Cory Abate-Shen - US grants

The murine gene encoding Hox 7.1 is expressed early in the developing nervous system and shares sequence similarity with a Drosophila gene, muscle segment homeobox (msh), that functions in transcriptional regulation during Drosophila embryogenesis. These properties have led to the proposal that the protein product of Hox 7.1 (Hox 7.1) plays an integral role in directing the expression of target genes that specify neuronal phenotype. The studies outlined in this proposal will characterize the features of the Hox 7.1 polypeptide that are required for its selective interaction with regulatory elements of neuronal target genes and for transcriptional regulation. Firstly, the DNA binding properties of Hox 7.1 will be characterized using purified Hox 7.1 polypeptides. The data obtained from these analyses will represent an initial step toward elucidating the basis for the selective interaction of Hox 7.1 with DNA regulatory elements of neuronal target genes. Secondly, the transcriptional properties of Hox 7.1 will be characterized using in vitro and in vivo assay systems. The findings obtained from these studies will provide fundamental information regarding the role of Hox 7.1 in transcriptional regulation during murine neuronal development. Thirdly, polyclonal antisera specific for Hox 7.1 will be generated and these will be used to characterize the expression and biochemical properties of Hox 7.1 in the developing mouse nervous system. These studies will permit a systematic analysis of the specific brain regions and cell types that express Hox 7.1 during murine development, will define the subcellular localization of Hox 7.1 in these cells, and will identify biochemical properties of Hox 7.1, such as post-translational modifications, that are likely to have functional significance. Finally, the ultimate goal to define the function of Hox 7.1 in neurogenesis is to identify potential target genes it regulates during neuronal development; this will be the focus of the last section of this proposal. Together, the findings obtained from these studies will provide fundamental information regarding the role of Hox 7.1 in the regulation of gene expression in the developing murine CNS and will also provide general insight into the molecular mechanisms that control gene transcription during mammalian neurogenesis.

The studies described in this proposal will characterize the function of HoxA7 in transcriptional regulation during murine embryogenesis. Transcriptional regulatory proteins play a pivotal role in embryogenesis by controlling the expression of phenotypic target genes therefore their, characterization is crucial for understanding the molecular processes that control mammalian development. The hox genes, are master regulators of murine embryogenesis that encode putative transcriptional regulatory proteins. Many elegant biological studies have elucidated the fundamental role of hox genes during development by detailing their coordinate expression, essential function, and mode of regulation. However, many issues remained unanswered regarding the molecular mechanisms by which Hox proteins exert their action as transcriptional regulators during murine embryogenesis. The goal of this proposal is to characterize the transcriptional properties of a prototype member of the Hox family. HoxA7. HoxA7 provides an appropriate model for this analysis since it has served at a prototype for many of the original studies which have examined the expression and function of hox genes during development and since its primary sequence contains features that are reminiscent of other key transcriptional regulatory proteins. Our strategy will be to perform a detailed biochemical analysis couched within the framework of relevant biochemical systems. Our specific plans are: (I) We will characterize the molecular bases of DNA recognition by HoxA7. These studies address the mechanisms by which HoxA7 interacts selectively with DNA regulatory elements and will facilitate subsequent identification of target genes. (II) Our second goal is to characterize the transcriptional regulatory properties of HoxA7. These studies provide the framework for addressing the function of HoxA7 in the variety of distinct cell populations in which is functional during embryogenesis. (IIl) Protein factors that associate with HoxA7 will be identified and their contribution to functional specificity of HoxA7 will be evaluated. Identification of protein partners is an important step towards recapitulating functional HoxA7 protein complexes, and will provide new insight as to HoxA7 function. (IV). Finally. we will characterize the expression of HoxA7 in the developing embryo using specific polyclonal antisera that we have generated. This analysis will provide a biological context for evaluating the functional significance of the genomic target sequences, transcriptional properties, and interacting proteins factors defined in the other specific Aims. Together, these studies provide an integrated approach which will impart real insight into the function of a prototype Hox protein in transcriptional regulation during embryogenesis.

DESCRIPTION: The goal of the work in this proposal is a characterization of the Msx1 gene in the mouse. Msx1 encodes a homeodomain protein which when deleted causes cranial facial defects in mice. During the past funding period, Abate identified a DNA binding consensus for Msx1 and produced convincing data that the gene functions as a repressor of transcription in vitro. Moreover, she showed that this repressor activity does not require Msx=s binding to its target DNA, although it does require an intact Msx homeodomain. Abate argues that Msx's transcriptional specificity reflects an interaction with other proteins in a large Msx complex and has identified several possible candidates. The future analysis of these interacting proteins and the resultant specificity of the complex provide the basis for the proposed experiments in this application. To evaluate the in vivo significance of newly identified potential candidates she has established cell lines in which Msx can be induced under TET control and its effects assayed by activation of MyoD. As a first step in developing a biologically relevant assay for Msx function, she has used the Tabin retroviral expression systems to overexpress Msx in the Chick limb. Overexpression at stage 10 causes wide spread defects in limb development. More useful for the present proposal is her observation that injection at stage 17 blocks differentiation of feather follicles. Abate proposes to develop this overexpression phenotype as a quantitative assay for function.

