|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Src Kinase Inhibitors for the Treatment of Sarcomas: Cellular and Molecular Mechanisms of Action by Audrey Cathryn Shor, M.P.H. A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Molecular Medicine College of Medicine University of South Florida Co-Major Professor: Richard Jove, Ph.D. Co-Major Professor: W. Jack Pledger, Ph.D. Denise R. Cooper, Ph.D. Larry P. Solomonson, Ph.D. Daniel M. Sullivan, M.D G. Douglas Letson, M.D. Carlos A. Muro-Cacho, M.D., Ph.D. Date of Approval: March 8, 2007 Keywords: tyrosine kinase inhibitor, da satinib, cancer therapy, microarray, gene expression profile Copyright 2007 Audrey Cathryn Shor
Dedication I dedicate this to my father, Richard C. Burns for his love and devotion inspiring my yearning to seek more. And to my wonderful, caring husband Adam For his endless love and support throughout our journey.
Acknowledgements I would like to acknowledge S ean Yoder and the rest of the members of the H. Lee Moffitt Microarray Core for completing the mi croarrays and aiding in the analysis. I would also like to graciously thank Samuel Falsetti for the coffee breaks and brainstorming sessions throughout all of our hurdles. Ralf Buettner for helping me get started in the beginning, layi ng the foundation and molding me into the scientist I am today. Bonnie Goodwin, for being there for me during the most difficult year of my life. Tania Mesa for keeping me sane w ith our long runs every weekend during the final stages. Lastly, I would like to thank my wonderful mentors, Drs. Rich Jove and Jack Pledger for having faith in me, Dr s. Doug Letson and Carlos Muro-Cacho for their constant support.
i Table of Contents List of Tables................................................................................................................. ....vi List of Figures..................................................................................................................viii List of Abbreviations.........................................................................................................x i Abstract....................................................................................................................... .....xiv Introduction................................................................................................................... .......1 Epidemiology of Sarcomas......................................................................................1 Risk Factors Investigated for the Development of Sarcoma...................................4 Limitations of Current Treatment Options..............................................................8 Role of Signal Transduction in Cancer....................................................................9 Applications of Tyrosine Ki nase Inhibitors in Sarcomas......................................14 c-KIT and PDGFR.....................................................................................14 EGFR/HER-2/neu......................................................................................17 VEGFR......................................................................................................19 Other TKIs.................................................................................................21 History of Src: 1911Present.................................................................................23 Development of Src Inhibitors for Clinical Trials.................................................31 Objectives..........................................................................................................................35 Aim 1: Evaluate the Biological Response of Three Src Kinase Inhibitors in Sarcoma Cell Lines.............................................................................................35
ii Aim 2: Identify a Molecular Signature that Predicts Response to Dasatinib by Induction of Apoptosis in Sarcoma Cell Lines...................................................36 Aim 3: Verify the Presence of the Mo lecular Signature in Primary Human Sarcoma Specimens............................................................................................37 Materials and Methods.......................................................................................................38 Cells and Reagents.................................................................................................38 Preparation of Cell Extr acts and Western Blotting................................................39 Immunohistochemistry..........................................................................................40 Src Family Kinase PCR.........................................................................................41 Wound Healing Assay for Cell Migration.............................................................41 Cell Invasion Assay...............................................................................................42 TUNEL Assay........................................................................................................42 siRNA Transfections..............................................................................................43 Statistical Analysis.................................................................................................43 Microarray Sample Preparation.............................................................................43 Isolation of RNA....................................................................................................44 Preparation of Labeled RNA Targets for Hybridization........................................44 Array Hybridization and Scanning........................................................................46 Data Analysis.........................................................................................................47 Preparation of Samples for Quan titative Real-Time PCR Analysis......................48 cDNA Reactions....................................................................................................48 Real-Time PCR Reactions.....................................................................................49 Personal Health Identifier......................................................................................49
iii Involvement of Human Subjects............................................................................50 Introductory Preclinical Data.............................................................................................51 Results....................................................................................................................53 Expression of SFKs in Sarcoma Cell Lines...............................................53 STAT3 is Activated in Human Sarcomas and Sarcoma Cell Lines...........53 PD180970 Inhibits Src and STAT3 Signaling in Human Sarcoma Cell Lines................................................................................................58 PD180970 Inhibits Cell Viability an d Induces Apoptosis in Sarcoma Cell Lines................................................................................................60 SKI-606 Does Not Inhibit Src-ST AT3 Signaling in Human Sarcoma Cell Lines................................................................................................63 Dasatinib Does Not Inhibit Src-STAT3 Signaling in Human Sarcoma Cell Lines................................................................................................64 Discussion..............................................................................................................66 Dasatinib Inhibits Migrati on and Invasion in Diverse Hu man Sarcoma Cell Lines and Induces Apoptosis in Bone Sarc oma Cells Dependent on Src Kinase for Survival................................................................................................................ .....69 Results....................................................................................................................71 Src Kinase is Activated in Human Sarcomas and Sarcoma Cell Lines................................................................................................71 Dasatinib Inhibits Src Kinase Ac tivity in Human Sarcoma Cell Lines.....73 Dasatinib Selectively Blocks Src Downstream Signaling.........................77 Dasatinib Blocks Cell Motility and Invasion by Sarcoma Cells................78
iv Dasatinib Induces Apoptosis of Bone Sarcoma Cell Lines.......................81 Src is required for survival of bone sarcoma cell lines..............................84 Discussion..............................................................................................................88 Gene Expression Profile of Sarcoma Cell Li nes Serves as Preliminary Signature Predictive of Response to Treatment with Dasatinib......................................................91 Results....................................................................................................................93 Unsupervised Clustering Identified Three Main Classes, with Five Subgroups in Relation to Cell Line Types ..............................................93 A Molecular Signature Distinguishes Res ponse to Dasatinib as Defined by Induction of Apoptosis............................................................................95 Identification of Two Probesets with Greater Fold Changes May Provide Further Insight into the Prediction of Response...................................105 Testing the Molecular Signature with Cell Lines of Unknown Response Reveals that the Molecular Sign ature Can Accurately Predict Response to Dasatinib in Cell Lines.....................................................................112 Discussion............................................................................................................126 Validation of Gene Expression Prof ile in Primary Human Sarcomas.............................129 Results..................................................................................................................131 Unsupervised Clustering Identifi ed Diversity Among Human Sarcoma Specimens.............................................................................................131 Identification of the Cell Line Molecular Signature in Human Sarcomas...............................................................................................131
v Molecular Signature that Predicts Response to Dasatinib in Cell Lines Can be Used to Group Tumors by Potential Response.........................137 Analysis of Two Probesets with Gr eater Fold Changes May Provide Further Insight into the Prediction of Response in Tumors..................141 Discussion............................................................................................................145 Conclusions......................................................................................................................148 Clinical Significance........................................................................................................15 7 References........................................................................................................................159 About the Author...................................................................................................End Page
vi List of Tables Table 1 Relative Frequency of Subtypes of Sarcomas.............................................3 Table 2 Incidence of Soft Tissue Sarcomas Diagnosed in the US by Race, SEER, 1998-2002.....................................................................................3 Table 3 Potential Risk Factors Eval uated by Epidemiological Studies....................6 Table 4 Development Stage of TKI Targeted Agents in Sarcoma.........................13 Table 5 SFK Gene Expression Status in Human Sarcoma Cell Lines....................55 Table 6 Summary of Cell Line IC 50 Values and Responses to Dasatinib...............87 Table 7 Fold Changes for Key Signatu re Genes in Sarcoma Cell Lines Compared to Median Intensity................................................................99 Table 8 Quantitative RT-PCR Expressi on of Key Signature Genes in Sarcoma Cells ......................................................................................101 Table 9 Fold Changes for Key Signature Genes in Sarcoma Cells Compared to Median Intensity of Non-Responders...............................................104 Table 10 Fold Changes for Ephrin-A1 and Dapper in Sarcoma Cell Lines Compared to Median Intensity..............................................................108 Table 11 Microarray Fold Changes for Ephrin-A1 and Dapper in Sarcoma Cell Lines Compared to Median Intensity of Non-Responders...................109 Table 12 Quantitative RT-PCR Values for Ephrin-A1 and Dapper in Sarcoma Cell Lines..............................................................................................111
vii Table 13 Fold Changes for Key Signature Genes in Test Cells Compared to Median Intensity...................................................................................117 Table 14 Relative qRT-PCR Fold Changes Key Signature Genes in Test Cells Compared to Median Intensity..............................................................117 Table 15 Fold Changes for Ephrin-A1 and Dapper in Test Cells Compared to Median Intensity...................................................................................125 Table 16 Microarray Fold Changes fo r Ephrin-A1 and Dapper in Test Cells Compared to Median Intensity of Non-Responders.............................125 Table 17 Classification of Sarcomas using Predictive Signature from Cell Lines......................................................................................................134 Table 18 Fold Changes for Key Signature Genes in Sarcomas..............................140 Table 19 Fold Changes for Ephrin-A1 and Dapper in Sarcomas...........................144
viii List of Figures Figure 1 Illustration of Associated Tyro sine Kinases and Targeted Therapies.......12 Figure 2 Organization of Src Kinase........................................................................26 Figure 3 Src Activation and Signaling Conformations............................................27 Figure 4 Structure of Three ATP-Co mpetitive Src Kinase Inhibitors......................33 Figure 5 Evaluation of SFK Expressi on in Human Sarcoma Cell Lines.................54 Figure 6 STAT3 Activation Status in Human Sarcoma Tissues and Cell Lines................................................................................................56 Figure 7 PD180970 Inhibits Src and STAT3 Signaling in Sarcoma Cell Lines................................................................................................59 Figure 8 PD180970 Inhibits Viability and Induces Apoptosis in Sarcoma Cells Lines..............................................................................................61 Figure 9 PD180970 Induces Apoptosis in Sa rcoma Cell Lines as Measured by TUNEL...................................................................................................62 Figure 10 SKI-606 Does Not Inhibit Src-STAT3 Signaling in Sarcoma Cell Lines................................................................................................63 Figure 11 STAT3 Signaling is Independent of Src Kinase Activity in Human Sarcoma Cell Lines.................................................................................65 Figure 12 Src Kinase is Activated in Sarcoma Tissues and Cell Lines.....................72
ix Figure 13 Dasatinib Inhibits Sr c Activation and Downstream Signaling in Sarcoma Cell Lines................................................................................................74 Figure 14 Dasatinib Does Not Induce c-Src mRNA Expression...............................76 Figure 15 Dasatinib Inhibits Cell Motility and Invasion............................................79 Figure 16 Dasatinib Induces Apoptosis in Bone Sarcoma Cell Lines.......................82 Figure 17 Src is Required for Bone Sarcoma Cell Line Survival..............................85 Figure 18 Hierarchical Clustering of Sarcoma Cell Lines Based on GEP.................94 Figure 19 Molecular Signature Generated Us ing the Median Value for Each Probe Set as Comparison Group.......................................................................97 Figure 20 Quantitative RT-PCR Validations of Histone H3, FAF1, -Catenin and -Catenin Genes.............................................................................100 Figure 21 Molecular Signature Generated Us ing the Median Value for Each Probe Set in the Non-Responders as Comparison Group...............................102 Figure 22 Expression of Ephrin-A1 and Da pper in Sarcoma Cell Lines Using the Median of All Samples as Comparison Group.....................................106 Figure 23 Expression of Ephrin-A1 and Da pper in Sarcoma Cell Lines Using the Median of the Non-Re sponders as Comparison Group........................107 Figure 24 Quantitative RT-PCR Validations of Ephrin-A1 and Dapper Genes.....................................................................................................110 Figure 25 Molecular Signature that Pred icts Response to Dasatinib Generated Using the Median Value for Each Probe Set as Comparison Group....................................................................................................114
x Figure 26 Dasatinib Inhibits Src Activ ation and Signaling in HOS and SW1353 Cells......................................................................................................118 Figure 27 Dasatinib Induces Apoptosis in HOS but Not SW1353 Sarcoma Cells......................................................................................................120 Figure 28 Expression of Ephrin-A1 and Dapper in HOS and SW1353 Cells Using the Median of All Samples as Comparison Group.....................121 Figure 29 Expression of Ephrin-A1 and Dapper in HOS and SW1353 Cells Using the Median of the Non-Responders Comparison Group............123 Figure 30 Hierarchical Clustering of Primary Sarcoma Specimens.........................132 Figure 31 Heatmap of Mo lecular Signature Expr ession in Sarcomas......................135 Figure 32 Expression of Histone H3, FAF1, -Catenin and -Catenin in Sarcomas...............................................................................................138 Figure 33 Expression of Ephrin-A1 and Dapper in Sarcomas.................................142
xi List of Abbreviations AJ Adherens Junctions AS Adenosarcoma ATP Adenosine Triphosphate CAS Crk and Src Associated Substrate Chk Csk homologous kinase CS Chondrosarcoma DFSP Dermatofibrosarcoma Protuberans DMSO Dimethyl sulfoxide EGFR Epidermal Growth Factor Receptor EMSA Electrophoretic Mobility Shift Assay EWS Ewings sarcoma FAF1 Fas Associated Factor 1 FAK Focal Adhesion Kinase FBS Fibrosarcoma Fig Figure GEP Gene Expression Profile GIST Gastroinstestinal Stromal Tumor h Hour HSP Heat Shock Protein
xii IC 50 50% Inhibitory Concentration ICD International Classification of Diseases IGFR-1 Insulin-like Grow th Factor Receptor-1 IHC Immunohistochemistry LMS Leiomyosarcoma LPS Liposarcoma MAS Affymetrix Microarray Suite MHF Malignant Fibrous Histiocytoma MFS Myxoid Fibrosarcoma NF1 Neurofibromatosis type 1 N/R No Response NRTK Non-receptor Tyrosine Kinase OSA Osteosarcoma PDGFR Platelet Derived Gr owth Factor Receptor qRT-PCR Quantitative Real-Time Polymerase Chain Reaction Rb Retinoblastoma RD Rhabdomyosarcoma RSV Rous sarcoma virus RTK Receptor Tyrosine Kinase RT-PCR Real-Time Polymerase Chain Reaction SAM Statistical Analysis of Microarrays Ser Serine SCS Synovial Cell Sarcoma
xiii SFK Src Family Kinase SH Src homology siRNA Silencing RNA STAT Signaling and Transducer of Activation Transcription STS Soft Tissue Sarcoma TK Tyrosine Kinase TKI Tyrosine Kinase Inhibitor Tyr Tyrosine U Units VEGFR Vascular Endothelial Growth Factor Receptor Y Ty
xiv Src Kinase Inhibitors for the Treatment of Sarcomas: Cellular and Molecular Mechanisms of Action Audrey Cathryn Shor ABSTRACT Sarcomas are rare mesenchymally-derived tumors with limited treatment options. Tyrosine kinases may serve as potential targets for sarcoma therapy because many are mutated or overexpressed in sarcomas and cel l lines. One potential molecular target for sarcoma treatment is the Src tyrosine kinase. Three independently synthesized Src kinase inhibitors were evaluated in human sarcoma cell lines. Of the three, dasatinib, provided promising results as a potential sarcoma therapy. Until this study, dasatinib activity had not been characterized in sarcoma cells. Base d on our previous findi ngs of Src activation in human sarcomas, we evaluated the effects of dasatinib in twelve sarcoma cell lines. Dasatinib inhibited Src activity and downstr eam signaling at nanomolar concentrations. Inhibition of Src signaling was accompanied by blockade of cell migration and invasion. Moreover, apoptosis was induced in a s ubset of bone sarcomas at nanomolar concentrations of dasatinib. Inhibition of Src protein expression by siRNA also induced apoptosis, indicating that these bone sarcoma cell lines are dependent on Src activity for survival. These results demonstrate that dasatinib inhibits migration and invasion of diverse sarcoma cell types, and selectively bl ocks the survival of bone sarcoma cells.
xv Therefore dasatinib may provide therapeu tic benefit by preven ting the growth and metastasis of sarcomas. Microarray analysis of the sarcoma cell lines lead to the identification of a molecular signature that successfully predic ts response to dasatinib by induction of apoptosis. Components of this molecular signature are expressed in primary human sarcomas. Furthermore, expression of th e molecular signature in sarcomas can be utilized to cluster tumors ba sed on theoretical response to da satinib. While the prediction of response in tumors is theoretical, there is encouraging evidence to support further endeavors into validating the potential of this molecular signa ture to predict response in patients. Together, these studies reveal that, in cell lines, bot h constitutive Src activation and the presence of a molecular signature th at predicts response to dasatinib are important parameters to consider when se lecting dasatinib as a treatment for. Furthermore, novel therapeutic approaches that inhibit Src signaling may selectively induce apoptosis in tumor cells and sensitize to chemothera py those tumors that contain the relevant molecular signature.
1 Introduction Epidemiology of sarcomas Sarcomas comprise a relatively rare and diverse group of malignant tumors that arise from mesenchymally-derived connectiv e tissues including bone, fat and muscle. There are more than 50 different subtypes of sarcomas, with approximately 12,000 new cases diagnosed nationwide each year (1, 2). While this represents a fraction of all cancers diagnosed in the United States (3), sarcomas account for approximately 20% of newly diagnosed pediatric solid tumor mali gnancies (4) and are among the cancers that pose the greatest risks of mort ality and morbidity in children and young adults (5-9). Histologically, sarcomas are divided into two sub-categories, soft-tissue and bone sarcomas. There are over 9,000 cases of soft tis sue sarcomas expected to be diagnosed in the United States this year and 5,500 Ameri cans are expected to die as a result this disease (10). Less than 1,000 soft tissue sarcom as are diagnosed in children each year in the United States, with an annual incidence of approximately 11 per million (11). The most commonly diagnosed soft tissue sarc omas (STS) include malignant fibrous histiocytomas (MFH), liposarcomas (LPS) a nd leiomyosarcomas (LMS) (Table 1) (1012). In 2006, over 2,760 new cases of bone sarc oma are expected to be diagnosed in the United States and 1,260 Americans are expected to die from complications from these tumors (13). Bone sarcomas may occur at any age and account for less than 0.2% of all cancers diagnosed each year, with 30% observe d in children and young adults (13).
2 Osteosarcoma (OSA) is the most commonly diagnosed primary bone tumor followed by Ewings sarcoma (EWS), chondrosarcoma (C S) and fibrosarcoma (FBS) (13, 14). In the United States, sarcomas affect men more often than women, and white men have the highest overall incidence among all ci tizens (Table 2) (4,17). There is a high frequency of sarcomas reported in Uganda and Zimbabwe which is related to the high prevalence of HIV/AIDS infections in thos e countries (18). Is rael has the highest prevalence of all non-African na tions with an elevated agestandardized incidence of diagnoses that is consistent w ith previous reports of increas ed incidence of sarcomas in men of Jewish background (19, 20). The detection of specific molecular abnor malities has contributed significantly to sarcoma classification while eliciting an interest in targeted therapies (1). Several welldescribed genetic mutations that predispose individuals to specific types of sarcomas have been described. The most common ar e mutations in the retinoblastoma (Rb) pathway, which increase the risk of devel oping osteogenic sarcomas (21), Li-Fraumeni syndrome, which is caused by a p53 mutation, increases the risk of a variety of sarcomas (22-25), deletion of the neurofibromatosis t ype 1 (NF1) gene increases the risk of malignant nerve sheath tumors (26) and c-kit gene mutations are responsible for the development of gastrointes tinal stromal tumors (GIST) (27).
3 Table 1: Relative frequency of subtypes of sarcomas Histological Subtypes % Malignant fibrous histiocytoma 28.15 Liposarcoma 15.16 Leiomyosarcoma 11.94 Synovial sarcoma 10.08 Malignant peripheral nerve sheath tumor 5.81 Rhabdomyosarcoma 4.84 Unclassified sarcoma 11.29 Fibrosarcoma 3.06 Ewings sarcoma 2.02 Angiosarcoma 2.02 Osteosarcoma 1.13 Epthelioid sarcoma 1.13 Chondrosarcoma 1.05 Clear cell sarcoma 0.97 Alveolar soft part sarcoma 0.56 Malignant hemangiopericytoma 0.40 Adapted from (15) Table 2: Incidence of soft tissue sarcomas diagnosed in the US by race, SEER, 1998-2002 Race/Ethnicity Men Women All Races 3.5 per 100,000 men 2.4 per 100,000 women White 3.6 per 100,000 men 2.4 per 100,000 women Black 3.5 per 100,000 men 2.6 per 100,000 women Asian/Pacific Islander 2.7 per 100,000 men 1.6 per 100,000 women Hispanic 3.0 per 100,000 men 2.3 per 100,000 women Adapted from (16)
4 Risk factors investigated fo r the development of sarcomas Few risk factors have been associated in the literature with the development of sarcomas (reviewed in Table 3). Epidemiologi c studies have identified radiation, certain chemotherapies and vinyl chloride as signifi cant factors while data is conflicting with regard to pesticides and he rbicides. Currently, phenoxyace tic acid and chlorophenol have been accepted as causal agents in the de velopment of sarcomas, although the data generated from epidemiologic studies investig ating this associati on do not support this conclusion, (28-38). Dioxin and Agent Orange more potent forms of pesticides and herbicides, have not been shown to play a si gnificant role in the induction of sarcomas (39-49). There have been no published re ports suggesting an association between viruses and the development of sarcomas in humans to date although ongoing investigations are trying to be tter understand the role of vi ruses in sarcomagenesis (50, 51). The relationship between traumatisms, bur ns and scars and sarcomagenesis is weak. Case series and isolated reports are the only studies available to assess the relevance of these risk factors and a measure of the magnitude of an association has not been established so far. Even in case-control studies a temporal relati onship is difficult to establish. Nevertheless, there have been numerous reports of sarcoma development in the location of previous injuries. Howeve r the lack of objective quantitation and the uncertainty in the temporal sequence of ev ents has prevented the establishment of a definitive exposure-outcome. Many difficulties plague epidemiological st udies of sarcoma. One example is the previous faults inherent of the International Classification of Diseases (ICD) system. The
5 identification of sarcomas was complicated by inefficient coding rules, which until recently, classified sarcomas according to visc eral organ or connective tissue of origin, rather than placing sarcomas in one specifi c ICD category. As a result, many sarcomas were not accurately reported and this has complicated retrospective analyses. It is expected that recent improvements in the methodology will improve recognition. Another concern of epidemiology studies has been the poor correlation between the diagnoses recorded on death cert ificates and hospital records. The rarity of sarcomas leads to a greater chance of misdiagnosis and misclassification. This is complicated by di scordant diagnostic opinions even among experts. Furthermore, methodological de ficiencies in study design to investigate potential risk factors have been common. An example is the difficulty in determining a precise exposure to a risk f actor and, often, surrogate meas urements are used to estimate average exposures. Also, in the case of pes ticides and herbicides, it has been difficult to precisely determine the exact chemical compos ition of the substances of interest in a given cocktail. Thus, it is speculated that that the observed risk from exposure to phenoxy herbicides may become insignificant when controlling fo r some of these concomitant exposures. Thus, further studies are needed to establish a relationship between exposure to phenoxyacetic acids and chlorophenols and development of sarcomas. The known epidemiologic risk f actors of sarcomagenesis have had major public health implications but more research is still needed. Sarcoma is an extremely rare cancer and there are insufficient cases to conduc t valid, classic epidemiologic studies.
6 Table 3: Potential risk factors evalua ted by epidemiological studies. Risk Factor Type of Study Conclusion Reference Radiation Case-series and Case-reports Definitive (38, 52-76) (77-80) Thorotrast Case-control Probable (81-84) Breast Conservation Therapy Case-series and Case-reports Probable (85, 86) Chemotherapy Retrospective cohort Definitive (87) Case-series Probable (88) Case-control Probable (73) Tamoxifen Case-series Possible (89-92) Vinyl Chloride Retrospective cohort Definitive (93-97) Retrospective cohort Unclear (98) Case-series Probable (99) Pesticides & Herbicides Retrospective cohort No Association (34, 37, 38) Case-control No Association (32, 36) Retrospective cohort Possible (35) Case-Control Possible (31, 33) Occupational Study Possible (28-30) Dioxin Retrospective cohort Possible (41, 44) Prospective cohort No Association (40) Agent Orange Case-control No Association (39, 42, 43, 45-49) Lymphadema Case-report Possible (100) Case-series Possible (101) Trauma Case-series Possible (102) Case-control Possible (103) Foreign Body Implantation Case-report Probable (104-111) Burns & Surgical Scars Case-series Possible (112-119) Viruses Retrospective cohort No Association (50, 51)
7 A molecular epidemiology approach guide d by classic principles may provide a more appropriate venue. The da ta reported in the literatu re supports causal roles for radiation, vinyl chloride, certain chemotherapi es and pesticides and herbicides containing phenoxyacetic acid and chlorophenol exposur es, as well as previous cases of lymphedema. This evidence has restricted o ccupational exposures to vinyl chloride, pestand herbicides. Radiation therapy has been re-evaluated and lesser doses are currently used to try to prevent the occurrence of future sarcomas. Physicians are monitoring patients who are receiving treatments such as tamoxifen in order to detect adverse effects before a sarcoma develops. Furthermore, the problems evident in the ICD coding of sarcomas have prompted a more sophisticated method of classifying ra re diseases. This will aid in resolving the difficulties in detecting each case and allow for more thorough and complete studies of rare disease such as sarcomas in the future. There have been recent advances in the understanding of sarcoma molecular biology. Employing what is currently known of the epidemiologic ri sk factors and the knowledge gained at the bench will help ta ilor design future studies. These studies should correlate environmental exposures with the molecular biology of the disease. Collectively, epidemiologists and basic scientists can achieve a more thorough understanding of sarcomagenesis. Together wi th a more efficient classification system, this will provide a more precise approach in the design of future clinical trials and treatments, as well as a greater understanding of potential risk factors of sarcomas yet to be identified.
8 Limitations of current treatment options The primary approach to treating sarcomas is surgery. Radiation therapy is also used following surgery for patients with unr esectable or residual tumors. Systemic chemotherapy, such as doxorubicin and ifosphami de are used to treat patients with metastatic disease. Pediatric and certain cases of chemosensitive sarcomas undergo chemotherapy treatment following surgery and/or radiation treatment as an effort to prevent recurrence or metastas is. However, because many sarcomas harbor Rb mutations and exposure to radiation increases the risk of developing sarcomas, variations in the method of treatment is required to prevent fu rther complications. While this approach has proven effective in treating patients with localized disease, in many cases only modest degrees of success are achieved. Fu rthermore, treatment for patients with advanced and/or metastatic disease remain s limited and chemotherapy is palliative in nature at best. Doxorubicin was first employe d as a treatment option for metastatic soft tissue sarcomas in 1972 at MD Anderson, Houston, TX (120) and ifosphamide was introduced in the late 1980s (121). The outcome for most adult patients with advanced sarcoma still remained bleak, however. Recently, the power of molecular biology was clearly demonstrated with the introduction of Gl eevec as a treatment option for GISTs (122, 123). Revealing a new method of approaching sarcoma treatment. To fully appreciate this, the role of signal transduc tion in cancer and most specifi cally in sarcomas must first be reviewed.
9 Role of signal transduction in cancer Protein phosphorylation plays a key role in nearly every aspect of cellular molecular biology. It is c ontrolled by kinases, which phosphorylate, and phosphates, which de-phosphorylate proteinserine, -threonine and -tyros ine residues. To date, 90 tyrosine kinases 58 transmembrane recepto rs and 32 non-receptor proteins (124) and 107 protein-tyrosine phosphatase genes (125) ha ve been identified in the human genome. Signals are communicated in the cell either throu gh external or internal stimuli. External stimulation is elicited through cell receptors and internal stimulati on is processed through non-receptors tyrosine kinases which activate sp ecific proteins and transmit the signal to other proteins within the cell, ultimately tran sducing the signal to the nucleus, regulating gene transcription. Cancer aris es as a result of aberrant ac tivation or stimulation of signal transduction pathways. Genetically, sarcomas fall into two su bgroups, those with complex karyotypes characteristic of severe genomic instabili ty and those characterized by near-diploid karyotypes. Sarcomas with simple, near-d iploid karyotypes usually possess specific chromosomal translocations. Sarcomas with complex karyotypes have high frequencies of p53 and Rb mutations as well as impairments in DNA repair and severe chromosomal instability. This group includes some of th e more commonly diagnosed sarcomas; EWS, LMS, rhabdomyosarcoma (RD) and OSA. Ma ny sarcomas that fall into both genetic subgroups also possess abnormalities in growth factor signaling and signal transduction pathways. Tyrosine kinases (TKs) make up th e majority of defectiv e signaling pathways in sarcomas, including mutations in the platel et-derived growth factor receptor (PDGFR), c-KIT, vascular endothelial growth factor (VEGF) and in sulin-like growth factor-1
10 receptor (IGF1-R) signaling pathways. For example, GISTs, EWS, dermatofibrosarcoma protuberans (DFSP), synovial sa rcoma and Kaposis sarcoma all have been shown to have mutations which elicit c-kit overexpression and/or P DGFR over-stimulation (1). Tyrosine kinase inhibitors (TKIs) are a class of novel therapeutics that are effective alone and in combination with c onventional chemotherapeutics in treating a variety of cancer subtypes. There are two ba sic classes of TKs, receptor tyrosine kinases (RTKs) and non-receptor tyrosine kinases (N RTKs). Upon TK activation, a cascade of protein interactions occur releasing signals of positive and negative regulators of a variety of cellular processes including cell cycle regul ation, proliferation, a dhesion, migration, invasion, transcription and survival. Unde r normal conditions, cellu lar signaling tightly regulates activated TKs. Th e induction of TK signaling in the oncogenic state overcomes controlled regulation and signaling become s activated by a myriad of cellular mechanisms including mutation and over-expre ssion of the TK receptors or receptor ligands. All TKs rely on adenosine triphospha te (ATP) to mediate the transfer of energy in the kinase domain and to elicit do wnstream signaling cascades via pathway intermediates. TKIs employ two strategi es: (1) antibodies which act as receptor antagonists or to sequester the TK ligand, pr eventing the ligand from binding to the RTK, and (2) small molecule inhibitors which ac t by competing for the ATP-binding domain in the catalytic site of the enzyme (Figure 1). The use of TKIs for the treatment of sarc omas is predicated on the hypothesis that malignant cells rely more heavily on TK si gnaling than neighboring normal cells. GIST tumors were previously untreatable with our current chemotherapeutic options. Advances in understanding the genetic nature of cancer have lead to the development of new
11 treatment for sarcoma. GISTs that harbor activating mutations in the c-kit tyrosine kinase are sensitive to treatment with imatinib mesy late (Gleevec) as well as sunitinib malate, whereas those without c-kit mutations are insensitive. Proof-of-principle of this concept has already been obtained in the sarcomas in the case of GISTs. TKIs currently under investigation as potential therapies for sarcomas are reviewed in Table 4.
