumor phenotype has been regarded as a hallmark of CSC. Clinical and histopathological data suggests that overt differentiation is evident in many cases of OS, and this tumor seems to exhibit a hierarchical organization. Thus, it is possible that canine OS follows CSC model, and marker(s) that distinguish tumorigenic from non-tumorigenic cells could be identified in the future. However, distinct tumor phenotypes induced by different subsets of D-l7 cells raise another possibility that different sub-populations of tumor-initiating cells (TICs) exist within the original cell population as a result of ongoing genetic changes, and these 189 tumor-initiating cells are themselves subjected to further stochastic progression. In this situation, the more aggressive tumor initiating subpopulation(s) of D-l7 cells dominate over OCT4- expressing CSC and induce poorly differentiated tumors. When OCT4P+ cells are purified from heterogeneous population and transplanted in mice, these cells may get the opportunity to re-establish the original tumor heterogeneity, leading to the development of OS phenotype. On the other hand, although OCT4P- cells can induce tumor with or without undergoing further genetic changes, the resultant tumor phenotype becomes different from OS and this suggests their origin from distinct subset of tumor initiating cells (either present in the original cell population or acquired during the course of further tumor progression). The higher tumorigenicity associated with parental D-17 might be the result of combined effect of several sub-populations of tumor cells in addition to the OCT4+ and OCT4- cells. It seems that some cells in the heterogeneous tumor population might provide suitable signals to the tumor initiating cells within the parental cell population, leading to the accelerated tumor development. To understand the possible interaction between different subsets of tumor cells, competitive repopulation assay should be carried out in future studies by reconstitution of different ratios of OCT4P+ and OCT4P- subsets with parental cells, followed by subsequent genetic tracking in xenograft model [76]. We evaluated the gene expression profile of all three subpopulations with a panel of stem cell-associated genes. We found considerable differences among the subgroups regarding the expression of many of these genes. OCT4P+ cells expressed significantly higher level of self-renewal gene-BMII, Wnt co-receptor-LRPS, and a limiting factor for 190 telomerase activity-TERT whereas expression of proliferation marker-MCM 7, and notch ligand JAG] were significantly downregulated. On the other hand, OCT4P— cells expressed significantly higher level of mRNAs for pluripotency-associated gene- NANOG, notch 1i gands- NOTHCI and DLL4, and ABC transporter gene MDR]. Moreover, both OCT4P+ cells and OCT4P- cells had significantly upregulated the expression of Wnt pathway component- WIF] and downstream component of sonic hedgehog pathway-GL1], and downregulated the expression of downstream effector of NOTCH pathway -HE Y 1. Intriguingly, expression of mRNA for anti-apoptotic protein BCL2 was dramatically downregulated in both of these subsets when compared to parental cells. We then looked upon the possibility of further changes in gene expression following xenotransplantation. There were indeed significant changes in expression of some of the genes, whereas expression of other genes remained fairly similar before and after transplantation. It is possible that those genes whose expressions were remarkably altered following xenotransplantation might have some important roles in tumor progression. It would be interesting to modulate the expression of these genes in experimental systems in future studies. We attempted to examine the expression of OCT4 protein in the subpopulations sorted on the basis of OCT4 promoter activity. Consistent with the experience of Levings et a1 [57], our efforts to evaluate the expression of endogenous OCT4 protein in canine D-l7 cells have remained inconclusive. It is possible that the commercially available antibodies against human or mouse OCT4 protein was not robust enough to detect the canine OCT4 protein. We have not formally ruled out the possibility that the activity of I91 exogenous OCT4 promoter may be related to the pseudogenes rather than the expression of functional isoforrn of OCT4 (OCT4A) that confers pluripotency [77]. In fact, pseudogenes of OCT4 have been shown to be transcribed in cancers, and they may paradoxically serve as informative markers, regardless of expression status of endogenous OCT4 [49]. However, the OCT4 positive and OCT4 negative cells used in this study are stably transfected cells which have ~3 kb exogenous OCT4 promoter integrated into their chromosome. Since the OCT4 promoter is regulated by a complex core of nuclear transcriptional factors including OCT4 itself [78], YFP expression from OCT4 promoter suggests the presence of functionally active and nuclear form of OCT4 in these cells. Development of canine-specific antibodies as well as OCT4-based loss-of- function studies in future may provide better insights on role of OCT4 in the process of tumorigenesis. In summary, isolation of discrete population of cancer cells based on OCT4 promoter activity alone could not enrich the tumor-initiating capacity of canine D-17 OS cell line. It is possible that OCT4 may not represent a universal C SC marker. However, only the cells that were positive for OCT4 promoter activity recapitulated the 08 tumor phenotype, suggesting that combination of OCT4 expression with other CSCs markers will enable to identify enriched population of CSCs in OS. This is a proof—of-concept study, and these observations made in D-l 7 08 cell line must be extended to larger sample sized primary tumors in future studies. Our current study underscores the heterogeneity of the tumor-initiating cell phenotype as well as the complexity of OS biology. 192 APPENDIX Table 6.1. Primer sets used in qRT-PCR of stemness-related markers Ampl Ann icon Tern Size p(°C Gene Primer sequence (5' - 3') (bp) ) Forward GCGTGTCTGGAGGAGAAAGA ABCG2 Reverse TTCAGGAGC AAAAGGACAGC 133 60 Forward ATCCCACTCGACCAGACATC MDR] Reverse GCTGAACAACTGTGCTCTTCC 120 60 Forward ATCTCCAGGC AC AGTCTCAG MRP] Reverse C AC GATGTTGGTC TC C ATCTC 120 60 Forward TCATCACCTCCCACAGCAT ABCA3 Reverse AGTAGCCGCTGCCAAACTT 126 60 Forward CAAGGC AC GC AGAAAGAAAT NOT CH] Reverse ATGGAGACGACAGCAGAGGT 140 60 Forward C ATGAACAACTTGTCGGACTTC DLL4 Reverse GCTC C TTC TTC TGG'ITI‘GTGTT 80 60 Forward TTGCTATGGACTATCGGAGTTTG HE Y 1 Reverse TAGTTGTTAAGGTGGGAGACCA 130 60 Forward TGACCC C TACTACCAGGATAA JAG] Reverse GTGACATCATCTCTI’TGTTGAAG 65 60 Forward AGTGGAAGATACGAAGGGAAGAT SHH Reverse ATATGATGTCGGGGTTGTAATTG 82 60 Forward ATCACAAGTCAGGCTCCTATCC GL1] Reverse TGCTCTATGGGAAGTCTTGTTG 83 60 Forward GATGTGTGCGACAGTGACTACA LRP5 Reverse AGGGGTCTGAGTCTGAGTTCAA 82 60 Forward C AAAC TGC TCAACTAC CTGCTT WIFI Reverse GGTTGTGGAC AC TTGC TTATTTC 1 12 60 Forward GTGGATGACTGAGTACCTGAAC BCL2 Reverse TCAGAGACAGCC AGGAGAAGT 130 60 Forward TC TGGTAC AC AATCAC GTC GTA T ERT Reverse C AC AGTAGAGTGGAC AGGATA 96 60 Forward AGGATGC C AC C TACAC ATCAG MC M 7 Reverse C TTCTC C AC TGTGTCC ACCAT 95 60 Forward GTCCATAAAGACGCAGCACTC SALL4 Reverse GCAAAGTC AC AGGGGTTCTC 138 60 Forward GAATAACCCGAATTGGAGCAG NANOG Reverse AGCGATI‘C C TCTTCAC AGTTG 141 - 60 Forward ATGGACTGACAAATGCTGGAG BMII Reverse GGAACTGAGGATGAGGAGAC 13 60 BZM Forward TCTACATTGGGCACTGTGTCAC Canine) Reverse TGAAGAGTTCAGGTCTGACCAAG 136 60 193 Table 6.2. In vitro growth properties of D-17 canine OS cell line, and its subpopulations sorted on the basis of OCT4 expression Serum dependence Population Cell lines doubling Saturation density time (hrs) (104 cells/cmz) 0.5% No serum D17 parental 28d: 0.03 42.5 :L- 2.1 6.2 :1: 1.9 5.3 :l: 0.8 D17 OCT4 P1 26 a 0.03 46 a: 1.4 2.7 a 0.8 2.9 a 0.6 D17 OCT4P' 28 i 0.04 43.8 d: 1.4 208 :t 3.9 210 :1: 4.2 Table 6.3. Tumorigenicity of various number of D-l 7 cells sorted on the basis of OCT4 expression D-17 Parental D-17 OCT4 P” D-17 OCT4 P' CC" Latency Latency Latency N umber Tumors (days) Tumors (days) Tumors (days) 5x106 5/5 25 5,53 82 5/5 74 5x105 5/5 29 5/5 32 5/5b 74 d 5x104 5/5 35 55° 94 4/5 97 5x103 5/5 51 5/5 94 3/5 97 5><102 3/5 79 1/5 127 1/5 124 1x102 0/5 130 0/5 127 0/5 124 a . . 3/5= 1nvasrve tumors b . 4/5= metastatlc tumors 0 . . 2/5= 1nvasrve tumors d . 2/5= metastatic tumors 194 Figure 6.1. Establishment of D- l 7 reporter cell lines. YFP-positive (A, B) and YFP- negative (C, D) stable clonal cell lines were derived from parental D-17 cell line. B and D represent phase contrast image of A and C, respectively. Both of these sub-populations remained neomycin resistant and maintained the pattern on fluorescence expression after several passages. Magnification = 100x. Figure 1E and IF shows heterogeneity in fluorescence between two clonal populations- clone 1.7D (1E) versus clone 4.2 (1F). Some of the YFP-positive cells consistently became YFP negative during the course of expansion (smaller peak in the P5 gate in IE), but we have not seen the appearance of YF P positive cells from YFP negative clones (we have not seen the cells in P8 gate from YFP negative population). (1G = clone 3 and 1H= clone 4, both being YFP negative clonal cell lines). 195 Count 450 400 300 300 250 200 150 100 50 102 P5 103 FL-l-A 1000 750 500 250 IllllllllLllllIlllI‘ll 102 0% P5 P8 900 800 700 600 500 400 300 200 100 n 103 lo4 FL-l -A 105 Figure 6.1. 196 FL-I-A FL-l -A Clone 1.6P+ Clone 2P+ Clone 4.2P+ Clone 9.1P+ Clone 1P- ,i'n? Fri? 3.9. Fri? 393-? 5 VV- V-V- VV- VV- V-V- 'URIN EKOKKDKRDRKU‘ gun uuzuuzouzouzg .400 QQmOOmQQmQQmA ‘2’ m assess ¥¥§ ass 11’. 1‘3. 9:1 r: E on E I: m 13. 1?. 