DESCRIPTION (adapted from investigator's abstract): Inductive signaling interactions between epithelial and mesenchymal cell layers represent one of the principal mechanisms of vertebrate organogenesis. In particular, the prostate arises late in gestation when signals from urogenital sinus mesenchyme induces the urogenital sinus epithelium to form the prostatic buds. Under the influences of androgens, these epithelia buds elongate and undergo ductal morphogenesis, thereby forming the ventral, dorsolateral, and anterior (coagulating gland) lobes of the rodent prostate. The prostate provides an excellent model system for studying inductive epithelial-mesenchymal interactions, the role of androgen signaling in the generation of sexual dimorphism, and the molecular mechanisms of ductal morphogenesis. Moreover, the analysis of prostate development is likely to provide insights into human disease, notably benign prostatic hyperplasia (BPH) and prostate carcinoma. The Principal Investigator has identified a novel homeobox gene, termed Nkx3.1 that appears to play an important regulatory role in prostate development. Nkx3.1 is one of the few transcriptional regulators known to be expressed specifically in the prostate during development and adulthood. Importantly, the preliminary analysis of Nkx3.1 null phenotype has demonstrated that Nkx3.1 is required for normal prostate development and function. The regulatory role of Nkx3.1 in prostate development will be investigated by performing a detailed analysis of its biochemical properties, expression pattern, cellular activities, and biological functions as follows: 1) the biochemical properties of the Nkx3.1 protein will be analyzed by examining its requirements for DNA recognition, its transcriptional properties, and its post-translational modification by phosphorylation; 2) the expression pattern and cellular functions of Nkx3.1 will be analyzed by examining the distribution of murine Nkx.3.1 transcripts and protein in the male urogenital system, and investigating the effects of Nkx3.1 on cell proliferation and transformation; and 3) the biological role of Nkx3.1 in development will be analyzed by investigating its role in the male urogenital system, by examining its cell autonomy, and its over-expression in reconstituted prostatic tissue.

DESCRIPTION (provided by applicant): We have assembled a team of highly talented individuals who collectively bring to the MMHCC their extensive expertise and/or unique resources for studying prostate cancer in mice, including: (i) generating and analyzing mutant mouse models (Michael Shen), (ii) non-invasive imaging to visualize tumors and metastases (Simon Cherry); (iii) comparative histopathology with human cancers (Robert Cardiff); (iv) analyzing cancer phenotypes (Cory Abate-Shen and Gerald Cunha); (v) functional genomics and proteomics (Peter Nelson); and (vi) translational research (William Nelson). Building upon our expertise and leveraging our previous studies that have generated mouse models that recapitulate stages of human prostate carcinogenesis, we are well-poised to implement an ambitious research program that should yield fundamental insights into basic molecular mechanisms of carcinogenesis and will have considerable clinical relevance. We propose the following projects: Project 1 is designed to produce a "next generation" of mouse models, focusing on advanced stages of prostate cancer. These studies will generate mouse models that combine temporal inducibility with indelible linage marking and multimodal imaging reporters to allow non-invasive whole-animal imaging of metastatic disease. Project 2 will identify novel components of the molecular pathways involved in prostate carcinogenesis, using functional generates and proteomics approaches, and may also lead to the generation of new serological tests for early detection of prostate cancer. Project 3 will address the significance of hormonal signaling for prostate carcinogenesis, and may provide insights into the efficacy and mechanisms of synthetic anti-androgens as chemopreventive agents. Project 4 will develop mouse models to examine the role of oxidative stress in prostate carcinogenesis, and will utilize these mice to test the efficacy of dietary anti-oxidants in a "pre-clinical" trial of chemoprevention. These studies will yield mouse models for longitudinal visualization of metastatic spread in vice, for examination of the temporal requirements of tumor suppressor function, and for investigation of chemoprevention; we will also exploit these mouse models to define prognostic indicators of cancer, and to address other clinically relevant issues that would be intractable using human subjects. Thus, we anticipate that our studies of prostate carcinogenesis in mice should also lead to new general paradigms and/or experimental model systems that arc broadly relevant for understanding human cancer biology.