12 Figure 1. Illustration of associated ty rosine kinases and targeted therapies. Several TKIs have been evaluated as potenti al therapeutics in sarcomas. TKI can inhibit TK signaling and activation by preventing ligand binding to re ceptors or by directly inhibiting the TK by binding in the catalytic domain of the kinase Here, the targets of several TKIs are depicted. RTKsPDGFR EGFR IGF-1R VEGFR Downstream Signaling RasNRTKsc-KIT Src Ligands Her-2/ neuBevacizumabInduction Inhibition Imatinib Sunitinib Semaxanib Gefinitib NVP-AEW541 Semaxanib SU6668 CEP-5214 CEP-7055 Sunitinib Sorafenib Dasatinib Imatinib Sunitinib VEGF Trastuzumab RTKsPDGFR EGFR IGF-1R VEGFR Downstream Signaling RasNRTKsc-KIT Src Ligands Her-2/ neuBevacizumabInduction Inhibition Imatinib Sunitinib Semaxanib Gefinitib NVP-AEW541 Semaxanib SU6668 CEP-5214 CEP-7055 Sunitinib Sorafenib Dasatinib Imatinib Sunitinib VEGF Trastuzumab
13 Table 4. Development stage of TKI targeted agents in sarcoma Agent Company Targets Stage of Development References Small Molecule Inhibitors Imatnib (Gleevec) Novartis c-KIT, PDGFR Phase III GIST, Phase II DFSP, Kaposis sarcoma Preclinical osteosarcoma, MFH, Ewings sarcoma (126, 127) (128) (129) Gefitinib (Iressa) AstraZeneca EGFR family Phase I Ewings sarcoma Preclinical osteosarcoma, rhabdomyosarcoma (130) (131, 132) Semaxanib Pfizer VEGFR, PDGFR Phase II STS Preclinical neurogenic sarcomas, Ewings sarcoma (132) (132, 133) SU6668 Sugen VEGFR Preclin ical Ewings sarcoma (133) CEP-5213 Cephalon VEGFR Prec linical angiosarcoma (134) CEP-7055 Cephalon VEGFR Prec linical angiosarcoma (135) NVP-AEW541 Novartis IGF1-R Preclinical osteosarcoma, Ewings sarcoma, rhabdomyosarcoma (136) Dastinib Brisol-Myers Squibb Src Phase II solid tumors (137, 138) Sorafenib Bayer Ras Phase II Kaposis sarcoma (127, 136) Perifosine Keryx Cell membrane signaling Phase II STS (139, 140) Sunitinib (Sutent) Pfizer Pan TKI, PDGFR, VEGFR Phase III refractory GIST (136, 141, 142) Antibody Targeted Herceptin (Tratusumab) Genentech HER-2/ neu Preclinical Ewings sarcoma (132) Bevacizumab (Avastin) Genentech VEGF Phase II osteosarcoma, STS, Ewings sarcoma, Kaposis sarcoma, alveolar soft part sarcoma Preclinical Ewings sarcoma (136, 141, 142) (132)
14 Applications of tyrosine kinase inhibitors in sarcomas c-KIT and PDGFR GISTs are one of the most common mesenchymal malignancies of the gastrointestinal system. Furthermore, a dvanced, unresectable or malignant GISTs are fatal and highly resistant to conventiona l chemotherapies (143, 144). Mutation in c-kit is a critical event in the development of this ma lignancy. It is estimated to occur in 75% to 92% of GISTs diagnosed (145, 146). Imatinib mesylate (Gleevec), a c-KIT, Bcr-ABL and PDGFR TKI, was originally u tilized as a therapy for ch ronic myelogenous leukemia and is now used to treat GISTs. The first GIST patient received imatinib treatment in 2000. Within weeks this patient demonstrated a clinical response and remained stable for 18 months (122). Clinical trials were initiated in the United States and Europe following the dramatic results of this case-report. The fi rst trials showed responses that ranged from 59 to 69%. However, approximately 12% of th e patients on the initial trials experienced resistance to imatinib (147-150). Duensing et al. and Heinrich et. al demonstrated that GISTs lacking c-kit mutations have a higher likeli hood of demonstrating imatinib resistance, while tumors with c-kit mutations experienced at least a 50% reduction of tumor volume (149, 150). A clinical trial consisting of 147 GIST patients randomly assigned to receive either 400 or 600 mg of imatinib daily, demons trated that 53.7% of patients had a partial response, 27.9% had stable disease, and 0% achieved a complete response (151). A follow-up randomized clinical trial conducted by Verweij et al ., consisted of 946 patients randomly allocated 400 mg of imatinib either once or twice a day. Patients remained enrolled for an average of 760 days, and 56% of the patients that re ceived imatinib once a
15 day progressed compared to the 50% of the patients who received the regiment twicedaily. The estimated hazard ratio for this study was 0.82 [95% CI 0.69-0.98]; p-0.026. There was a similar response with no signifi cant difference between treatment arms: 5% of patients achieved a comple te response, 47% achieved a partial response, and 32% achieved stable disease. This study conc luded that the once daily 400 mg dose of imatinib is a sufficient treatment option for tr eating advanced or metastatic GISTs (152). Together, these studies provi de the necessary evidence to recommend treating patients with advanced or metastatic GISTs containing c-kit and PDGFR mutations with imatinib for extended periods. There are numerous ongoing studies evaluating the role TKI; imatinib and sunitinib can be used in neoadj uvant, adjuvant, and metastatic settings alone or in combination with other agents. DFSP is a slow growing sarcoma that is locally aggressive and has a high likelihood of local recurrence. Most DFSPs ha ve a characteristic translocation, t(17;22). This places the regulation of PDGF a ligand for PDGFR, under the control of the collagen 1A1 promoter, induci ng overexpression of PDGF In vivo experiments conducted in nude mice carrying tumors induced by DFSP-transformed cells treated with imatinib demonstrate significant in hibition of tumor growth (153). A case-series of ten patients with advanced DFSP with the characteristic t(17;22) translocation were evaluated for clinical res ponse to imatinib. Patients were treated with 400 mg of imatinib two times daily. Eight pa tients with local di sease experienced a clinical response, four of wh ich had a complete clinical response of an average duration of 220 days. The patients with metastatic disease had more complex karyotypes than those of the localized DFSPs. One patient with metastatic dis ease and the t(17;22)
16 translocation achieved a partia l response (198 of 383 days of follow-up) but experienced disease progression seven months following treat ment. The other patient with metastatic disease did not have the t(17;22 ) translocation and did not expe rience a clinical response. Based on this, imatinib is a viable option for the treatment of DFSP. AIDS-related Kaposis sarcoma is associ ated with HIV and herpes virus/human herpes virus-8 co-infections usually found on the skin (126). Kaposis sarcoma has been shown to express high levels of both c-KIT a nd PDGFR, and imatinib has been evaluated as a potential therapeutic option for this sarc oma. In a study ten male patients with AIDS-related Kaposis sarcoma received 300 mg of imatinib twice daily for four weeks. Of the ten patients, five had pa rtial clinical responses. Th e remaining five participants had stable disease at the end of the st udy, two of them with histological disease regression. This study demons trates that imatinib has potential effectiveness as a treatment option for patients with AIDS-related Kaposis sarcoma (126). TKIs inhibit proliferation over a ra nge of sarcoma-derived cell lines in vitro and in vivo as measured by xenograft models in nude mi ce. Studies using imatinib have been completed in rat OSA and MFH cell lin es expressing high levels of PDGFR Imatinib inhibited 20% and 40% of ce llular proliferation, respectiv ely when OSA and MFH cell lines were treated with 10 M imatinib (154). OSAs have been shown to express high levels of PDGFR. However, preclinical studies have not sh own imatinib to achieve antitumor activity within clinically releva nt or achievable doses (155, 156). The effects of imatinib were evaluated on a panel of eight different EWS cell lines and found that imatinib inhibited prolifer ation by 50% and indu ced apoptosis at IC50s ranging from 10-12 M. Furthermore, in xenograft models, imatinib treatment resulted
17 in regression and/or stabilization of primary Ewings tumors (157). Druker et al ., also determined a synergistic effect on growth inhibition when 10 M imatinib was combined with increasing doses of doxorubicin (134). Phase I clinical trials have demonstrated that the maximally tolerated dose of imatin ib is 1000 mM/day (148, 158, 159), which corresponds to a concen tration between 6-10 M in the blood (160). This concentration is below the IC50 required to inhibit th e proliferation and induce apoptosis of EWS cell lines (161). However, a study conducted on a panel of eight EWS cell lines with high levels of c-KIT expression, found the cell lin es to be resistant to imatinib at concentrations ranging from 0.1 to 10 M as determined by pro liferation assays (162). The preclinical data is conflicting in EWS and may be caused by inconsistencies among the basic science experiments performed or may be inherent in the na ture of the cell lines utilized in these studies. Therefore, the pr eclinical data suggests that c-KIT is not a critical target for EWS survival and thus ma y not serve a single agent treatment option. EGFR/HER-2/neu Gefitinib (Iressa), an epidermal growth factor receptor (EGFR) inhibitor, has shown potential anti-tumor effects in se veral sarcoma cell lines when used in combination with irinotencan, a topoisomer ase I inhibitor (129). Murine xenograft models have been used to assess the effect s of gefitinib on implanted tumor activity. Gefitinib was orally admini stered at 100 mg/kg either once or twice daily for 5 days/week. The anti-tumor activity of the co mbination was greater than additive in one of the seven cell line-derived xenografts and enha nced the activity of irinotecan in three of the seven xenografts. These data also reported an increased bioavailability of irinotecan when used in combination with gefitinib (129).
18 A phase I trial of gefitinib was completed on 25 children with refractory solid tumors. Twelve of the 25 patients enrolle d were sarcoma patients. Gefitinib was administered once daily for a course of 28 consecutive days starting at 150 mg/m2 and escalated to 500 mg/m2 a median of 54 courses were de livered. A recurrent EWS patient experienced a partial response af ter one course of treatment wh ich lasted for ten weeks. The maximum tolerated dose in this study was 400 mg/m2/day and was tolerated well (128). There is conflicting evidence in the literature evaluating HER-2/ neu as an effective target in treating sarcoma, thus the potential role of HER-2/neu as a therapeutic target for sarcoma is controversial. On e group evaluated five EWS cell lines and 13 archival primary EWS samples for HER-2/ neu gene amplification a nd protein expression. While several of the EWS cell lines and tumor samples had high HER-2/ neu protein expression, none had HER-2/ neu gene amplification. In addition, upon assessment of trastuzumab, (Herceptin) an anti-HER-2/ neu monoclonal antibody that inhibits HER2/ neu expression and blocks tumo rigenesis induced by HER-2/ neu, the group determined that trastuzumab had minimal inhibitory effects on cell growth, survival or colony formation of the EWS cell lines evaluated ( 140). Furthermore, others have evaluated trastuzumab alone and in combination with co nventional chemotherape utics used to treat sarcomas in two EWS cell lines. These st udies have not been presented anti-tumor activity at clinically achievable doses (139) and it was concluded that HER-2/ neu was not a major therapeutic target fo r the treatment of EWS. These results argue that HER-2/ neu is not a critical pathway for the pathogenes is of EWS and targeting this pathway may have little therapeutic benefit in EWS patients.
19 VEGFR Semaxanib is a dual VEGFR and PDFGR i nhibitor that has been evaluated in neurogenic sarcoma cell lines and human tumo r explants. Semaxanib has no effect on cell lines in vitro at doses up to 200 M; however, it reduces proliferation in tumor explants by 54.8% in mice treated with 25 mg /kg/day for eight days. This growth reduction was due to decreased tumor angiogene sis, which led to decreased proliferation and increased apoptosis (131). These data suggest that VEGF may serve as a viable target in preventing angiogene sis in neurogenic sarcomas. VEGFR inhibitors have been shown to be potent inhibitors of EWS growth in mouse models. Several inhibito rs of VEGF signaling have been evaluated in EWS and have shown significant reduction of tumo r growth in mouse models. The VEFGR inhibitors, SU6668 at 25 mg/kg/day and se maxanib at 100 mg/kg/day and anti-VEGF agents bevacizumab (avastin) at 10 mg/kg twice weekly and VEGF Trap at 2.5 or 25 mg/kg twice weekly significantly reduced tu mor growth at clinically achievable doses (132). Another VEGF signaling inhibitor, CEP-5214, and its pro-drug CEP-7055 inhibit tumor growth in a mouse angiosarcoma model. CEP-5214 administration at 1-3 mg/kg/day for 10 days achieved the minimu m effective dose and the maximum efficacy was observed by treating with 23.8 mg/kg CEP-7055 for 10 days (133). These preliminary data demonstrate the anti-tumor ac tivity of these drugs and offer evidence to further investigate the possible therapeutic pote ntial of these inhibitors for the treatment of angiosarcoma. Therapeutics that inhibit TK signaling through the use of antibodies are currently undergoing clinical trials in sarcomas. Bevacizumab, a recombinant VEGF monoclonal
20 antibody, has shown clinical promise by signifi cantly increasing dise ase-free progression and overall survival in several malignancies including breast, renal cell and colorectal cancers (136). A phase II tria l combining doxorubicin at 75 mg/m2 with 15mg/kg bevacizumab intravenously every three weeks wa s completed in patients with metastatic soft tissue sarcomas. A 12% response rate was observed in this study, and 65% of the patients experienced stable disease fo r four or more cycles (142). A case-report of a nine-yea r-old boy demonstrated that bevacizumab has potential activity for treating alveolar soft part sarcom a. The patient was treated with bevacizumab intravenously for 1 hour at 5 mg/kg biweekly for four cycles, then increased to 10 mg/kg with no substantial side effects. He experi enced a reduction of tumor load at all sites, including lung metastases and the patient was in good condition after 26 cycles of treatment (141). Additionally, there are seve ral ongoing clinical tria ls investigating the use of bevacizumab as a possible therapy fo r sarcoma. The National Cancer Institute (NCI) is sponsoring clinical trials evalua ting bevacizumab as a single agent and in combination with sorafenib (another TKI) in Kaposis sarcoma. In addition, there is an open-label phase II trial evaluating bevacizuma b as a single agent in the treatment of angiosarcoma (136). Further investigati on of bevacizumab in combination with conventional chemotherapies for the treatm ent of metastatic sarcoma is therefore warranted.
21 Other TKIs Several additional TKIs are currently under inves tigation. IGF-1R is a target of great interest for the treatment of sarcomas The therapeutic pot ential of NVP-AEW541, an IGF-1R inhibitor, was evaluated in a pa nel of 8 OSA, 10 EWS and 5 RD cell lines. NVP-AEW541 induced cell cycle arrest in a dose-dependent manner in the cell lines tested when treated with 300 nM, 1 M and 3 M for 24 hours. Apoptosis was induced in all of the EWS cell lines and in many of the OSA and RD cell lines, which correlated with IGF-1R inhibition. (134). Perifosine targets cell membrane sign aling pathways and has undergone phase I and phase II clinical trials in patients with advanced, metastatic solid tumors, including sarcomas. In the phase I study, 5 of 42 patie nts enrolled were sarcoma patients. One LMS patient was the only patient of the 42 tota l to experience a partial response. Six months following therapy, the patient reporte d elevated energy levels and symptom relief, and she experienced no significant disease progression once off all anticancer therapies (138). A phase II clinical trial consisting of 16 adult patients with locally advanced or metastatic STS were treated with perifosine. There were no clinical responses observed in this study. Four patien ts had stable disease which lasted 1.3 to 8.2 months (137). The preliminary clinical da ta may have suggested potential therapeutic benefit from treatment with perifosine; however, the phase II trial does not provide compelling evidence to suggest a high degree of anti-tumor activity. The Raf kinase inhibitor sorafenib is cu rrently under inves tigation in National Cancer Institutes sponsored st udies for the treatment of metastatic and locally advanced sarcoma (136). Sunitinib (Sutent), a pan-TKI that elicits inhibitory effects on a variety
22 of TKs including PDGFR and and VEFGR 1-3 has recently been approved for refractory GIST patients (136) A double-blind, phase III trial conducted to determine the efficacy of sunitinib in patients with GISTs that failed imatinib treatment demonstrated that patients had a significan t increased progression free duration of 6.3 months compared to 1.5 months in the place bo controls (127). Semaxanib, a reversible VEGFR2 and PDGFR inhibitor, underwent phase II clinical trials inve stigating its use in patients with relapsed soft tissue sarcoma. Anti-tumor activity was encouraging with 23% observed responses when treated with 145 mg/m2 twice weekly (130). There is considerable evidence to suggest that TKIs may serve as effective treatments options for sarcoma either alone or in combination with other chemotherapies. A protein kinase that serves as a point of convergence for many of these targeted TK signaling pathways is Src kinase. Our pr eliminary data has shown that Src is constitutively activated in sarcoma tissues. Th erefore, Src may be a worthwhile target for treating sarcomas.
23 History of Src: 1911present Nearly a century Peyton Rous discovered a filterable agent, or virus, that caused sarcoma in chickens (163, 164). This virus, later called Rous sarcoma virus (RSV), has remained in mainstream cancer research ever since its identifica tion. However, it was some 30 to 40 years after its initial discovery that the scientific community would come to accept that idea that a virus could contain genetic information which induced tumors. The road to establishing the foundation of identifying v-Src, the oncogene that causes the RSV has been long. In 1941 it was shown th at chicken embryo fibroblasts growing in culture infected with RSV adopted round and spindle-shaped morphology, similar in characteristic to Rous sarcoma cells (165). The focus assay was developed to quantify the number of cells infected with RSV in th e early 1950s. By the late 1950s, RSV was shown to induce tumors in mammals by many groups (166-170). These findings later, proved significant in iden tifying the transforming abil ity of Src protein. The transformation abilities of RS V mutants were identified in the 1960s, in which mutations induced spindle-shaped cells rather than rounded cells (171). This discovery, among others, lead to the 1970 identification of the oncogene responsible for cellular transformation of RSV, v-src Then in 1977, a 60 kDa phosphoprotein immunoprecipitated from RSV tumor-bea ring rabbits was identified as the src gene product (172). A year later two independent groups working on Src demonstrated that Src was a protein kinase (173, 174) and more specifically, a tyrosine kinase in 1979 by Tony Hunter (175). The discovery that RSV required a specifi c gene for its transforming capability shed light on a new way of studying the role genes play in cancer. The idea of the
24 existence of oncogenes, transforming genes found in RNA tumor viruses, lead to the idea that proto-oncogenes may exist as norma l cellular counterparts. Proto-oncogenes are precursors of viral transforming genes, wh ich they themselves lack the ability to transformation cells unless mutated or overexpress ed. The first of these to be identified was c-src (176, 177), the discovery of which initiated a fury of research into the role of proto-oncogenes in cancer. Upon sequencing c-src it was revealed that the cellular and viral versions differ in that v-src has carboxy-terminal substitutions and de letions, which were later linked to the aberrant activity of v-Src pr otein (176). Upon cloning of the c-src gene lead to comparing the biological activity of both fo rms of the gene. Except when overexpressed, c-Src has lower transforming activity in ma mmalian and avian fibr oblast cell lines (178180). Determining how the transforming ab ility of c-Src was activated lead to characterization of the c-src protein product. The human Src gene encodes a 536 amino acid protein, three more amino acids than the chicken ortholog. It is now known that Src is one of 11 members of the human Src-family Kinases (SFK), which also includes Blk, Brk, Frk, Fyn, Hck, Lck, Lyn, Srm and Yes (124). Fyn, Src and Yes are expressed in all cell types, Srm is found in keratinocytes and Blk, Fgr, Hck, Lck and Lyn are primarily found in hemat opoietic cells. While Frk is expressed in bladder, brain, breast colon and lymphoid cells and Brk is expressed in colon prostate and the small intestine (181). SFKs share similar homology and protein or ganization. The organization of Src protein from Nto Ctermini contains a 14-carbon myristoyl group, which facilitates the attachment of Src to membranes and is requ ired for normal function, is attached to an
25 SH4 (Src Homology) domain. This unique domain is followed by an SH3 and SH2 domain, an SH2 kinase linker, a protein kina se domain containing the active site, and the C-terminal regulatory domain (Fig. 2). The SH2 and SH3 domains have several important functions, including constraint of en zyme activity via intramolecular contacts during periods of inactivity and binding to lig ands with SH2 and SH3 domains to attract them to specific locations within the cell. There are two vital ty rosine phosphorylation sites on Src, Tyr530 and Tyr419. When phos phorylated, Tyr530 binds intramolecularly to the SH2 domain, stabilizing the restrained enzyme and preventing interaction of SH2 and SH3 with other ligands (Fi g. 3). In normal cells Src activ ity is tightly regulated, 9095% of Src is phosphorylated on Try530 and in an inactive state and becomes transiently activated during specif ic cellular event such as mitosi s (182). Csk and Csk homologous kinase (Chk) are two cytosolic tyrosine kinases responsible for phosphorylation of Tyr530. Tyr419 is an autophosphorylation site th at is located in the active site of the enzyme and promotes Src kinase activity. Phosphorylation of Try419 stabilizes the active confirmation. Thus Tyr 419 is the activation phosphorylat ion site and Tyr530 is the inhibitory phosphorylation site of Src which is missing from the v-Src protein (183). When phosphorylated on Tyr530, Src cannot undergo autophosphorylation, Tyr530 must first be dephosphorylated. However, when phosphorylated on Tyr419 and Tyr530 simultaneously, the Tyr419 phosphorylation overr ides the inhibitory Tyr530 and remains active (184). Src activation is initiated by several cellular mechanisms including receptor tyrosine kinases, integrin receptors, cytokine receptors, G-protein receptors and steroid hormone receptors (185). Once activated, Src si gnals to the nucleus through a variety of
26 downstream effectors including STAT3 (186, 187) PI3K (188, 189) and Ras (190). Src plays a role in a myriad of cellular functions including proliferation, survival, cell adhesion, migration, cellular morphology and bone resorbtion. For example, under normal conditions, when epidermal growth factor receptor (EGFR) is activated, Src becomes activated during G2/M transition through the cell cycl e. Src activity has also been shown to be an important characterist ic of osteoporosis. Mice with Src -/null mutations die within weeks of birth and e xpress an osteoporosis phenotype, resulting from increased bone resorbtion cau sed by a defect in osteoclasts (191). Thus, Src kinase inhibitors currently serve as targeted therapy optio n for treating osteoporosis. Figure 2. Organization of Src kina se. Structural features of c-Src compared to v-Src protein. Both proteins are myristoylated at Gly2 contain membrane binding (MB), unique (U), SH3, SH2, linker and catalytic domains. Notice only c-Src possesses the inhibitory Tyr530, or regulatory (Reg) domai n at the C-termini. c-Src is usually phosphorylated on Tyr530 and in the inhi bitory conformation, while v-Src is phosphorylated at Tyr419 and constitutively active because the regulatory domain is missing. Other differences between the two pr oteins include amino acid substitutions in v-Src. c-Src v-Src Myr MyrGly Gly H2N H2N MB USH3SH2LinkerCatalytic DomainRegP P Tyr Tyr419529 530 536 2 2COOH COOH c-Src v-Src Myr MyrGly Gly H2N H2N MB USH3SH2LinkerCatalytic DomainRegP P Tyr Tyr419529 530 536 2 2COOH COOH
27 Figure 3. Src activation and si gnaling conformations The left panel represents the inactive conformation of the c-Src protein. When phosphorylated, Tyr530 binds to the SH2 domain and prevents phosphorylation of the active site, Tyr419, while the SH3 domain interacts with the linker domain, fu rther stabilizing the conformation. Upon activation of the active site the SH2 doma in release Tyr530, SH3 releases the linker domain and they are now available to interact with other proteins containing SH domains. Furthermore, the active site is open or active, available to elicit si gnals (right side of panel). SH3 SH2 PTyr 419 Tyr 530K i n a s e D o m a i n SH3 SH2 PTyr 419 Tyr 530K i n a s e D o m a i nActive Conformation Inactive Conformation Dasatinib (BMS 354825) Src Stimulation P Binding Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n P Signaling Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n Downstream Signaling Cascade P Binding Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n P Signaling Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n Downstream Signaling Cascade SH3 SH2 PTyr 419 Tyr 530K i n a s e D o m a i n SH3 SH2 PTyr 419 Tyr 530K i n a s e D o m a i nActive Conformation Inactive Conformation Dasatinib (BMS 354825) Dasatinib (BMS 354825) Src Stimulation P Binding Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n P Signaling Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n Downstream Signaling Cascade P Binding Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n P Signaling Protein SH3 SH2 Tyr 419 Tyr 530K i n a s e D o m a i n Downstream Signaling Cascade
28 Considerable evidence suggests that Sr c may serve as an option for targeted therapeutics in cancer as well. Src kinase activity has been shown to be constitutively activated in several types of human cancers including breast, prostate, colon, pancreatic and just recently sarcoma as compared to normal tissues. For example, Src is overexpressed in ~70% of breas t cancers (192). Furthermore, the increased activity of Src is strongly implicated in the developm ent, growth, progression and metastasis of cancer. Mutations in the c-Src gene are very rare, theref ore overexpression may be induced by increased activation and associati on with RTKs. Reduced levels of Csk or Chk activity and decreased phosphatase activit y within the cell may also be potential mechanisms for increased Src activity in cancer (193). The increased activity of Src in cancer i nduces dramatic alterations in signal transduction of downstream tr anscriptional events. Many of these downstream targets have been linked to tumorigenicity and meta stasis (194). One widely studied pathway involved in tumorigenesis and downstream of Src activation is STAT 3, a member of the signal transducers and activator s of transcription (STAT) fa mily. STATs are mediators of Src induced cellular transformation. This model of evidence introduced in 1995, showed that the DNA-binding activity of STATs was enhanced in cells transformed with v-Src (187). Upon activa tion, Src phosphorylates STAT 3, which homodimerizes and translocates to the nucleus where it regulat es the transcription of genes involved in cellular proliferation, survival and migration. It is not surprising considering that STAT3 is activated by Src to learn that STAT 3, along with other STAT proteins, are constitutively activated in diverse human can cer cell lines and tissues. In cancer, STAT3
29 regulates the transcription of genes involved in uncontro lled tumor cell proliferation, resistance of apoptosis, evasion of the im mune system and angiogenesis, among other malignant phenotypes (195). Consequently, ST AT proteins and their activators have emerged as potential targets for cancer therapy (196, 197). In addition to STATs, increased Src activity has been is associ ated with adhesion changes characteristic of mesenchymal transi tion and is suggested to promote cancer cell migration and metastasis. Focal adhesion kinase (FAK), along with Crk-associated substrate (CAS) are substrates of Src and serve as focal adhesion proteins vital for integrin signaling. FAK, also a tyrosine ki nase, modulates the formation and turnover of focal adhesions, the intracellular structures that link the extra cellular matrix to the actin cytoskeleton (198). Src and FAK associ ate upon autophosphorylation of the Tyr397 residue on FAK, after which Src beco mes autophosphorylated and can then phosphorylate several residues on FAK, including Tyr576/577 and Tyr925. Phosphorylation of FAK by Src enhances FAK activity. Mutational studies have shown that the inactivation of FAK is embryonic lethal (199) and selective inhibition of fak expression in skin ke ratinocytes suppressed chemically induced tumor formation in vivo (200). Others have shown that inhibition of fak expression induced apoptosis in vivo and in vitro in keratinocytes (200) and endothelial cells in vitro (201). Furthermore, the interplay between Src and FAK is important for modulating the gene expression of transcription fact ors and cell motility proteins, as elucidated in de letion and rescue experiments (202). Overexpression of FAK suppresses p53-mediated apoptosis by bindi ng to the transactivation domain of p53 and limiting transcription activity (203). While, FAK itself does not function as an oncogene,
30 elevated expression in cancer correlates w ith increased cell motil ity, invasiveness and proliferation (202, 204). RNAi-mediated knockd own of FAK in an aggressive breast carcinoma cell line did not affect cellular pr oliferation, bu t cells lacking FAK lacked invasive characteristics in vitro and lacked spontaneous lung metastasis in vivo (205). In addition, FAK expression, along with Src, in creases with colon cancer invasiveness (206). The FAK/Src complex also phosphorylates CAS, a cytoskeletal adapter protein. Upon Src-mediated phosphorylation, CAS acts as docking protein for multiple protein protein interaction domains a nd is important for the recruitment of adapter proteins, including Crk, to the substrate domain of CA S (207). CAS activity has been shown to promote the invasiveness of Src-transformed cells (208) and play a vital role in cell motility and survival. While the precise mechanism of CAS mediated apoptosis is unknown, phosphorylation of CAS and its associa tion with FAK have been shown to be essential in the survival pathway. Furtherm ore, expression of an tisense CAS partially reversed the transformation activ ity of v-Src, suggesting that the inhibition of CAS plays a direct role in cellula r transformation (209).
31 Development of Src inhibitors for clinical trials Models have demonstrated that overexpres sion and activation of Src is associated with many oncogenic characteristics such as disruption of the cell cycle, increased migration and invasion and prot ection from apoptotic stimuli (193). Given the history of Src and because it plays such a pivotal role in many oncogenic processes, it is surprising that Src has only recently been widely consid ered as a potential therapeutic target for cancer therapy. There are thre e variety of Src inhibitors; SH2/SH3 blocking inhibitors, Src destabilizing agents and ATP-competitive inhibitions. SH2/SH3 blocking inhibitors prevent the SH domain mediated interac tions and prevents Src protein-protein interactions. These inhibitors have demons trated poor transport and uptake properties and only inhibit a subset of Sr c protein interactions, limiting the clinically efficacy of this class of Src inhibitors (210-212) Src destabilizing agents inte rfere with the association of Src and its molecular chaperone Heat Shock Protein (HSP) 90. These agents have proven to be non-specific and results in the disrupt ion of many HSP interactions, which could result in multiple adverse side-effects in pa tients (213, 214). ATP competitive Src kinase inhibitors target the active site and prevent ATP from binding and initiating the phosphotransferase activity of the enzyme. Ther e is significant homology in the structure and sequence of ATP-binding domains of ki nases, which has made the narrowing the specificity of ATP competitors difficult. However, several ATP competitive inhibitors are currently under various stages of investigation as potential therapies for cancer. PP1 and PP2 were among the first Src kinase inhibitors utilized to study the role of Src activity in cellular events. The use of PP2 has shown that inhibition of Src resu lts in a loss of downstream Src signaling
32 through pathways including FAK, Akt a nd STAT3 and elicits anti-tumorigenic phenotypes (193). These findings were later proven to be through Src inhibition using siRNA to knockdown Src expression (215). Si nce the discovery that inhibition of Src can lead to anti-tumorigenic effects in cancer models several ATP-co mpetitive Src kinase inhibitors have been synthesized and studi ed. Three such inhi bitors include PD180970 (Pyridol[2,3-d] pyrimidine), SKI-606 (4-ani lino-3-quinolinecarbonitr ile) and dasatinib (BMS-354825, [N-(2-chloro-6-methylphenyl)-2-( 6-(4-(2-hydroxyethyl) piperazin-1-yl)2-methylpyrimidin-4-ylamino)thiazole-5-carboxa mide) (Fig. 4), all of which have been thoroughly studied in epithelial and hematopoietic cancers. PD180970 was the first of the three to unde rgo investigation as a Src kinase inhibitor. Several groups have shown that PD180970 can inhibit Src at nanomolar doses (216-219). Most significantl y, others have shown that the inhibition of Src-STAT3 signaling in cancer cells can l ead to the induction of apopt osis (216, 219). However it is a non-specific kinase inhibitor, only solubl e in DMSO and is highly unstable, which leaves this compound with little to no clini cal future. SKI-606 is an orally available compound with dual specificity for SFKs and Abl tyrosine kinase (220 ). Studies have shown that SKI-606 decreases growth a nd motility of colorectal cancer cells by preventing activation of Src and downstream signaling (221). Furthermore, SKI-606 has been shown to have anti-proliferative eff ects in colon cancer and myelogenous leukemia cancer models (220, 222). These data have lead to the initiat ion of clinical trials in patients with hematopoietic and epithelial cancers. Dasatinib, is also an orally available Src inhibitor that is currently undergoing clinical trial in ep ithelial cancers. Preclinical data have shown dasatinib to be a potent inhi bitor of Src activity in prostate, head and
33 Figure 4. Structure of three ATP-competitive Src kinase inhibitors. A, PD180970, B, SKI-606 and C, dasatinib were independently devel oped and provided by three different pharmaceutical companies. All three inhibitors act as ATP-competitors for the catalytic domain of Src kinase and prevent the activation of Sr c by phosphorylation of Tyr419. A. C. B. A. C. B.