99 L) D L) L) D L) L) L) Q Q Q O O O O Q Clone 3P- Clone 4P- Testes blank Bone Figure 6.2. mRNA expression of 0C T 4 in different D-l7 OCT4 reporter cell lines. Endogenous expression of 0C T 4 mRN A was examined with primers specific to exon 1 of canine 0C T 4 (0C T 4-P] ), using BZ-MICROGLOBULIN as a endogenous reference gene. 0C T 4-P2 is an intronic primer set that would amplify OCT4 genomic DNA, but not OCT4 cDNA, and thus will be useful to detect any contaminating genomic DNA. cDNA from canine testis was used as a positive control to validate the OCT4 primer, and no template control (water instead of cDNA) was used as a negative control. Bone cDNA was included for comparison. Expected product size: OCT 4P1 = 145 bp, 0C T 4P2= 287 bp, BMG = 138 bp. P+ and P- indicates the positive and negative clones for OCT4 promoter aetivity respectively. White arrows indicate the clones chosen for firrther studies. 197 O 70 50 40 30 I >200 pm 20 I>100 um 10 I<100 am % AIG colonies D-l 7 parental D-l7 Oct4P+ D-l7 Oct4P- Cell lines Figure 6.3. Anchorage independent growth of D-17 OC T4 reporter cell lines. Anchorage independent growth of D—l7 parental (A, B), D-17 OCT4P+ (C, D), and D-l 7 OCT4P- (E, F) cell lines are shown. B, D, and F represent fluorescent images of A, C, and E respectively. Magnification = 40x. A total of 5 X 104 cells were seeded on soft agar in 60 mm plates, and cultured for 2 weeks. At the end, the numbers of colonies were quantified after counting them in 1cm2 area, and colonies were grouped according to their sizes. Bar diagram on the right (G) shows the frequencies of different sized colonies (n=3 for each group) from three experimental groups. 198 90- 80- 70- 60- 50- 40- 30- 20- 10- % Invasion D-17 Parean D-17 OCT 4P+ D-17 OCT4P- Canine D-l7 Cell line Figure 6.4. Matrigel invasion assay of D-l 7 OCT4 reporter cell lines. A total of 5 x 104 cells from parental, SP and non-SP fractions were seeded in on the upper side of Matrigel Matrix and incubated for 24 hrs using 10% FBS as a chemoattractant. Cells that had invaded the Matrigel were quantified colorimetrically after cresyl violet staining. Data are represented as meanisd of triplicates. 100 80 60 5' = E 40 > ,\° 20 Cisplatin D-17 OCT4 reporter cell lines I Untreated I D-17 Parental I D- 1 7 OCT4P+ I D-l 7 OCT4P- Doxorubicin Figure 6.5. Chemosensitivity assay of D-l7 OCT4 reporter cell lines. Cells from all experimental groups were treated with cisplatin and doxorubicin for 48 hrs, and viability was determined by MTS assay. Data are represented as mean i sd (n =3). 199 100 90 80 70 60 50 D17 OCT4 reporter cell lines -0—D-l7 Parental +D—I7 OCT4P+ +D—l7 OCT4P- % Survival 30 20 10 Dose (Gy) Fibroblasts -0-Regis-N +Reed +Raleigh Dose (Gy) Figure 6.6. Radiosensitivity assay of D-17 OCT4 reporter cell lines. The difference in sensitivity among parental, OCT4P+, and OCT4P- cells were measured by clonogenic survival assay. Cells were seeded in 6-well plates, and colonies were counted after 10 days of exposure to single graded dose of radiation. Survival curve was established for each cell line. Fibroblasts from neonatal SCID dogs and their normal littermate were included in the assay to compare the intrinsic radioresistance of OS cells. Results are expressed as mean i sd of triplicate wells. 200 Figure 6.7. SP phenotype in D-17 OCT4 reporter cell lines. (a) SP profile of parental D- 17 cells (A, B), OCT4 P+ clone 4.2 (C, D), and OCT4P- clone 3 (E, F) are shown after staining with Hoechst 333342 in the presence (B, D, F) or absence (A, C, E) of verapamil. (b) Bar diagram shows the effect of OCT4 reporter activity on the enrichment of SP fraction. Data are represented as mean :1: sd of three different clonal populations of both OCT4P+ and OCT4P' cells. 201 ’t 0 E. .1: ‘6: .= 8 9 = i Z/ SP= 0% Hoechst red C 0.6 0.5 .21 To 0.4 O a 0.3 °\° 0.2 0.1 0 D-l7 parental D-17 OCT4 P+ D-l7 OCT4 P- Cell lines Figure 6.7. 202 u‘ Figure 6.8. Tumorigenicity of D-17 cells sorted on the basis of OC T4 expression and the resultant tumor phenotypes. Tumors (arrows) were seen in mouse injected with parental D-17 cell line (A, B) as well as those injected with both OCT4-promoter activity positive cells (C, D) and OCT4-promoter activity negative cells (E, F). Differences in tumor types and behavior among the experimental groups were revealed by histopathological examination (G—L= H&E staining). Injection of parental D-17 led to the development of anaplastic sarcoma (G) which were highly invasive and locally destructive (skeletal muscle invasion is shown in H). Interestingly, many of the tumors formed by OCT4- promoter activity positive cells were typical OS (I), which were highly invasive and locally destructive (J). On the other hand, cells negative for CC T4 promoter activity gave rise to anaplastic sarcoma (K) which were highly invasive with distant metastatic lesions (L = kidney; M = intestine). 203 Figure 6.9. Expression of stemness-associated genes in D-l 7 OC T4 reporter cell lines. Relative mRNA expressions of NANOG and BMI] (A) as well as Notch (B), Hedgehog and Wnt (C) family genes were compared by qRT-PCR among three experimental groups (parental, OCT4P+, and OCT4P-). Expression levels were normalized to endogenous reference gene [FD-MICROGLOBULIN, and expressed as fold change relative to unfractionated cells of each cell line. Normalization to second reference gene GAPDH showed very similar results (data not shown). 204 A 12 ‘ lD-l7 Parental 9 _ ID-I7OCT4P+ “all I D-17 OCT4P- = 6 _ a .= U E 3 7 O In 0 .. 3 NANOG BMII ' Gene I D-17 Parental I D-l7 OCT4 P+ I D-17 OCT4 P- G) M fl 5 n: .: NOTCH] DLL4 'e‘ In. -20 a -30 ~ ~40 '1 _ l 50 Gene C 12 I D— l 7 Parental gas 10 g 8 lD—l7OCT4 P+ f. 6 ID-17OCT4P- E 4 O “I" 2 0 GL1 LRP5 WIFI Gene Figure 6.9. 205 N Ur A En 20 I D-17 Parental g 15 ID-17OCT4P+ G K 5 0 ABCGZ MDRI MRP] ABCA3 Gene B 5 _ Parental — OCT4 P+ _ OCT4P- 0 - €200 / III “-250 - -300 - -350 . . . TERT MC M 7 BC L2 Gene Figure 6.10. Expression of ABC transporters and stemness-related genes in D-17 OC T4 reporter cell lines. Relative mRNA expressions of ABC transporters (A) as well as stemness-relevant genes such as T ERT , MC M 7 and BCL2 (B) were compared by qRT- PCR among three experimental groups (parental, OCT4P+, and OCT4P-). Expression levels were normalized to endogenous reference gene 112-MIC ROGLOBULIN, and expressed as fold change relative to unfractionated cells of each cell line. Normalization to second reference gene GAPDH showed very similar results (data not shown). 206 Figure 6.11. Re-establishment of OS cell lines from mouse tumor xenografts. Cell lines were re-established from all 3 experimental groups including D17 parental (A, B), D17 OCT4P+ (C, D), D17 OCT4P- (E, F) cells. OCT4P+ and OCT4P- cell lines were monitored for their ability to grow in the presence of neomycin, and expression of fluorescence. B, D, and F represent fluorescence images of A, C, and E respectively. Magnification = 100x. 0 M 5 ID17 '5 '3 ID17OCT4+ a". IDI7OCT4- Gene Figure 6.12. Comparison of expression of stemness markers in original versus re- established cell lines. Quantitative RT-PCR was carried out to evaluate the expression of same set of stemness-associated genes in each subpopulation of cells isolated from tumor xenografts by comparing the expression level against respective original cell lines used for xenotransplantation. Expression levels were normalized to endogenous reference gene fiZ-MICROGLOB ULIN, and fold changes were calculated by 2M” method. 208 10. ll. 12. REFERENCES . F ialkow PJ. Clonal origin of human tumors. Annu Rev Med 1979;30:135-143. Heppner GH, Miller BE. Tumor heterogeneity: biological implications and therapeutic consequences. Cancer Metastasis Rev l983;2:5-23. Nowell PC. The clonal evolution of tumor cell populations. Science 1976;194:23- 28. Campbell LL, Polyak K. Breast tumor heterogeneity: cancer stem cells or clonal evolution? Cell Cycle 2007;6z2332-2338. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med 1997;3:730-737. Reya T, Morrison SJ, Clarke MF et al. Stem cells, cancer, and cancer stem cells. Nature 2001;414:105-111. Clarke MF, Dick JE, Dirks PB et al. Cancer Stem Cells--Perspectives on Current Status and Future Directions: AACR Workshop on Cancer Stem Cells. Cancer Res 2006;66:9339-9344. . Ward RJ, Dirks PB. Cancer stem cells: at the headwaters of tumor development. Annu Rev Pathol 2007;22175-189. Visvader JE, Lindeman GJ. Cancer stem cells in solid tumours: accumulating evidence and unresolved questions. Nat Rev Cancer 2008;8:755-768. Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer 2005;52275-284. Li X, Lewis MT, Huang J et al. Intrinsic resistance of tumorigenic breast cancer cells to chemotherapy. J Natl Cancer Inst 2008;100:672-679. Baumann M, Krause M, Hill R. Exploring the role of cancer stem cells in radioresistance. Nat Rev Cancer 2008;8:545-554. 209 13. 14. 15. 16. l7. 18. 19. 20. 21. 22. 23. 24. Bao S, Wu Q, McLendon RB et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 2006;444:756-760. Thompson KG, Pool RR. Tumors of bone. In: Meuten DJ, ed. Tumors in Domestic Animals. Iowa State Press, Ames, IA 2002:245—83. Barroga EF, Kadosawa T, Okumura M et al. Establishment and characterization of the grth and pulmonary metastasis of a highly lung metastasizing cell line from canine osteosarcoma in nude mice. J Vet Med Sci 1999;61:361-367. Withrow SJ, Powers BE, Straw RC et al. Comparative aspects of osteosarcoma. Dog versus man. Clin Orth0p Relat Res 1991;159-168. Vail DM, MacEwen EG. Spontaneously occurring tumors of companion animals as models for human cancer. Cancer Invest 2000;18:781-792. Hansen K, Khanna C. Spontaneous and genetically engineered animal models; use in preclinical cancer drug development. Eur J Cancer 2004;40:858-880. Mueller F, Fuchs B, Kaser-Hotz B. Comparative biology of human and canine osteosarcoma. Anticancer Res 2007;27:155-164. Sutter NB, Ostrander EA. Dog star rising: the canine genetic system. Nat Rev Genet 2004;5:900-910. Straw RC. Tumors of the skeletal system. In: Withrow SJ, MacEwen GE, eds. Small Animal Clinical Oncology. W.B. Saunders Co. Philadelphia 1996:287-305. Haydon RC, Luu HH, He TC. Osteosarcoma and osteoblastic differentiation: a new perspective on oncogenesis. Clin Orthop Relat Res 2007;454:237-246. Thomas D, Kansara M. Epi genetic modifications in osteogenic differentiation and transformation. J Cell Biochem 2006;982757-769. Rosner MH, Vigano MA, Ozato K et al. A POU-domain transcription factor in early stem cells and germ cells of the mammalian embryo. Nature 1990;345:686- 692. 