34 neck and lung cancers (223-225). Nam et al ., showed dasatinib inhibition of Src at nanomolar concentrations inhibited down stream signaling of STAT3 and FAK which subsequently inhibited the migration and inva sive properties of pr ostate cells (223). While others have shown that in addition to inhibiting migration and invasion, dasatinib also induces cell cycle arrest and apoptosis in head a nd neck (224) and lung cancer cell lines (225). There is remarkable evidence to suggest that Src kinase activity may serve as an important target for the treatment of cancer. Until the culmination of this dissertation research, the role of Src activ ation in human sarcoma had not yet been elucidated. In addition, the responses and mechanism of action of ATP-competitive Src kinase inhibitors in mesenchymally-derived tumor cell lines have not been described previously. There is considerable evidence to suggest a role for Src in sarcomagenesis. Vigneron et al ., have shown that increased Src expression in sarcoma cell lines reduced sensitivity to conventional chemotherapies used to trea t sarcomas. This chemo-resistance was associated with a failure of cells to up-regul ate p21 in response to adriamycin (226). Furthermore, increased Src activity has also been shown to cont ribute to anoikis resistance in human OSA cell li nes. Inhibition of Src activit y restored anoikis sensitivity and cells underwent apoptosis upon detachment (227). Thus, these three independently synthesized compounds were evaluated in sarcom a cell lines to identify the role of Src kinase activation in the sarcoma cell lines. The objective of this dissertation was to determine the role of Src activation in sarc omas and determine whether Src may serve as a viable option for the treatment of sarcoma patients.
35 Objectives The overall objective of the studies c onducted for this dissertation was to investigate the role of Src kinase activati on in the survival and malignant phenotype of sarcomas. Src kinase was the original oncoge ne identified in Rous Sarcoma Virus (172). Previous research has suggested a possible role for Src activation in tumor survival, in that Src is involved in a myri ad of cellular mechanisms such as survival, migration and invasion. However, the role of Src in human sarcomas had not been determined. Results from preliminary experiments suggested that Src kinase is activated in a variety of human sarcomas. The recent development a nd interest of Src kinase inhibitors aided in evaluating Src as a possible target of sarc oma treatment. These findings in addition to the overwhelming evidence suggesting that Src may serve a possible target essential for the survival of sarcomas lead to the developm ent of the hypothesis that Src is required for the migration, invasion and surv ival of sarcomas. To verify this hypothesis, the following aims were pursued. Aim 1: Evaluate the biological response of three Src kinase inhibitors in sarcoma cell lines. A. To determine the role of Src kinase activation in sarcomas cells lines. Src kinase has been shown to be activated to varying degrees in primary sarcomas. However, the degree of activation and si gnificance of Src activation in sarcoma
36 cell lines has not been determined. Becau se Src was originally identified as an oncogene significant in the development of sarcoma in chickens, we sought to determine to role of Src kinase activity in sarcoma cell lines using three different Src kinase inhibitors. B. To determine if STAT3 activation is required Src activated migration, invasion and survival of sarcomas. STAT3 is among multiple downstream signals initiated by Src activation. STAT 3 was evaluated as a possible signaling protein involved in the cellular events elicited by Src activation in sarcoma cell lines. Aim 2: Identify a molecular signature th at predicts response to dasatinib by induction of apoptosis in sarcoma cell lines. A. Identify a gene expression profile uni que to sarcoma cell lines that undergo apoptosis when treated with dasatinib. Src kinase has previously been shown to be essential for tumor survival in epithelial derived tumors (185, 207, 228-230). The data from aim 1 has shown that Src kina se activity is required for survival in a subset of sarcoma cell lines. Therefor e, microarray analys is was performed on sarcoma cell lines to identify a molecula r signature that predicts response to dasatinib by induction of a poptosis in sarcoma cells. B. Confirm the predictability of the molecu lar signature to predict response to dasatinib in an independent dataset. It is hypothesized that a molecular signature may be identified in tumors a nd tumor cell lines which may predict the response to specific therapies. To te st the effectiveness of the molecular
37 signature that predicts response to dasatinib two new cell lines of unknown response status were utilized. The cell lines were used to determine whether the molecular signature extracted from the responsive cell lines was successful in identifying new cell lines that would or would not respond to dasatinib by induction of apoptosis. Aim 3: Verify the presence of the molecu lar signature in primary human sarcoma specimens. A. Determine if the molecular signature id entified in the sa rcoma cell lines is present in untreated, primar y human sarcoma specimens. Cell lines lack the microenvironment interactions of human tumors and one could not be certain that the molecular signature identified in the cell lines is expressed in human tumors. Untreated sarcoma specimens were utili zed to establish whether the molecular signature unique to the ce ll lines that respond to dasa tinib was present in human tumors. Microarray analysis was perfor med on a diverse set of human sarcomas to validate the expression of the cell line signature in tumors. B. Establish whether the molecular signature can cluster sarcomas based on expression and theoretically p redict response to dasatinib. Once the molecular signature was identified in prim ary sarcomas, the signature was used to predict theoretical response to dasatini b. The molecular signature was further tested to determine if tumors clustered into the respective theoretical categories based on the expression of the molecular signature.
38 Materials and Methods Cells and reagents SaOS-2, U 2 OS, MG-63, SK-ES-1, A673, RD, SK-LMS-1, HT-1080, SW-872, HOS and SW1353 sarcoma cell lines were obtai ned from the American Type Culture Collection. The LM2 and LM7 OSA cell lines were provided by Dr. Eugenie S. Kleinerman of MD Anderson Cancer Cent er (Houston, TX). The TC-71 EWS cell line was provided by Dr. Timothy Triche of the Un iversity of Southern California (Los Angeles, CA). The SaOS-2, MG-63, LM2, LM7, TC-71, SK-LMS-1, HT-1080 and SW872 cell lines were maintained in MEM suppl emented with Eagles salts, 10% fetal bovine serum, 2-fold MEM vitamins, 1 mM sodium pyruvate, 1 mM non-essential amino acids and 2 mM L-glutamine. The SK-ES-1 and U 2 OS cell lines were maintained in McCoys 5A medium supplemented with 10% fetal bovine serum. The A673 cell line was maintained in DMEM supplemented w ith 10% fetal bovine serum. The RD and RD18 cell lines were maintained in DMEM /F12 (1:1) supplemented with 10% fetal bovine serum. All cells were maintained at 37C in 5% CO2. All experiments were performed using exponentially prolifer ating cells unless otherwise noted. Polyclonal antibodies to phosphorylat ed p-Src (Y419), p-FAK (Y576/577), pp130CAS (Y410), pSTAT3 (Y705) proteins, and to total FAK and PARP proteins were obtained from Cell Signaling Tech nologies (Cambridge, MA). Polyclonal antibodies to p130CAS and -actin were obtained from Santa Cr uz Biotechnology (Santa Cruz, CA). Monoclonal antibodies to total Src (clone GD11) and hILP/XIAP were obtained from
39 Upstate Biotechnology (Lake Placid, NY) a nd BD Biosciences (San Diego, CA), respectively. Dasatinib was provided by Bris tol-Myers Squibb Pharmaceutical Research Institute (Princeton, NJ). Dasatinib was synthe sized by the addition of methylprimidine to the 2-amino group of thiazole, followed by a reaction with hydroxyethyl piperazine (231). Preparation of cell extracts and Western blotting Adherent sarcoma cells were washed w ith ice cold 1x PBS, followed by washing and scraping from plates in ice cold 1x PBS containing 5 mM sodium fluoride and 1 mM sodium orthovanadate (1x PBS + inhibitors ). Cells were harvested by centrifugation, washed with 1x PBS + inhibitors and harv ested again by centrifugation. Pellets were resuspended in RIPA buffer containing 0.1 mM Na3VO4, NaF, DTT and 1:100 fold dilution of protease inhibitor cocktail obtai ned from Sigma (St. Louis, MO). Samples were vortexed for 30 min at 4C and insol uble material was removed by centrifugation. For Western blotting, 50 g of protein was resolved by SDS-PAGE and transferred to nitrocellulose membranes. Blots were blocked in 1x TBS co ntaining 0.01% Tween-20 and 5% non-fat dry milk for 30 min, and then incubated with prim ary antibody in 1x TBS containing 0.01% Tween-20 a nd 5% BSA obtained from Sigma (St. Louis, MO) overnight at 4C with rocking. Protein-bound primary antibodies were detected using respective horseradish peroxidase-coupled secondary antibodies (Amersham anti-rabbit for polyclonal and anti-mouse for monoclonal, obtained from GE Healthcare Unlimited, Buckinghamshire, UK) diluted 1:10,000 in 1x TBS containing 0.01% Tween-20 and 5% non-fat dry milk and incubated for 2 h at r oom temperature. Bound secondary antibodies
40 were detected using Amersham ECL PLUS We stern blotting detecti on reagents obtained from GE Healthcare Unlimited. Densitome try was performed on p-Src (Y419) Western blots performed for dose responses in each of the cell lines and then analyzed using ImageQuant 5.2 (Molecular Dynamics) softwa re. Percent inhibition of Src kinase activity as measured by Src (Y419) phos phorylation was determined by nonlinear regression analyses and data were reported as the inhibitory concentration required to achieve 50% inhibition rela tive to control reactions (IC50) in Table 1. Data are the averages of triplicate determinations. Immunohistochemistry Human tissues were obtained through the Moffitt Cancer Center Tumor Bank using IRB-approved protocols. Tissues were fixed in forma lin within 15-20 min from the moment of surgical excision to preserve the phosphorylation status of proteins such as Src and STAT3. Formalin-fixed, paraffin-embedded tissue sections of 3 m thickness were deparaffinized by an initial warming to 60C, followed by two xylene changes of 10 min each, two series of 30 dips in absolute alcohol, 30 dips in 95% alcohol, and 20 dips in deionized water. Slides were placed for 5 min in TBS/Tween and processed on a DAKO Autostainer using the Dako LSAB+ peroxidase detecti on kit (Carpinteria, CA). Endogenous peroxidase was blocked with 3% aqueous hydrogen peroxide followed by 20 dips in deionized water. Th e anti-p-Src (Y419) or pSTAT3 (Y705) was applied at 1:100 dilution for 30 min after microwave antigen retrieval with 0.1 mol/L citrate buffer (pH 6.0; Emerson 1,100 W microwave, high to boiling, then 20 min on power level 5). The chromogen 3,3'-diaminobenzidine was used for detection. Counterstain was done with
41 modified Mayer's hematoxylin. Slides were dehydr ated through graded alcohol, cleared with xylene, and mounted with resinous mounting medium. Src Family Kinase PCR A multiGene-12 RT-PCR Profiling Kit was obtained from SuperArray Bioscience Corporation to determine the status of ge ne expression for 11 SFK members in eight sarcoma cell lines; SaOS-2, LM-2, LM-7, U2 OS, MG-63, SK-LMS-1, HT1080 and SW872 cell lines. RNA was isolated from exponentially growing cells using TRIzol Reagent (Invitrogen, Carlsbad, CA) according to the manufacturers protocol. First strand cDNA synthesis was completed usi ng the SuperScript Fi rst Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA) according to the manufacturers protocol. The remainder of the RT-PCR protocol was completed according to the SuperArray protocol. Once resolved in a 1% agarose gel, photographs were taken for each of the cell lines to compare gene expression. Wound healing assay for cell migration Cells were plated in 24 well tissue culture plates, grown to confluency and serum starved in 0.1% FBS overnight. Monolayer w ounds were produced using a pipette tip scratched through the center of the well. Phot omicrographs were taken of the initial wound for comparison. Cells were then treate d with either DMSO alone, as vehicle control or escalating dos es of dasatinib and allowed to migrate into the denuded areas for 24 h. Following incubation, cells were briefl y stained with Coomassie blue. Cell migration was visualized at 10x magnificat ion and digitally photographed. The distance
42 of migration was measured as pixel units and compared to time zero. The average number of pixel units measured in th e denuded area were determined from wound healing assays for dose responses in each of th e cell lines analyzed. Percent inhibition of migration was determined by nonlinear regression analyses and data were reported as the inhibitory concentration required to achieve 50% inhibition relative to control reactions (IC50). Data are the averages of triplicate determinations. Cell invasion assay Cell invasion assays were performed following the BioCoat Matrigel Invasion Chambers protocol, obtained from BD Biosci ences (Bedford, MA). Briefly, cells were trypsinized and washed once with 1x PB S and twice using serum-free medium. Cell suspensions were prepared at 5 x 104 cells/mL in 0.5 mL containing DMSO alone as vehicle control or escalating doses of dasati nib in serum-free medium and added to the chamber insert. Chambers were incubated for 22 h at 37 C and 5% CO2. Following incubation, non-invading cells we re removed using a cotton-ti pped swab. The cells that invaded to the lower surface of the membrane were stained using a Diff Quick Staining Kit from Fisher Scientific (Pittsburgh, PA), digitally pho tographed and counted. Each experiment was completed in triplicate. TUNEL assay Apoptosis was detected by the TUNEL assay using the In Situ Cell Death Detection Kit obtained from Roche Molecula r Biochemicals (Indianapolis, IN) according
43 to the manufactures protocol. Cells were tr eated with either DMSO alone as vehicle control or escalating doses of dasatinib for 48 h. siRNA transfections siRNA directed specifically against c-Src and a non-targeting siRNA control were obtained from Dharmacon RNA Technologies (Chicago, IL). Ce lls were plated on 6 cm tissue culture plates in complete media (5 x 105 cells per plate) and allowed to attach overnight. siRNA was transfected in es calating doses (50 nM and 100 nM) using Oligofectamine obtained from Sigma-Aldric h (St. Louis, MO). The transfection incubation time for the siRNA/Oligofectamin e complexes was 24 h, and total incubation time before harvesting cell lysates was 72 h. Statistical analysis Descriptive statistics, such as mean valu es and standard devi ation, were calculated for the biological effects of dasatinib on invasion by dose levels (nM). To determine statistical significance between pair-wise dose levels, the exact Wilcoxon two-sample test or T-test were used, considering the small sa mple sizes. One-sided te sts at a significance level of 0.05 were examined. All data were analyzed using the SA S software (version 9.1, SAS Institute, Cary, NC). Microarray sample preparation Three consecutive passages of untreat ed, exponentially growing SaOS-2, LM2, LM7, U 2 OS, MG-63, SK-ES-1, SK-LMS-1, HT1080, A673, RD, RD|18, HOS and
44 SW1353 were utilized for microarray analysis. Sarcoma tissue was collected from the H. Lee Moffitt Tumor Bank. Tissues were unt reated primary malignancies that were snap-frozen in liquid nitrogen within 15 minut es following removal from the patient to minimize cellular degradation. Isolation of RNA Total RNA was excised from sarcoma tissue sp ecimens and cell lines using the TRIzol Reagent (Invitrogen, Carlsbad, CA) according to the manufacturers protocol. The aqueous phase containing the RNA separa ted from the TRIzol reagent was further purified using the RNeasy cleanup procedure (Qiagen Inc., Valencia, CA). The quality of total RNA was assessed by agarose gel electrophoresis and the total RNA from the tissues was further analyzed on a Agilent 2100 Bioanalyzer. Preparation of labeled RNA targets for hybridization Within the total RNA pool the poly(A) RNA was specifically converted to cDNA and then amplified and labeled with biotin following the proce dure initially described by Van Gelder et al. (232). First-strand cDNA synthesis was carried out using the Superscript Choice System (I nvitrogen, Carlsbad, CA) and the T7 promoter/oligo (dT) primer (5-GGCCAGTGAATTGTAAT ACGACTCACTATAGGGAGGCGG-(dT)24-3), (Genset Corp., La Jolla, CA). Following a nnealing the rest of the cDNA synthesis reaction was prepared such that the final reaction contains 5 g RNA, 100 pmol T7-(T)24 primer, 500 M each dNTP, 10 mM DTT, 50 mM Tr is-HCl, pH 8.3, 75 mM KCl, 3 mM MgCl2, and 200 U of Superscript II reverse tr anscriptase (Invitr ogen Corporation,
45 Carlsbad, CA). The reaction was incubated for 1 hr at 42C. A second-strand cDNA synthesis was performed at 16C fo r 2 hr in a total volume of 150 L, using 10U of E.coli DNA ligase, 40 U of E. coli DNA polymerase I, and 2 U of E. coli RNase H in the presence of 200 M of each dNTP, 10 mM (NH4)SO4, 1.3 mM DTT, 26.7 mM Tris-HCl, pH 7.0, 100 mM KCl, 5 mM MgCl2, and 150 M -NAD+ (Invitrogen). Following the second-strand DNA synthesis, 10 U of T4 D NA Polymerase (Invitrogen) was added and the samples incubated an additional 5 min at 16C. The reaction was stopped by the addition of 0.5 M EDTA and subsequently extracted with an equal volume of phenol/chloroform/isoamyl alcohol. The double-stranded DNA (dsDNA) will then be precipitated with the addition of 0.5 volumes of 7.5 M NH4 Acetate and 2.5 volumes of ice-cold 100% ethanol. The ds DNA then serves as a template for a transcription reaction performed with the GeneChip IVT Labeling k it according to manufacturer's instructions (Affymetrix Corp., Santa Clar a, CA) which incorporates biotinylated UTP into the transcripts. The Biotin-labeled RNA was purified using RNeasy co lumns (Qiagen) and fragmented to a size of 35 to 200 bases by incubating at 940 C for 35 minutes in fragmentation buffer (40 mM Tris-acetate pH 8.1/100 mM potassium acetate/30 mM magnesium acetate). The integrity of the st arting material and the products of each reaction were monitored on agarose gels to a ssess the size distributi on of the products and compare them to the starting material.
46 Array hybridization and scanning The hybridization solution consisted of 20 g of fragmented RNA and 0.1 mg/ml sonicated herring sperm DNA, in 1 x MES buffer (containing 100 mM MES, 1 M Na+, 20 mM EDTA, and 0.01% Tween 20). In addition the hybridization solutions were spiked with known concentrations of RNA from the bacterial genes, BioB, BioC, and BioD, and one phage gene, Cre, as hybridization standards. The hybridization mixtures were heated to 99C for 5 min followed by incubation at 45C for 5 min before injection of the sample into a probe array cartridge. A ll hybridizations were ca rried out at 45C for 16 h with mixing on a rotisserie at 60 rpm. Following hybridization, the solutions were removed and the arrays were rinsed with 1 x MES. Subsequent washing and staining of the arrays was carried out using the GeneChip Flui dics station protoc ol EukGE_WS2, which consists of 10 cycles of 2 mixes per cycle with non-stringent wash buffer (6 x SSPE, 0.01% Tween 20) at 25C followed by 4 cycles of 15 mixes per cycle with stringent wash buffer (100 mM MES, 0.1 M Na+, and 0.01% Tween 20) at 50C. The probe arrays were then stained for 10 min in streptavidin-phycoerythrin solution (SAPE) [1 x MES solution, 10 g/ml SAPE (Molecular Probes, Eugene, OR), and 2 g/l acetylated BSA (Invitrogen)] at 25C. The pos t-stain wash was 10 cycles of 4 mixes per cycle at 25C. The probe arrays were treated for 10 min with an antibody solution [1 x MES solution, 2 g/l acetylated BSA, 0.1 g/l normal goat IgG (Sigma Chemical, St. Louis, MO), 3 g/l biotinylated goat-antistreptavidin antibod y, (Vector Laboratories, Burlingame, CA)] at 25C. The final wash consisted of 15 cycles of 4 mixes per cycle at 30C. Following washing and staining, probe arrays were scanned once at 1.5-m resolution using the Affymetrix GeneChip Scanner 3000.
47 Data analysis Scanned output files were visually inspect ed for hybridization artifacts and then analyzed by using Affymetrix Micro Array Suite (MAS) 5.0 soft ware. Arrays were scaled to an average intensity of 500 and analyzed independently. The MAS 5.0 software uses a statistical algorithm to determine the signal inte nsity of a transcript from the behavior of 11 different oligonucleotide probe s designed to detect the sa me gene (233). Probe sets that yield a change p-value less than 0.05 were identified as changed. Gene changes were selected using one of two methods the Si gnificance Analysis of Microarrays (SAM) technique of Tusher et al. (234) or by filteri ng the probe sets to contain only probe sets with a maximum/minimum intensity value ratio of less then 2.0 for responders and nonresponders. Probe sets that fit both criteria were furthe r examined for use in the molecular signature. To test the reliability of the signatur e and determine the response status of two cell lines with unknown response to dasatinib th e fold change of the probe sets that comprise the molecular signature was determ ined as compared to either the median intensity for each probe set for all samples as the comparison group. A decision on response was determined based on how the fold change compared to the fold changes for the molecular signature in the original dataset. Hierarchical clustering analysis and da ta transformation for microarray analysis were performed using Cluster version 2.11 (235) Microarray data visualization using
48 heatmaps was prepared using Java Treevie w (236), where red depicts an increased expression and green depicts a decrea sed expression in a given gene. Preparation of samples for Quanti tative Real-Time PCR analysis For src analysis SaOS-2, U 2 OS and SK-L MS-1 cells were treated with 30, 100, 300 and 1000 nM of dasatinib or DMSO for six hours. Cells were washed with ice cold 1x PBS, RNA was isolated using the TRIz ol Reagent (Invitrogen, Carlsbad, CA) according to the manufacturers protocol The aqueous phase containing the RNA separated from the TRIzol reagent was fu rther purified using the RNeasy cleanup procedure (Qiagen Inc., Valencia, CA). The quantitative Real-Time PCR analysis for histone H3, FAF1, -catenin, -catenin, ephrinA1 and dapper was completed using the same RNA samples used for the microarray analysis. cDNA reactions Reverse Transcriptase reactions were random hexamer-primed using Applied Biosystems (Foster City, CA) High Capacity cDNA Archive Kit. (All Reverse Transcriptase reactions were done at the same time so that the same reactions could be used for all gene studies.) For the construc tion of standard curves serial dilutions of pooled sample RNA were used (50, 10, 2, 0.4, 0.08, and 0.016 ng) per reverse transcriptase reaction. One no RNA cont rol was included. Additionally, one no reverse transcriptase control was included for the standard curve and for each sample.
49 Real-Time PCR reactions TaqMan Gene Expression Assays (Applie d Biosystems) were used. The assay primer & probe sequences are proprietar y. The probe is labeled with 6-carboxyfluorescein as the reporter on the 5 end, and a non-fluorescent quencher plus a minorgroove binder on the 3-end. Each assay is supp lied as a 20X mix of primers and probe. Real-time quantitative PCR analyses we re performed using the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). All standards, the no template control (H2O), and the no amplification control (Bluescr ipt plasmid) were tested in quadruplicate wells (2 wells/plate x 2 plates). All samples were tested in triplicate wells. The no RT controls were tested in duplicate wells. PCR was carried out with the TaqMan Universal PCR Master Mix (Applied Biosystems) using 2 l of cDNA and 1X primers & probe in a 20-l final reaction mixture. After a 2-mi n incubation at 50C, AmpliTaq Gold was activated by a 10-min incubation at 95C, fo llowed by 40 PCR cycles consisting of 15 s of denaturation at 95C and hybridization of pr obe and primers for 1 min at 60C. Data were analyzed using SDS software version 2.2.2 and exported into an Excel spreadsheet. The 18s data were used for normalizing th e gene values ng gene/ng 18s per well. Personal health identifier Tissues from patients treated at the Sa rcoma clinic and collected by the tissue procurement lab at the H. Lee Moffitt Cancer Center were used for this study. The specimens were handled and stored by the tissue procurement lab under the supervision of Dr. T. Hoover. Specimens were released for RNA extraction and microarray analysis
50 after the removal of all PHIs. The specimens were coded and could not be linked to the donor any individual involved in the analyses of the said samples. Involvement of human subjects No procedures were performed on patients. Only pathologic material that was previously collected for research purposes and that was not necessary for a complete pathologic diagnosis was utilized in this study. A separate informed consent beyond what is routine for surgery at H. Lee Moffitt Cancer Center was not necessary.
51 Introductory Preclinical Data Studies have demonstrated that STAT pr oteins participate in essential cellular functions including cellular immune function, developm ent, differentiation and proliferation (237-242). STATs are a family of latent cytoplasmic transcription factors that associated with RTKs and NRTKs. Upon association with activated tyrosine kinases and activation by tyrosine phosphorylati on they can either homdimerize or heterodimerize with other activated STAT proteins via SH2 domain-phosphotyrosine interacting with other STAT family members. Activated STAT dimers translocate to the nucleus and regulate gene tran scription by binding to elements within gene promoters (243, 244). An accumulation of evidence has demonstrat ed a critical role of STAT3 in the molecular pathology of cancer, including tumor formation (245, 246). Constitutive activation of STATs has been shown to pl ay a significant role in the malignant transformation of cells (245, 247, 248). Specifi cally, constitutive activation of STAT3 is associated with transformation by v-Src a nd other viral oncopro teins that activate tyrosine kinase signaling pathways (187, 217, 249, 250) and is required for the transcription of genes involved in v-Srcinduced cellular transf ormation (186, 251). STATs have been demonstrated to be c onstitutively activated in epithelial and hematopoietic human cancers. However, de spite overwhelming evidence to suggest a potential role of STATs in sarcomagenesis, few direct studies have been completed to
52 evaluate the role of STAT3 ac tivation in human sarcomas. The most significant evidence to imply a role for STATs in sarcomagenes is is the overexpressi on and/or mutation of several TKs in sarcomas. Recall that PDGF R, c-KIT, IGFR-1 and VEGFR have been shown to be overexpressed in many sarcomas of various histologies. STATs are involved in transducing TK signaling for all of these pathways under normal and malignant conditions. Furthermore, Src, another essen tial protein involved in activating STATs and eliciting the activation of the aberrantly activ ated TKs in sarcomas also activates STAT3. Src was originally identified as the viral oncoprotein res ponsible for RSV, however the role of Src has yet to be esta blished in human sarcomas. Th erefore, the evaluation of the Src-STAT3 signaling pathway in human sarc oma cell lines was initiated. Signaling by STAT proteins is associated with activation of Src tyrosine kinase (187, 249), thus Immunohistochemi stry (IHC) was utilized to determine activation status of STAT3 in human sarcoma specimens a nd STAT3 DNA-binding activity in nuclear extracts from human sarcoma cell lines. Pharm acological inhibitors of Src were used to evaluate its role in STAT3 activation in human sarcoma cell lines.
53 Results Expression of SFKs in sarcoma cell lines RTPCR was performed to determine which SFKs were expressed in eight sarcoma cell lines. Of the 11 recognized SFK members, three including Src, Fyn and PTK2 were expressed in all eight cell lines examined (Fig. 5A-H; Table 5). Yes1 was expressed in all but the HT-1080 cell line (Fig. 5G). Hck was not expressed in any of the cells lines, while only barely detectable leve ls of Lck were expressed in U 2 OS cells (Fig. 5D). The remaining SFK members were ex pressed in some cell li nes, but not others without an apparent preference fo r tumor type, i.e. STS or OSA. STAT3 is activated in human sarcomas and sarcoma cell lines Immunohistochemistry for the activated form of STAT3 protein was performed on human sarcoma specimens using antibod ies specific for phosphorylated STAT3 Tyr705 (pSTAT3). STAT3 is activated upon phosphorylation of Tyr705, which induces dimerization, translocation to the nucl eus and DNA binding (248). Results show activated STAT3 in sarcomas of diverse subt ypes (Fig. 6), including high grade OSA (A), pleomorphic LPS (B) and pleomorphic, undiffere ntiated high grade sarcoma (C). Since activation of STAT3 was found in the majority of sarcoma tissues examined (Fig. 6D), we determined the status of STAT3 activati on in seven sarcoma cell lines using EMSA. Results show STAT3 was activated in all seve n cell lines to varying degrees (Fig. 6E); MG-63, an OSA cell line, had the greatest and HT1080, a FBS cell line, had the least amount of STAT3 activation.
54 Figure 5. Evaluation of SFK expression in human sarcoma cell lines. The expression of SFKs were determined in eight sarcoma cell lines using the SFK Superarray multiGene12 RT-PCR Profiling Kit. The kit was perf ormed on eight cell lines and the PCR products were resolved on 1% agarose gels fo r SaOS-2 (A), LM-2 (B), LM-7 (C), U 2 OS (D), MG-63 (E), SK-LMS-1 (F), HT1080 (G) and SW-872 (H). The lane numbers correspond to the specific SFK for that we ll and the numbering system for can be found on Table 5. A B C D F E G H 123456789101112123456789101112 A B C D F E G H 123456789101112123456789101112
55 Table 5. SFK gene expression status in human sarcoma cell lines SFK Gene Expression Status Cell Line BLK (1) FGR (2) FYN (3) HCK (4) LCK (5) LYN (6) PTK2 (7) PTK2B (8) SRC (9) SYK (10) YES1 (11) GAPD (12) SaOS-2 x x x x x x x LM-2 x x x x x x x LM-7 x x x x x x x x U 2 OS x x x x x x x x MG-63 x x x x x x x x x SK-LMS x x x x x x x x HT1080 x x x x x x SW872 x x x x x x x x x Table 5. PCR was performed for 11 SFK family members in eight sarcoma cell lines. The presence of an x on the table denotes expression of the co rresponding SFK. The SFK genes are listed at the top of the table, the number in parent heses corresponds to the lane number in Figure 6; panels A-H.