210 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Loh YH, Wu Q, Chew JL et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat Genet 2006;38:431- 440. Tai MH, Chang CC, Kiupel M et al. Oct4 expression in adult human stem cells: evidence in support of the stem cell theory of carcinogenesis. Carcinogenesis 2005;26:495-502. J iang Y, Jahagirdar BN, Reinhardt RL et al. Pluripotency of mesenchymal stem cells derived from adult marrow. Nature 2002;418:41-49. Kucia M, Reca R, Campbell FR et al. A population of very small embryonic-like (VSEL) CXCR4(+)SSEA-1(+)Oct-4+ stem cells identified in adult bone marrow. Leukemia 2006;20:857-869. Matthai C, Horvat R, Noe M et al. Oct-4 expression in human endometrium. Mol Hum Reprod 2006;12:7-10. Monk M, Holding C. Human embryonic genes re-expressed in cancer cells. Oncogene 2001;20:8085-8091. Hochedlinger K, Yamada Y, Beard C et al. Ectopic expression of Oct-4 blocks progenitor-cell differentiation and causes dysplasia in epithelial tissues. Cell 2005;121:465-477. Looijenga LH, Stoop H, de Leeuw HP et a1. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003;63:2244- 2250. Ezeh UI, Turek PJ, Reijo RA et al. Human embryonic stem cell genes OCT4, NANOG, STELLAR, and GDF3 are expressed in both seminoma and breast carcinoma. Cancer 2005;104:2255-2265. Atlasi Y, Mowla SJ, Ziaee SA et al. OCT-4, an embryonic stem cell marker, is highly expressed in bladder cancer. Int J Cancer 2007;120:1598-1602. Webster JD, Yuzbasiyan-Gurkan V, Trosko JE et al. Expression of the embryonic transcription factor Oct4 in canine neoplasms: a potential marker for stem cell subpopulations in neoplasia. Vet Pathol 2007;44:893-900. 211 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. Chen YC, Hsu HS, Chen YW et al. Oct-4 expression maintained cancer stem-like properties in lung cancer-derived CD133-positive cells. PLoS ONE 2008;32e2637. Chiou SH, Yu CC, Huang CY et al. Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and high-grade oral squamous cell carcinoma. Clin Cancer Res 2008;14:4085-4095. Hu T, Liu S, Breiter DR et al. Octamer 4 small interfering RNA results in cancer stem cell-like cell apoptosis. Cancer Res 2008;68:6533-6540. Iki K, Pour PM. Expression of Oct4, a stem cell marker, in the hamster pancreatic cancer model. Pancreatology 2006;62406-413. Karoubi G, Gugger M, Schmid R et al. OCT4 expression in human non-small cell lung cancer: implications for therapeutic intervention. Interact Cardiovasc Thorac Surg 2009;8z393-397. Gibbs CP, Kukekov VG, Reith JD et al. Stem-like cells in bone sarcomas: implications for tumori genesis. Neoplasia 2005;72967-976. Wilson H, Huelsmeyer M, Chun R et a1. Isolation and characterisation of cancer stem cells from canine osteosarcoma. Vet J 2008;175:69-75. Yamaguchi S, Yamazaki Y, Ishikawa Y et al. EWSRI is firsed to POUSFI in a bone tumor with translocation t(6;22)(p21;q12). Genes Chromosomes Cancer 2005 ;43 :2 1 7-222. Gidekel S, Pizov G, Bergman Y et al. Oct-3/4 is a dose-dependent oncogenic fate determinant. Cancer Cell 2003;42361-370. Cantz T, Key G, Bleidissel M et al. Absence of OCT4 expression in somatic tumor cell lines. Stem Cells 2008;26:692-697. Lengner CJ, Camargo FD, Hochedlinger K et al. Oct4 expression is not required for mouse somatic stem cell self-renewal. Cell Stem Cell 2007;1:403-415. Liedtke S, Stephan M, Kogler G. Oct4 expression revisited: potential pitfalls for data misinterpretation in stem cell research. Biol Chem 2008;389:845-850. 212 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. Lin H, Shabbir A, Molnar M et al. Stem cell regulatory function mediated by expression of a novel mouse Oct4 pseudogene. Biochem Biophys Res Commun 2007;355:111-116. Suo G, Han J, Wang X et a1. Oct4 pseudogenes are transcribed in cancers. Biochem Biophys Res Commun 2005;337:1047-1051. Zangrossi S, Marabese M, Broggini M et al. Oct-4 expression in adult human differentiated cells challenges its role as a pure stem cell marker. Stem Cells 2007;25: 1675-1680. Giudice A, Trounson A. Genetic modification of human embryonic stem cells for derivation of target cells. Cell Stem Cell 2008;2z422-433. Gerrard L, Zhao D, Clark AJ et al. Stably transfected human embryonic stem cell clones express OCT4-specific green fluorescent protein and maintain self-renewal and pluripotency. Stem Cells 2005;23:124-133. Eiges R, Schuldiner M, Drukker M et al. Establishment of human embryonic stem cell-transfected clones carrying a marker for undifferentiated cells. Curr Biol 2001;11:514-518. Singh RN, Nakano T, Xuing L et al. Enhancer-specified GFP-based F ACS purification of human spinal motor neurons from embryonic stem cells. Exp Neurol 2005;196:224-234. Huber I, Itzhaki I, Caspi O et al. Identification and selection of cardiomyocytes during human embryonic stem cell differentiation. FASEB J 2007;21:2551-2563. Lavon N, Yanuka O, Benvenisty N. Differentiation and isolation of hepatic-like cells from human embryonic stem cells. Differentiation 2004;72:230-238. Levings PP, McGarry SV, Currie TP et al. Expression of an exogenous human Oct-4 promoter identifies tumor-initiating cells in osteosarcoma. Cancer Res 2009;69:5648-5655. Rozen S, Skaletsky HJ. Primer3 on the W for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics Methods and 213 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. Protocols: Methods in Molecular Biology. Totowa, NJ: Humana Press 2000:365- 86. Livak KJ, Schnrittgen TD. Analysis of relative gene expression data using real- time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402-408. Schnrittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C ('1) method. Nat Protoc 2008;3z1 101-1 108. Lin SL, Chang D, Ying SY. Hyaluronan stimulates transformation of androgen- independent prostate cancer. Carcinogenesis 2007;28:310-320. Kruger JA, Kaplan CD, Luo Y et al. Characterization of stem cell-like cancer cells in immune-competent mice. Blood 2006;108:3906-3912. Chen MF, Lin CT, Chen WC et al. The sensitivity of human mesenchymal stem cells to ionizing radiation. Int J Radiat Oncol Biol Phys 2006;66:244-253. Fitzpatrick CL, Farese JP, Milner RJ et al. Intrinsic radiosensitivity and repair of sublethal radiation-induced damage in canine osteosarcoma cell lines. Am J Vet Res 2008;69:1197-1202. Hu Y, Smyth GK. ELDA: extreme limiting dilution analysis for comparing depleted and enriched populations in stem cell and other assays. J Irnrnunol Methods 2009;347:70-78. Hadnagy A, Gaboury L, Beaulieu R et al. SP analysis may be used to identify cancer stem cell populations. Exp Cell Res 2006;312:3701-3710. Quintana E, Shackleton M, Sabel MS et al. Efficient tumour formation by single human melanoma cells. Nature 2008;456:593-598. Wu C, Wei Q, Utomo V et al. Side population cells isolated fiom mesenchymal neoplasms have tumor initiating potential. Cancer Res 2007;67:8216-8222. Vlashi E, Kim K, Lagadec C et al. In vivo imaging, tracking, and targeting of cancer stem cells. J Natl Cancer Inst 2009;101:350-359. 214 70. 71. 72. 73. 74. 75. 76. 77. 78. Hamburger AW, Salmon SE. Primary bioassay of human tumor stem cells. Science 1977;197:461-463. Shackleton M, Quintana E, Fearon ER et al. Heterogeneity in cancer: cancer stem cells versus clonal evolution. Cell 2009;138:822-829. Masters JR, Koberle B. Curing metastatic cancer: lessons from testicular germ- cell tumours. Nat Rev Cancer 2003;3:517-525. Kelly PN, Dakic A, Adams JM et al. Tumor growth need not be driven by rare cancer stem cells. Science 2007;317:337. Williams RT, den BW, Sherr CJ. Cytokine-dependent imatinib resistance in mouse BCR-ABL+, Arf—null lymphoblastic leukemia. Genes Dev 2007;21:2283- 2287. Nieves MA, Vahle J, Ackennann M et al. Production and characterization of canine osteosarcoma cell lines that induce transplantable tumors in nude mice. Am J Vet Res 1998;59:359-362. Schatton T, Murphy GF, Frank NY et al. Identification of cells initiating human melanomas. Nature 2008;451:345-349. Liedtke S, Enczrnann J, Waclawczyk S et a1. Oct4 and its pseudogenes confuse stem cell research. Cell Stem Cell 2007;] :364-366. Pei D. Regulation of pluripotency and reprogramming by transcription factors 5. J Biol Chem 2009;284:3365-3369. 215 CHAPTER 7 Side population phenotype in canine osteosarcoma 216 SUMMARY Osteosarcoma (OS) is the most common primary bone tumor of human and dogs. The tumor is highly resistant to conventional chemo-and radiotherapies, and is composed of a varying proportion of undifferentiated and differentiated cell types of mesenchymal lineages. These attributes make OS as a plausible candidate for being a cancer driven by stem cells. Several studies have suggested that a subpopulation of cells having characteristic Hoechst dye efflux capacity, known as side population (SP), are greatly enriched for tumor-initiating stem cells in a given tumor cell population. The purpose of this study was to isolate and characterize putative tumor stem cell fraction from canine OS using such functional assay. Primary cultures of OS cells were established in-house from several different tumor samples. We developed and optimized SP assay for canine OS, and used this approach to examine the SP cells in several canine OS specimens, including primary culture of canine 08 as well as established OS cell lines. Our findings revealed the presence of SP cells in all of the primary isolates and established cell lines of canine OS that we examined. However, the frequency of SP cells varied widely among the different OS samples (0.1% to 1.6%). We then sorted the SP and non-SP fraction of these OS cells, and evaluated them for their in vitro tumorigenic phenotype by a panel of surrogate assays, as well as for characteristic gene expression profile associated with stemness and drug resistance. Our results indicate that differences among various OS cell lines were much higher than the differences within the sub-populations of cells sorted on the basis of Hoechst efilux. Although SP cells in some cell lines behaved differently than non-SP fraction and vice versa, we could not find consistent and predictable differences between 217 SP and non-SP fractions for their in vitro tumorigenic phenotypes, and characteristic gene expression profile. In vivo tumorigenicity assays will be conducted in future studies to further evaluate any potential differences in tumorigenic behavior between these subsets. In conclusion, we have validated the SP isolation technique for canine OS, and provided a general protocol for isolating SP cells from canine tumors. Isolation of SP may represent a simpler strategy to enrich putative CSCs, study their biology and therapeutic response in a clinically feasible manner. However, our data suggested that SP analysis alone is not sufficient to accurately define CSC population in canine OS. The relevance of SP phenotype and stemness could vary among different tumor types and species, and so does the expression and function of transporter proteins. Our results supported and extended the concept of OS being a highly heterogeneous tumor type, and these findings will be useful to design more effective strategies to treat OS in future. 