56 D A B C A B CpStat3 Expression Index 80 20 10 240 180 150 100 100 80 60 60 40 5 270 240 180 160 160 140 60 40 30 80 40 140 5 120 20 5 270 120 270 240 200 160 140 050100150200250300Osteosarcoma Liposarcoma Chondrosarcoma EWS MPNST Angiosarcoma Undifferentiated High Grade Sarcoma Leiomyosarcoma 80 20 10 240 180 150 100 100 80 60 60 40 5 270 240 180 160 160 140 60 40 30 80 40 140 5 120 20 5 270 120 270 240 200 160 140 050100150200250300Osteosarcoma Liposarcoma Chondrosarcoma EWS MPNST Angiosarcoma Undifferentiated High Grade Sarcoma Leiomyosarcoma pStat3 Expression Index 80 20 10 240 180 150 100 100 80 60 60 40 5 270 240 180 160 160 140 60 40 30 80 40 140 5 120 20 5 270 120 270 240 200 160 140 050100150200250300Osteosarcoma Liposarcoma Chondrosarcoma EWS MPNST Angiosarcoma Undifferentiated High Grade Sarcoma Leiomyosarcoma 80 20 10 240 180 150 100 100 80 60 60 40 5 270 240 180 160 160 140 60 40 30 80 40 140 5 120 20 5 270 120 270 240 200 160 140 050100150200250300Osteosarcoma Liposarcoma Chondrosarcoma EWS MPNST Angiosarcoma Undifferentiated High Grade Sarcoma Leiomyosarcoma
57 E Figure 6. STAT3 activation status in huma n sarcoma tissues and cell lines. A-C immunohistochemistry for p-STAT3 (Y705) re veals that Src is activated in human sarcoma tissues; high grade OSA ( A ), pleomorphic LPS ( B ) and pleomorphic, undifferentiated high grade sarcoma ( C ). D, STAT3 is activated in several diverse sarcomas, this graph depicts the averag e intensity of STAT3 phosphorylation as measured by IHC. E Cell-free extracts were prepared from untreated cells grown in 10% FBS and EMSA was performed to determine th e activation status of STAT3 in several sarcoma cell lines. Supershift was completed on all cell lines in prev ious experiments to prove that STAT3, specifically, is activated in the cell lines. Supershift was only completed on SaOS-2 in lane one of this experiment and is denoted by the *. LM-2 SaOS-2 SW872 M-G63 STAT3 *SaOS-2 Supershift U 2 OS SK-LMS-1 HT1080 LM-2 SaOS-2 SW872 M-G63 STAT3 *SaOS-2 Supershift U 2 OS SK-LMS-1 HT1080
58 PD180970 inhibits Src and STAT3 signaling in human sarcoma cell lines PD180970 has previously been shown to dir ectly inhibit the kinase activity of purified Src protein in vitro with an IC50 of 16.8 nM (252). To evaluate the effect of PD180970 on Src kinase activity in intact sarcoma cells, we treated the SK-LMS-1, SaOS-2, LM2 and U2 OS cell lines with escalating doses of PD180970 (125, 250 and 500 nM) for 24 h. Western blot analysis was pe rformed to evaluate p-Src (Tyr419) levels and EMSA was performed to determine the effect on STAT3 activation. Dose-response results for p-Src are shown in Figure 7A fo r a representative cell line (SK-LMS-1) and STAT3 activation for SK-LMS-1, SaOS-2 and LM2 are shown in Figure 7B. Src phosphorylation is diminished with 125 nM and completely in hibited with 250 nM PD180970. The kinetics of STAT3 inhibition is not as straight forward as the Src response to PD180970. EMSA shows that STAT3 activation is partia lly inhibited with all doses of PD180970 in three of the four cell lines displayed (Fig. 7B). SaOS-2 is the most sensitive with STAT3 inhibition ma rkedly decreased with 125 nM PD180970. SKLMS-1 and LM2 demonstrate maximum inhi bition of STAT3 activation with 500 nM PD180970. U 2 OS cells had no response to PD180970 as measured by Src phosphorylation and STAT3 activation. The activ ity of Src (data not shown) and STAT3 remained constitutively activated at all dos es of PD180970 (Fig. 7B). Time course experiments were performed to determine the inhibitory kinetics of PD180970 on STAT3 activation. As shown in representative resu lts with SK-LMS-1 cells, inhibition of STAT3 activation is reduced at 3 h and is completely inhibited at 24 h with 500 nM treatment of PD180970 (Fig. 8A). However, STAT3 activation increases at 36 h and later time points, although activation does not return to untreated levels.
59 Figure 7. PD180970 inhibits Src and STAT3 signaling in sarcoma cell lines. A, SKLMS-1 cells were treated with PD180970 in a dose-dependent manner for 6 h. Cell-free extracts were immunoblotted with an tibodies specific to p-Src (Y419) and -actin. B SKLMS-1, SaOS-2, LM-2 and U 2 OS cells were treated with PD180970 in a dose response for 24 h. EMSA was performed to evaluate ST AT3 activation. Western blot analysis and EMSA were performed as described. DMSO was used as a vehicle control in all experiments. pSrcTyr419 ActinDMSO 125 nM 250 nM 500 nM PD180970A B DMSO 125 nM 250 nM 500 nM PD180970 SaOS-2 LM-2 SK-LMS-1 U2 OS pSrcTyr419 ActinDMSO 125 nM 250 nM 500 nM PD180970A B DMSO 125 nM 250 nM 500 nM PD180970 SaOS-2 LM-2 SK-LMS-1 U2 OS
60 PD180970 inhibits cell viability and indu ces apoptosis in sarcoma cell lines To determine the effect of PD180970 on sarcoma cell surviv al over time, we performed growth curve analyses in cell lin es treated with 500 nM PD180970 in a time course. These assays suggested that the sarcoma cell lines responded to PD180970 by induction of apoptosis in a tim e-dependent manner (Fig. 8B). SaOS-2 cells experienced reduced viability with increased exposure to 500 nM PD180970. By 36 h, less than 55% of the cells were viable (p < 0.05) and by 72 h less than 10% of the cells were viable (p<0.001). To further validate the induction of apoptosis, Western blot analysis was performed for PARP cleavage. SK-LMS-1 a nd SaOS-2 cells were treated with 500 nM of PD180970 increasing periods of time and PARP cleavage was evaluated in both cell lines (Fig. 8C). PARP cleavage, an indicator of apoptosis, was evident in both cell lines after 8 h of treatment and increased with tim e. Moreover, TUNEL assays performed on SK-LMS-1, SaOS-2 and LM2 cells confirme d that increasing numbers of the cells underwent apoptosis 48 h after treatment with 500 nM of PD180970 (Fig. 9A-C respectively). Therefore, PD180970 induces ap optosis in sarcoma cell lines with doses that correspond to the in hibition of Src and STAT3 activation by PD180970.
61 Figure 8. PD180970 inhibits viability and induces apoptosis in sarcoma cell lines. A PD180970 inhibits STAT3 activation in a time-d ependent manner. SK-LMS-1 cells were treated with 500 nM PD18097 for increasing peri ods of time. Cell-free extracts were utilized for EMSA to determine STAT3 activa tion status. B, SaOS-2 were treated with 500 nM PD18097 in a time course and cell viability was determined by trypan blue exclusion assays in triplicate. C, SK-LMS-1 and SaOS-2 cells were treated with 500 nM PD18097 in a time course. Cell-free extrac ts were immunoblotted with antibodies specific to PARP to measure induction of apoptosis. DMSO wa s used as vehicle control for all experiments. Stat3 EMSA 8 DMSO 1224364872 3HStat3 EMSA 8 DMSO 1224364872 3HAB 0 20 40 60 80 100 120 DMSO81224364872 Hours% Viabilit y * ** 0 20 40 60 80 100 120 DMSO81224364872 Hours% Viabilit y * **SK-LMS-1 PARP SaOS-2 8 DMSO 1224364872 3HPARP Cleavage PARP PARP Cleavage SK-LMS-1 PARP SaOS-2 8 DMSO 1224364872 3HPARP Cleavage PARP PARP CleavageC
62 Figure 9. PD180970 induces apoptosis in sarcoma cell lines as measured by TUNEL. AC, apoptosis was further verified by TUNEL assay as described. SK-LMS-1, SaOS-2 and LM-2 cells were plated in 12-well tissue cu lture plates and treated with DMSO or 500 nM PD180970 for 48 h. After completing the T UNEL assay, cells were visualized using light microscopy and photogra phed at 20x magnification. DMSO48 Hr A B C
63 SKI-606 does not inhibit Src-STAT3 signaling in human sarcoma cell lines SKI-606 has previously been shown to di rectly inhibit the kinase activity of purified Src protein in vitro with an IC50 of 3.8 nM (253) and 100 nM for intact cells (254). To evaluate the effect of SKI-606 on Src kinase activity and STAT3 activation in intact sarcoma cells, we treated several sarcoma cell lines with escalating doses of SKI606 (300, 1000, 3000 and 10,000 nM) for 24 h. Western blot analysis was performed to evaluate p-Src (Tyr419) levels and EMSA was performed to determine the effects on STAT3 activation. There were no apparent eff ects on pSrc (Y419) in any of the cell lines evaluated (data not shown). Furthermore, dose-response resu lts for STAT3 activation for SK-LMS-1 and U2 OS are shown in Figur e 10. There was no inhibition of STAT3 activation by SKI-606 in any of the cell lines examined. Furthermore, there were no observed anti-prolifera tive or pro-apoptotic effects fr om SKI-606 at any of the doses tested. Figure 10. SKI-606 does not inhibit Src-STAT3 signaling in sarcoma cell lines. SKLMS-1 and U 2 OS cell were treated with SKI-606 in a dose-response for 24 h. EMSA was performed to measure STAT3 activation in response to SKI-606 treatment. DMSO was used as vehicle control. NIN3T3 vSRC SKLMS-1U2 OS DMSO0.31M 310DMSO0.313 Stat3NIN3T3 vSRC SKLMS-1U2 OS DMSO0.31M 310DMSO0.313 Stat3
64 Dasatinib does not inhibit Src-STAT3 signaling in human sarcoma cell lines Dasatinib has previously been shown to di rectly inhibit the enzymatic activity of Src with an IC50 of 0.5 nM (231). The effects of Src activity are described in the next chapter of this dissertation. However, to evaluate the effect of dasatinib on STAT3 activation in sarcoma cells, se veral sarcoma cell lines were treated with escalating doses of dasatinib (30, 100, 300 and 1000 nM) for 6 h. EMSA was performed to determine the effects on STAT3 activation si x cell lines. There was no i nhibition of STAT3 activation by dasatinib in any of the cell lines examin ed Fig. 11A. Western blot analysis was performed to confirm these observations Figure 11B and C show that pSTAT3 expression does not decrease with increased dasatinib concentrati on in SaOS-2 and U 2 OS cell lines, respectively. Furthermore, there were no observed effects of STAT3 activation by dasatinib in SK-LMS-1 cell lines when the doses escalated to 3 and 10 M for 24 h (Figure 11D).
65 Figure 11. STAT3 signaling is indepe ndent of Src kinase activity in human sarcoma cell lines. A, SaOS-2, LM-2, U 2 OS, MG-63, SK-LMS-1 and HT1080 were treated with dasatinib in a dose response for 6 h. EMSA was performed to evaluate STAT3 activation status in response to treatment with dasatinib. B and C, SaOS-2 and U 2 OS cells were treated with escalating doses of dasatini b for 6 h and cell free extracts were immunoblotted for pSTAT3 (Y705) and -actin as loading control. D, SK-LMS-1 cells were treated with dasa tinib in a dose response that achie ved higher doses of dasatinib for 24 h. EMSA analysis was performed to evalua te STAT3 activation status in response to treatment with dasatinib. MG-63 HT1080 SaOS-2 LM-2 U 2 OS SK-LMS-1 DMSO30 1003001000 nMA B C DMSO 301003001000 pStat3 (Y705) -actin nM pStat3 (Y705) -actin DMSO 301003001000nM DMSO1252505001000300010,000 1102550100nM Stat3DMG-63 HT1080 SaOS-2 LM-2 U 2 OS SK-LMS-1 DMSO30 1003001000 nMA B C DMSO 301003001000 pStat3 (Y705) -actin nM pStat3 (Y705) -actin DMSO 301003001000nM DMSO1252505001000300010,000 1102550100nM Stat3 DMSO1252505001000300010,000 1102550100nM Stat3D
66 Discussion Constitutive activation of STAT3 has been observed and demonstrated to play an essential role for the tumori genesis in many solid tumor and hematopoietic malignancies (195, 196, 245, 246). The status of STAT3 activa tion in sarcomas has not been explored previously. Our results show that STAT3 is constitutively activated in many of the human sarcoma specimens analyzed and most of the sarcoma cell lines examined. To elucidate the role of STAT3 activation in hum an sarcoma cell lines, we investigated the activity of three different Sr c kinase inhibitors. We ha ve shown the PD180970 inhibits Src and STAT3 signaling in several of the cell lines utiliz ed. Although the IC50 for STAT3 inhibition is markedly higher than the IC50 required to inhibit Src activity for most of the responsive cell lines (Fig. 7). However, cell viability was decreased and apoptosis was induced by 500 nM PD108970 trea tment in a time-dependent fashion. Furthermore, the dose of PD180970 corresponded to complete inactivation of Src and STAT3 kinases is all responsive cell lines. These data highlight the potential promise of employing a Src or STAT3 inhibitors for the treatment of sarcomas. There is little in centive to perform additional experiments with PD180970 because the stab ility of this compound is limited and obtaining reproducible data is difficult. This compound also has little to no future in the clinic as it is only soluble in DMSO and is highly nonspecific. In addition, several more stable, more specific, bioavailable Src kinase inhibitors with more promising clinical futures have been made available. Two such Src kinase inhi bitors are SKI-606 and
67 dasatinib were further investig ated to determine a more precise role for Src and STAT3 activity in sarcomas. The Src kinase inhibitor, SKI-606, is currently undergoi ng several phase 2 clinical trials for malignancies including breast, panc reatic, colon and non-small cell lung cancers and chronic myelogenous leukemia (255). We evaluated the activity of this compound in sarcoma cell lines to inve stigate the effects of Src and STAT3 signaling. To our disappointment, SKI-606 had no inhibitory ac tivity on either of these proteins in sarcomas. There were no apparent prolifera tion or survival effect s generated by exposure to this compound either. One possible explan ation for these results may be that the compound is not crossing the cell membrane and gaining access to the cell. Src kinase may also have a greater affinity for PD 180970 or dasatinib, which may further explain the lack of response to SKI-606. Dasatinib was the last Src inhibitor to be obtained and evaluated, interestingly enough it was also the compound that exhib ited the most potential as a possible therapeutic option for sarcomas. However, th e promising anti-tumor effects of dasatinib do not appear to be mediated via a Src-STAT3 pathway. Dasatinib was shown to potentially inhibit Src activ ity in sarcoma cell lines, as will be more thoroughly demonstrated in the following chapter. Surprisingly, inhibition of Src activity by dasatinib had no effect on STAT3 activation in sarcoma cell lines (Fig. 11). These results do not rule out a possible role for constitutive STAT3 activation in sarcomas, but suggests that the Src and STAT3 signaling path ways may be uncoupled in sarcomas. The data generated using PD180970 suggested a higher IC50 was required to inhibit STAT3 as
68 compared to Src activity, suggesting that Sr c and STAT3 may be cross-talking with other pathways in sarcoma cell lines. In spite of this, these results suppor t the possibility of a dependence of Src activity for survival in sa rcomas. Therefore, we next set out to determine the role of Src activity in sarc omas by investigating other Src dependent pathways.
69 Dasatinib Inhibits Migration and Invasio n in Diverse Human Sarcoma Cell Lines and Induces Apoptosis in Bone Sarcoma Cells Dependent on Src Kinase for Survival One potential molecular target for sarcom a treatment is the Src tyrosine kinase. Dasatinib, a small-molecule inhibitor of Src kinase activity, is a promising cancer therapeutic agent with oral bioavailability. Dasatinib exhibits anti-tumor effects in cultured human cell lines derived from epith elial tumors including prostate and lung carcinomas. However, the action of dasatinib in mesenchymally-derived tumors has yet to be demonstrated. Dasatinib was originally selected as a Src kinase inhibitor and then shown to inhibit Bcr-Abl as we ll as other tyrosine kinases. There have been several studies demonstrating the activ ity of dasatinib against Bc r-Abl-positive leukemic cell lines as well as epithelial tumor cell lines (4348). In addition, early phase clinical trials have established the safety and efficacy of dasatinib for treatment of imatinib-resistant chronic myelogenous leukemia patients. However, the responses and mechanisms of action of dasatinib in mesenchymally-derived tumor cell lines have not been described previously. We report that dasatinib inhibits Src and downstream FAK signaling at nanomolar concentrations, blocks cell mi gration and invasion in many diverse human sarcoma cell lines and induces apoptosis in bone sarcomas. Furthermore, knockdown of Src expression by siRNA in bone sarcoma cells also induces apoptos is, suggesting that the observed response to dasatinib in these cells is conveyed through inhibition of Srcmediated signaling. Together these findings indicate that dasatinib is a promising
70 therapeutic agent for preventing growth and meta stasis of a wide diversity of soft-tissue and bone sarcomas. Based on our previ ous findings of Src activation in human sarcomas, we evaluated the effects of dasati nib in twelve cultured human sarcoma cell lines derived from bone and soft-tissue sarcomas.
71 Results Src kinase is activated in human sarcomas and sarcoma cell lines Immunohistochemistry for activated Src protein was performed on human sarcoma specimens for activated Src protei n using antibodies to phospho-Src (p-Src). Levels of p-Src on tyrosine residue 419 (Y 419) due to autophosphorylation reflect Src kinase activities in intact cells and tissues. Results show activated Src in sarcomas of diverse subtypes (Fig. 12), including leiom yosarcoma (A), high-grade OSA (B), and LPS (C). Since autophosphorylated Src was found in a majority of the human sarcomas examined, including diverse soft-tissue a nd bone sarcomas (data not shown), we determined the level of Src activation in a panel of human sarcoma cell lines by Western blot analysis for p-Src (Y419) and total Src pr otein levels. Src was detectably activated in all but one (HT-1080) of the cell lines examined, albeit to di fferent extents (Fig. 12D). Total Src protein expression does not correlate with levels of phosphorylated Src in every case, indicative of different levels of Src kinase activation among the sarcoma cell lines. In addition, the level to which Src kinase is activated (p-Src levels ) does not correlate with specific sarcoma histological sub-types (Fig. 12 and Table 6).
72 Figure 12. Src kinase is activated in hu man sarcoma tissues and cell lines. A-C immunohistochemistry for p-Src (Y419) reveals that Src is activated in human sarcoma tissues (leiomyosarcoma A high grade osteosarcoma B and liposarcoma C ). D, Src is activated in all but one (HT-1080) of the cell lines utilized for these experiments. Cellfree extracts were prepared from untreated cells grown in 10% FBS and immunoblotted with antibodies specific fo r p-Src (Y419), total Src or -actin. D p-Src (Y419)LM2 MG-63 U-2 OS SK-LMS-1 HT-1080 LM7 SaOS-2Total Src -ActinSK-ES-1 TC-71 A673 RD RD18 p-Src (Y419)LM2 MG-63 U-2 OS SK-LMS-1 HT-1080 LM7 SaOS-2Total Src -ActinSK-ES-1 TC-71 A673 RD RD18 LM2 MG-63 U-2 OS SK-LMS-1 HT-1080 LM7 SaOS-2Total Src -ActinSK-ES-1 TC-71 A673 RD RD18 ABC ABC
73 Dasatinib inhibits Src kinase acti vity in human sarcoma cell lines Dasatinib has previously b een shown to directly inhi bit the kinase activity of purified Src protein in vitro with an IC50 of 3 nM (223). To evaluate the effect of dasatinib on Src kinase activity in intact sarcoma cells, we treated the above cell lines with escalating doses of dasatinib (30, 100, 300 and 1000 nM) for 6 h and Western blot analysis was performed to evaluate p-Src levels. Dose-response results for the two representative cell lines (SaOS-2 and U 2 OS ) are shown in Figure 13 (A and B), the IC50 values for inhibition of p-Src by dasatinib range from 3 to 68 nM for all cell lines analyzed (summarized in Table 6). Time course experiments were performed to determine the kinetics at which Src phosphor ylation is inhibited by dasatinib. As shown in representative results with U 2 OS cells, inhibition of Sr c phosphorylation is complete by 15 min following treatment with 100 nM of dasatinib and persists for at least 24 h (Fig. 13C). Interestingly, total Src protein expression was increased in a doseand timedependent manner in a subset of the cell lines tr eated with dasatinib. In particular, all but one bone sarcoma cell line (MG-63) produced in creases in total Src protein expression, yet this effect was not observed in the soft -tissue sarcoma cell lines (Fig. 13A-C, and data not shown). However, this in creased protein expression of Src was not accompanied by an increase of c-Src mRNA, as shown in both cell lines which an increased expression of total Src was and was not observed (Fig. 14). These data suggest a positive feedback mechanism for compensation of Sr c kinase inhibition with incr eased levels of Src protein expression in the bone sarcoma cells.
74 B C0.25.502 DMSO4 6 12 24 H Src -Actin p-Src (Y419) 1 301003001000 DMSO p-Src (Y419) Src -Actin nM A p-Src (Y419) Src -Actin 301003001000 DMSOnM B C0.25.502 DMSO4 6 12 24 H Src -Actin p-Src (Y419) 1 0.25.502 DMSO4 6 12 24 H Src -Actin p-Src (Y419) 1 301003001000 DMSO p-Src (Y419) Src -Actin nM 301003001000 DMSO p-Src (Y419) Src -Actin nM A p-Src (Y419) Src -Actin 301003001000 DMSOnM p-Src (Y419) Src -Actin 301003001000 DMSOnM
75 Figure 13. Dasatinib inhibits Src activation a nd downstream signaling in sarcoma cell lines. A and B SaOS-2 and U 2 OS cells were treate d with dasatinib in a dose-dependent manner for 6 h. Cell-free extracts were immuno blotted with antibodies specific to p-Src (Y419) and total Src. C U 2 OS cells were treated with 100 nM of dasatinib in a timedependent manner. Western blot analysis wa s performed as described. DMSO was used as a vehicle control and -actin was immunoblotted for as a loading control in all experiments. D dasatinib specifically blocks ty rosyl phosphorylation of FAK (Y576/577, Y925) and 130CAS (Y410), but not FAK Y397. SaOS-2 ce lls were treated with dasatinib for 6 h in a dose-dependent manner. Ce ll-free extracts were immunoblotted with antibodies specific to p-FAK (Y 397, Y576/577, Y925), total FAK, p130CAS and pp130CAS (Y410). D 301003001000 DMSOnM p-FAK (Y576/577) FAK p-p130CASp130CAS-Actin p-FAK (Y925) p-FAK (Y397) 301003001000 DMSOnM p-FAK (Y576/577) FAK p-p130CASp130CAS-Actin p-FAK (Y925) p-FAK (Y397)
76 Figure 14. Dasatinib does not induce c-Src mRNA expression. Quantitative RT-PCR was performed on SaOS-2, U 2 OS and SK-LMS-1 ce lls to determine the effect of dasatinib on c-Src mRNA expression. Cells were treated with dasatinib in a dose-response and RNA was isolated, purified and qRT-PCR for c-Src was performed as described. 0 20 40 60 80 100 120 140 160 0301003001000 dasatinib, nM% Contro l SaOS-2 U 2 OS SK-LMS-1
77 Dasatinib selectively blocks Src downstream signaling Src kinase has been shown to regulate cellular activities through a number of downstream signaling pathways. One such pathway is FAK, a non-receptor tyrosine kinase found to be increased in a variety of epithelial cancers including those arising from prostate, cervical and colon (256-259). Fu rthermore, increased FAK expression is associated with tumor progression in a mouse model of skin carcinogenesis (200). FAK, in turn, has been implicated in the ac tivation of CRK-associated substrate, p130CAS, which together with Src and FAK plays a vital role in cell adhesion, migration, proliferation and survival (209). To inves tigate the effect of dasatinib on these Src downstream signaling pathways, sarcoma cell line s were treated in culture with escalating doses of dasatinib for 6 h. As representa tive results using Sa OS-2 cells, Figure 13D shows Western blot analysis performed using antibodies to total FAK protein, phosphorylated FAK (Y397, Y576/Y577, and Y925), p130CAS and phosphorylated p130CAS (Y410). The IC50 values for inhibition of phosphorylated FAK (Y576/Y577 and Y925) and p130CAS were between 30 to 100 nM, consistent with the IC50 values for inhibition of Src kinase activity in th ese cells (Fig. 13 and Table 6). FAK autophosphorylation (Y397) was not substa ntially inhibited until higher doses of dasatinib (1000 nM), indicating that dasatini b does not directly inhibit FAK kinase activity. Surprisingly, total p130CAS protein was diminished with dasatinib treatment while total FAK protein was not aff ected (Fig. 13D), suggesting that p130CAS protein is subject to negative feedback re gulation in these cells. By contrast, dasatinib did not inhibit STAT3 signaling in sarcoma cell lines (data not shown), another signaling pathway that has been shown to act downstr eam of Src in cells of other tumor types
78 (260). Thus, dasatinib selectively blocks FAK and p130CAS signaling downstream of Src in sarcoma cell lines. Dasatinib blocks cell motility and invasion by sarcoma cells Both FAK and p130CAS activity are involved in re gulating cell migration and invasion downstream of Src kinase. The e ffect of dasatinib on cell migration was evaluated using wound healing assays (by sc ratching cell monolayers with a pipette tip) and treating with drug. Cells were plated in 0.1% serum me dium prior to inducing the wound to ensure that migration rather than cell growth was measure d. The width of the wound was determined at T0 and then cells were treate d with escalating doses of dasatinib in 0.1% serum medium for 24 h. Representative results are shown with the SaOS-2 cells, which were digitally photogr aphed and the width of denuded area in the wound was measured in pixels (Fig. 15A and B). Wound healing was dramatically inhibited by dasatinib in a dos e-dependent manner, with dete ctable inhibition at 30 nM and substantial inhibition at 100 nM dasatinib. To evaluate th e effect of dasatinib on cell invasion, SaOS-2 and U 2 OS cells were trea ted with dasatinib in a dose-response manner for 22 h in Matrigel Invasion Chambers. Dasatinib significantly inhibited cellular invasion in both cell lines (Fig. 15C). Th e SaOS-2 cell line was more sensitive to inhibition of invasion by dasatinib comp ared to the U 2 OS cell line. The IC50 values for inhibition of tumor cell invasion in this a ssay ranged from 30 to 100 nM. These IC50 values for blockade of cell migration a nd invasion are consistent with the IC50 values for inhibition of Src kinase as we ll as downstream FAK and p130CAS signaling (compare with Table 6 and Fig. 13D).
79 DMSO30 nM 100 nM 300 nM1 M T0 DMSO30 nM 100 nM 300 nM1 M T0 A 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 T00301003001000 dasatinib, nMPixe* * 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 T00301003001000 dasatinib, nMPixe* * B
80 Figure 15. Dasatinib inhibits cel l motility and invasion. A, wound healing assays were performed to determine the effects of dasatin ib on inhibiting cell migration. SaOS-2 cells were plated in 12-well tissu e culture plates, grown to c onfluency and serum-starved overnight in medium containing 0.1% FBS. Wounds were introduced on cell monolayers using a pipette tip. Cells were washed with 1x PBS to remove non-adherent cells and treated with DMSO or dasati nib in a dose-dependent manner for 24 h. Cell migration was visualized at 10x magnification by light mi croscopy and photographed with a digital camera. B width of voided area versus dasatini b dose concentration was graphed to express the degree of inhibi tion of cell migration. The num ber of pixels within the denuded area were the units used to demons trate inhibition of ce ll migration induced by dasatinib (*p<0.001, n=3). C matrigel Invasion Chambers were used to measure dasatinib inhibition of cellula r invasion. SaOS-2 (solid bars ) and U 2 OS (shaded bars) were treated with da satinib in a dose-dependent manne r. Cells were diluted in 500 L of serum-free medium and placed over the inner ch amber of the insert in a 24-well tissue culture plate and 500 L of complete medium was placed in the lower chamber of the insert. After incubating for 22 h, cells that invaded through the Matrigel were stained, visualized using light microscopy, photogra phed and counted. Each experiment was performed in triplicate. The m ean values of invasive cells were graphed versus dasatinib concentration (*p<0.001, **p<0.01, n=3). To conf irm that invasion was measured rather than induction of apoptosis, trypan blue excl usion assays were performed with SaOS-2 and U 2 OS cells treated in a dose-dependent manner with dasatinib for 24 h. Greater than 90% of the cells were viable afte r dasatinib treatmen t (data not shown). C 0 20 40 60 80 100 120 0301003001000 dasatinib, nMInvasion (% control)** ** 0 20 40 60 80 100 120 0301003001000 dasatinib, nMInvasion (% control)** **
81 Dasatinib induces apoptosis of bone sarcoma cell lines To determine the effect of dasatinib on sarcoma cell survival, we performed growth curve analyses in cell lines treated wi th increasing concentra tions of dasatinib. These assays suggested that the subset of cell lines derived from bone sarcomas responded to dasatinib by induction of apopt osis in a dose-dependent manner (data not shown). To further validate the induction of apoptosis in this subset of cell lines, Western blot analysis was performed for apoptotic markers. SaOS-2 and U 2 OS cells were treated with escalating doses of dasatini b for 72 h, and PARP cleavage and XIAP expression were evaluated in both cell lines (Fig. 16A and C). PARP cleavage, an indicator of apoptosis, is evid ent at 30 nM dasatinib and in creased with escalating doses of dasatinib. Furthermore, expression of XI AP, an inhibitor of apoptosis, was diminished by dasatinib treatment with IC50 values ranging from 30 nM to 100 nM (Fig. 16A and C). A time-course analysis with 100 nM dasatini b was performed in th e SaOS-2 cell line to determine when apoptosis was induced as measured by PARP cleavage. PARP cleavage is evident as early as 6 h following treatment with dasatinib and increased with time (Fig. 16B). Moreover, TUNEL assays performed on SaOS-2 cells confirmed that increasing numbers of the cells were undergoing apoptos is by 48 h after treatm ent with escalating doses of dasatinib (Fig. 16D). Therefore, dasatinib i nduces apoptosis in the bone sarcoma subset with IC50 values corresponding to those of inhibition of Src kinase and downstream signaling by dasatinib (Table 6).
82 PARP cleavage -Actin PARP XIAP 301003001000 DMSOnM A B0.25.50 DMSO4 6 12 24 48 72 H PARP cleavage -Actin PARP 1 2 36 -Actin PARP XIAP PARP Cleavage 30 1003001000 DMSO nM C PARP cleavage -Actin PARP XIAP 301003001000 DMSOnM PARP cleavage -Actin PARP XIAP 301003001000 DMSOnM A B0.25.50 DMSO4 6 12 24 48 72 H PARP cleavage -Actin PARP 1 2 36 0.25.50 DMSO4 6 12 24 48 72 H PARP cleavage -Actin PARP 1 2 36 -Actin PARP XIAP PARP Cleavage 30 1003001000 DMSO nM -Actin PARP XIAP PARP Cleavage 30 1003001000 DMSO nM C
83 Figure 16. Dasatinib induces apoptosis in bone sarcoma cell lines. A and C dasatinib induces apoptosis in a dose-dependent manne r. SaOS-2 and U-2 OS cells were treated with dasatinib for 72 h with escalating doses. Cell-free extr acts were immunoblotted with antibodies specific to XIAP and PARP. B, SaOS-2 cells were tr eated with 100 nM of dasatinib in a time-dependent manner and immunoblotted with an antibody specific for PARP. D, apoptosis was further verified by TUNEL analysis as described. SaOS-2 cells were plated in 12-well tissue culture plates and treated with dasatinib for 48 h in a dosedependent manner. After completing the TUNEL assay, cells were visualized using light microscopy and photographed at 20x magnification. D DMSO30 nM 100 nM300 nM D DMSO30 nM 100 nM300 nM DMSO30 nM 100 nM300 nM
84 Src is required for survival of bone sarcoma cell lines To determine if inhibition of Src kina se by dasatinib is sufficient to induce apoptosis in the bone sarcoma cell lines, we transfected these cell lines with siRNA to cSrc Two representative bone sarcoma cell lines, SaOS-2 and U 2 OS, underwent induction of apoptosis in a dos e-dependent manner as meas ured by PARP cleavage in response to siRNA against c-Src but not to control siRNA. Src protein expression was inhibited by transfection with 50 nM and 100 nM siRNA against c-Src corresponding to induction of PARP cleavage in both cell lines (Fig. 17A and B). MG-63, an OSA cell line that does not undergo apoptosis when trea ted with dasatinib, also does not undergo apoptosis when transfected with siRNA to Sr c (data not shown). These data demonstrate that a subset of the bone sarcoma cell lines rely on Src kinase for survival, indicating that inhibition of Src kinase activ ity by dasatinib is sufficient to induce apoptosis in these cells.
85 B A Src -actinBuffer Control siControl50100 siSrc PARP PARP Cleavage nM Src -actinBuffer Control siControl50100 siSrc PARP PARP Cleavage nM Src -actinBuffer Control siControlPARP PARP Cleavage nM 50100 siSrc Src -actinBuffer Control siControlPARP PARP Cleavage nM 50100 siSrc
86 Figure 17. Src is required for bone sarcoma cell line survival. A and B, siRNA to c-Src induces apoptosis in bone sarcoma cell lines. SaOS-2, A, and U 2 OS, B, cells were plated in 6 cm tissue culture plates, transf ected with 50 and 100 nM of siRNA to c-Src (si Src) and harvested after 72 h. Cell-free extracts we re immunoblotted with antibodies specific for Src and PARP to measure the efficiency of knockdown by siRNA and to determine if apoptosis was induced upon deplet ion of Src protein expression. C, Src activation and signaling are inhibited by da satinib. Inhibition of Src si gnaling by dasatinib prevents cellular migration and invasion in sarc oma cell lines. Upon inhibition of Src phosphorylation or c-Src expression, a sub-set of bone sarcoma cell lines undergo an induction of apoptosis. C p130CAS Survival Migration InvasionGrowth Factor Receptor Growth Factors dasatinibY419 Extracellular Cellular Membrane Induction of Apoptosis ExtracellularMatrix P FAK Src p130CAS p130CAS Survival Migration InvasionGrowth Factor Receptor Growth Factors dasatinibY419 Extracellular Cellular Membrane Induction of Apoptosis ExtracellularMatrix P FAK FAK Src Src
87 Table 6. Summary of cell line IC50 values and responses to dasatinib IC50, nM Cell Line Tumor Type pSrc (Y419) Expression pSrc (Y419) Migration Induction of Parp Cleavage, nM SaOS-2 Osteosarcoma ++ 46 65 30 LM2 Osteosarcoma ++ 26 35 30 LM7 Osteosarcoma ++ 68 24 100 U 2 OS Osteosarcoma ++++ 57 4 30 MG-63 Osteosarcoma +++ 28 58 N/R SK-ES-1 Ewings Sarcoma +++ 11 4 30 TC-71 Ewings Sarcoma + 3 4 30 SK-LMS-1 Leiomyosarcoma +++ 46 44 N/R HT-1080 Fibrosarcoma N/R N/R N/R A673 Rhabdomyosarcoma ++++ 26 23 N/R RD Rhabdomyosarcoma ++ 45 29 N/R RD18 Rhabdomyosarcoma + 50 277 N/R +/depicts relative p-Src (Y419) expression N/R No Response Table 6. Summary of responses to dasatinib in human sarcoma cell lines. With the exception of one soft-tissue sarcoma cell lin e (HT1080) all of the cell lines examined respond to dasatinib by inhibition of Src phosphorylation on Y419 and migration at IC50 values consistent with Src kinase inhibiti on. A subset of bone sarc oma cell lines respond to dasatinib treatment by induction of apoptos is. Induction of apoptosis was not observed in the soft-tissue sarcomas cell lines and one osteosarcoma (MG-63) cell line.
88 Discussion After nearly a century since the discov ery of the Rous sarcoma virus, which subsequently was shown to induce sarcomas by capturing and mutationally activating the cellular gene encoding the Src tyrosine kinase targeted Src kinase inhibitors are now entering clinical trials for solid tumors. Sa rcomas comprise a highly diverse set of human tumors that frequently occur among pediatri c cancer patients and for which there are limited treatment options. Based on our observa tion of Src kinase activation in sarcoma clinical specimens, we sought to determine the action of dasatini b, a potent and orally bioavailable inhibitor of Src ki nase, on human sarcoma cell lines. Our findings demonstrate that dasatinib i nhibits Src kinase activity, as measured by autophosphorylation at Tyr419, in a dos e-dependent manner in sarcoma cells. Furthermore, in 11 out of 12 sarcoma cell lines examined dasatinib inhibits cell migration and invasion. The single cell line that did not respond to da satinib (HT-1080) was also the only one that lacked detectab le Src kinase activity in this panel (Table 1). Moreover, suppression of cell migration and invasion was associated with inhibition of downstream Src signaling through FAK and p130CAS, proteins known to be involved in mediating these cellular processes (261-264). The IC50 values for inhibition of Src/FAK/p130CAS signaling as well as migration a nd invasion are all in the range of approximately 30 nM to 100 nM regardless of histological type (Table 1). Taken together, our findings suggest a model for the mechanism of dasatinib action in which blockade of Src and downstream signaling suppresses migration and invasi on of sarcoma cells (Fig. 5C). Significantly, dasatinib induces apoptosis in the majority of bone sarcoma cells lines, including OSA and EWS, but not in any of the soft tissue sarcoma cell lines in our
89 panel. Genetic inhibition of Src using siRNA also induced apoptosis in bone sarcoma cell lines that respond to dasatinib with a poptosis, but not in th e only OSA cell line (MG63) in which dasatinib did not induce apoptos is. Thus, dasatinib induces apoptosis in bone sarcoma cell lines dependent on Src kina se for survival. A major Src signaling pathway involved in preventing apoptosis in other cellular contexts is STAT3 (245, 265). While most of the sarcoma cell lines in this study harbor activated STAT3, with the sole exception of SK-ES-1, dasatinib did not inhi bit STAT3 activation, in dicating that this pathway is not involved in the dasatinib-med iated apoptosis response in sarcomas (our unpublished results). On the other hand, the IC50 values for induction of apoptosis by dasatinib are in the same range re quired for blockade of FAK and p130CAS signaling in these cell lines. Because FAK and p130CAS have been implicated in tumor cell survival, in addition to cell migration and invasion, it is possible that these pathways are involved in dasatinib-mediated apoptos is in sarcoma cells. It is notable that levels of Sr c activation do not correlate with IC50 values of dasatinib responses in terms of cell migrati on, invasion or apoptosis (Table 1). This finding may be explained by the possibility that low levels of Src kinase activation are sufficient to induce these biological propert ies. Alternative explanations are the possibilities that other SFK members or unident ified targets are involved in the responses to dasatinib. Similar results have been obser ved for other molecular targeted-therapeutic agents, such as Iressa, where c linical response to this EGFR inhibitor is not correlated with levels of EGFR expressi on or activation (225). In th e specific case of Iressa, EGFR mutations have been shown to influence res ponse to Iressa; however, mutations in the cSrc gene are extremely rare in human cancers (230). Thus, selection of patients for
90 dasatinib treatment on the basis of Src expressi on or activation levels may not predict the optimal clinical responses. It remains to be determined whether any of the known genetic sub-types of sarcomas are more sensitive to dasatinib than others. Earlier preclinical laborato ry studies pointed to the pr omise of dasatinib in the treatment of Gleevec-resistant chronic myeloid leukemia, a prediction that has been borne out in clinical trials (266-272) On the basis of more recent pr eclinical laboratory studies, several human solid tumor sites have shown pr omise for clinical trials, including prostate, lung, pancreatic, and head and neck cancers ( 223-225). We have established that Src is activated in a wide variety of human sarcoma clinical spec imens, including STS and bone sarcomas. Furthermore, our data demonstrat e that dasatinib inhi bits Src kinase and downstream signaling, leading to blockade of cell migration and invasion of sarcoma cell lines of diverse origins. In the subset of bone sarcomas, dasatinib also induces apoptosis. Taken together, our results suggest that dasa tinib will provide clinical benefit to soft tissue and other sarcomas by preventing metast asis, which may be further augmented in bone sarcomas by induction of apoptosis.
91 Gene Expression Profile of Sarcoma Cell Lines Serves as Preliminary Signature Predictive of Response to Treatment with Dasatinib It has recently been shown that unlike epithelial cancers, sarcomas are better defined by their molecular pathology rather th an the organ of origin (1). Identified molecular alterations and cyt ogenetic analysis of specific subtypes of sarcomas have proven that previous classifica tions based on the site of the tumor are less important than the molecular phenotype of the tumor. Target ed therapies have aided in reaching this conclusion. Patients with sp ecific molecular phenotypes have been shown to respond better to specific targeted therapies, as was the case with Gleevec and c-kit mutations in GISTs. These findings have had an importa nt impact on the appr oaches in treating sarcomas. However, there are still many sarc omas subtypes with little to no improved treatment options. Microarray analysis can help iden tify significant gene s involved in sarcomagenesis and better classify tumo r subtypes by molecular phenotype. Gene expression profiles (GEP) generated from microarray analysis have proven to successfully identify signatures that can predict prognosis and response to chemotherapies for breast cancer (273-276). Microarray analysis has recently provided insight into receptor tyrosine kinase expre ssion patterns in sarcomas that may serve as potential targets for novel therapeutics The examination of GISTs by microarray analysis lead to the identification of c-KIT and PDGFR as potential targets of therapy and
92 provided prognostic signatures that presented a biological basis for the differential responses exhibited to Gleevec. Furthermor e, molecular signatures predicting response to therapy have been published for EWS a nd OSA (277). Consequently, the use of gene expression profiling by microarray analysis cannot only rapidly provide potential therapeutic targets, but may also serve as a screening mechanism used to predict response to a specific therapy, ther eby preventing unnecessary patient exposure to various chemotherapeutics. The second aim of this dissertation resear ch was to identify a candidate molecular signature that predicts response to dasatini b by induction of apoptos is in human sarcoma cell lines. The response status of 12 cell lines was determined in aim one and was used to classify the cells into responsive and non -responsive categories. RNA was extracted from three consecutive passages of each cell line and purified for microarray analysis. A total of 36 Human Genome U133 Plus 2.0 Arrays were utilized for the initial analysis and six more for the second phase of this aim. Once a molecular signature was identified and validated, the accuracy of the signature to pr edict response in cell lines was tested using two new cell lines of unknown response status, to prevent bias. The same approach for RNA collection and purification was used for th e test cell lines. Microarray analysis was performed independently of the molecular analysis and verification of response to dasatinib for each cell line. Once both analyses were completed, the results were compared. Amazingly, the results from both an alyses were the identical; one cell line responded by induction of apoptosis to dasatinib while the other did not. Therefore, a molecular signature that successfully predic ts response to dasatinib by induction of apoptosis was identified in sarcoma cell lines.
93 Results Unsupervised clustering identified three main classes, with five s ubgroups in relation to cell line types Using an unsupervised hierarchical clus tering approach, we tried to identify natural subclasses of cell lines as determined by gene expression prof iles. We performed unsupervised clustering on a low-level filter ed probe list using genes with median intensities of greater than or equal to 2000 to avoid us ing background noise in this analysis. The clustering resu lts are shown in Figure 18A. Interestingly, three main classes were defined; the first included the EWS and RD cell lines, the second consisted of the STS and two OSA cell lines, while the third included three OSA cell lines. The first class is further sub-grouped by EWS and RD Intriguingly, the A673 cell line, which was originally characterized as a RD, but later identified to possess an EWS translocation, was sub-grouped with the EWS ar m. The second class consists of two subgroups, one arm with two STS cell lines, the other arm with two OSA cell lines. Not surprising, the third class includ es three cell lines wh ich were derived from one of the cell lines in the sub-group. A second hierarchical clustering approach was completed on the data set to determine if the cell lines could group accordi ng to response. This cluster was performed using a filtered probe of genes with median intensities greater than or equal to 2000 and had a significant (p-value 0.05) difference fold change between responders and nonresponders. To calculate the fold change the median signal intensity of the nonresponders was compared to the average signal intensity of each probe set and the cluster is shown in Figure 18B. This analysis provide d two classes, one with all but one of the
94 non-responsive cell lines (A673) and the other with all but one responsive cell line (U 2 OS), suggesting that the cell lines can be se parated into two distinct groups based on GEP. A B Figure 18 Hierarchical clustering of sarcoma cell lines based on GE P. Microarray analysis was performed on 12 sarcoma cell lines in triplicate. The GEP of the cell lines was determined by calculating the average intensity per triplicate sample set and the fold change for each probe set using the median in tensity for each probe set as the comparison group. The data were filtered for probe sets with 2000 intensity values, the remaining gene fold changes were transformed using Log2 and an unsupervised cluster was completed. A shows the dendogram associated with this cluster. B another hierarchical clustering analysis was performed, but the median intensity of the non-responders was used as the comparison group rather than the median of all samples. Data was further filtered for genes with fold changes that were significantly (p-value 0.05) different between responders and non-responders and had median signal intensities 2000.
95 A molecular signature distinguishes response to dasatinib as defined by induction of apoptosis In order to stringently iden tify genes that might repres ent the molecular signature that predicts response to dasatinib, name ly to select genes which have higher discrimination between the lists previously identified by the unsupervised analysis, we employed two different approaches of an alysis. Our first approach was a SAM (statistical analysis of microa rrays) analysis of the respons ive versus the non-responsive cell lines as two groups. To prevent ex tracting a bone sarcoma specific signature, because those are the only cell lines that res ponded to dasatinib, we filtered out the bone sarcoma specific overlapping genes from the list. This provided a list of more than 1000 genes. Our second approach was to comple te the standard MAS5 analysis. Once the signal intensities were determined for each probe the ratio for the maximum to minimum value for each probe set and each group wa s determined. Ratios less than 2.0 were included for additional analysis. This was comp leted to utilize the most consistent data for further analysis. Next, the fold change of responders was calculated. To complete this, the ratio of the average of each sample set for the resp onders to the median for all samples was calculated for each probe set. Probe sets with a significant fold change (pvalue 0.05) were identified and selected to be compared to the probe sets which appeared on list generated using SAM. Ther e were 26 probe sets that appeared on both lists, representing 22 different ge nes. Theses probe sets were selected as the molecular signature that predicts response to dasatinib (Figure 19A). Probe sets for four genes: Histone H3, Fas associated factor-1 (FAF1), -catenin and -catenin were further validated using qR T-PCR because more than one probe set
96 appeared on the molecular signature for these genes. The fold change values for these four genes are shown in Table 7 and Log2 transformed fold changes are graphically depicted in Figure 19B and validated using qR T-PCR, the results of which are displayed in Table 8 and Log2 fold changes are graphed in Fi gure 20. Histone H3 and FAF1 had higher fold expression, while -catenin and -catenin had lower fold expression in the responsive cell lines. The initial analysis of the 22 gene list provides the fold change of the responders compared to the median of all the cell lines analyzed and may dilute the precise fold change between responders and non-responders for this molecular signature. To determine the exact fold change between th e two groups we also calculated the fold change using the ratio of the average res ponder signal for each cell line to the median non-responder signal intensity us ing the same probe set list. The heatmap generated for this analysis is shown in Figure 21A. This analysis pr ovides a more pronounced fold change between the two groups for the mol ecular signature. Furthermore, the fold change for Histone H3, FAF1, -catenin and -catenin are also shown in Table 9 and the Log2 transformed fold changes are graphically de picted in Figure 21B and the trends of expression are upheld.
98 B RD|18 RD A673 HT1080 SK-LMS TC-71 SK-ES-1 MG-63 U 2 OS LM-7 LM-2 SaOS-2 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2Relative Fold Change Histone H3 Fas Associated Factor -Catenin -Catenin Figure 19. M olecular signature was generated using the median value for each probe set as comparison group. A, heatmap of 26 different probe se ts representing 22 genes, the average was determined for each sample set and fold change was calculated using median of all samples for the comparison group. The pr obe sets on this list were significantly different among responders and non-responders (p-value 0.05) and appeared on the list for both analyses. B graph expressing fold changes of four genes, Histone H3, FAF1, catenin and -catenin that appeared several times on the molecular signature. The fold changes calculated for A were transformed by Log2 and the values graphed to depict accurate fold changes between samples.
99 Table 7. Fold changes for key signature genes in sarcoma cell lines compared to median intensity Cell Line Histone H3 Fas Associated Factor -Catenin -Catenin SaOS-2 1.14 0.21 1.62 0.15 0.982 0.13 0.640 0.21 LM-2 1.30 0.088 1.62 0.089 0.925 0.14 0.653 0.024 LM-7 1.46 0.11 1.49 0.35 0.865 0.10 0.478 0.072 U 2 OS 1.09 0.17 0.905 0.14 0.909 0.069 0.569 0.089 MG-63 0.919 0.11 0.947 0.12 1.71 0.19 1.27 0.051 SK-ES-1 1.51 0.12 0.965 0.17 0.908 0.080 0.732 0.088 TC-71 1.33 0.39 1.00 0.12 0.684 0.30 1.00 0.0053 SK-LMS 0.530 0.03 0.823 0.057 1.25 0.20 1.77 0.25 HT1080 0.641 0.10 0.771 0.057 1.26 0.40 1.35 0.45 A673 1.01 0.13 0.678 0.037 1.52 0.13 1.48 0.27 RD 0.680 0.12 1.03 0.078 1.01 0.15 1.29 0.17 RD|18 0.690 0.23 1.03 0.25 0.913 0.12 1.32 0.44
100 RD-18 RD A673 HT1080 SK-LMS TC-71 SK-ES-1 MG-63 U2OS LM-7 LM-2 SaOS-2 -3 -2 -1 0 1 2 3 4Relative Fold Change (Log2) Histone H3 Fas Associated Factor -Catenin -Catenin Figure 20. Quantitative RT-PCR validation of Histone H3, FAF1, -catenin and catenin gene expression. Bars represent Log2 fold changes for the selected genes. Positive fold change represents up-regulated, and nega tive fold change represents down-regulated in sarcoma cells.
101 Table 8. Quantitative RT-PCR expr ession of key signature genes in sarcoma cells Cell Line Histone H3 Fas Associated Factor -Catenin -Catenin SaOS-2 0.865 0.016 1.04 0.14 1.10 0.13 1.10 0.0058 LM-2 1.45 0.010 2.13 0.12 1.08 0.19 1.14 0.0070 LM-7 3.43 0.025 4.17 0.18 1.13 0.062 1.04 0.025 U2OS 0.227 0.018 0.187 0.18 1.13 0.037 1.03 0.036 MG-63 0.500 0.0089 0.483 0.18 1.13 0.10 1.07 0.015 SK-ES-1 1.33 0.0049 0.637 0.20 1.15 0.015 1.01 0.016 TC-71 2.62 0.010 1.21 0.082 1.06 0.17 1.13 0.020 SK-LMS 2.95 0.0077 4.29 0.15 1.11 0.15 1.11 0.014 HT1080 0.160 0.0043 0.334 0.19 1.14 0.12 1.08 0.035 A673 0.812 0.0073 0.551 0.19 1.14 0.10 1.07 0.016 RD 0.337 0.020 0.559 0.23 1.17 0.013 1.01 0.054 RD|18 0.818 0.087 1.23 0.011 1.01 0.28 1.22 0.053
103 B Figure 21. Molecular signature generated using the median value for each probe set in the non-responders as comparison group. A, heatmap of 26 different probe sets representing 22 genes, the average was determined for each sample set and fold change was calculated using median of the non-re sponders as comparison group for each probe set. The probe sets on this list were significantly different among responders and nonresponders (p-value 0.05) and appeared on th e list for both analyses. B graph expressing fold changes of f Histone H3, FAF1, -catenin and -catenin. The fold changes calculated for A were transformed by Log2 and the values graphed to depict accurate fold changes between samples. RD|18 RD A673 HT1080 SK-LMS-1 TC-71 SK-ES-1 MG-63 U 2 OS SaOS-2LM-2LM-7-1.5 -1 -0.5 0 0.5 1 1.5Relative Fold Change (Log2) Histone H3 Fas associated factor -Catenin -Catenin
104 Table 9. Fold Changes for key signature genes in sarcoma cells compared to median intensity of non-responders Cell Line Histone H3 Fas Associated Factor -Catenin -Catenin SaOS-2 1.63 0.31 1.95 0.18 0.854 0.11 0.482 0.16 LM-2 1.86 0.12 1.96 0.10 0.804 0.13 0.493 0.018 LM-7 2.09 0.15 1.81 0.42 0.752 0.090 0.360 0.054 U 2 OS 1.56 0.24 1.10 0.16 0.792 0.061 0.429 0.067 MG-63 1.311 0.15 1.15 0.15 1.49 0.17 0.955 0.038 SK-ES-1 2.16 0.18 1.18 0.21 0.789 0.069 0.552 0.067 TC-71 1.91 0.56 1.20 0.14 0.594 0.26 0.756 0.0040 SK-LMS 0.756 0.037 0.993 0.069 1.08 0.18 1.33 0.19 HT1080 0.915 0.15 0.931 0.067 1.10 0.35 1.02 0.34 A673 1.44 0.18 0.820 0.046 1.32 0.12 1.12 0.21 RD 0.970 0.17 1.26 0.090 0.877 0.13 0.971 0.13 RD|18 0.984 0.32 1.25 0.30 0.793 0.10 0.997 0.33
105 Identification of two probe sets with greater fold changes may provide further insight into the prediction of response To identify possible probe sets that ma y convey a greater fold change between response groups, we altered our search criter ia by increasing the ratio of the maximum to minimum ratio value for each probe set to 10 and searched for probe sets with 5.0 or 5.0 fold change between groups. Of the genes provided from this search two probe sets with interesting potential were followed-up; Ephrin-A1, the ligand of a known target of dasatinib, EphA2 receptor, and Dapper, an antagonist of catenin, both of which have recently been shown to play unique roles in the malignant phenotype. When clustered based on the expression of these two genes, the heatmap generated from this search shows that the responders and non-responders co mpletely cluster apart on separate arms when the fold change is calculated using the median probe set intensity as comparison or the median non-responder probe set intensity (Figures 22A and 23A). In addition, a graphic depiction of the Log2 transformed fold changes for these two genes shows greater expression of both Ephrin-A1 and Dapper in th e responsive cell lines as compared to the non-responsive cell lines using either median comparison group (Figures 22B and 23B). The values of relative fold change are also shown on Tables 10 and 11 for both analyses. Quantitative RT-PCR was completed on the cell lines for ephrin-A1 and dapper to validate the microarray results. Figure 24 s hows the graphic expression of these two genes compared to the median expression of all samples, and Table 11 shows the absolute values of the average expression for ephrin-A1 and dapper in the cell lines generated from the analysis. The validation concludes that the microarrays accurately revealed an increased expression of Ephrin -A1 and Dapper in the responsive cell lines.
106 RD|18 RD A673 HT1080 SK-LMS TC-71 SK-ES-1 MG-63 U 2 OS LM-7 LM-2 SaOS-2-5 -4 -3 -2 -1 0 1 2 3Relative Fold Change (Log2) Ephrin-A1 Dappe r A B Figure 22. Expression of Ephrin-A1 and Dapper in sarcoma cell lines using the median of all samples as comparison group. A heatmap of Ephrin-A1 and Dapper fold changes in 12 sarcoma cell lines. The average was determined for each sample set and the fold change was calculated by using the median of all samples as the comparison group for each probe set. B graph of Ephrin-A1 and Dapper fold change between sample sets. The fold changes were calculated by transfor ming the fold changes calculated in A by Log2. A
107 RD|18 RD A673 HT1080 SK-LMS TC-71 SK-ES-1 MG-63 U 2 OS LM-7 LM-2 SaOS-2-4 -3 -2 -1 0 1 2 3 4 5Relative Fold Change (Log2) Ephrin-A1 Da pp e r B Figure 23. Expression of Ephrin-A1 and Dapper in sarcoma cell lines using the median of the non-responders as comparison group. A heatmap of Ephrin-A1 and Dapper fold changes in 12 sarcoma cell lines. The aver age was determined for each sample set and the fold change was calculated by using the median of the non-responders as the comparison group for each probe set. B graph of Ephrin-A1 and Dapper fold change between sample sets. The fold changes were calculated by transforming the fold changes calculated in A by Log2.
108 Table 10. Fold changes for ephrin-a1 a nd dapper in sarcoma cell lines compared to median intensity Cell Line Ephrin-A1 Dapper SaOS-2 3.71 0.66 5.39 1.5 LM-2 3.60 0.93 3.54 0.44 LM-7 2.94 0.87 5.63 1.1 U 2 OS 1.13 0.82 3.02 0.86 MG-63 0.354 0.13 0.252 0.25 SK-ES-1 3.26 0.78 1.06 0.21 TC-71 1.27 0.78 0.914 0.23 SK-LMS 0.136 0.029 0.374 0.047 HT1080 0.159 0.17 0.328 0.092 A673 0.055 0.17 0.954 0.18 RD 0.237 0.31 0.440 0.30 RD|18 0.388 0.078 0.423 0.16
109 Table 11. Microarray fold changes for ephrin -a1 and dapper in sarcoma cell lines compared to median intensity of non-responders Cell Line Ephrin-A1 Dapper SaOS-2 10.3 1.8 14.9 4.3 LM-2 10.0 2.6 9.77 1.2 LM-7 8.19 2.4 15.5 3.0 U 2 OS 3.16 2.3 8.33 2.4 MG-63 1.38 0.35 0.921 0.69 SK-ES-1 9.10 2.1 2.92 0.13 TC-71 3.55 2.2 2.52 0.62 SK-LMS 4.67 0.081 2.12 0.13 HT0180 0.983 0.47 0.790 0.25 A673 0.705 0.48 3.05 0.49 RD 1.52 0.88 1.02 0.83 RD|18 1.10 0.22 0.772 0.45
110 Figure 24. Quantitative RT-PCR validations of Ephrin-A1 and Dapper genes. Bars represent fold changes for the selected genes. Positive fold change represents upregulated, and negative fold change repr esents down-regulated in sarcoma cells. SaOS-2 LM-2 LM-7 U 2 OS MG-63 SK-ES-1 TC-71 SK-LMS HT1080 A673 RD RD|18-6.000 -5.000 -4.000 -3.000 -2.000 -1.000 0.000 1.000 2.000 3.000 4.000 5.000Relative Fold Change (Log2) Ephrin A1 Da pp e r
111 Table 12. Quantitative RT-PCR values for ephrin -a1 and dapper in sarcoma cell lines Cell Line Ephrin-A1 Dapper SaOS-2 1.74 0.11 1.64 0.16 LM-2 4.44 0.28 2.90 0.19 LM-7 5.58 0.40 11.7 0.31 U 2 OS 1.30 0.16 1.76 0.11 MG-63 0.197 0.028 0.228 0.017 SK-ES-1 3.21 0.12 0.603 0.057 TC-71 2.25 0.13 1.68 0.14 SK-LMS 0.474 0.018 1.05 0.022 HT1080 0.212 0.028 0.156 0.011 A673 0.603 0.075 0.954 0.11 RD 0.406 0.077 0.061 0.019 RD|18 0.700 0.12 0.189 0.017
112 Testing the molecular signature with cell lin es of unknown response reveals that the molecular signature can accurately predict response to dasatinib in cell lines To test the molecular signature of response, two new cell lines of unknown response status were acquired. Microarray and molecular analysis were independently performed and the response status generated from the two analyses were compared. To determine response status of the new cell lin es the fold change of each probe set was calculated by taking the ratio the average inte nsity for each probe set in the new cell lines to the median intensity of the original dataset and compared the fold change to the molecular signature extracted from the original group of cell lines. Upon further examination of the gene expression profile of the two cell lines and comparison to the molecular signature, HOS was characterized as a responsive cell line and SW1353 was characterized as a non-responsive cell line. Each cell line also clustered with its respective group as is observed in Figure 25A. In addition, the expression of Histone H3, FAF1 and -catenin were significantly different in each of the cell lines. Histone H3 and FAF1 had greater expression, while catenin had a decreased expression in the HOS cell line (Figure 25B, Table 13). The expression of these three genes correlates with the expected expression patterns according to the predictive molecular signature. However, when q-RT-PCR was performed to validate the expression of th ese genes in the HOS and SW1353 cell lines, Histone H3 and FAF1 were significantly upr egulated in the HOS cell line compared to the SW1353 cell line. The expression of -catenin was not significantly different, although it was upregulated in SW1353 as co mpared to HOS. While, the expression
113 catenin was also sign ificantly different in the two cell lines, the opposite expression pattern was observed than expected. To validate the characterization of the cel l lines dose response to dasatinib was completed in both cell lines at 6, 24 and 72 h. Dasatinib inhibited Src activation, signaling and cellular migration in both cell lines (Figure 26A-E). However, HOS was the more sensitive of the two cell lines, in th at a lower dose of dasatinib was required to inhibit Src activation. Furthermore, dasatinib also induced apoptosis in the HOS cell line and not the SW1353 cell line (Figure 27A and B). These results confirm the characterization of the cell li nes by microarray analysis. T hus, these data successfully validated the use of our signature in the abil ity to predict the response of human sarcoma cell lines to dasatinib. On the other hand, Ephrin-A1 and Dapper expression did not correlate with the expected expression pattern in accordance w ith response. HOS was observed as having significantly lower expression of both genes as compared to SW1353 (Figure 28A, C and 29A, C). Furthermore, when clustered base d on the expression of these two genes, HOS and SW1353 clustered with the opposite res ponse groups (Figures 28B and 28B). The values of Ephrin-A1 and Dapper can be furt her assessed in Tables 15 and 16. Several possibilities could explain the lack of validation of Ephrin -A1 and Dapper as predictive genes and will be further articulated in the discussion.