218 INTRODUCTION Various in vitro and in vivo assays have demonstrated that tumors are composed of heterogeneous population of cells with differences in proliferation, differentiation, and tumor initiation potential despite their monoclonal origin [1-3]. These observations led to the idea that tumors are initiated and maintained by a distinct subpopulation of tumor cells, the so-called cancer stem cells (CSCs) [4-9]. CSCs are defined as the “cells within a tumor that possesses the capacity to self-renew and to cause the heterogeneous lineages of cancer cells that comprise the tumor” [10]. First identified in 1994 in acute myeloid leukemia, C SCs have been documented in several solid cancers including brain, breast, prostate, melanoma, lung, colon, head and neck, and pancreatic cancers (reviewed in [11,12]). Importantly, many of these studies have shown that these CSCs are generally resistant to conventional chemotherapy and radiotherapy regimens and are responsible for relapse of tumor after completion of conventional therapy [13-16]. Osteosarcoma (OS) is the most common primary bone tumor of human and dogs [17,18]. It is a highly malignant tumor with poor clinical prognosis. Canine OS closely resembles human OS in terms of histopathological appearance and biological behavior [19], and is an excellent model to study tumor biology and therapeutic intervention [20]. Osteosarcoma accounts for 4% of all malignancies, 85% of all skeletal malignancies, and 98% of appendicular primary canine bone tumors [17]. It is quite resistant to conventional chemotherapy and radiotherapy, and contains varying proportion of undifferentiated cells, osteoblasts, chondroblasts, and fibroblasts [21]. These attributes make canine OS as a plausible candidate for cancer derived from and/or driven by stem cells. 219 Although a few linage specific cell surface markers have been used to enrich and prospectively isolate cancer stem cells/tumor initiating cells in humans (reviewed in [11,22], these markers are not CSC specific. Thus, there is an acute need for marker(s) which link surface markers to functional assays. Furthermore, only a limited number of lineage specific cell surface markers are available for the dog. Several studies have suggested that a subpopulation of cells, called the side population (SP) which are defined by their characteristics in flow cytometry in the presence of Hoechst 33342 vital dye, to harbor and be greatly enriched for the stem cells in a given tumor cell population [23]. We propose that in the absence of well-defined surface markers for tumor initiating cells, SP analysis is a feasible and clinically useful approach to isolate tumor-initiating cells from a heterogeneous population of cells. SP assay has been demonstrated to enrich primitive and undifferentiated cells (reviewed in [24]), including mouse embryonic stem cells [25]. In recent studies on human malignancies, SPs have been identified in several tumors and cancer cell lines (reviewed in [26]), including human mesenchymal tumors such as OS [27,28] and Ewing’s sarcoma [29]. Originally described by Goodell et al [30,31], SP cells are typically identified on lower left quadrant of FACS dot plot as a population of cells displaying low blue and low red fluorescence, after incubation of target cells with the fluorescent dye Hoechst 33342 and subsequent analysis of dual-wavelength fluorescence [32]. These low-staining SP cells disappear from the F ACS profile after treatment with ATP-binding cassette (ABC) transporter inhibitors such as veraparnil, reserpine or fumitremorgin C (FTC). Chemotherapeutic resistance has remained a major cause of tumor relapse and therapeutic failure. One mechanism by which cancer cells become resistant to 220 chemotherapy is the expression of ABC transporter proteins which use the energy of ATP hydrolysis to actively pump out a wide variety of substrates, including drugs as well as fluorescent dyes such as Hoechst, across the cell membrane [33]. Thus, Hoechst efflux— based SP analysis seems to be an attractive surrogate system to increase our understanding of therapeutic failure and tumor recurrence. In the present study, we have isolated and characterized SP cells for the first time in any canine tumors-more specifically, in canine OS. We have optimized the SP assay for canine OS cells, isolated the subpopulations by FACS and then evaluated them for expression of characteristic genes and tumorigenic phenotypes in vitro. 221 MATERIALS AND METHODS Materials: All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise stated. Establishment of canine OS cell lines: Samples were obtained from 9 different canine patients (8 primary tumors and . lmetastatic tumor) with histopathologically confirmed cases of OS (Veterinary Teaching Hospital, Michigan State University). Tissue pieces were minced with a sterile scalpel blade, and digested at 37°C for 2 hours with collagenase type IA (1 mg/ml) in Minimmn Essential Medium-alpha (tr-MEM) containing 10% FBS, 100 U/ml penicillin, 100 ug/ml streptomycin and 0.25 jig/ml amphotericin. At the end of incubation, cells were further triturated by passing the cell suspension through 20—gauge needle, filtered through a 70 um nylon mesh cell strainer (BD biosciences, San Jose, CA), washed with phosphate buffered saline (PBS) containing 10% FBS, and resuspended in a-MEM. Cells were then counted in hemocytometer using trypan blue, and plated in u—MEM with 10% FBS. Unattached cells were removed on the next day by washing with PBS. Adherent cells were expanded in a-MEM with 10% F BS, subcultured, and cryopreserved for further studies. Early passage cells (passage 2 to passage 4) were used for all experiments related to current study. D17 cell line was purchased from ATCC, and two other OS cell lines- ABRAMS and Gracie were obtained from Dr. Elizabeth McNeil (Michigan State University). All cell lines and primary culture isolates from OS tumors were maintained in a-MEM with 10% FBS. SP analysis: 222 Hoechst 33342 staining: Cells from log-phase cultures (~70% confluence) were harvested with 0.05% trypsin-EDTA, and washed twice with pre-warmed PBS. In preliminary studies, 1 X 10‘5 cells/m1 of cell suspension were stained with several concentrations of Hoechst 33342 dye (1, 3, 5, 7.5, and 10 rig/ml) for various time length (45, 60, 70, 90, and 120 min) using different staining media (rt-MEM, PBS or HBSS with or without 10 mM HEPES and/or 2% F BS) at 37°C. For inhibition of ABC transporters, these cells were stained with Hoechst 33342 in the presence of verapamil (50, 100, 150, 200, and 250 pM), reserpine (5, 50, and 100 pM), fumitremorgin-C (FTC) (1, 5, and 10 pM), or 2-deoxyglucose (50 mM) and sodium azide (15 mM). After initial optimization, cells were resuspended at 1 X 106 cells/ml in pre-warmed a-MEM (Invitrogen Corporation, Carlsbad, CA) with 2% F BS, stained with Hoechst 33342 dye at a final concentration of 3 rig/m1 in the presence or absence of 100 uM verapamil (added 15 min before Hoechst staining), and incubated at 37°C for 70 min in dark with intermittent mixing. At the end of the incubation, cells were washed with ice—cold PBS, centrifuged (800X g) at 4°C, resuspended in ice-cold PBS containing 2% F BS, and kept in ice at dark until analysis. The cells were filtered through a 70-um cell strainer (BD Biosciences, San Jose, CA) to obtain single cell suspension before analysis and sorting. Just before sorting, propidium iodide (PI) (Invitrogen) was added at a final concentration of 1 jig/ml to identify non-viable cells. Flow cytometry: SP cells were analyzed and sorted on a FACSVantage SE (Becton Dickinson, San Jose, CA). Hoechst dye was excited with 365 nm UV laser (70 mW), and dual wave-length fluorescence was captured simultaneously with 450/50 band pass (BP) filter for Hoechst blue and a 660 long pass (LP) filter for Hoechst red on a linearly 223 amplified fluorescence scale. A 485 long pass dichroic mirror (DMSP) was used to separate the emission wavelengths. A second 488 nm argon laser (300 mW) was used to excite PI, and fluorescence emission was measured with a 630/28 BP filter. Dead cells and cellular debris were excluded based on scatter signal and PI fluorescence. Fifty to one hundred thousand live cells were analyzed from each sample. The SP gate was defined as the diminished region on the dot plot in the presence of veraparnil. The sorting gates for SP were established as described by Goodell et a1 [34], and results were analyzed using F lowJo software version 6.3.3 (Tree Star Inc., Ashland, OR). All the experiments were repeated thrice (representative profiles shown). Sorted cells were used within one passage (less than one week of culture) for all the experiments. RT-PCR: Total RNA was isolated from all experimental groups, using Versagene RNA isolation kit (5 Prime Inc, Gaithersburg, MD). RNAs were treated with TURBO DNA- free DNase I (Ambion, Austin, TX) to remove any contaminating genomic DNA. One ug total RNA was reverse transcribed in 30 ul reaction volume, using 10 units of Superscript III reverse transcriptase (Invitrogen, Carlsbad,CA), and final concentrations of 5 11M anchored oligo-dT primers, 0.5 mM dNTPs, 5 mM DTT, and 1x first strand buffer supplied by the manufacturer (Invitrogen, Carlsbad,CA) . The reverse transcription reaction was performed at 50°C for 1 hr, and then inactivated at 70°C for 15 min. The cDNAs were purified by Qiaquick PCR purification kit (Qiagen) before using them for real time PCR. Validation of primers: All primers were designed using Primer 3 software [35]. Primers derived from coding regions of respective genes in the canine genome were used to 224 amplify the target sites (Table 7.1). All of the primers were first validated with canine genomic DNA. Presence of single amplicon of expected product size was verified by agarose gel electrophoresis following conventional PCR. Before using AACT method, reaction efficiencies were determined by standard curves generated from canine genomic DNA. A ten—fold serial dilution of 100 ng total genomic DNA at five different points confirmed that both the slope and R2 values were close to the theoretical values (slope= - 3.8 to -3.3; R220.99). A validation experiment was done to demonstrate the similarities in efficiency of target amplification and reference amplification (slope difference <0.l). Dissociation curve analysis was performed to confirm the specificity of each product and absence of primer-dimer complexes. Real time PCR: Both side population (SP) and non-side population (non-SP) cells as well as parental cell lines were used for quantitative RT-PCR studies. Quantitative RT- PCR was carried out with 2X SYBR Green master mix (Applied Biosystems, Foster city, CA) and 100 nM primers in 15 ul reaction using ABI 7700 sequence detection system (Applied Biosystems, Foster city, CA) on the thermal profile of 50°C for 2 min, 95°C for 10 nrin, 40 cycles at 95°C for 15 sec, 60°C for 1 min. All reactions were performed in triplicate, and reaction mixture with RNA instead of cDNA (no reverse transcriptase control) was used in each run to ensure the absence of genomic DNA contamination. Expression of each target gene was normalized against ,BZ-MICROGLOBULIN (endogenous reference gene), and relative quantification was carried (Z’MCT method) by comparing the expression against unsorted parental cells from each cell line [36,37]. Characterization for tumorigenic phenotypes in vitro: 225 (i) Determination of population doubling time and saturation density: To determine the growth properties, 1 x 10" cells were seeded in each of multiple 60-mm dishes on day 0, and grown in a-MEM with 10% FBS. For doubling time (DT) calculation, the cultures were re-fed every 3 days, and cells were counted from triplicate dishes on days 2-10. The numbers of cells per 60-mm dish were plotted against the days of culture, and the slope of the linear region of each growth curve was used to calculate the doubling time. For saturation density determination, the cultures were re-fed every day for 10 days, and cells were counted from triplicate dishes on days 4-10. The saturation density was reported as the maximum number of cell per cmz. (ii) Anchorage independent growth: In order to assess the ability of AIG, 5 x 104 cells in 3 ml of 0.33% agarose medium were plated on top of 3 ml pre-hardened 0.5% agarose medium in each triplicate 60-mm dishes with grids to aid colony counting. 2.5 ml medium (or-MEM with 10% F BS) was then added and renewed every 3 days. The numbers of colonies developed in soft agar were scored after 2 weeks. Colonies were stained by overnight incubation at 37°C in iodonitrotetrazolium chloride (1mg/ml in 0.9% NaCl), destained in ethanol, and photographed. (iii) Matrigel invasion assay: For invasion assay, 5 x 10" cells were resuspended in 500 pl of serum free a-MEM medium, and seeded on the upper side of Matrigel Matrix (with 8 pm pore size PET membrane) of each well of 24-well cell culture insert (BD Biosciences, San Jose, CA). The lower chamber of the transwell insert were filled with 750 pl a-MEM containing 10% FBS as a chemoattractant. After 24 hrs of incubation, the filters were removed and cells that invaded the Matti gel and attached to the lower chamber of the transwell were fixed with 100% methanol for 10 min, and stained with 226 0.2% cresyl violet for 20 min. Cresyl violet stain on the membranes was eluted with 100% ethanol/0.2 M sodium citrate (1:1), and absorbance was read at 570 nm using SpectraMax M5 microplate reader (Molecular Devices, Sunnyvale, CA). The percentage of invading cells was calculated by comparison of absorbance in cells attached to the underside of the membrane (topside wiped with cotton wool) against absorbance of total cells in the membrane (topside not wiped) [38]. (iv) Serum dependence assay: Cells were seeded at 2 x 105 per well in 24-well plates in a-MEM medium with 10% FBS. After overnight attachment, the cells were washed twice with PBS, and cultured in low-serum (0.1% FBS) or serum-free medium. Cell proliferation and viability was monitored for 2 weeks by counting the cells from triplicate wells with a hemocytometer using trypan blue [39]. (v) Chemoresistance and Radioresistance: Chemosensitivity assay: Cells were seeded in 96-well plates at 4000 cells per well in 100 pl a-MEM with 10% FBS, and cultured for 24 hrs. Five replicates of each experimental group were tested with doxorubicin (1.5 pg/ml) and cisplatin (4 jig/ml), using 0.1% DMSO as a vehicle control [40]. Cell viability was determined 48 hrs later by a colorimetric 3-(4,5-dimethylthiazol-2-yl)-5-(3 carboxymethoxyphenyl)-2-(4- sulfophenyl)-2H-tetrazolium, inner salt (MTS) cell proliferation assay [CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (MTS), Promega, Madison, WI] according to the instructions of the manufacturer (Promega). Briefly, 20 ul of substrate solution was added to each well, and the cells were incubated for 2 hrs at 37°C. Absorbance was measured at 490 nm wavelength for each well using Spectramax M5 microplate reader (Molecular Devices, Sunnyvale, CA). Drug resistance was represented 227 as percentage viability which was calculated as: (mean absorbance of the test well) / (mean absorbance of the control) X 100%. Radiosensitivity assay: Cells were seeded in triplicate in 6-well plates (100 cells per well for 0 and 2 Gy dose, 200 cells per well for 5 Gy dose, and 400 cells per well for 10 Gy dose), and cultured for 24 hrs in a-MEM with 10% FBS. Early passage neonatal fibroblasts from two SCID dogs (with mutation in DNA-PKC) and one of their normal littermate (gift from Dr. Kathy Meek, Michigan State University) were included for assessing the intrinsic radiosensitivity of canine OS cells. Radiation survival was defined as the ability of cells to maintain clonogenic capacity and form colonies after radiation exposure. Cells from each experimental group were treated with single graded dose (300 cGy/nrin) of X-ray radiation (0, 2, 5, 10 Gy), using a 6 MV linear accelerator (Clinac 2100 C, Varian Inc., Palo Alto, CA) and incubated for 10 days (until most cell clones reached >50 cells, but colonies were still distinctly apart). Crystal violet (1%) staining was carried out to examine the number of surviving colonies and survival curve was established for each cell line [41,42]. Results are expressed as mean :1: sd of triplicate wells. Statistical Analysis: All quantitative data were presented as the mean i standard deviation. Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparison test, using SigmaStat version 2.03 (SPSS Inc., Chicago, IL). A p-value of less than 0.05 was considered significant. 228 RESULTS Establishment of primary culture of canine OS: Nine different primary culture isolates were successfully established from tissue samples obtained from spontaneously occurring cases of canine OS (designated as MSU- DOS). Cell morphologies were considerably different among the isolates, reflecting the usually noticed heterogeneity between individual OS tumors. Representative morphologies included polygonal/cuboidal cells and elongated spindle-shaped cells as well as cells with round, loosely attached spindle-shaped morphology (Figure 7.1). Expression of ABC transporters in canine OS cells: We investigated the expression of transcripts for three ABC transporters that are commonly implicated in multi-drug resistance i.e. ABCGZ, MDR-l and MRP-I . Our results from conventional RT-PCR results revealed that canine OS cell lines expressed the mRNAs for all of these three transporters (Figure 7.2). Evaluation of SP phenotype in canine OS cells: We first sought to determine whether the SP assay is applicable to canine OS tumors or not. To that end, we first validated the standard SP parameters in our experimental system by using A549 human pulmonary adenocarcinoma cell line as a positive control (data not shown). We then established the SP assay protocol for canine OS tumors, based on the modification of original protocol of Goodell et al [43,44]. Primary isolates from 9 different canine OS patients and 3 different established cell lines (12 in total) were used for SP analysis. After preliminary testing of several different concentrations of Hoechst for different staining periods in different staining media (see 229 materials and methods), we confirmed that staining with 3 rig/ml Hoechst for 70 min in a-MEM + 2% FBS was optimum for canine OS cells in our system. Then, we used this protocol to survey all of the canine OS cells, and we detected a SP of 0.8 :1: 0.4 % (mean i sd) in canine OS cells, ranging from 0.1 % to 1.6 % (n = 12) (Figure 7.3 and Figure 7.4). Figure 7.3 shows typical SP profiles of canine OS cells. We tested 3 different ABC transporter inhibitors (veraparnil, reserpine and FTC) at various concentrations (see materials and methods), and found 100 pM veraparnil (added 15 min before Hoechst staining) to be optimum for all inhibition studies. Interestingly, incubation of these cells with veraparnil revealed two distinct groups of respondent cell lines. In first group of cells (4 primary tumor isolates and 2 established cell lines), veraparnil reduced the SP robustly (92.58 35.03% reduction; n = 6; p<0.05, paired-t test) (Figure 7.3). In second group (5 primary tumor isolates and 1 established cell line), verapamil was able to reduce the SP only partially (32.6 a: 1 1.7% reduction; n = 6; p<0.05, paired-t test) (Figure 7.4). Incubation with higher concentrations of verapamil, or use of other pump-inhibitor drugs (reserpine and FTC) gave similar results (Figure 7.4), and none of these drugs at any concentration were able to completely eliminate SP fraction in these cell lines. Thus, some of the OS cell lines were more responsive to commonly used ABC-transporter inhibitors than other cell lines. Interestingly, in two of the cell lines (MSU-DOS-Gl and MSU-DOS- 10M), we consistently noticed another subpopulation of Hoechst low cells on the lower right side of FACS plot (Figure 7.3, P2 gate), which was insensitive to veraparnil inhibition (Figure 7.3) unlike the cells in P5 gate which disappeared after verapamil treatment. We did not characterize this sub-population in the current study. 230 Thus, we were able to identify SP in both heterogeneous isolates of primary cultures and monoclonal cell environment of established cell lines from canine OS. For the purpose of current study, we decided to characterize two each of primary culture isolates and established cell lines of canine OS (4 cell lines in total; 3 of them were robustly responsive to veraparnil inhibition and 1 cell line was incompletely responsive to verapamil inhibition). Evaluation of tumorigenic phenotypes in vitro: To gain insights into the possible differences between the roles of Hoechst- excluding and the Hoechst-retaining subpopulations of canine OS in tumor behaviors, we evaluated these cells with a set of in vitro assays for tumorigenic phenotypes. Cell growth analysis: To establish doubling time (DT), cells from all the experimental groups (unsorted and sorted subpopulations of 4 different canine OS cell lines) that were plated at low density were harvested for cell count at regular intervals over a period of 10 days. Our results (summarized in Table 7.2) indicated that although there was some difference among the cell lines, the differences among the unsorted, SP and non-SP subpopulations within the cell line was not significant in any of the OS cell lines we evaluated (p>0.05; one-way ANOVA). Similarly, there were some differences in saturation density, but these variations were not consistently related to the ability of dye-efflux. The saturation density of SP cells was significantly higher than that of both non-SP cells and unfractionated cells in MSU-DOS-IOM cell line (p<0.05; Tukey’s post hoc test). In contrast, the saturation density of non-SP cells were significantly higher than both SP cells and unfractionated cells (p<0.05; Tukey’s post hoc test). On the other hand, the 231 differences were not significant among these subpopulations in two other cell lines, D-l 7 and MSU-DOS-Gl (p>0.05; Tukey’s post hoc test). Thus, we conclude that the phenotype of Hoechst efflux is not consistently correlated with the phenotype of cell growth. Anchorage independent growth (AIG): The cells were assayed for their ability to form anchorage-independent colonies in soft agar. The results of this experiment are shown in Figure 7.5. We did not find a consistent relationship between SP phenotype and AIG, although there were obvious within-group (unfractionated, SP, and non-SP) differences as well as between-group (different cell lines) differences. For instance, non-SP cells displayed significantly higher level of anchorage independent grth than SP cells and unfractionated cells from D— l 7 and Abrams cell line (p<0.001; Tukey’s post hoc test), whereas SP cells exhibited significantly higher level of AIG than non-SP and unfractionated cells isolated from MSU-DOS-Gl and MSU-DOS-IOM. Moreover, there was no consistent pattern of colony size and SP phenotype. These results suggest that SP phenotype and AIG are not necessarily correlated, and other changes in the parental tumor cells might have an important role for determining AIG. Matrigel invasion assay: To investigate the possible differences in invasiveness between SP and non-SP cells, we performed matrigel invasion assay by adapting the method reported by Lin et a1 [45]. We first validated the assay using HT-1080 human fibrosarcoma cells as a positive control and NIH-3T3 mouse embryonic fibroblast cells as a negative control (data not shown). As shown in Figure 7.6, we found that SP cells from all of the 4 cell lines were 232 significantly more invasive than non-SP counterpart or original unfractionated cells (p<0.001; Tukey’s post hoc test) and (p<0.01 for MSU-DOS-MIO SP and non-SP). Serum dependence assay: When grown in low serum (0.1%) or serum-free medium, majority of cells from all of the experimental groups underwent death by day 12, irrespective of Hoechst efflux status of the plated cells (Table 7.3). A few non-SP cells from MSU-DOS-Gl were alive after 12 days. D-17 cells showed typical “capillary morphogenesis” (a reticular network of cord of cells), as reported earlier as a phenotype of tumorigenic MSCs after serum withdrawal [46], but this structure was deteriorated after day 7. Chemosensitivity assay: We determined the viability of cells from all experimental groups after 48 hrs of treatment with cisplatin and doxorubicin, the two commonly used drugs in the treatment of canine OS. Although the SP cells from all four cell lines had statistically significant survival advantage when compared to non-SP or unfractionated cells (p<0.001; Tukey’s post hoc test), the difference seemed too little to be biologically significant (Figure 7.7). Radiosensitivity assay: The difference in sensitivity among unfiactionated, SP and non-SP cells was measured by clonogenic survival assay. We counted the colonies after 10 days of exposure to single graded dose of radiation, and determined the colony forming efficiency of each treatment group. All the canine OS cell lines and most of their subsets were significantly more radioresistant than normal canine fibroblasts (p<0.001), which were in turn more radioresistant than intrinsically radiosensitive SCID fibroblasts (see 233 Figure 7.8). However, the Hoechst efflux-based grouping did not show consistent difference among the sub-groups in terms of their clonogenic survival response following radiation treatment. Although SP cells were slightly more radioresistant than non-SP cells in general, the differences were not significant in most cases (p>0.05), and in some cell lines, unfractionated cells were more radioresistant than either SP or non-SP (Figure 7.8). Gene expression studies: To evaluate the expression pattern of ABC transporters, we looked at 3 different OS cell lines for the expression of mRN A for 4 different transporter proteins-ABCG2, MDR-I, MRP-I, and ABCA3 (Figure 7.9). We found that expression of MDRI was higher in SP cells when compared to non-SP cells or parental cells (p<0.05 or p<0.001), but there was a considerable variation among the cell lines (less than 1 fold to more than 15 fold). Expression level of MRP1 was higher in SP cells than in non-SP cells from all 3 cell lines (p<0.05), but both of these fractions had lower level of expression when compared to their unsorted parental cells (in 2 out of 3 cell lines). On the other hand, ABCGZ expression was modestly higher in SP cells when compared to non-SP cells in only D-17 cell line (p<0.001), but it was downregulated in the SP fraction of two other cell lines (p<0.001). Similarly, expression of an intracellular ABC transporter-ABC A3 was significantly higher in SP fraction than in non-SP fraction of all cell lines (p<0.05 for D-I7 and p<0.001 for ABRAMS and MSU-DOS-IOM), although the expression level in SP cells were somewhat lower than that of parental cells in all of the 3 cell lines. We examined the gene expression profile of all three experimental subgroups of parental cells as well SP and non-SP cells using a panel of stemness-related genes, including pluripotency-associated gene NANOG and self-renewal-associated gene BMII 234 (Figure 7.10 A), sonic hedgehog family gene-GL1] and WNT family genes-LRP5 and WIFI (Figure 7.10 B), notch family genes-NOT CH1, DLL4, HE Y1, and JAGGED] (Figure 7.11) and, telomerase catalytic subunit T ERT , anti-apoptotic gene BCL2, minichromosome maintenance protein-MGM 7 (Figure 7.12). As shown in Figures 7.10, 7.11, and 7.12, there were marked variations among the sub-groups in terms of expression of different genes, but the differences were not always consistent among the subgroups. It is noteworthy that in many cases, expressions of these genes were higher in unsorted parental cells than in either SP or non-SP cells. For instance, expression of notch ligand-NOT CH1 was somewhat higher in SP cells than in non-SP cells from all 3 cell lines, but the expression level was lower than that of parental cells in 2 out of 3 cell lines (Figure 7.11). On the other hand, another cell line MSU-DOS-IOM had significantly higher level of NOT CH1 expression (~20 fold) when compared to parental cell line (p<0.001). Similarly, expression of two other notch ligands-JA G 1 and DLL4 were higher in SP fraction in 2 out of 3 cell lines when compared to non-SP cells (Figure 7.11). HE Y 1 , which is a downstream effector of notch signaling pathway, was also expressed in modestly higher level in SP cells (Figure 7.11). On the other hand, expression of GL1], a major transcriptional effector of sonic hedgehog pathway, was downregulated in SP cells than in non-SP cells in 2 out of 3 cell lines (Figure 7.10B). Expression of Wnt co-receptor-LRP5 was not considerably different between SP and non-SP cells, whereas Wnt inhibitory factor- WIFI was expressed at higher level in SP cells than in non-SP cells from 2 out of 3 cell lines (Figure 7.10B). Out of the 3 different cell lines examined, expression of pluripotency marker- NANOG was noticed only in subgroups of D-l7 cell line (Figure 7.10A). Similarly, when 235 compared to parental population, expression of mRNA for anti-apoptotic protein BCL2 was remarkably downregulated in non-SP fraction of D-17 cell line (>600 fold; p<0.001) and SP fraction of ABRAMS cell line (>10 fold, p<0.001) (Figure 7.12). Moreover, expression of proliferation marker-MCM 7 was significantly downregulated in SP cells (p<0.001) in only 1 out of 3 cell lines (MSU-DOS—lOM), when compared to either non- SP cells or parental cells. In the same way, expression of T ERT was higher in SP cells (p<0.001) than in non-SP cells in only 1 out of 3 cell lines (D-17) (Figure 7.12). 