115 B SW1353 HOS -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1Relative Fold Change (Log2) Histone H3 Fas Associated Factor -Catenin -Catenin* **
116 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 HOSSW1353Relative Fold Change (Log 2) Histone H3 Fas Associated Factor -Catenin -Catenin C Figure 25. Molecular signature that predicts resp onse to dasatinib generated using the median value for each probe set as comparison group. A, the same molecular signature was used to create this heatmap as for figur e 20A. The only difference in this heatmap is the presence of two new cell lines, HOS a nd SW1353. The average was determined for each sample set and fold change was calculated using median of all samples for the comparison group to calculate the fold changes utilized for this heatmap. Samples were clustered based on the GEP of the 26 probe set list. B graph expressing fold changes of Histone H3, FAF1, -catenin and -catenin in HOS and SW1353 cells. The fold changes calculated for A were transformed by Log2 and the values graphed to depict accurate fold changes between samples. C graph expressing the relativ e qRT-PCR fold changes of Histone H3, FAF1, -catenin and -catenin in HOS and SW1353 cells using the median expression for comparison. (* p-value 0.01, ** p-value 0.0001) ** ** **
117 Table 13. Fold changes key signature genes in test cells compared to median intensity Cell Line Histone H3 Fas Associated Factor -Catenin -Catenin HOS 1.16 0.14 0.907 0.072 0.93 0.12 1.04 0.12 SW1353 1.05 0.17 0.508 0.040 0.88 0.086 1.20 0.063 Table 14. Relative qRT-PCR fold changes ke y signature genes in test cells compared to median intensity Cell Line Histone H3 Fas Associated Factor -Catenin -Catenin HOS 3.05 0.035 2.46 0.11 2.51 0.12 1.21 0.045 SW1353 1.91 0.017 1.21 0.15 1.77 0.085 1.62 0.0023
118 p-Src (Y419) Src -Actin 301003001000 DMSOnM p-Src (Y419) Src -Actin 301003001000 DMSOnM A B 301003001000 DMSOnM p-FAK (Y576/577) FAK p-p130CASp130CAS-Actin p-FAK (Y925) p-FAK (Y397) 301003001000 DMSOnM p-FAK (Y576/577) FAK p-p130CASp130CAS-Actin p-FAK (Y925) p-FAK (Y397) 301003001000 DMSOnM p-FAK (Y576/577) FAK p-p130CASp130CAS-Actin p-FAK (Y925) p-FAK (Y397) 301003001000 DMSOnM p-FAK (Y576/577) FAK p-p130CASp130CAS-Actin p-FAK (Y925) p-FAK (Y397) p-Src (Y419) Src -Actin 301003001000 DMSOnM p-Src (Y419) Src -Actin 301003001000 DMSOnM C D
119 E Figure 26. Dasatinib inhibits Src activation and signaling in HOS and SW1353 cells. A and C HOS and SW1353 cells were treated with dasatinib in a dos e-dependent manner for 6 h. Cell-free extracts were immunoblotted with antibodies specif ic to p-Src (Y419) and total Src. B and D dasatinib specifically blocks tyrosyl phosphorylation of FAK (Y576/577, Y925) and 130CAS (Y410), but not FAK Y397. HOS and SW1353 cells were treated with dasatinib for 6 h in a dose-de pendent manner. Cell-free extracts were immunoblotted with antibodies specific to p-FAK (Y397, Y576/577, Y925), total FAK, p130CAS and p-p130CAS (Y410). -actin was blotted for as a loading control in all experiments. Dasatinib inhib its cell motility and invasion. E wound healing assays were performed on HOS and SW1353 to determine th e effects of dasatinib on inhibiting cell migration as completed previously. Cell mi gration was visualized at 10x magnification by light microscopy and photographed with a di gital camera. The width of voided area versus dasatinib dose concentration was gra phed to express the degree of inhibition of cell migration. The number of pi xels within the denuded area were the units used to demonstrate inhibition of cell migrat ion induced by dasa tinib (*p<0.001, n=3). -20.0 0.0 20.0 40.0 60.0 80.0 100.0 120.0 T00301003001000 dasatinib, nMMigration (% Control ) HOS SW1353 *
120 Figure 27. Dasatinib induces apoptosis in HOS but not SW1353 sarcoma cells. A dasatinib induces apoptosis in a dose-depe ndent manner in HOS cell only. Cells were treated with dasatinib for 72 h with es calating doses. Cell-free extracts were immunoblotted with antibodies sp ecific to XIAP and PARP. -actin was blotted for as a loading control in all experiments. PARP Cleavage PARP -Actin 301003001000 DMSOnM XIAP A B PARP Cleavage PARP -Actin 301003001000 DMSOnM XIAP PARP Cleavage PARP -Actin 301003001000 DMSOnM XIAP
121 A B
122 C Figure 28. Expression of Ephrin-A1 and Dapper in HOS and SW1353 cells using the median of all samples as comparison group. A heatmap of Ephrin-A1 and Dapper fold changes in HOS and SW1353 cells. The average was determined for each sample set and the fold change was calculated by using the median of all samples as the comparison group for each probe set. B integration of HOS and SW1353 into the original heatmap of Ephrin-A1 and Dapper expression. C, graph of Ephrin-A1 and Dapper fold changes in HOS and SW1353. The fold changes were calc ulated by transforming the fold changes calculated in A by Log2. (* p-value 0.01) SW1353 HOS -2 -1 0 1 2 3 4Relative Fold Change (Log2) Ephrin-A1 Dappe r
123 A B
124 C Figure 29. Expression of Ephrin-A1 and Dapper in HOS and SW1353 cells using the median of the non-responders comparison group. A heatmap of Ephrin-A1 and Dapper fold changes in HOS and SW1353 cells. The average was determined for each sample set and the fold change was calculated by us ing the median of the non-responders as the comparison group for each probe set. B integration of HO S and SW1353 into the original heatmap of Ephrin-A1 and Dapper expression. C, graph of Ephrin-A1 and Dapper fold changes in HOS and SW1353. The fold changes were calculated by transforming the fold changes calculated in A by Log2. (* p-value 0.001) SW1353 HOS -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4Relative Fold Change (Log2) Ephrin-A1 Da pp er*
125 Table 15. Fold changes for ephrin-a1 and dapper in test cells compared to median intensity Cell Line Ephrin A1 Dapper HOS 0.526 0.10 1.03 0.23 SW1353 0.926 0.25 10.1 1.2 Table 16. Microarray fold changes for e phrin-a1 and dapper in test cells compared to median intensity of non-responders Cell Line Ephrin A1 Dapper HOS 0.534 0.30 1.48 0.31 SW1353 1.33 0.38 4.80 0.17
126 Discussion Microarray analysis of ge ne expression profiles in sarcomas is a valuable technique that provides a comprehensive survey of activated molecular pathways and has been shown to successfully identify potential therapeutic targets. In this aim, we attempted to identify a molecular signatur e that will predict re sponse to dasatinib by induction of apoptosis in sarcoma cell lines. As a model, we initially compared the GEP of 12 sarcoma cell lines characterized by response to dasatinib. The unsupervised analysis of GEP data clearly separated the cell lines by tumor type (Figure 19A). Using statistical filtering to identify genes signi ficantly different between the responders and non-responders, the GEP data analyzed created two distinct classes, with all but one cell line from each group clustering within their corresponding classes (Figure 19B). Then, using two robust statistical filtering pro cedures, a 22 gene signature with a high discrimination power among the resistant and non-resistant groups wa s identified (Figure 20A and 21A). Interestingly, many of the gene s present in the signa ture are found within close proximity on chromosome one, particul arly Histone H3 and FAF1 are located on chromosome 1q21, which is an area that has been shown to be amplified in OSA. Of the 22 genes in this signature four genes were selected for further analysis because multiple probe sets for these genes appeared on the si gnature with consistent fold changes between classes; Histone H3, FAF1, -catenin and -catenin. Histone H3 and FAF1 were upregulated, while -catenin and -catenin were downregulated in the responsive group as a whole (Figure 19B and 21B; Tables 7 and 9) and these findings were further validated by qRT-PCR (Table 8, Figure 20).
127 Two other probe sets were identified as potential predictors of response using a variation of our initia l analysis; Ephrin-A1 and Dapper. Ephrin-A1, a known target of dasatinib, and Dapper, an antagonist of catenin s, were both upregulated in the responsive cell lines (Figures 22A and 23A) and this upregulation was validated by qRT-PCR (Figure 24, Table 12). Upon identification of an RNA signature, our next goal was to test whether this signature can successfully charact erize cell lines into their respective groups based on response to dasatinib. To test th e signature we acquired two new cell lines; HOS, an OSA, and SW1353, a CS cell line of unknown response to dasatinib. Microarray analysis was performed i ndependently as the molecular biology assays were performed to determine the respon se status of the two cell lines. The results of both analyses were compared once the resu lts for each were determined. The results of both analyses agreed; the microarray an alysis grouped the HOS with the responders and SW1353 with the non-responders based on th e expression of the molecular signature (Figure 25A). Furthermore, Histone H3, FAF1 and -catenin expression were significantly different in the two cell lines and corresponded w ith the expression patterns of their respective classes. The molecula r biology analysis determined that HOS responded to dasatinib by induction of a poptosis, while SW1353 merely responded by inhibition of Src signaling and migrati on (Figure 26A-E and Figure 27A-B). Alternatively, prediction of response ba sed on the expression of Ephrin-A1 and Dapper did not successfully classify the cell li nes into the correct classes (Figures 28 and 29; Table 15 and 16). In fact, the cell lines were characteri zed into the co mplete opposite groups based on the expression of these two ge nes. One possible explanation for this finding could be that only two cell lines were used to test the predictive characterization
128 of these genes, perhaps if more cell lines we re utilized for the validation of Ephrin-A1 and Dapper their ability to be used in pr edicting response would be more promising. Another explanation may be that these tw o genes alone, are not sufficient enough to predict response, but when applied together within the context of the signature may possess more significance. Nevertheless, the class prediction analysis performed using these 26 probe sets as a predictor gene list for two cell lines of unknown response successfully classified the cell lines with their respectiv e response groups. These findi ngs lend evidence to suggest that this molecular signature can to predict re sponse to dasatinib in cell lines and aid in identifying cell lines that required Src activity for survival. To further validate this signature, our next aim was to establish wh ether the signature was present in primary human sarcomas. Should the signature be expr essed in sarcomas, futu re clinical trials with dasatinib could include a component to test the validity of the molecular signature in predicting response in sarcoma patients.
129 Validation of Gene Expression Pr ofile in Primary Human Sarcomas The primary objective of this aim is to determine whether the molecular signature that predicts response to dasa tinib by induction of apoptosis in cell lines is expressed in human sarcomas. Validating the expression of the signature predic ative of response to Src inhibition may lay the founda tion for designing future clini cal trials that will help tailor design treatment options for sarcoma pa tients. In addition to designing trials, the results gained from this research will be usef ul for the impending clinical trial to evaluate dasatinib as a potential treatme nt option for sarcomas. If present, the molecular signature could be tested using samples gathered in th e future trial and validated as a potential predicative signature. In addition, the resu lts of this study will contribute to our understanding of the response to treatment of tumors where histological evaluation and ancillary studies have proven to be insufficient predictors. In the present study, we applie d microarray gene expression profiling to establish the presence of a candidate molecular signature from the cell lines that predicts response to dasatinib, in human sarcomas. Our goal was to obtain untreated, primary human sarcoma specimens collected under a general tissue banking consent protocol. The tumor histologies utilized fo r this aim corresponded to the tumo r types represented in the cell line component of the study, including OSA, LMS, FBS and RD. Unfortunately, there was no RNA available for the EWS that were collected under this protocol. To compensate for the lack of samples, sarcom as representing other hi stologies were also
130 utilized for these investigations. GEPs were established for each specimen and compared to the molecular signature extracte d from the cell lines. The comparison was completed to determine whether components of the cell line derived molecular signature were present in human sarcomas. To our astonishment, 10 or the 22 sarcomas used for this study expressed significant components of the molecular signature from the cell lines. The expression of the signature also provided a theoretical assi gnment of potential res ponse to dasatinib. Furthermore, the expression of the molecula r signature was significant enough to cluster many of the tumors based on their GEPs into their theoretical response groups. These data lend evidence to suggest that at leas t components of the cell line signature are expressed in sarcomas. These results are encouraging and support the implementation of further investigation into the expression and predictability of this signature to predict response in primary sarcomas.
131 Results Unsupervised clustering identified diversity among human sarcoma specimens Using an unsupervised hierarchical clus tering approach, we tried to identify natural subclasses of tumor specimens as dete rmined by gene expression profiles. We performed the unsupervised clustering on an unfiltered probe list and have shown that sarcomas are not easily groupe d into tumor subclasses base d on gene expression alone (Figure 30A). In fact, the only subgroup th at clustered togeth er was the two OSA specimens. The remaining sarcomas clustered with apparently no preference for tumor subgroup. Identification of the cell line mo lecular signature in human sarcomas Our next goal was to determine whether components of the 26 gene molecular signature are present in human sarcoma specimens. To complete this we compared the intensity of each probe set for the tumors calculated fold change by comparing the intensity for each probe set to the median inte nsity of the dataset and compared the fold change to the molecular signature extracted from the cell lines. Of the 22 tumors analyzed, 10 were characterized as potentia l responders and 12 we re characterized as potential non-responders based on the fold changes of the genes that comprise the molecular signature that predicts response in cell lines (Table 16) A breakdown of the percentage responders and non-responders by tumor subgroups is illustrated in Figure 30B. Figure 30A identifies th e potential responsiv e tumors with red stars above the sample label. Clearly the tumors do not clus ter by potential respons e to dasatinib when unsupervised and unfiltered.
132 A ** * * * * ** * * *
133 B Figure 30. Hierarchical clustering of primar y sarcoma specimens. The GEP of the tumors was determined by calculating fold change for each probe set, using the average intensity of each probe set across samples as the comparison group. An unsupervised cluster was performed to determine i nherent similarities between tumors, A Sarcomas categorized as potential res ponders are indicated with B, breakdown of tumors categorized as potential responders and non -responders by molecular signature according to sarcoma type. Predicted Tumor Response Responders 45% NonResponders 55% Responders AS 0% CS 10% LMS 20% LPS 10% MFH 20% OSA 20% RD 0% SCS 10% FBS 10% Non-Responders AS 8% CS 17% FBS 17% LMS 34% LPS 0% RD 8% SCS 8% OSA 0% MFH 8%
134 Table 17. Classification of sarcomas using predictive signature from cell lines Tumor Responder Non-Responder AS T98 R CS T2355 NR CS T2744 NR CS T3109 R FBS T3946 NR FBS T4436 NR FBS T6390 NR LMS T174 R LMS T191 R LMS T294 R LMS T374 NR LMS T466 R LMS T3863 NR LPS T2500 R MFH T1913 NR MFH T5087 NR MFS T6251 NR OSA T1898 R OSA T6357 R RD T395 NR SCS T548 NR SCS T7431 R Total 10 12
135 A Responders Non-Responders Responders Non-Responders
136 Figure 31. Heatmap of molecular signature expr ession in sarcomas. Fold changes for the molecular signature were calculated by us ing the average intensity for each probe set as the comparison group. A heatmap of unclustered tumors arranged by response category. B heatmap of tumors clustered based on the expression of the molecular signature used to define theo retical response status. Dendogr am at top of heatmap shows the relationship of tumors based of the GE P of the molecular signature. Sarcomas categorized as potential res ponders are indicated with a B * * * * * * * * * *
137 Molecular signature that predicts response to das atinib in cell lines can be used to group tumors by potential response An unclustered heatmap of the molecula r signature expression in the tumors arranged by potential response does not pres ent a clear separation by expression between the two groups (Figure 31A). Furthermor e, when clustered based on the GEP of the molecular signature, the tumors do not comple tely cluster into their respective groups (Figure 31B). However, there is some promis e to a trend of consistency in the clustering, in that some of the tumors did cluster in to their respective groups ; 6 out of 10 for the responders clustered together and 8 out of 12 of the non-resp onders. There are several possibilities to explain this inconsistency and will be a ddressed in the discussion. A heatmap was also prepared for the tumo r expression of the four genes validated from the predictive molecular signature, Histone H3, FAF1 -catenin and -catenin, When clustered, again the tumors classified as potential responde rs and non-responders did not cluster in two separate groups. Although, 70% of responsive tumors clustered together and 50% of the non-responsive tumors clustered together (Figure 31A). The expression patterns of Histone H3, FAF1 -catenin and -catenin are also inconsistent when are pulled from the GEP and compared to the cell line signa ture (Figure 32A-B, Table 17). Again, there is a trend of recu rrent expression patterns between responsive verses the non-responsive groups, although it is not uniform across all samples.
138 A * * * * * * * * * *
139 B T548 T395 T6251 T5087 T1913 T3863 T374 T6390 T4436 T3946 T2744 T2355 T7431 T6357 T1898 T2500 T466 T294 T191 T174 T3109 T98-1.5 -1 -0.5 0 0.5 1 1.5Relative Fold Change (Log2) Histone H3 Fas Associated Factor -Catenin -Catenin Figure 32. Expression of Histone H3, FAF1, -Catenin and -Catenin in sarcomas. A, heatmap of Histone H3, FAF1, -catenin and -catenin extracted from molecular signature and clustered based on GEP of these f our genes. B, graphic depiction of fold changes associated with Histone H3, FAF1, -catenin and -catenin in tumors. Sarcomas categorized as potential res ponders are indicated with a
140 Table 18. Fold changes for key signature genes in sarcomas Tumor Histone H3Fas Associated Factor Catenin Catenin AS T98 1.22 1.17 0.969 1.17 CS T3109 0.919 0.753 0.875 0.834 LMS T174 1.405 1.15 0.861 0.718 LMS T191 1.24 0.721 1.49 1.59 LMS T294 1.18 0.804 1.09 1.56 LMS T466 1.57 1.31 0.996 1.43 LPS T2500 0.901 0.818 1.06 1.25 OSA T1898 1.23 0.712 1.03 0.965 OSA T6357 1.42 1.24 1.13 0.781 SCS T7431 0.769 0.798 1.84 1.43 CS T2355 1.12 0.661 1.18 0.793 CS T2744 0.611 1.105 0.973 0.902 FBS T3946 0.881 0.940 0.964 1.04 FBS T4436 0.947 1.07 1.40 1.20 FBS T6390 0.859 0.955 0.856 0.693 LMS T374 1.23 1.73 1.83 1.66 LMS T3863 1.04 1.90 0.664 0.898 MFH T1913 0.732 1.87 0.983 0.894 MFH T5087 0.832 0.716 0.963 1.06 MFS T6251 0.611 1.13 0.674 0.686 RD T395 0.789 1.25 1.11 0.901 SCS T548 1.14 1.24 0.740 1.01
141 Analysis of two probe sets with greater fold changes may provide further insight into the prediction of response in tumors The expression of Ephrin-A1 and Dapper were also analyzed in the tumors specimens. A heatmap with hierarchical cl ustering produced three classes of tumors based on Ephrin-A1 and Dapper gene expr ession. While none of the three groups contained all of one response group, the clus tering is promising. One group contained one responder and three non-re sponders, the second group ha d seven responders and two non-responders, while the thir d group included two responders and seven non-responders (Figure 33A). The fold change and e xpression patterns of Eprhin-A1 and Dapper correlate more with the cell lines, in that more of the tumors that express the potentially responsive signature had greater fold change s for Ephrin-A1 and Dapper as compared to the potentially non-responsive tumors as a whole (Figure 33B, Table 18).
142 A * * * * * * * * * *
143 B Figure 33. Expression of Ephrin-A1 and Dapper in sarcomas. The fold changes of Ephrin-A1 and Dapper were calcu lated using the median intensity of all samples as the comparison group. A Hierarchical clustering was performed based on the GEP of Ephrin-A1 and Dapper in sarcomas. B graphic depiction of Ephrin-A1 and Dapper fold changes between primary sarcoma specimens as compared to the median intensity for each probe set. Sarcomas categorized as potential responders ar e indicated with a T548 T395 T6251 T5087 T1913 T3863 T374 T6390 T4436 T3946 T2744 T2355 T7431 T6357 T1898 T2500 T466 T294 T191 T174 T3109 T98-2 -1 0 1 2 3 4 5Relative Fold Change (Log 2 Ephrin-A1 Da pp er* * * * * T548 T395 T6251 T5087 T1913 T3863 T374 T6390 T4436 T3946 T2744 T2355 T7431 T6357 T1898 T2500 T466 T294 T191 T174 T3109 T98-2 -1 0 1 2 3 4 5Relative Fold Change (Log 2 Ephrin-A1 Da pp er* * * * *
144 Table 19. Fold changes for ephrin-a1 and dapper in sarcomas Tumor Ephrin-A1 Dapper AS T98 1.48 2.68 CS T3109 2.75 0.953 LMS T174 1.25 1.05 LMS T191 2.67 1.07 LMS T294 0.848 0.469 LMS T466 0.807 3.53 LPS T2500 2.05 3.14 OSA T1898 2.31 3.02 OSA T6357 2.47 0.943 SCS T7431 0.973 0.178 CS T2355 6.46 6.56 CS T2744 0.316 0.666 FBS T3946 3.76 0.707 FBS T4436 0.849 1.05 FBS T6390 1.81 0.447 LMS T374 0.580 1.76 LMS T3863 0.275 0.391 MFH T1913 0.952 0.341 MFH T5087 0.389 1.49 MFS T6251 0.966 1.09 RD T395 1.03 0.627 SCS T548 0.804 0.208
145 Discussion Traditionally, sarcoma therapy has been selected on the basis of non-molecular considerations such as tumor t ype, grade and stage. As a re sult, treatments for sarcomas have not developed at the same pace as t hose for other more common tumors, and today we still accept with fatalism that a frac tion of sarcoma patients will respond to a particular therapy, whereas others will not. We also accept that we have little control over which outcome will prevail in a given case. Ne w approaches are desperately required to tailor treatments to the indi vidual patient, select from curre nt therapeutic possibilities, predict therapeutic response and develop novel targeted therapeutic modalities. Previous studies conducted in our laborat ory have demonstrated a GEP in sarcoma cell lines that is unique to cell lines, which undergo apoptosis when treated with dasatinib at low nanomolar doses. In addition, we have also identified a molecular signature that successfully predicts response to dasatinib in sarcoma cell line s. In this third aim, we further investigated the clinical significance of thes e findings by demonstrating the expression of the molecular signature in patient tissue samples from 22 diverse sarcoma specimens. Intriguingly, this signature theoreti cally predicted 10 out of 22 tumors to be responders (Table 16). The 22 gene molecula r signature was not able to successfully cluster all of the tumors according to poten tial response, however there is promise because many of the tumors classified with their respective response groups. The fact that a signature, which was derived from ce ll lines, can be found in the tumors lends significant evidence for further research into the possibility of this si gnature to serve as a predictor of response in sarcomas.
146 While the focus of this dissertation ha s been placed on the expression of a molecular signature that predicts response to dasatinib, the reverse could also be tested. A signature that predicts non-response has equally been generated and proven in cell lines. Furthermore, it is expressed in primar y sarcomas as well. The flipside to these studies could provide just as much help in designing future studies and identifying other potential therapeutic targets. Either way, the molecular si gnature of the responders or non-responders can be used to theoretically pr edict tumors that require Src activity for survival and/or metastasis. There are several theories that could e xplain the discrepancies between the GEP of the cell lines and tumors and why there is not a complete correlation with potential response and GEP. One possibility to expl ain this discrepancy includes the immense differences between gene and protein expression that are inherent when comparing cell lines and tumors. Cell lines are grown i ndependent of a microenvironment and lack interaction with othe r biological components. Tumo rs interact with a complex microenvironment and are exposed to a consid erable amount of stimuli on an infinite level. These interactions alone (or the lack of) may account for the discrepancies between the GEP. In addition, cell lines ma y lack the expression of metabolic enzymes present in tumors. Tumors have been shown to become resistant to therapy with time, as in the case with gleevec and GISTs. Cell lines may lack the ability to increase the metabolism of a drug of choice because of the inability of the cells to adapt to their environment as is the case with some tumors. Moreover, it is important to account for the ability of a tumor to become resistant to a therapy as a consequence of altering the expression of genes encoding metabolizing proteins even when a signature is expressed.
147 The purity of the tumors specimens utilized for the microarrays are also of some concern. While, we tried to utilize samples that are as pure as possible, sarcomas by nature are heterogeneous both within the tu mor sample and among patients. Sarcomas can be comprised of a variet y of cellular types including st romal cells, blood vessels and other undifferentiated cells. Th e lack of homogeneity of cell type can dilute the signature in the tumor specimens. There have been advancements in technology which aim at improving the purity of patient specimens, su ch as laser capture microdissection. This would be a wonderful technique to utilize, however it just wa s not feasible for this study. The heterogeneity within tumors between sarc oma patients is also a possible explanation for the molecular signature discrepancies. While, there have been many consistent mutations and translocations identified in sarcoma specimens, there are still a considerable amount of diversity among sarcom as, even within the same tumor type. Plus, it is very difficult to compare tumors among different patients. The diversity among tumor GEP is as great as the diversity am ong people in general. However, when a molecular signature of commonality identified in cell lines can be extracted from a tumor set as diverse as ours, there is great promise and encouraging ev idence to suggest that this signature may be upheld in sarcomas.
148 Conclusions Three independently synthesized Src kinase inhibitors were evaluated in these studies. While the three compounds were each designed to target th e catalytic domain of Src, the data generated usi ng dasatinib have presented th e most convincing preclinical data to complete further studies with this compound. Dasatinib inhibits Src mediated cellular migration and invasion in all of the Src activated cell lines examined in these studies, however only a subset of bone derive d sarcoma cell lines underwent apoptosis in response to dasatinib. These observations sugg est that Src associates and interacts with different molecular pathways in sarcoma cell lines. In the bone sarcoma cell lines, Src mediates an induction of a FAK-Cas-migrati on/invasion/survival pathways, while in the STS cell lines Src initiates FAK-Cas-migrati on/invasion only pathways. These findings lend evidence to suggest that Src interacts with diverse signaling molecules which varies with sarcoma type. The dasatinib data have shown cell lines that respond to dasatinib by induction of apoptosis can be identified using a 22 gene molecular signature. Eleven of 12 human sarcoma cell lines express constitutively activ ated Src kinase and re spond to dasatinib by inhibition of Src activation and signaling as measured by Western blot analysis and migration and invasion studies. Of these cell li nes, six, which were all derived from bone sarcomas, were identified as responders to da satinib as defined by induction of apoptosis.
149 Moreover, blockade of c-Src expression in this subset of sarcoma cells by silencing RNA induces apoptosis, consistent with the critic al role of Src-mediated sarcoma cell survival. Microarray analysis and hierarchical clus tering allowed us to identify a 26 probe set list that successfully predicts response to dasatinib. To test the reliability of the molecular signature two cells lines with unknown response to dasatinib were used. Independent analysis of both the GEP and the molecular response to da satinib of the cells were compared. The molecular signature that predicts response successfully characterized the cell lines into their respective groups one cell line, HOS, was a responder and the other, SW1353 was a non-re sponder. Here, we show a molecular signature predicting response to dasatinib by induction of ap optosis in cell lines. Furthermore, we performed microarra y analysis on 22 human sarcomas of varying sarcoma histology to determine whether components of the molecular signature, Ephrin-A1 or Dapper expression could be us ed to characterize the tumors based on potential response to dasatini b. A list of possible respons ive and non-responsive tumors was generated based on their GEP. When cl ustered based on the expression of the 26 probe set molecular signature or Ephrin -A1 and Dapper, the clustered provided encouraging results to suggest that the si gnature has promise to predict response in tumors. While not all of the sarcomas cluste red with their respectiv e groups (potentially, responsive or non-responsive), many of the tu mors did cluster with their groups and the predictive trends of Ephrin-A 1, Dapper and the molecular si gnature are up-held. These data suggest that our signature could become useful for the iden tification of patients eligible for new therapeutic options.
150 Three of the four genes from the mol ecular signature further validated by qRTPCR analysis were upheld as reliable pred ictors of response; Histone H3, FAF1 and catenin. However, Ephrin-A1 and Dapper were less reliable as predictive markers as reversed expression pattern was evident in these cell lines. The six genes that were validated by qRT-PCR; Histone H3, FAF1, -catenin, -catenin, Ephrin-A1 and Dapper, are all thought to play unique roles in cancer. The functi ons these genes play in the malignant phenotype lend evidence to further investigate the role of these genes in sarcomas. Histone H3 and FAF1 were upregulated in the cell lines that re spond to dasatinib. Many types of cancer are associ ated with translocations or mutations in chromatinmodifying enzymes and regulat ory proteins. Histones are in tegral components of the machinery responsible for regulating gene tr anscription. Phosphoryl ation of Histone H3 plays important regulatory signaling roles in the processes of ch romosome condensation during mitosis and transcription following extern al stimulation of gene expression growth factors or stress. Histone H3 is phosphoryl ated in serine 10 and 28. Fyn, one of many kinases that phosphorylate Ser 10. The phosphorylation of Ser10 on Histone H3 is critical for EGF-induced neoplastic cell transformation. This eviden ce suggests that Histone H3 may serve as a critical target for cancer ther apy (278). Interestingl y, the overexpression of Histone H3 was upheld in many of the tumors categorized as potential responders. Histone H3 overexpression in the responsive group may not be easil y explained in the context of these studies. However these studies have identified Histone H3 as an overexpressed gene which provides evidence to further investigate its activity in sarcomas and possibly evaluated it as a therapeutic target.
151 While FAF1 overexpression was only upheld in the cell lines, this may be caused by the inherent difference in phenotype between cell lines and tumors. FAF1 is a proapoptotic factor that is invol ved with Fas-mediated caspase 8 cleavage and induction of apoptosis (279). The FAF1 pr otein expression has been dem onstrated to be reduced in gastric carcinomas (280). One possible explan ation for the overexpression of FAF1 in responsive cell lines may cause th e cells to be more sensitive to induction of apoptosis as compared to the non-responsive cell lines. Fu rthermore, the function of FAF1 has been evaluated only in the context of cell lines, th e role of FAF1 may not be consistent in tumors. Two catenins were underexpressed in the responsi ve cell lines; and -catenin. The expression of these two catenins were less clear cut in the tumors, however there is evidence to suggest that they may play im portant function in tumorigenesis. The formation of tissues and organs largely de pends on interactions between neighboring cells. These associations orchestrate the asse mbly of diverse cell types into organized patterns to establish a complex organism. These interactions also permit adult tissues to perform unique functions, preserve architectur al integrity and precisely coordinate the events that enable cells to remodel tissues during normal homeostasi s and to synchronize in response to injury to repair tissues. By communicating signals through adhesion receptors, cells can respond and elicit the spa tially coordinated events needed to maintain tissue homeostasis. The imbalance of which is frequently observed in the malignant phenotype. Catenins are an intrace llular anchor proteins that attach cadherins to actin fibers within the cell and aid in the forma tion of adherens junctions (AJ), the building blocks of cellula r architecture. and -catenin are two important AJ catenins that
152 integrate cell to cell juncti ons and regulate cytoskeletal dynamics by governing signaling pathways involved in morphogenesis, homeo stasis and intercellular communication between different cell types w ithin a given tissue (281). -catenin binds indirectly to ca dherins via interactions with -catenin, a protein commonly mutated in cancers. Recently, -catenin has also been found to be an essential in coordinating actin dynamics and inversely correlating cell adhesion with proliferation (282). It is hypothe sized that loss of -catenin may account for the absence of calciumdependent cell-cell adhesion in cancers with intact E-cadherin expression. In addition, CTNNA1 the gene that encodes -catenin, has been shown to be a tumor suppressing gene (283-285). Furthermore, inactivating mutations of CTNNA1 have been demonstrated in lung, prostate, ovarian, a nd colon cancer cell lines (286). These cell lines lack normal cell-cell adhesion becau se the cadherin-catenin interactions are disrupted. IHC has also demonstrated a loss of -catenin expression in primary tumors (286). Mutational analysis has also shown that -catenin expression is decreased in synovial sarcomas (287). Therefore, -catenin is considered as an invasion suppressor molecule in cancers. The expression of -catenin was relatively low in all of the sarcoma specimens. This finding may provide evid ence to further explore the role of -catenin in sarcomas. Thus far, -catenin mutations have been id entified in synovial sarcomas, perhaps the expression of -catenin is so low in the microarray analysis because more sarcoma possess mutations in this gene. While the responsive cell lines had low levels of -catenin expression, the responsive tumors had higher levels as comp ared to the non-responsive tumors. These findings could lend evidence to suggest differential roles for -catenin in cell lines and
153 tumors and could be do to the lack of micr oenvironment in the cell lines. A role for catenin in cancer has been demonstr ated in the literature. Unlike -catenin whose sequence is considerably diffe rent from other catenins, -catenin is closely resembles catenin. Originally identified as a Src substrate, -catenin associates with cadherinmediated AJs, thus stabilizing them at the membrane. IHC studies have reported loss of delta-catenin in some primary tumors, however mutation of the CTNND1 gene is extremely rare (286). Ephrin-A1 and Dapper were two other gene s identified as potential biomarkers that predict response. Both genes were upregulated in the re sponsive cell lines. However, the predictability of response governed by Ephrin-A1 and Dapper expression was not successful when they were tested using the HOS and SW1353 cell lines. On the other hand, Ephrin-A1 and Dapper were overexp ressed in the responsive group of tumors compared to the non-responsive group. Furthe rmore, a role of Ephrin-A and Dapper in cancer has been described. The Eph family of receptor tyrosine kina ses and their cell-presented ligands, the ephrins, are overexpressed in a variety of can cers, including breast and gastrointestinal cancers, melanomas and neuroblastomas. Upon ephrin binding, the Eph receptors are phosphorylated at specific tyrosine residues in the cytoplasmic region, which then serve as docking sites for various signaling molecu les. Many SH2 domain -containing proteins have been found to interact with the phos photyrosines of activat ed Eph receptors, including Src and Fyn. These interactions ha ve been implicated in cell adhesion, cell motility, and cytoskeleton rearrangement. However, the mechanisms by which
154 individual molecules and signaling pathways exert specific functions has yet to be described (288). Ephrin-A1 is a ligand of EphA2 and is involved in the initiation of cellular migration, functions as an angiogenic factor in vitro and is essential for TNF-induced angiogenesis. Its expression in tumors corr elates with initiating invasion by attracting endothelial cells. In mice models using Ka rposi-sarcoma-derived tumors as xenografts, the vasculature invading the tumors expr essed increased ephrin -A1 and EphA2 (289). Further, ephrin-A1 and EphA2 can be detect ed on the surface of endothelial cells in a variety of human tumors. Dapper1, gene mapped to human chro mosome 14q22.3 and is deleted in astrocytoma, was identified as an intera cting protein for Disheveled, a Wnt signaling mediator, and modulates Wnt signaling. WNT signals play key roles in carcinogenesis and embryogenesis through the specification of cell fate and polarity. Dapper has been described as an antagonizer Wnt signaling by inducing Disheveled degradation (290). Consequently, based on the importance of the WNT signaling pathway in cancer, Dapper is predicted to be potent cancer-associated gene (291). While Ephrin-A1 and Dapper were not succes sful at predicting the response of HOS and SW1353 to dasatinib, the expression of these genes in sarcomas may be of interest for further studies. Both have significant roles in cancer, and may have a vital function in sarcomas. Furthermore, EphA1, the receptor for Ephrin-A1, is a known target of dasatinib. Perhaps the cell lines have provi ded a vital clue into a significant biomarker of predicting dasatinib prediction not only in cell lines, but in tumors as well. Only future studies can tell.