236 DISCUSSION There is an increasing evidence that several cancers contain stem-like cells which are largely responsible for tumorigenicity and malignancy [47]. These cells are more likely to be accountable for tumor recurrence after treatment. Therefore, identification and characterization of CSCs from various tumor types will contribute to improved therapeutic outcome. Active efflux of Hoechst 33342 dye by a subpopulation of tumor cells forms the basis for the SP phenotype, and this phenotype of dye-efilux has been regarded as a surrogate marker for drug-efflux following chemotherapy. Several recent studies have attempted to exploit this functional phenotype of a subset of tumor cells to identify and enrich CSCs. In the current study, we have successfully isolated SP cells from several canine OS specimens, including primary culture of canine OS tumors as well as established OS cell lines. We developed and optimized SP assay for canine OS, and used this approach to enrich and Ma characterize the Hoechst efflux-based phenotypes of subpopulations of canine OS cells. Our findings revealed the presence of SP cells in all of the primary isolates and established cell lines of canine OS that we examined. However, the frequency of SP cells varied widely among the different OS samples (0.1% to 1.6%). Veraparnil sensitivity of the cells in SP gate also differed widely (see results). We tested three different ABC transporter inhibitors at several concentrations. Interestingly, SP fraction disappeared almost completely from some of the cell lines after incubation with commonly used pump-inhibitors, where as the inhibition remained only partial in other cell lines we examined. 237 Although we did not formally rule out the possibilities of mutations and/or epigenetic modifications in the ABC transporter genes, these observations rather underscore the heterogeneity noticed in OS tumors and can have two possible explanations. Side population is a function of the ratio between passive influx and active efflux of Hoechst [48]. It is likely that multiple transporters are expressed and responsible for SP phenotype in canine OS, and these transporters are insensitive to commonly used pump-inhibitors (issue of “active efflux”). Alternatively, a subpopulation of cells within the SP fraction might have different cell membrane composition or limited surface membrane [49] which makes them refractory to Hoechst dye uptake (issue of “passive influx”). In any case, our results are consistent with another study on canine hepatic progenitor cells, where authors had noticed similar incomplete disappearance of SP cells from FAC S plot after verapamil treatment [50]. SP phenotype is not exclusively responsible for the prolonged in vitro life-span of cells, since not all the established cell lines contain SP cells [51]. In support of this notion, growth rate or saturation density of canine OS subsets did not depend upon Hoechst efflux phenotype. Moreover, there was no difference in susceptibility to death from long term serum deprivation. Only a fraction of cells in a tumor cell culture can form a colony in soft agar [52], and this ability of cancer cells has been regarded as one of the surrogate phenotypes of their tumorigenic ability. It has been postulated that this phenotype of AIG is related to stem or progenitor cell activity of cells. However, we found that many of the canine OS cells were capable of growing in soft agar at high fi'equency, and we could not see a consistent association of SP phenotype with anchorage independent growth (both in 238 terms of frequency and colony size) in soft agar. On the other hand, we found that SP cells from all cell lines were significantly more invasive than their non-SP counterpart or unfractionated parental cells. In earlier studies, SP cells from human lung cancer cell lines [53] and primary astrocytoma [54] were found to be more invasive than non-SP cells, and our current finding is similar to these observations. Increased expression of ABC transporters in SP fraction of tumors have been implicated in Chemoresistance [55,56]. Therefore, we examined the sensitivity of SP and non-SP subsets of canine OS cells to two commonly used drugs in the treatment of canine OS, doxorubicin and cisplatin. However, our findings indicated that SP cells did not show considerably different survival advantage over non-SP cells or parental cell population upon treatment with standard doses of either MDRI transporter-dependent doxorubicin or ABC transporter-independent cisplatin. It has been argued that CSCs are not always more resistant than their nontumorigenic progeny. In testicular cancer, for instance, undifferentiated cells are more sensitive to cisplatin therapy than the differentiated cells that they form [57]. Canine OS cells have been reported as fairly radioresistant cells [58]. When we compared their clonogenic survival differences with neonatal fibroblasts from normal and SCID dogs following X-ray irradiation, the OS cells were significantly more radioresistant than both normal fibroblasts and radiosensitive SCID fibroblasts. We wondered whether this difference was related to the differences in Hoechst efflux-based SP phenotype. However, our results showed that the differences between SP and non-SP cells were not significant and, in some cases unsorted cells were even more resistant than 239 SP and non—SP. Thus, we conclude that canine OS cells are intrinsically more radioresistant, but this cannot be attributed to their differences in Hoechst efflux capacity. We evaluated the gene expression profile of SP, non-SP, and unsorted parental cells from 3 different canine OS cell lines, with a panel of stem cell-associated genes. Although there were differences among sub-populations regarding the expression of one gene or another, these differences were not consistently related to Hoecsht efflux phenotype. In many cases, expression level was even higher in unsorted parental cells than cells isolated by SP assay. Our results indicate that differences among the OS cell lines were much higher than the differences within the sub-populations of cells sorted on the basis of Hoechst efflux. We could not find consistent differences between SP and non-SP fractions for the expression of majority of genes across the three 08 cell lines that we examined in this study. A number of studies have challenged the existence of stem cells within SP fraction. For instance, although SP cells were absent in ABCG2-deficient mice, these mice were viable and they had no hematopoietic abnormalities [59]. In addition, mature cells expressing lineage specific markers could be detected in SP fiaction of murine bone marrow [60]. Triel et a] had shown earlier that cells within SP fraction did not express stem cell markers [61]. Furthermore, SP fraction was found to be only partially enriched in stem cells in thyroid cancer cell line [62], and it was not a major CSC marker in adrenocortical carcinoma cell line [63]. Conversely, all stem cells may not display SP phenotype, as evidenced by the recent demonstration of absence of SP in human ES cells [64]. Our study underscores the fact that mere presence of SP within a tumor sample does not indicate the presence of stem-like cells within that tumor. However, further detailed 240 functional studies need to be carried out to evaluate the stem cell capacity of SP in any tumor type. We have validated the SP isolation technique for canine OS, and provided a general protocol for isolating SP cells from canine tumors. Isolation of SP may provide a simpler strategy to enrich putative C SC 5, study their biology and therapeutic response. However, our data suggested that SP analysis alone is not sufficient to accurately define C SC population in canine OS. The relevance of SP phenotype and stemness could vary among different tumor types and species, and so does the expression and function of transporter proteins. In fact, discrepancies have been noted in the identification of SP cells in human OS. Although Wu et a1 [65] had documented the presence of SP cells in human OS, another earlier study on two most widely used human OS cell lines (U208 and SaOSZ) had demonstrated the absence of SP cells [66]. It is noteworthy that identification of murine HSCs, but not human HSCs, was better achieved by Hoechst exclusion, and aldehyde dehydrogenase (ALDH)-based assay was found to be a method of choice for isolation of human HSCs [67]. Moreover, a recent study demonstrated that human ES cells do not express ABC G2 and also lack SP, which is in stark contrast to mouse ES cells which express ABCG2 as well as possess SP [68]. OS has been regarded as a highly heterogeneous tumor, and heterogeneity has been found both within cells of the same OS tumor and between cells from different OS tumors [69,70]. Our results supported and extended this concept, and this information will be useful to design more effective strategies to treat 08 in future. We suggest that a combination of blockers of ABC-transporter activity and cytotoxic agents may lead to a better clinical outcome for OS. 24] In conclusion, our study suggests SP assay as a simple strategy for characterization of subpopulations of tumor cells from canine OS which can then be further studied for their tumorigenic phenotypes. Given the disappointing results of past clinical trials on P-glycoprotein inhibitors [71], future efforts should be directed towards targeting multiple drug transporter proteins. We postulate that further characterization of ABC transporters and SP phenotype in the context of OS and careful selection of chemotherapeutic agents may lead to the development of more effective therapies. 