155 One must keep in mind that this signatur e that predicts response to dasatinib by induction of apoptosis in sarc oma cell lines is completely based on a genomic expression analysis. The information gained from a ge nomic analysis using mi croarrays is highly informative, however does not portray the func tional events occurring within the system analyzed. A GEP provides a snapshot of the genes which are expressed at a given point of time. In comparison, a proteomic anal ysis would provide a more informative understanding of the actual even ts occurring within a system as dictated by the gene products. Identifying a protein expression profile would present a functional analysis of the genes expressed and more di rectly translate the relevant overall effects of the biology at play within a system and more precisel y predict cellular behavi or. The information lost in the translation betw een gene expression and prot ein expression may account for many of the inconsistencies observed throughout the signature analysis within the cell lines and tumors analyzed in these studies Nevertheless, the identification of a molecular signature that has been proven to pr edict response to dasatinib in cell lines is an excellent starting point fo r generating more sophisticated analyses aimed at further identifying a more precise and accurate signat ure that effectively predicts response to dasatinib in human sarcoma patients. In conclusion, Src activity is essential for the survival of a subset of sarcomas. We have identified a molecular signature that can predict the res ponse of cell lines to dasatinib. The molecular signature can succe ssfully identify cell lin es that will undergo apoptosis upon inhibition of Src activity by dasa tinib. This molecular signature is also present to some degree in primary human sarc omas and should be further validated to determine the efficiency of this signature to predict patient re sponse to dasatinib.
156 Together, these studies reveal that both const itutive Src activation a nd the expression of a molecular signature that predicts response to dasatinib in cell lines may potentially play a critical role in the response of sarcomas to dasatinib. Thus, novel therapeutic approaches that inhibit Src signaling in sarcoma that express the molecular signature may have the potential to induce apoptosis and sensitize tumors to chemotherapy.
157 Clinical Significance TKIs have become a viable therapeutic option for the trea tment of sarcomas eliciting aberrantly activated tyrosine ki nase signaling cascades. Cell line based in vitro and in vivo models have demonstrated mixed, yet promising potential uses of TKIs as single agent or in combination with current tr eatment options. Proof-of-principle of this concept has been obtained for sarcomas in the case of Gleevec treatment of GISTs and DFSPs. Our data offers preclinical evidence to further investigate the response of sarcomas to dasatinib in clinical trials. In that, dasatinib is a promising therapeutic option for treating sarcomas with activated Src ki nase because it may prevent metastasis by inhibiting tumor cell migration and invasion, as well as induce apoptosis in a subset of bone derived sarcomas. A molecular signature in the bone subset of sarcomas that predicts apoptosis response to dasatinib ha s been identified in these studies. The predictive molecular signature of apoptosis re sponse to dasatinib re presents a testable hypothesis in clinical trials TKIs currently in early phase trials ha ve provided promising therapeutic options for treating sarcomas. Further work is required to delineate the role TKs play as effectors of tumorigenesis in sarcoma and differentiat e between relevant and irrelevant mutations with regards to upstream and downstream signali ng. To be effective, a TKI must target a
158 pathway which is essential and vital for the su rvival, as in the case of c-KIT in GIST and PDGFR in DFSP and Src in OSA cell lines.
159 References 1. Helman LJ, Meltzer P. Mechanisms of sarcoma development. Nat Rev Cancer 2003;3:685-694. 2. Jemal A, Tiwari RC, Murray T, et al. Ca ncer statistics, 2004. CA Cancer J Clin 2004;54:8-29. 3. U.S. Cancer Statistics Working Group [database on the Internet]. United States Cancer Statistics: 1999-2002 Incidence a nd Mortality Webbased Report Version. Atlanta (GA): Department of Health a nd Human Services, Centers for Disease Control and Prevention, and National Ca ncer Institute (US). c2005 [cited 2006 Oct 01]. Available from: www.cdc.gov/cancer/npcr/uscs/ 4. Ries L, Harkins D, Krapcho M, et al SEER Cancer Statistics Review, 1975-2003 [database on the Internet]. Bethesda (MD): National Cancer Institute (US). c2005 [cited 2006 Oct 01]. Available from: http://seer.can cer.gov/csr/1975_2003/ 5. Albritton KH. Sarcomas in adolescents and young adults. Hematol Oncol Clin North Am 2005;19:527-546, vii. 6. Bleyer A, Montello M, Budd T, Saxm an S. National survival trends of young adults with sarcoma: lack of progress is associated with lack of clinical trial participation. Cancer 2005;103:1891-1897. 7. Bleyer A MM, Budd T., et al. Young adults with sarcoma: lack of clinical trials participation and lack of survival prol ongation [abstract 3280]. Proc Am Soc Clin Oncol 2003;22:816.
160 8. Bleyer W. The adolescent gap in can cer treatment. J Registry Management 1996;23:114-115. 9. Bleyer WA, Tejeda H, Murphy SB, et al National cancer clini cal trials: children have equal access; adolescents do no t. J Adolesc Health 1997;21:366-373. 10. Society AC What are the key statistics about soft tissue sarcomas? MedlinePlus, 2006. http://www.cancer.org/docr oot/cri/content/cri_2_4_1x_wha t_are_the_key_statisti cs_for_sarcoma_38.asp?sitearea=cri 11. Gurney JG YJ, Roffers SD, Smith MA, Bunin GR Soft Tissue Sarcomas. SEER Pediatric Monograph, 1999. 12. Coindre JM, Terrier P, Guillou L, et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of 1240 patients from the French Federati on of Cancer Centers Sarcoma Group. Cancer 2001;91:1914-1926. 13. Society AC. What are the key st atistics about bon e cancer? 2006. 14. Gurney JG SA, Bulterys M Malignant Bone Tumors. SEER Pediatric Monograph, 1998. 15. American Cancer Society. What are the key statistics about bone cancer? American Cancer Society, Inc. c200 6 [cited 2006 May 16]. Available from: http://www.cancer.org/docroot/CRI/c ontent/CRI_2_4_1X_What_are_the_key_sta tistics_for_bone_can cer_2.asp?sitearea= 16. Cancer of soft tissue including hear t. SEER [cited 2007 Feb 6], 2002. Available from: https://seer.cancer.gov/ statfacts/html/soft.html
161 17. Pollock R. Soft Tissue Sarcomas. London: Hamilton; 2002. 18. Thomas JO. Acquired immunodeficiency syndrome-associated cancers in SubSaharan Africa. Semin Oncol 2001;28:198-206. 19. Stiller CA, Parkin DM. Geographic and ethnic variations in the incidence of childhood cancer. Br Med Bull 1996;52:682-703. 20. Zahm SH, Fraumeni JF, Jr. The epidem iology of soft tissue sarcoma. Semin Oncol 1997;24:504-514. 21. Abramson DH, Ellsworth RM, Kitchin FD Tung G. Second nonocular tumors in retinoblastoma survivors. Are they radiation-induced? Ophthalmology 1984;91:1351-1355. 22. Li FP, Fraumeni JF, Jr. Soft-tissue sa rcomas, breast cancer, and other neoplasms. A familial syndrome? Ann Intern Med 1969;71:747-752. 23. Li FP, Fraumeni JF, Jr. Prospective study of a family cancer syndrome. Jama 1982;247:2692-2694. 24. Malkin D The Li-Fraumeni syndrome. In: PPO Updates. Philadelphia: JB Lipincott, 1995. 25. Malkin D, Li FP, Strong LC, et al. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 1990;250:1233-1238. 26. King AA, Debaun MR, Riccardi VM, Gutmann DH. Malignant peripheral nerve sheath tumors in neurofibromatosis 1. Am J Med Genet 2000;93:388-392. 27. Nishida T, Hirota S, Taniguchi M, et al Familial gastrointestinal stromal tumours with germline mutation of the KIT gene. Nat Genet 1998;19:323-324.
162 28. Blair A, Malker H, Cantor KP, Burmeist er L, Wiklund K. Cancer among farmers. A review. Scand J Work Environ Health 1985;11:397-407. 29. Blair A, Zahm SH. Agricultural exposu res and cancer. Environ Health Perspect 1995;103 Suppl 8:205-208. 30. Dirix LY, Van Oosterom AT. Diagnosis a nd treatment of soft tissue sarcomas in adults. Curr Opin Oncol 1994;6:372-383. 31. Hardell L, Sandstrom A. Case-control study: soft-tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols. Br J Cancer 1979;39:711-717. 32. Kelly S, Guidotti, TL. Phenoxyacetic aci d herbicides and chlorophenols and the etioloty of lymphoma and soft tissue neoplasms. JAMA 1990;256:1141-1147. 33. Merlo F, Puntoni R. Soft-tissue sa rcomas, malignant lymphomas, and 2,3,7,8TCDD exposure in Seveso. Lancet 1986;2:1455. 34. Riihimaki V, Asp S, Hernberg S. Mortality of 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid herbicide app licators in Finland: first report of an ongoing prospective cohort study. Sca nd J Work Environ Health 1982;8:37-42. 35. Saracci R, Kogevinas M, Bertazzi PA, et al. Cancer mortality in workers exposed to chlorophenoxy herbicides and ch lorophenols. Lancet 1991;338:1027-1032. 36. Smith JG, Christophers AJ. Phenoxy herbic ides and chlorophenols: a case control study on soft tissue sarcoma and maligna nt lymphoma. Br J Cancer 1992;65:442448. 37. Wiklund K, Holm LE. Soft tissue sarc oma risk in Swedish agricultural and forestry workers. J Natl Cancer Inst 1986;76:229-234.
163 38. Wiklund TA, Blomqvist CP, Raty J, et al. Postirradiation sarcoma. Analysis of a nationwide cancer registry ma terial. Cancer 1991;68:524-531. 39. The association of selected cancers with service in the US military in Vietnam. II. Soft-tissue and other sarcomas. The Se lected Cancers Cooperative Study Group. Arch Intern Med 1990;150:2485-2492. 40. Bertazzi PA, Consonni D, Bachetti S, et al. Health effects of dioxin exposure: a 20-year mortality study. Am J Epidemiol 2001;153:1031-1044. 41. Bertazzi PA, Zocchetti C, Guercilena S, et al. Dioxin exposure and cancer risk: a 15-year mortality study after the "S eveso accident". Epidemiology 1997;8:646652. 42. Brann E. The association of selected can cers with service in the US military in Vietnam. II. Soft tissue and other sarcomas. Arch Intern Med 1990;150:24852492. 43. Constable JD, Timperi R, Clapp R, Antm an K, Boynton B. Vietnam veterans and soft tissue sarcoma. J Occup Med 1987;29:726. 44. Fingerhut MA, Halperin WE, Marlow DA, et al. Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzop-dioxin. N Engl J Med 1991;324:212-218. 45. Greenwald P, Kovasznay B, Collins DN, Th erriault G. Sarcomas of soft tissues after Vietnam service. J Na tl Cancer Inst 1984;73:1107-1109. 46. Kang H, Enziger F, Breslin P, al e. So ft tissue sarcoma and military service in Vietnam: a case control study. J Nat Cancer Inst 1987;79:693-699.
164 47. Kang HK, Weatherbee L, Breslin PP, Lee Y, Shepard BM. Soft tissue sarcomas and military service in Vietnam: a cas e comparison group analysis of hospital patients. J Occup Med 1986;28:1215-1218. 48. Pierce S. The purported link between ag ent orange and cancer. Hum Pathol 1995;26:693-695. 49. Sarma PR, Jacobs J. Thoracic soft-tissue sarcoma in Vietnam veterans exposed to Agent Orange. N Engl J Med 1982;306:1109. 50. Engels EA, Katki HA, Nielsen NM, et al Cancer incidence in Denmark following exposure to poliovirus vaccine contaminated with simian virus 40. J Natl Cancer Inst 2003;95:532-539. 51. McClain KL, Leach CT, Jenson HB, et al. Association of Epstein-Barr virus with leiomyosarcomas in children w ith AIDS. N Engl J Med 1995;332:12-18. 52. Amendola BE, Amendola MA, McClatchey KD, Miller CH, Jr. Radiationassociated sarcoma: a review of 23 patie nts with postradiation sarcoma over a 50year period. Am J Clin Oncol 1989;12:411-415. 53. Aoki T, Ozeki Y, Watanabe M, et al. Development of primary leiomyosarcoma of the sternum postirradiation: report of a case. Surg Today 1998;28:1326-1328. 54. Brady MS, Gaynor JJ, Brennan MF. Radia tion-associated sarcoma of bone and soft tissue. Arch Surg 1992;127:1379-1385. 55. Cahan W, Woodard, HQ, Higinbotham, NL, et al. Sarcoma arising in irradiated bonereport of 11 cases. Cancer 1948;1:3-29. 56. Chang SM, Barker FG, 2nd, Larson DA, Bollen AW, Prados MD. Sarcomas subsequent to cranial irradiat ion. Neurosurgery 1995;36:685-690.
165 57. Hatfield P, Schultz, MD. Postirradiation sarcoma, including five cases after x-ray therapy of breast carci noma. Radiology 1970;96:593-602. 58. Hawkins MM. Second primary tumors following radiotherapy for childhood cancer. Int J Radiat Onco l Biol Phys 1990;19:1297-1301. 59. Huvos AG, Woodard HQ, Cahan WG, et al Postradiation osteogenic sarcoma of bone and soft tissues. A clinicopathologic study of 66 patients. Cancer 1985;55:1244-1255. 60. Kim JH, Chu FC, Woodard HQ, et al. Radiation-induced soft-tissue and bone sarcoma. Radiology 1978;129:501-508. 61. Lagrange JL, Ramaioli A, Chateau MC, et al. Sarcoma after radiation therapy: retrospective multiinstitutional study of 80 histologically confirmed cases. Radiation Therapist and Pathologist Gr oups of the Federation Nationale des Centres de Lutte Contre le Cancer. Radiology 2000;216:197-205. 62. Laskin WB, Silverman TA, Enzinger FM. Postradiation soft tissue sarcomas. An analysis of 53 cases. Cancer 1988;62:2330-2340. 63. Mark RJ, Bailet JW, Poen J, et al. Pos tirradiation sarcoma of the head and neck. Cancer 1993;72:887-893. 64. Mark RJ, Poen J, Tran LM, et al. Po stirradiation sarcomas. A single-institution study and review of the lite rature. Cancer 1994;73:2653-2662. 65. Martland H. Occupational Poisoning in Manufacture of Luminous Watch Dials. JAMA 1929;92:466-473.
166 66. Menu-Branthomme A, Rubino C, Sham saldin A, et al. Radiation dose, chemotherapy and risk of soft tissu e sarcoma after solid tumours during childhood. Int J Cancer 2004;110:87-93. 67. Mermershtain W, Cohen AD, Koretz M, Cohen Y. Cutaneous angiosarcoma of breast after lumpectomy, axillary lym ph node dissection, and radiotherapy for primary breast carcinoma: case report a nd review of the literature. Am J Clin Oncol 2002;25:597-598. 68. Miracco C, Materno M, De Santi MM Pirtoli L, Ninfo V. Unusual second malignancies following radiation ther apy: subcutaneous pleomorphic rhabdomyosarcoma and cutaneous melanoma. Two case reports. J Cutan Pathol 2000;27:419-422. 69. Phillips TL, Sheline GE. Bone Sarcomas Following Radiation Therapy. Radiology 1963;81:992-996. 70. Scanlon E, Berk, RS, Khandekar, JD. Postirradiation Neoplasia: A Symposium. Curr Prob Cancer 1978;111:4-35. 71. Taghian A, de Vathaire F, Terrier P, et al. Long-term risk of sarcoma following radiation treatment for br east cancer. Int J Radiat Oncol Biol Phys 1991;21:361367. 72. Tountas AA, Fornasier VL, Harwood AR, Leung PM. Postirradiation sarcoma of bone: a perspective. Cancer 1979;43:182-187. 73. Tucker MA, D'Angio GJ, Boice JD, Jr., et al. Bone sarcomas linked to radiotherapy and chemotherapy in children. N Engl J Med 1987;317:588-593.
167 74. Weatherby RP, Dahlin DC, Ivins JC. Post radiation sarcoma of bone: review of 78 Mayo Clinic cases. Mayo Clin Proc 1981;56:294-306. 75. Wei LH, Huang CY, Cheng SP, Chen CA Hsieh CY. Carcinosarcoma of ovary associated with previous radiothera py. Int J Gynecol Cancer 2001;11:81-84. 76. Zejda JE, McDuffie HH, Dosman JA. Epidem iology of health and safety risks in agriculture and relate d industries. Practical applica tions for rural physicians. West J Med 1993;158:56-63. 77. Huang J, Mackillop WJ. Increased risk of soft tissue sarcoma after radiotherapy in women with breast carcino ma. Cancer 2001;92:172-180. 78. Karlsson P, Holmberg E, Samuelsson A, Johansson KA, Wallgren A. Soft tissue sarcoma after treatment for breast cancer --a Swedish population-based study. Eur J Cancer 1998;34:2068-2075. 79. Rubino C, Shamsaldin A, Le MG, et al. Radiation dose and risk of soft tissue and bone sarcoma after breast cancer trea tment. Breast Cancer Res Treat 2005;89:277-288. 80. Le Vu B, de Vathaire F, Shamsaldin A, et al. Radiation dose, chemotherapy and risk of osteosarcoma after solid tumours during childhood. Int J Cancer 1998;77:370-377. 81. dos Santos Silva I, Jones M, Malvei ro F, Swerdlow A. Mortality in the Portuguese thorotrast stu dy. Radiat Res 1999;152:S88-92. 82. Mori T, Kido C, Fukutomi K, et al. Summary of entire Japanese thorotrast followup study: updated 1998. Radi at Res 1999;152:S84-87.
168 83. Platz EA, Wiencke JK, Kelsey KT, et al. Chromosomal aberrations and hprt mutant frequencies in long-te rm American thorotrast surv ivors. Int J Radiat Biol 2000;76:955-961. 84. Ron E. Cancer risks from medical radiation. Health Phys 2003;85:47-59. 85. Fodor J, Orosz Z, Szabo E, et al. Angi osarcoma after conservation treatment for breast carcinoma: our experience and a re view of the litera ture. J Am Acad Dermatol 2006;54:499-504. 86. Tomasini C, Grassi M, Pippione M. Cutaneous angiosarcoma arising in an irradiated breast. Case report and re view of the literature. Dermatology 2004;209:208-214. 87. Neglia JP, Friedman DL, Yasui Y, et al Second malignant neoplasms in five-year survivors of childhood cancer: childhood cancer survivor study. J Natl Cancer Inst 2001;93:618-629. 88. Saggia C, Forti G, Biaggi G, et al. Tw o cases of secondary soft tissue sarcomas after radiotherapy and radioche motherapy. Tumori 2004;90:622-624. 89. Fotiou S, Hatjieleftheriou G, Kyrousis G, Kokka F, Apostolikas N. Long-term tamoxifen treatment: a possible aetiological factor in the deve lopment of uterine carcinosarcoma: two case-reports and revi ew of the literatu re. Anticancer Res 2000;20:2015-2020. 90. Kloos I, Delaloge S, Pautier P, et al Tamoxifen-related uterine carcinosarcomas occur under/after prolonged treatment: repor t of five cases and review of the literature. Int J Gyneco l Cancer 2002;12:496-500.
169 91. Martin-Loeches M, Rius J, Orti RM. Ut erine sarcoma associated with tamoxifen use: case report. Eur J Gynaecol Oncol 2003;24:202-203. 92. Yildirim Y, Inal MM, Sanci M, et al Development of uterine sarcoma after tamoxifen treatment for breast cancer: repo rt of four cases. In t J Gynecol Cancer 2005;15:1239-1242. 93. Belli S, Bertazzi PA, Comba P, et al. A cohort study on vinyl chloride manufacturers in Italy: study design and preliminary results. Cancer Lett 1987;35:253-261. 94. Buffler PA, Wood S, Eifler C, Suarez L, Kilian DJ. Mortal ity experience of workers in a vinyl chloride mono mer production plant. J Occup Med 1979;21:195-203. 95. Byren D, Engholm G, Englund A, Wester holm P. Mortality and cancer morbidity in a group of Swedish VCM and PCV production workers. Environ Health Perspect 1976;17:167-170. 96. Elliott P, Kleinschmidt I. Angiosarcoma of the liver in Great Britain in proximity to vinyl chloride sites. Occup Environ Med 1997;54:14-18. 97. Evans DG, Baser ME, McGaughran J, et al. Malignant peripheral nerve sheath tumours in neurofibromatosis 1. J Med Genet 2002;39:311-314. 98. Cooper WC. Epidemiologic study of vinyl chloride workers: mortality through December 31, 1972. Environ Health Perspect 1981;41:101-106. 99. Makk L, Delorme F, Creech JL. [Angiosarcomas of the liver in workers having had prolonged contact with vinyl chloride : epidemiology and program of research in the workers]. Union Med Can 1975;104:1833-1835.
170 100. Muller R, Hajdu SI, Brennan MF. Lympha ngiosarcoma associated with chronic filarial lymphedema. Cancer 1987;59:179-183. 101. Stewart F, Treves N. Lymphoangiosarcoma in postmastectomy lymphedema: a report of six cases of elephantia sis chirurgica. Cancer 1948;1:64. 102. Brennan M, Lewis J. Diagnosis and Management of Soft Tissue Sarcoma. London: Martin Dunitz; 2002. 103. Operskalski EA, Preston-Martin S, Henderson BE, Visscher BR. A case-control study of osteosarcoma in young persons. Am J Epidemiol 1987;126:118-126. 104. Ben-Izhak O, Kerner H, Brenner B, al e. Angiosarcoma of the colon developing in a capsule of a foreign body. Am J of Clin Path 1991;97:416-420. 105. Ben-Izhak O, Vlodavsky E, Ofer A, et al. Epithelioid angios arcoma associated with a Dacron vascular graft. Am J Surg Pathol 1999;23:1418-1422. 106. Brand K Foreign body induced sarcomas In: B. FF (ed.), Cancer. New York: Plenum, 1975. pp. 485-489. 107. Burns W, Kanhouwa S, Tillman L, Saini N, Herrmann J. Fibrosarcoma occurring at the site of a plastic vasc ular graft. Cancer 1971;29:66-72. 108. Dube VE, Fisher DE. Hemangioendothel ioma of the leg following metallic fixation of the tibia Cancer 1972;30:1260-1266. 109. Fyfe BS, Quintana CS, Kaneko M, Grie pp RB. Aortic sarcoma four years after Dacron graft insertion. Ann Thorac Surg 1994;58:1752-1754. 110. Hayman J, Huygens H. Angiosarcoma developing around a foreign body. J Clin Pathol 1983;36:515-518.
171 111. Lindeman G, McKay MJ, Taubman KL Bilous AM. Malignant fibrous histiocytoma developing in bone 44 years after shrapnel trauma. Cancer 1990;66:2229-2232. 112. Aboulafia AJ, Brooks F, Piratzky J, Weiss S. Osteosarcoma arising from heterotopic ossification after an electrical burn. A case report. J Bone Joint Surg Am 1999;81:564-570. 113. Can Z, Yilmaz S, Riza A, Apaydin EI, Kuzu I. Sarcoma developing in a burn scar: case report and review of the literature. Burns 1998;24:68-71. 114. Davis GB, Blanchard DK, Hatch GF, 3rd, et al. Tumors of the stomach. World J Surg 2000;24:412-420. 115. Ju DM. Fibrosarcoma arising in surgi cal scars. Plast R econstr Surg 1966;38:429437. 116. Micali G, Gasparri O, Nasca MR, Sapuppo A. Kaposi's sarcoma occurring de novo in the surgical scar in a heart tran splant recipient. J Am Acad Dermatol 1992;27:273-274. 117. Nakanishi H, Tomita Y, Yoshikawa H, et al. Frequent p53 gene mutations in soft tissue sarcomas arising in burn scar. Jpn J Cancer Res 1999;90:276-279. 118. Ozyazgan I, Kontas O. Burn scar sarcoma. Burns 1999;25:455-458. 119. Ugurlu K, Turgut G, Kabukcuoglu F, et al. Malignant fibrous histiocytoma developing in a burn sc ar. Burns 1999;25:764-767. 120. Gottlieb JA, Baker LH, Quagliana JM, et al. Chemotherapy of sarcomas with a combination of adriamycin and dimethyl triazeno imidazole carboxamide. Cancer 1972;30:1632-1638.
172 121. Bramwell VH, Mouridsen HT, Santoro A, et al. Cyclophosphamide versus ifosfamide: final report of a randomized phase II trial in adult soft tissue sarcomas. Eur J Cancer Clin Oncol 1987;23:311-321. 122. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metast atic gastrointestinal stromal tumor. N Engl J Med 2001;344:1052-1056. 123. Blanke C, con Mehren H, Joensuu P, al. e. Evaluation of the safety and efficacy of an oral molecularly-targeted therapy STI571, in patients with unresectable or metastatic gastrointestinal stromal tumors (GISTS) expressing c-kit (CD117). ASCO Proc. 2001;20. 124. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science 2002;298:1912-1934. 125. Alonso A, Sasin J, Bottini N, et al. Protein tyrosine phosphatases in the human genome. Cell 2004;117:699-711. 126. Koon HB, Bubley GJ, Pantanowitz L, et al. Imatinib-induced regression of AIDSrelated Kaposi's sarcoma. J Clin Oncol 2005;23:982-989. 127. McArthur GA, Demetri GD, van Oosterom A, et al. Molecular and clinical analysis of locally advanced dermatof ibrosarcoma protuberans treated with imatinib: Imatinib Target Explora tion Consortium Study B2225. J Clin Oncol 2005;23:866-873. 128. Daw NC, Furman WL, Stewart CF, et al Phase I and pharmacokinetic study of gefitinib in children with refractory solid tumors: a Children's Oncology Group Study. J Clin Oncol 2005;23:6172-6180.
173 129. Stewart CF, Leggas M, Schuetz JD, et al. Gefitinib enhances the antitumor activity and oral bioavailability of ir inotecan in mice. Cancer Res 2004;64:74917499. 130. Kuenen BC, Tabernero J, Baselga J, et al. Efficacy and toxicity of the angiogenesis inhibitor SU5416 as a single ag ent in patients with advanced renal cell carcinoma, melanoma, and soft tissue sarcoma. Clin Cancer Res 2003;9:1648-1655. 131. Angelov L, Salhia B, Roncari L, McMahon G, Guha A. Inhibition of angiogenesis by blocking activation of the vascular endot helial growth factor receptor 2 leads to decreased growth of neurogenic sarcomas. Cancer Res 1999;59:5536-5541. 132. Dalal S, Berry AM, Cullinane CJ, et al Vascular endothelial growth factor: a therapeutic target for tumors of the Ew ing's sarcoma family. Clin Cancer Res 2005;11:2364-2378. 133. Gingrich DE, Reddy DR, Iqbal MA, et al. A new class of potent vascular endothelial growth factor re ceptor tyrosine kinase inhi bitors: structure-activity relationships for a series of 9alkoxymethyl-12-(3-hydroxypropyl)indeno[2,1a]pyrrolo[3,4-c]carbazole-5 ones and the id entification of CEP-5214 and its dimethylglycine ester prodrug clinic al candidate CEP-7055. J Med Chem 2003;46:5375-5388. 134. Scotlandi K, Manara MC, Strammiello R, et al. C-kit receptor expression in Ewing's sarcoma: lack of prognostic valu e but therapeutic ta rgeting opportunities in appropriate conditions. J Clin Oncol 2003;21:1952-1960.
174 135. Shor AC, Keschman EA, Lee FY, et al. Dasatinib inhibits migration and invasion in diverse human sarcoma cell lines and induces apoptosis in bone sarcoma cells dependent on Src kinase for survival. Cancer Res 2007;In Press. 136. Balasubramanian L, Evens AM. Targeting angiogenesis for the treatment of sarcoma. Curr Opin Oncol 2006;18:354-359. 137. Knowling M, Blackstein M, Tozer R, et al. A phase II study of perifosine (D21226) in patients with previously untreate d metastatic or loca lly advanced soft tissue sarcoma: A National Cancer Institute of Canada Clinical Trials Group trial. Invest New Drugs 2006;24:435-439. 138. Van Ummersen L, Binger K, Volkman J, et al. A phase I trial of perifosine (NSC 639966) on a loading dose/maintenance dose sc hedule in patients with advanced cancer. Clin Cancer Res 2004;10:7450-7456. 139. Guan H, Jia SF, Zhou Z, Stewart J, Kleinerman ES. Herceptin down-regulates HER-2/neu and vascular endot helial growth factor expr ession and enhances taxolinduced cytotoxicity of human Ewing's sa rcoma cells in vitro and in vivo. Clin Cancer Res 2005;11:2008-2017. 140. Ye D, Maitra A, Timmons CF, et al. The epidermal growth factor receptor HER2 is not a major therapeutic target in Ewing sarcoma. J Pediatr Hematol Oncol 2003;25:459-466. 141. Azizi AA, Haberler C, Czech T, et al. Vascular-endothelial-growth-factor (VEGF) expression and possible response to angiogenesis inhibitor bevacizumab in metastatic alveolar soft part sarcoma. Lancet Oncol 2006;7:521-523.
175 142. D'Adamo DR, Anderson SE, Albritton K, et al. Phase II study of doxorubicin and bevacizumab for patients with metastatic soft-tissue sarcomas. J Clin Oncol 2005;23:7135-7142. 143. DeMatteo RP, Lewis JJ, Leung D, et al Two hundred gastrointestinal stromal tumors: recurrence patterns and prognos tic factors for survival. Ann Surg 2000;231:51-58. 144. Goss GA MP, Manola K, SInger S, Fletcher CD, Demetri GD. Clinical and pathological characteristics of gastrointestinal stromal tumors (GIST). Prog Proc Am Soc Clin Oncol 2000;19:abstract. 145. Joensuu H. Gastointestinal stromal tumor (GIST). Annals of Oncology 2006;17:x280-x286. 146. Rubin BP, Singer S, Tsao C, et al. KI T activation is a ubi quitous feature of gastrointestinal stromal tu mors. Cancer Res 2001;61:8118-8121. 147. Blanke CD, Eisenberg BL, Heinrich MC Gastrointestinal stromal tumors. Curr Treat Options Oncol 2001;2:485-491. 148. van Oosterom AT, Judson I, Verweij J, et al. Safety and efficacy of imatinib (STI571) in metastatic gast rointestinal stro mal tumours: a phase I study. Lancet 2001;358:1421-1423. 149. Duensing A, Heinrich MC, Fletcher CD, Fl etcher JA. Biology of gastrointestinal stromal tumors: KIT mutations and beyond. Cancer Invest 2004;22:106-116. 150. Heinrich M, Corless C, Blanke C, at. e. Muational status of the KIT oncoprotein correlates with the clincinal responses to Imatinib mesylate (Gleevec) in patients with metastatic gastroin testinal stromal tumors (GISTs). ASCO Proc. 2002;21.
176 151. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472-480. 152. Verweij J, Casali PG, Zalcberg J, et al. Progression-free survival in gastrointestinal stromal tumours with highdose imatinib: randomised trial. Lancet 2004;364:1127-1134. 153. Greco A, Roccato E, Miranda C, et al. Growth-inhibitory effect of STI571 on cells transformed by the COL1A1/PDGFB rearrangement. Int J Cancer 2001;92:354-360. 154. Yoshitani K, Honoki K, Morishita T, et al. Growth inhibition of rat osteosarcoma and malignant fibrous histiocytoma cells by tyrosine kinase inhibitor STI571. In Vivo 2003;17:255-258. 155. McGary EC, Weber K, Mills L, et al. Inhi bition of platelet-der ived growth factormediated proliferation of osteosarcoma ce lls by the novel tyrosine kinase inhibitor STI571. Clin Cancer Res 2002;8:3584-3591. 156. Matsuyama S, Iwadate M, Kondo M, et al. SB-431542 and Gleevec inhibit transforming growth factor-beta-induced proliferation of human osteosarcoma cells. Cancer Res 2003;63:7791-7798. 157. Merchant MS, Woo CW, Mackall CL, Thie le CJ. Potential use of imatinib in Ewing's Sarcoma: evidence for in vitro and in vivo activity. J Natl Cancer Inst 2002;94:1673-1679.
177 158. Druker BJ, Talpaz M, Resta DJ, et al. Effi cacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chr onic myeloid leukemia. N Engl J Med 2001;344:1031-1037. 159. Druker BJ, Sawyers CL, Kantarjian H, et al Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast cr isis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 2001;344:1038-1042. 160. Peng B HM, Racine-Poon A, Druker BJ, Ta lpaz M, Sawyers CL, et al. CLinical investigtion of the pharm acokinetic/pharmacodynamic relationship for Glivec (STI571): a novel inhibiotr of signal transduction. Proc ASCO 2001;20:280a. 161. Druker BJ. Taking aim at Ewing's sarcoma: is KIT a target and will imatinib work? J Natl Cancer Inst 2002;94:1660-1661. 162. Hotfilder M, Lanvers C, Jurgens H, Boos J, Vormoor J. c-KIT-expressing Ewing tumour cells are insensitive to imatinib mesylate (STI571). Cancer Chemother Pharmacol 2002;50:167-169. 163. Rous P. A transmissible avian neoplasm (sarcoma of the common fowl). J. Exp. Med. 1910;12:696-705. 164. Rous P. A sarcoma of the fowl transm issible by an agent separable from the tumor cells. J Exp Med 1911;13:397-411. 165. Halberstaedter L, Doljanski L, Te nenbaum E. Br. J. Exp. Pathol. 22 1941. 166. Schmidt CG, Vermathen H, Ritter S. [El ectrophoretic and total-protein studies in Yoshida sarcoma.]. Z Gesamte Exp Med 1959;131:396-406.
178 167. Svet-Moldavsky GJ. Development of multiple cysts and of haemorrhagic affections of internal organs in albino rats treated during the embryonic or newborn period with Rous sarcoma virus. Nature 1957;180:1299-1300. 168. Svet-Moldavsky GJ. Sarcoma in albino rats treated during the embryonic stage with Rous virus. Nature 1958;182:1452-1453. 169. Zilber L, Kryukova IN. Vopr. Virusol. 1958;3:166-170. 170. Zilber LA, Kryukova IN. Haemorrhagic dis ease of rats due to the virus of chick sarcoma. Acta Virol 1957;1:156-160. 171. Temin HM. The control of cellular mor phology in embryonic cells infected with rous sarcoma virus in v itro. Virology 1960;10:182-197. 172. Brugge JS, Erikson RL. Identification of a transformation-specific antigen induced by an avian sarcoma virus. Nature 1977;269:346-348. 173. Collett MS, Erikson RL. Protein kinase act ivity associated with the avian sarcoma virus src gene product. Proc Natl Acad Sci U S A 1978;75:2021-2024. 174. Levinson AD, Oppermann H, Levintow L, Varmus HE, Bishop JM. Evidence that the transforming gene of avian sarcoma virus encodes a protein kinase associated with a phosphoprotein. Cell 1978;15:561-572. 175. Hunter T, Sefton BM. Transforming gene product of Rous sarcoma virus phosphorylates tyrosine. Proc Natl Acad Sci U S A 1980;77:1311-1315. 176. Shalloway D, Zelenetz AD, Cooper GM Molecular cloning and characterization of the chicken gene homologous to the tran sforming gene of Rous sarcoma virus. Cell 1981;24:531-541.
179 177. Takeya T, Hanafusa H. Structure and sequence of the cellular gene homologous to the RSV src gene and the mechanism for generating the transforming virus. Cell 1983;32:881-890. 178. Iba H, Takeya T, Cross FR, Hanafusa T, Hanafusa H. Rous sarcoma virus variants that carry the cellular src gene instead of the viral src gene cannot transform chicken embryo fibroblasts. Proc Natl Acad Sci U S A 1984;81:4424-4428. 179. Parker RC, Varmus HE, Bishop JM. Expre ssion of v-src and chicken c-src in rat cells demonstrates qualitative differe nces between pp60v-src and pp60c-src. Cell 1984;37:131-139. 180. Shalloway D, Coussens PM, Yaciuk P. Overexpression of the c-src protein does not induce transformation of NIH 3T 3 cells. Proc Natl Acad Sci U S A 1984;81:7071-7075. 181. Roskoski R, Jr. Src kinase regulat ion by phosphorylation and dephosphorylation. Biochem Biophys Res Commun 2005;331:1-14. 182. Bjorge JD, Jakymiw A, Fujita DJ. Sele cted glimpses into the activation and function of Src kinase Oncogene 2000;19:5620-5635. 183. Roskoski R, Jr. Src protein-tyrosine kinase structure a nd regulation. Biochem Biophys Res Commun 2004;324:1155-1164. 184. Sun G, Sharma AK, Budde RJ. Autophosphor ylation of Src and Yes blocks their inactivation by Csk phosphoryla tion. Oncogene 1998;17:1587-1595. 185. Thomas SM, Brugge JS. Cellular functions regulated by Src family kinases. Annu Rev Cell Dev Biol 1997;13:513-609.
180 186. Turkson J, Bowman T, Garcia R, et al. Stat3 activation by Src induces specific gene regulation and is required fo r cell transformation. Mol Cell Biol 1998;18:2545-2552. 187. Yu CL, Meyer DJ, Campbell GS, et al. Enhanced DNA-binding activity of a Stat3-related protein in cells tran sformed by the Src oncoprotein. Science 1995;269:81-83. 188. Macara IG, Marinetti GV, Balduzzi PC. Transforming protein of avian sarcoma virus UR2 is associated with phosphatidy linositol kinase activity : possible role in tumorigenesis. Proc Natl Acad Sci U S A 1984;81:2728-2732. 189. Sugimoto Y, Whitman M, Cantley LC, Erikson RL. Evidence that the Rous sarcoma virus transforming gene produc t phosphorylates phos phatidylinositol and diacylglycerol. Proc Natl Acad Sci U S A 1984;81:2117-2121. 190. Smith MR, DeGudicibus SJ, Stacey DW. Requirement for c-ras proteins during viral oncogene transformation. Nature 1986;320:540-543. 191. Martin GS. The road to Src. Oncogene 2004;23:7910-7917. 192. Biscardi JS, Ishizawar RC, Silva CM, Pa rsons SJ. Tyrosine kinase signalling in breast cancer: epidermal growth factor re ceptor and c-Src interactions in breast cancer. Breast Cancer Res 2000;2:203-210. 193. Chen T, George JA, Taylor CC. Src tyrosi ne kinase as a chemotherapeutic target: is there a clinical case? Anticancer Drugs 2006;17:123-131. 194. Irby RB, Yeatman TJ. Role of Src expr ession and activation in human cancer. Oncogene 2000;19:5636-5642.
181 195. Yu H, Jove R. The STATs of cancer -new molecular targets come of age. Nat Rev Cancer 2004;4:97-105. 196. Buettner R, Mora LB, Jove R. Activ ated STAT signaling in human tumors provides novel molecular targets for ther apeutic intervention. Clin Cancer Res 2002;8:945-954. 197. Turkson J, Jove R. STAT proteins: novel molecular targets for cancer drug discovery. Oncogene 2000;19:6613-6626. 198. Burridge K, Chrzanowska-Wodnicka M. Focal adhesions, contractility, and signaling. Annu Rev Cell Dev Biol 1996;12:463-518. 199. Mitra SK, Schlaepfer DD. Integrin-re gulated FAK-Src signaling in normal and cancer cells. Curr Opin Cell Biol 2006;18:516-523. 200. McLean GW, Komiyama NH, Serrels B, et al. Specific deletion of focal adhesion kinase suppresses tumor formation and blocks malignant progression. Genes Dev 2004;18:2998-3003. 201. Shen TL, Park AY, Alcaraz A, et al Conditional knockout of focal adhesion kinase in endothelial ce lls reveals its role in angiogenesis and vascular development in late embryogene sis. J Cell Biol 2005;169:941-952. 202. Parsons JT. Focal adhesion kinase: the first ten years. J Cell Sci 2003;116:14091416. 203. Golubovskaya VM, Finch R, Cance WG. Di rect interaction of the N-terminal domain of focal adhesion kinase with the N-terminal trans activation domain of p53. J Biol Chem 2005;280:25008-25021.
182 204. Owens LV, Xu L, Craven RJ, et al. Over expression of the focal adhesion kinase (p125FAK) in invasive human tu mors. Cancer Res 1995;55:2752-2755. 205. Mitra SK, Lim ST, Chi A, Schlaepfer DD. Intrinsic focal adhesion kinase activity controls orthotopic breast carcinoma metastasis via th e regulation of urokinase plasminogen activator expression in a syngeneic tumor model. Oncogene 2006;25:4429-4440. 206. Frame MC. Src in cancer: deregulation and consequences for cell behaviour. Biochim Biophys Acta 2002;1602:114-130. 207. Playford MP, Schaller MD. The interpla y between Src and integrins in normal and tumor biology. Oncogene 2004;23:7928-7946. 208. Brabek J, Constancio SS, Shin NY, et al. CAS promotes invasiveness of Srctransformed cells. Oncogene 2004;23:7406-7415. 209. Bouton AH, Riggins RB, Bruce-Staskal PJ. Functions of the ad apter protein Cas: signal convergence and the determinati on of cellular responses. Oncogene 2001;20:6448-6458. 210. Bradshaw JM, Waksman G. Calorimetr ic investigation of proton linkage by monitoring both the enthalpy a nd association constant of binding: application to the interaction of the Src SH2 domain with a high-affinity tyrosyl phosphopeptide. Biochemistry 1998;37:15400-15407. 211. Gilmer T, Rodriguez M, Jordan S, et al. Peptide inhibitors of src SH3-SH2phosphoprotein interactions. J Biol Chem 1994;269:31711-31719.
183 212. Oneyama C, Nakano H, Sharma SV. UCS15A, a novel small molecule, SH3 domain-mediated protein-protein in teraction blocking drug. Oncogene 2002;21:2037-2050. 213. Banerji U, Judson I, Workman P. The clin ical applications of heat shock protein inhibitors in cancer present and fu ture. Curr Cancer Drug Targets 2003;3:385390. 214. Uehara Y. Natural product origins of Hsp90 inhibitors. Curr Cancer Drug Targets 2003;3:325-330. 215. Pengetnze Y, Steed M, Roby KF, Terranova PF, Taylor CC. Src tyrosine kinase promotes survival and resistance to chem otherapeutics in a mouse ovarian cancer cell line. Biochem Biophys Res Commun 2003;309:377-383. 216. Song L, Turkson J, Karras JG, Jove R, Haura EB. Activation of Stat3 by receptor tyrosine kinases and cytokines regulate s survival in human non-small cell carcinoma cells. Oncogene 2003;22:4150-4165. 217. Garcia R, Yu CL, Hudnall A, et al. C onstitutive activation of Stat3 in fibroblasts transformed by diverse oncoproteins and in breast carcinoma cells. Cell Growth Differ 1997;8:1267-1276. 218. Wang YZ, Wharton W, Garcia R, et al Activation of Stat3 preassembled with platelet-derived growth factor beta re ceptors requires Src kinase activity. Oncogene 2000;19:2075-2085. 219. Niu G, Bowman T, Huang M, et al. Role s of activated Src an d Stat3 signaling in melanoma tumor cell growth. Oncogene 2002;21:7001-7010.
184 220. Golas JM, Arndt K, Etienne C, et al SKI-606, a 4-anilino-3-quinolinecarbonitrile dual inhibitor of Src and Abl kinases, is a potent antiproliferative agent against chronic myelogenous leukemia cells in culture and causes regression of K562 xenografts in nude mice. Cancer Res 2003;63:375-381. 221. Coluccia AM, Benati D, Dekhil H, et al SKI-606 decreases growth and motility of colorectal cancer cel ls by preventing pp60(c-Src)-dependent tyrosine phosphorylation of beta-catenin and its nuclear signaling. Cancer Res 2006;66:2279-2286. 222. Golas JM, Lucas J, Etienne C, et al. SKI-606, a Src/Abl inhibitor with in vivo activity in colon tumor xenograft models. Cancer Res 2005;65:5358-5364. 223. Nam S, Kim D, Cheng JQ, et al. Action of the Src family kinase inhibitor, dasatinib (BMS-354825), on human pros tate cancer cells. Cancer Res 2005;65:9185-9189. 224. Johnson FM, Saigal B, Talpaz M, Donato NJ. Dasatinib (BMS-354825) tyrosine kinase inhibitor suppresses invasion and induces cell cycl e arrest and apoptosis of head and neck squamous cell carcinoma and non-small cell lung cancer cells. Clin Cancer Res 2005;11:6924-6932. 225. Song L, Morris M, Bagui T, et al. Da satinib (BMS-354825) selectively induces apoptosis in lung cancer cells dependent on epidermal growth factor receptor signaling for survival. Cancer Res 2006;66:5542-5548. 226. Vigneron A, Roninson IB, Gamelin E, Coqueret O. Src inhibits adriamycininduced senescence and G2 checkpoint arrest by blocking the induction of p21waf1. Cancer Res 2005;65:8927-8935.
185 227. Diaz-Montero CM, Wygant JN, McInty re BW. PI3-K/Akt-mediated anoikis resistance of human osteosarcoma cells requires Src activation. Eur J Cancer 2006;42:1491-1500. 228. Alvarez RH, Kantarjian HM, Cortes JE. The role of Src in solid and hematologic malignancies: development of newgeneration Src inhibitors. Cancer 2006;107:1918-1929. 229. Trevino JG, Summy JM, Le sslie DP, et al. Inhibiti on of SRC expression and activity inhibits tumor progression a nd metastasis of human pancreatic adenocarcinoma cells in an orthot opic nude mouse model. Am J Pathol 2006;168:962-972. 230. Yeatman TJ. A renaissance for SRC. Nat Rev Cancer 2004;4:470-480. 231. Lombardo LJ, Lee FY, Chen P, et al Discovery of N-(2-chloro-6-methylphenyl)-2-(6-(4-(2-hydroxyet hyl)piperazin-1-yl)-2-methylpyrimidin-4ylamino)thiazole-5-carboxamide (BMS354825), a dual Src/Abl kinase inhibitor with potent antitumor activity in pr eclinical assays. J Med Chem 2004;47:66586661. 232. Van Gelder R, von Zastrow M, Yool A, et al. Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proc Natl Acad Sci U S A 1990;Mar;87:1663-1667. 233. Liu WM, Mei R, Di X, et al. Analysis of high density expression microarrays with signed-rank call algorithms. Bioinformatics 2002;18:1593-1599. 234. Tusher VG, Tibshirani R, Chu G. Signif icance analysis of mi croarrays applied to the ionizing radiation response. Pr oc Natl Acad Sci U S A 2001;98:5116-5121.
186 235. Eisen MB, Spellman PT, Brown PO, Botste in D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 1998;95:1486314868. 236. Saldanha AJ. Java Treeview--extensi ble visualization of microarray data. Bioinformatics 2004;20:3246-3248. 237. Akira S. Functional roles of STAT family proteins: lessons from knockout mice. Stem Cells 1999;17:138-146. 238. Dearolf CR. JAKs and STATs in inverteb rate model organisms. Cell Mol Life Sci 1999;55:1578-1584. 239. Hirano T, Ishihara K, Hibi M. Roles of STAT3 in mediating the cell growth, differentiation and survival signals relaye d through the IL-6 family of cytokine receptors. Oncogene 2000;19:2548-2556. 240. Levy DE. Physiological significance of STAT proteins: inve stigations through gene disruption in vivo. Cell Mol Life Sci 1999;55:1559-1567. 241. Mui AL. The role of STATs in prolifer ation, differentiati on, and apoptosis. Cell Mol Life Sci 1999;55:1547-1558. 242. Smithgall TE, Briggs SD, Schreiner S, et al. Control of myeloid differentiation and survival by Stats. Oncogene 2000;19:2612-2618. 243. Decker T, Kovarik P. Transcription fact or activity of STAT proteins: structural requirements and regulation by phosphorylati on and interacting proteins. Cell Mol Life Sci 1999;55:1535-1546. 244. Darnell JE, Jr. STATs and gene regulation. Science 1997;277:1630-1635.
187 245. Bowman T, Garcia R, Turkson J, Jo ve R. STATs in oncogenesis. Oncogene 2000;19:2474-2488. 246. Bowman T, Jove R. STAT Proteins and Cancer. Cancer Control 1999;6:615-619. 247. Catlett-Falcone R, Landowski TH, Oshiro MM, et al. Consti tutive activation of Stat3 signaling confers re sistance to apoptosis in human U266 myeloma cells. Immunity 1999;10:105-115. 248. Garcia R, Jove R. Activation of STAT transcription factors in oncogenic tyrosine kinase signaling. J Biomed Sci 1998;5:79-85. 249. Cao X, Tay A, Guy GR, Tan YH. Activati on and association of Stat3 with Src in v-Src-transformed cell lines. Mol Cell Biol 1996;16:1595-1603. 250. Chaturvedi P, Sharma S, Reddy EP. Ab rogation of interle ukin-3 dependence of myeloid cells by the v-src oncogene requires SH2 and SH3 domains which specify activation of STATs. Mol Cell Biol 1997;17:3295-3304. 251. Bromberg JF, Fan Z, Brown C, Mendelsohn J, Darnell JE, Jr. Epidermal growth factor-induced growth inhibition requir es Stat1 activation. Cell Growth Differ 1998;9:505-512. 252. Kraker AJ, Hartl BG, Amar AM, et al. Biochemical and cellular effects of c-Src kinase-selective pyrido[2, 3d]pyrimidine tyrosine kina se inhibitors. Biochem Pharmacol 2000;60:885-898. 253. Boschelli DH, Wang YD, Johnson S, et al. 7-Alkoxy-4-phenylamino-3quinolinecar-bonitriles as dual inhibitors of Src and Abl kinases. J Med Chem 2004;47:1599-1601.
188 254. Boschelli DH, Ye F, Wang YD, et al. Optimization of 4-phenylamino-3quinolinecarbonitriles as potent inhibitors of Src kinase activity. J Med Chem 2001;44:3965-3977. 255. Wyeth Research and Clinical Trials: Project Listing. 2007. http://www.wyeth.com/research/projects 256. Tremblay L, Hauck W, Aprikian AG, et al. Focal adhesi on kinase (pp125FAK) expression, activation and association with paxillin and p50CSK in human metastatic prostate carcinom a. Int J Cancer 1996;68:164-171. 257. McCormack SJ, Brazinski SE, Moore JL, Jr., Werness BA, Goldstein DJ. Activation of the focal adhesion kinase signal transduction pathway in cervical carcinoma cell lines and human genital epit helial cells immortalized with human papillomavirus type 18. Oncogene 1997;15:265-274. 258. Cance WG, Harris JE, Iacocca MV, et al. Immunohistochemical analyses of focal adhesion kinase expression in benign and malignant human breast and colon tissues: correlation with pr einvasive and invasive phe notypes. Clin Cancer Res 2000;6:2417-2423. 259. Lark AL, Livasy CA, Calvo B, et al. Ov erexpression of focal adhesion kinase in primary colorectal carcinomas a nd colorectal liver metastases: immunohistochemistry and real-time PCR analyses. Clin Cancer Res 2003;9:215222. 260. Yu H JR. The STATs of cancer -new mo lecular targets come of age. Nat Rev Cancer 2004;4:97-105.
189 261. McLean GW, Carragher NO, Avizienyte E, et al. The role of focal-adhesion kinase in cancer a new therapeu tic opportunity. Nat Rev Cancer 2005;5:505515. 262. Defilippi P, Di Stefano P, Cabodi S. p130Cas: a versatile scaffold in signaling networks. Trends Cell Biol 2006;16:257-263. 263. Hanks SK, Ryzhova L, Shin NY, Brabek J. Focal adhesion kinase signaling activities and their implications in the control of cell survival and motility. Front Biosci 2003;8:d982-996. 264. Schaller MD. Biochemical signals and bi ological responses e licited by the focal adhesion kinase. Biochim Biophys Acta 2001;1540:1-21. 265. Silva CM. Role of STATs as downstream signal transducers in Src family kinasemediated tumorigenesis. Oncogene 2004;23:8017-8023. 266. Oral BMS-354825 rescues Gleevec-resista nt CML. Cancer Biol Ther 2004;3:700. 267. Copland M, Hamilton A, Elrick LJ, et al. Dasatinib (BMS-354825) targets an earlier progenitor populati on than imatinib in pr imary CML but does not eliminate the quiescent fraction. Blood 2006;107:4532-4539. 268. Doggrell SA. BMS-354825: a novel drug with potential for the treatment of imatinib-resistant chronic myeloid leuk aemia. Expert Opin Investig Drugs 2005;14:89-91. 269. O'Hare T, Corbin AS, Druker BJ. Targeted CML therapy: controlling drug resistance, seeking cure. Cu rr Opin Genet Dev 2006;16:92-99. 270. Shah N. Loss of response to imatinib: mechanisms and management. Hematology (Am Soc Hematol Educ Program) 2005:183-187.
190 271. Shah NP, Lee FY, Luo R, et al. Dasatinib (BMS-354825) inhibits KITD816V, an imatinib-resistant activating mutation that triggers neoplastic growth in most patients with systemic ma stocytosis. Blood 2006;108:286-291. 272. Tokarski JS, Newitt JA, Chang CY, et al. The structure of Dasatinib (BMS354825) bound to activated ABL kinase domain elucidates its i nhibitory activity against imatinib-resistant ABL mutants. Cancer Res 2006;66:5790-5797. 273. Chang JC, Wooten EC, Tsimelzon A, et al. Gene expression profiling for the prediction of therapeutic response to do cetaxel in patients with breast cancer. Lancet 2003;362:362-369. 274. Sorlie T, Tibshirani R, Parker J, et al. Repeated observa tion of breast tumor subtypes in independent ge ne expression data sets. Proc Natl Acad Sci U S A 2003;100:8418-8423. 275. Sotiriou C, Neo SY, McShane LM, et al. Breast cancer classification and prognosis based on gene expression profile s from a population-based study. Proc Natl Acad Sci U S A 2003;100:10393-10398. 276. van 't Veer LJ, Dai H, van de Vijver MJ, et al. Expression profiling predicts outcome in breast cancer. Br east Cancer Res 2003;5:57-58. 277. Nielsen TO. Microarray analysis of sarcomas. Adv Anat Pathol 2006;13:166-173. 278. Dong Z, Bode AM. The role of histon e H3 phosphorylation (Ser10 and Ser28) in cell growth and cell transforma tion. Mol Carcinog 2006;45:416-421. 279. Ryu SW, Lee SJ, Park MY, et al. Fas-asso ciated factor 1, FAF1, is a member of Fas death-inducing signaling comp lex. J Biol Chem 2003;278:24003-24010.
191 280. Bjorling-Poulsen M, Seitz G, Guerra B, Issinger OG. The pro-apoptotic FASassociated factor 1 is specifically re duced in human gastric carcinomas. Int J Oncol 2003;23:1015-1023. 281. Perez-Moreno M, Fuchs E. Catenins: k eeping cells from getting their signals crossed. Dev Cell 2006;11:601-612. 282. Kobielak A, Fuchs E. Alpha-catenin: at the junction of interc ellular adhesion and actin dynamics. Nat Rev Mol Cell Biol 2004;5:614-625. 283. Fearon ER, Vogelstein B. A genetic m odel for colorectal tumorigenesis. Cell 1990;61:759-767. 284. Ewing CM, Ru N, Morton RA, et al. Chro mosome 5 suppresses tumorigenicity of PC3 prostate cancer cells: correlation w ith re-expression of alpha-catenin and restoration of E-cadherin func tion. Cancer Res 1995;55:4813-4817. 285. Van Aken E, De Wever O, Correia da Rocha AS, Mareel M. Defective Ecadherin/catenin complexes in human cancer. Virchows Arch 2001;439:725-751. 286. Hajra KM, Fearon ER. Cadherin and caten in alterations in human cancer. Genes Chromosomes Cancer 2002;34:255-268. 287. Saito T, Oda Y, Sakamoto A, et al. Prognostic value of the preserved expression of the E-cadherin and catenin families of adhesion molecules and of beta-catenin mutations in synovial sarc oma. J Pathol 2000;192:342-350. 288. Holder N, Klein R. Eph receptors and ephrins: effectors of morphogenesis. Development 1999;126:2033-2044.
192 289. Ogawa K, Pasqualini R, Lindberg RA et al. The ephrin-A1 ligand and its receptor, EphA2, are expressed during tumor neovascularization. Oncogene 2000;19:6043-6052. 290. Zhang L, Gao X, Wen J, Ning Y, Chen YG. Dapper 1 antagonizes Wnt signaling by promoting dishevelled degrada tion. J Biol Chem 2006;281:8607-8612. 291. Katoh M, Katoh M. Identification and characterization of human DAPPER1 and DAPPER2 genes in silico. Int J Oncol 2003;22:907-913.
About the Author Audrey Shor was born in Pennsylvania and lived most of her life in the Northeast. She received an academic scholarship to attend We st Chester University, Pennsylvania, where she obtained a Bachelors of Science in Ce llular and Molecu lar Biology as well as minors in Chemistry and Studio Arts in 2002. Audrey was among the first students to pursue the PhD PLUS program at the University of Sout h Florida. This pr ogram offered her the opportunity to obtain a Master of Public H ealth in Epidemiology from the College of Public Health in 2006 and a Doctoral Degree in Molecular Medicine from the College of Medicine in 2007. Audrey has completed her doctoral st udies under the guidance of Doctors Richard Jove and W. Jack Pledger. She has been fortunate enough to have the opportunity to present her research findings both nationally and internationally. This research was published in Ca ncer Research in March 2007.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001925983
007 cr mnu|||uuuuu
008 080205s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001906
Shor, Audrey Cathryn.
Src kinase inhibitors for the treatment of sarcomas :
b cellular and molecular mechanisms of action
h [electronic resource] /
by Audrey Cathryn Shor.
[Tampa, Fla] :
University of South Florida,
ABSTRACT: Sarcomas are rare mesenchymally-derived tumors with limited treatment options. Tyrosine kinases may serve as potential targets for sarcoma therapy because many are mutated or overexpressed in sarcomas and cell lines. One potential molecular target for sarcoma treatment is the Src tyrosine kinase. Three independently synthesized Src kinase inhibitors were evaluated in human sarcoma cell lines. Of the three, dasatinib, provided promising results as a potential sarcoma therapy. Until this study, dasatinib activity had not been characterized in sarcoma cells. Based on our previous findings of Src activation in human sarcomas, we evaluated the effects of dasatinib in twelve sarcoma cell lines. Dasatinib inhibited Src activity and downstream signaling at nanomolar concentrations. Inhibition of Src signaling was accompanied by blockade of cell migration and invasion. Moreover, apoptosis was induced in a subset of bone sarcomas at nanomolar concentrations of dasatinib.^ ^Inhibition of Src protein expression by siRNA also induced apoptosis, indicating that these bone sarcoma cell lines are dependent on Src activity for survival. These results demonstrate that dasatinib inhibits migration and invasion of diverse sarcoma cell types, and selectively blocks the survival of bone sarcoma cells. Therefore dasatinib may provide therapeutic benefit by preventing the growth and metastasis of sarcomas. Microarray analysis of the sarcoma cell lines lead to the identification of a molecular signature that successfully predicts response to dasatinib by induction of apoptosis. Components of this molecular signature are expressed in primary human sarcomas. Furthermore, expression of the molecular signature in sarcomas can be utilized to cluster tumors based on theoretical response to dasatinib.^ ^While the prediction of response in tumors is theoretical, there is encouraging evidence to support further endeavors into validating the potential of this molecular signature to predict response in patients.Together, these studies reveal that, in cell lines, both constitutive Src activation and the presence of a molecular signature that predicts response to dasatinib are important parameters to consider when selecting dasatinib as a treatment for. Furthermore, novel therapeutic approaches that inhibit Src signaling may selectively induce apoptosis in tumor cells and sensitize to chemotherapy those tumors that contain the relevant molecular signature.
Dissertation (Ph.D.)--University of South Florida, 2007.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 192 pages.
Adviser: Richard Jove, Ph.D.
Tyrosine kinase inhibitor.
Gene expression profile.
x Molecular Medicine
t USF Electronic Theses and Dissertations.