242 APPENDIX Table 7.1. Primer sets used in qRT-PCR of stemness-related markers Ann Product Temp Gene Primer sequence (5' - 3') (13;) (°C) Forward GCGTGTCTGGAGGAGAAAGA ABC 62 Reverse TTCAGGAGCAAAAGGACAGC 133 60 Forward ATCC C AC TC GACC AGACATC MDR] Reverse GCTGAACAACTGTGCTCTTCC 120 60 Forward ATC TC C AGGC ACAGTCTCAG MRPI Reverse CACGATGTTGGTCTCCATCTC 120 60 Forward TC ATCACCTCC CACAGCAT ABCA3 Reverse AGTAGCCGCTGCCAAACTT 126 60 Forward C AAGGC AC GC AGAAAGAAAT NOTCH] Reverse ATGGAGACGACAGCAGAGGT 140 60 Forward CATGAACAACTTGTCGGACTTC DLL4 Reverse GCTCCTTCTTCTGGTT'TGTGTT 80 60 Forward TTGCTATGGACTATCGGAGTTTG HE Y 1 Reverse TAGTTGTTAAGGTGGGAGACCA 130 60 Forward TGAC C C CTAC TACCAGGATAA JA G] Reverse GTGACATCATCTCTT'TGTTGAAG 65 60 Forward AGTGGAAGATACGAAGGGAAGAT SHH Reverse ATATGATGTCGGGGTTGTAATTG 82 60 Forward ATCACAAGTCAGGCTCCTATCC GL1] Reverse TGC TC TATGGGAAGTCTI’GT‘TG 83 60 Forward GATGTGTGCGACAGTGACTACA LRP5 Reverse AGGGGTCTGAGTCTGAGT‘TCAA 82 60 Forward C AAAC TGCTCAACTACCTGCTT WIFI Reverse GGTTGTGGACACTTGCTTATTTC l 12 60 Forward GTGGATGACTGAGTACCTGAAC BCL2 Reverse TC AGAGACAGC C AGGAGAAGT I30 60 Forward TCTGGTACACAATCACGTCGTA T ERT Reverse C AC AGTAGAGTGGACAGGATA 96 60 Forward AGGATGCCACCTACACATCAG MCM 7 Reverse C TTC TC C ACTGTGTC C ACCAT 95 60 Forward GTCCATAAAGACGCAGCACTC SALL4 Reverse GC AAAGTC ACAGGGGTTC TC 138 60 Forward GAATAAC C CGAATTGGAGCAG NANOG Reverse AGCGATTCCTCTTC AC AGTTG 141 60 Forward ATGGACTGACAAATGCTGGAG BMII Reverse GGAACTGAGGATGAGGAGAC 13 60 BZM Forward TCTACATTGGGCACTGTGTCAC Reverse TGAAGAGTTCAGGTCTGACCAAG I36 60 243 Table 7.2. Growth properties of cells sorted on the basis of Hoechst dye efflux Population Saturation density Serum Serum doubling time dependence dependence (10" cells/cmz) 0 (hr) (0.] /0 serum) (no serum) CF“ ur" SP NS UF SP NS UF SP NSP UG SP NS lines P P P D-l 7 282‘: 29 27 42.5 46.5 44 d: 6.2 3.1 2.4 5.3 2.4 2.6 0.0 :0 :1: :t i 1.5 :t :t :1: :1: :t :1: 3 .05 0.0 2.1 1.9 1.9 0.9 0.8 0.8 0.7 0.9 5 ABRA 25 26 24 20 :1: 17.5 25 :1: 2.4 1.4 5.2 2.9 1.6 4.3 Ms :1: a: :1: 1.5 a: 1,9 :1: :1: :1: :t i :1: 0.0 0.0 0.0 1.2 0.7 0.7 0.9 0.7 0.4 0.8 5 6 3 MSU- 38 40 36 11.5 10 :1: 12.5 8.3 8.6 10.1 8.7 8.3 9.2 008- :t i i :t 1.6 d: :t :1: i :1: :1: :1: GI 0.0 0.0 0.0 1.2 1.4 1.2 1.3 2.3 1.5 1.2 1.8 8 7 8 MSU- 25 24 27 20 i 30 :1: 15 :h 5.7 7.3 4.3 5.4 7.6 4.1 DOS- :t i :t 1.8 2.1 1.9 :t i :1: :1: :1: :t 10M 0.0 0.0 0.0 0.5 0.8 0.7 1.2 0.5 0.9 4 6 5 °° UF=Unfractionated parental cells 244 Figure 7.1. Representative morphologies of canine OS cells. Polygonal cells (MSU DOS- 1) (A), loosely attached round to spindle-shaped cells (MSU DOS-2) (B), as well as elongated spindle-shaped cells (MSU DOS-7) (C) were evident in the primary culture. Magnification = 100X MSU DOS-8 MSU DOS-9 MSU DOS-10 Dog Liver l . Abrams ’ MSU DOS-3 :3 MSU DOS-5 , MSU DOS-6 « MSU DOS-7 32M MRPI WK] ABCGZ Figure 7.2. Expression of ABC 02, MDR-], and MRP-l in canine OS cell lines. Nine different canine OS cell lines were evaluated by RT-PCR for the expression of ABC transporters associated with drug resistance and stem cell phenotypes. Dog liver cDNA was used as a positive control and 32M CID-MICROGLOBULIN) was used a house- keeping gene. M represents 100 bp DNA ladder, with the lowest band representing 100 bps. 245 Figure 7.3. SP profile of canine OS cell lines (Groupl tumors). Typical profiles from one well-known, established canine OS cell line ABRAMS (A, B), and two in-house established primary culture isolates of OS tumors-MSU-DOS-IOM (C, D) and MSU- DOS-Gl (E, F) are shown. The P5 and P4 gates show Hoechst negative/dim SP cells, and Hoechst positive non-SP cells respectively- in the absence (A, C, E,) or presence (B, D, F) of efflux pump inhibitor-verapamil. The SP gate was set using the Hoechst-stained cells incubated with verapamil, and the gate was defined at the position which showed the maximal reduction of dye efflux in the presence of veraparnil. In this group of tumor samples, verapamil was able to robustly reduce the SP from P5 gate (92.58 :t5.03% reduction; n = 6; p<0.05, paired-t test). In some tumor samples (C-F), we consistently noticed another subpopulation of Hoechst dim cells on the lower right side (P2 gate), which was insensitive to veraparnil inhibition. Shown are representative results of I out of 3 independent experiments. 246 Hoechst blue llllllllllllllllIlllllllll IrlrllrlrlrILILllrrLrJLrl Hoechst red Figure 7.3. 247 rrrrllrralrrrrlral Hoechst blue SP=0.5% Hoechst red Figure 7.4. SP profile of canine OS cell lines (Group 2 tumors). In this subgroup of tumors, veraparnil was able to reduce the SP only partially (32.6 :t 11.7% reduction; n = 6; p<0.05, paired-t test). A typical profile of established cell line Gracie is shown here (A-D). Incubation with higher concentrations of verapamil (B), or use of other pump- inhibitor drugs such as reserpine (C) and FTC (D) gave similar results, and none of these drugs at any concentration were able to completely eliminate SP fraction in these cell lines. Shown are representative results of I out of 3 independent experiments. 248 60 I Parental 50 I SP 30 I NSP AIG% >0 e 0‘ $15 o ,9“ 00% 0°? 99' s59 5“ 45° Canine OS cell lines Figure 7.5. Anchorage independent growth of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. A total of 5 X 104 cells (parental, SP, and non-SP cells from 4 different canine OS cell lines) were seeded on soft agar in 60 mm plates, and cultured for 2 weeks. At the end, the numbers of colonies were quantified after counting them in 1cm2 area. A, B = D17; C, D = ABRAMS; E, F = MSU-DOS-GI; G, H =MSU- DOS-IOM. l, 2, and 3 represents parental, SP, and non-SP cells respectively. B, D, F, H corresponds to A, C, E, G respectively. Scale bars = 100 pm. Bar diagram shows the total frequency of AIG colonies. Data are represented as mean i sd (n = 3). 249 °/olnvaslon I Parental I SP I NSP Canine OS cell lines Figure 7.6. Matrigel invasion assay of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. A total of 5 x 10" cells from parental, SP and non-SP fractions were seeded in on the upper side of Matrigel Matrix and incubated for 24 hrs using 10% FBS as a chemoattractant. Cells that had invaded the Matrigel were quantified colorimetrically after cresyl violet staining. Data are represented as meanisd of triplicates. 250 Cisplatin E‘ E .2 > °\e I Untreated I Parental I SP 0 ‘13: ,o‘ 65“ Q vggy 00 co ,\ l Nsp 9 $9 Canine OS Cell lines Doxorubicin a? = .e .2 > ,\° I Untreated I Parental I SP I NSP Canine OS Cell lines Figure 7.7. Chemosensitivity assay of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. Cells from all experimental groups were treated with cisplatin and doxorubicin for 48 hrs, and viability was determined by MTS assay. Data are presented as mean at sd (n =3). 251 Figure 7.8. Radiosensitivity assay of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. The difference in sensitivity among unfractionated, SP, and non-SP cells were measured by clonogenic survival assay. Cells were seeded in 6-well plates, and colonies were counted after 10 days of exposure to single graded dose of radiation. Survival curve was established for each cell line. Results are expressed as mean i sd of triplicate wells. 252 % Survival 100 —o—Parental +SP -a- NSP O I I I D_l7 % Survival + Parental -I- SP +NSP ABRAMS % Survival -0-Parental ° -I- SP + NSP 0 2 5 10 MSU-DOS-Gl % Survival + Parental + SP + NSP MSU-DOS-IOM 10 2 5 Dose (Gy) % Survival -0-Regis-N -I-Reed Fibroblasts _ + Raleigh 10 2 5 Dose (Gy) Figure 7.8. 253 Figure 7.9. Relative mRNA expression of ABC transporter genes in of subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. Three different canine OS cell lines (D-l7, ABRAMS, and MSU-DOS-IOM) were evaluated for the expression of ABCGZ, MDR], MRP] , and ABCA 3 among three experimental groups (parental, SP, and NSP). Expression levels were normalized to endogenous reference gene ,82— MICROGLOBULIN, and expressed as fold change relative to unfractionated cells of each cell line. Normalization to second reference gene GAPDH showed very similar results (data not shown). 254 Fold change 0 ABC G2 MSU-DOS-IOM ABRAMS Cell lines I Parental I SP I NSP MDRI 80 E0 60 a I .5 40 Parental E 20 ISP 0 [g 0 D-17 ABRAMS INSP _20 MSU-DOS-IOM Cell lines 4 MRPI MSU-DOS-IOM g, 2 D-17 ABRAMS 5 IParental 5 0 E ISP 1?. -2 INSP .4 Cell lines 10 ABCC3 MSU-DOS-IOM 0 3’» 5 -10 IParental 5 g _20 ISP Er. INSP -30 -40 Cell lines Figure 7.9. 255 Figure 7.10. Expression of stemness-associated genes in subpopulations of canine OS cells sorted on the basis of Hoechst dye efflux. Relative mRNA expressions of NANOG and BMI] (A), as well as Hedgehog family gene GL1] and Wnt family genes LRP5 and WIFI (B) were compared by qRT-PC R among three experimental groups (parental, SP, and NSP) from three different canine OS cell lines (D-17, ABRAMS, and MSU-DOS- IOM). Expression levels were normalized to endogenous reference gene ,82- MIC ROGLOBULIN, and expressed as fold change relative to unfractionated cells of each cell line. Normalization to second reference gene GAPDH showed very similar results (data not shown). 256 I Parental I SP I NSP BMII N/D N/D NANOG 2o _ -wOQ-DmE m2