DESCRIPTION (provided by applicant): We have been investigating the relationship between the molecular pathways that underlie embryogenesis and those that are deregulated in carcinogenesis using the mammalian prostate as a model system. Our studies of the Nkx3.1 homeobox have elucidated its functions for development and carcinogenesis of the prostate. We have found that Nkx3.1 mutant mice display defects in prostatic epithelial differentiation and are predisposed to prostate cancer initiation. Moreover, loss-of-function of Nkx3.1 collaborates with that of the Pten and p27 Kip_ tumor suppressor genes in prostate carcinoma in compound mutant mice (i.e., Nkx3.1+/-; Pten+/-; p27+/-). Finally, epigenetic loss of Nkx3.1 protein (but not its mRNA) is a hallmark of prostate carcinogenesis, and loss of its expression is accompanied by upregulation of androgen receptor. Thus, we have hypothesized that Nkx3.1 represents a key regulator of prostatic epithelial differentiation as well as a modulator of androgen signaling. We further hypothesize that cancer susceptibility due to Nkx3.1 loss-of-function is a consequence of defects in prostatic epithelial differentiation and androgen signaling. We will now investigate: (Aim 1) The functions of Nkx3.1 in specification of prostatic epithelial differentiation using a tissue recombination approach, which provides a model system for studying epithelial and mesenchymal tissue requirements for prostatic growth and carcinogenesis. (Aim 2) The relationship of Nkx3.1 and androgen receptor for prostate differentiation and carcinogenesis by investigating: (i) the coordinate expression of Nkx3.1 and androgen receptor in the prostate; (ii) the role of the androgen receptor as a downstream mediator of Nkx3.1 function in prostate differentiation and carcinogenesis; and (iii) the consequences of altered androgen signaling for prostate carcinogenesis in mutant mouse models. (Aim 3): The mechanisms underlying the cooperativity and tissue-specificity of loss-of- function of Nkx3.1, Pten, and p27 kip1 prostate carcinogenesis, by studying the contributions of candidate effectors as well as pursuing exploratory approaches to identify novel mediators of Nkx3. I, Pten, and p27 kip1 cooperativity. (Aim 4) The regulation of Nkx3.1 protein expression in prostate development and cancer, by defining the regulatory sequences controlling Nkx3.1 protein expression, with the ultimate goal of developing a mouse model that lacks this mode of regulating Nkx3.1 protein expression. These mice will provide a valuable model for exploring the reversibility of prostate carcinogenesis in vivo.

DESCRIPTION (provided by applicant): Msx homeobox genes provide an ideal model system to address the broad issue of how homeoproteins achieve functional specificity in vivo, since they have been particularly well studied with regard to their expression patterns, biological functions, and biochemical properties. Our extensive analyses of the biochemical properties of Msx homeoproteins have established their functions as transcriptional repressors and have determined that their transcriptional activities are mediated through selective protein-protein interactions. In recent studies, we have been utilizing "discovery" approaches to identify physiologically relevant protein partners for Msx1, as well as downstream target genes. We have established: (i) a yeast two-hybrid screen to isolate Msx1-interacting proteins, (ii) a co-immunoprecipitation approach to isolate Msx1-protein complexes from living cells, and (iii) a microarray approach to identify differentially regulated genes. These strategies have led to three key observations: (i) that the activity of Msx1 is regulated by post-translational modification by sumoylation; (ii) that histone H1, a transcriptional repressor, interacts with Msx1 in living cells; and (iii) that genes which are differentially regulated by Msx1 include those involved in cellular differentiation. These "discovery" approaches serve as the basis for our proposed hypothesis-driven experiments, in which we will address the complex issue of how Msx homeoproteins achieve functional specificity in vivo by integrating a variety of biochemical and biological approaches. Our hypothesis is that the functional specificity of Msx homeoproteins in vivo represents the outcome of a complex interplay of several components, including post-translational modifications, protein-protein interactions, and cellular context. Thus, in Specific Aim 1, we will investigate the functional consequences of post-translational modification by sumoylation through our analysis of the biochemical properties, sub-cellular distribution, and biological activities of the sumoylated and non-sumoylated forms of Msx1. In Specific Aim 2, we will analyze protein partners for Msx1 in living cells by investigating the biochemical and biological consequences of Msx1-histone H1 interaction, focusing on their mutual ability to repress target genes, such as MyoD. These studies will provide a framework for our subsequent analysis of endogenous Msx1-protein complexes in living cells and embryos, which will be pursued using an epitope-tagged Msx1 "knock-in" allele. In Specific Aim 3, we will use a tamoxifen-inducible expression system (Msx1-ERtm) to identify direct target genes for Msx1, which will facilitate the subsequent identification of physiologically relevant "Msx1-responsive elements." Finally, in Specific Aim 4, we will analyze the limb phenotype of conventional and conditional mutant mice to address the biological functions of Msx genes in the limb. Taken together, our focused analyses of Msx homeoproteins will provide general paradigms for addressing the functional specificities of other homeoproteins in vivo.

DESCRIPTION (provided by applicant): The prostate is critically dependent on androgen receptor (AR) signaling for all aspects of its growth, as well as all stages of carcinogenesis. Indeed, androgen-deprivation therapy remains the most prevalent treatment for men with advanced prostate cancer. However, androgen-deprivation ultimately results in the emergence of "androgen-independent" (or hormone-refractory) disease, which counter-intuitively continues to rely on AR signaling despite the fact that androgens are absent or at least limiting. We have been investigating androgen-independence in a mutant mouse model based on the loss-of-function of two genes that are important for human prostate cancer, namely, the Nkx3.1 homeobox gene and the Pten tumor suppressor. These Nkx3.1;Pten mutant mice recapitulate stages of prostate tumorigenesis as occur in humans, including PIN, adenocarcinoma, and metastatic disease and the molecular pathways of cancer progression are partially conserved with the human disease. Most relevant for this proposal, androgen-deprivation of these mutant mice leads to androgen-independent phenotypes, which are evident from the survival and proliferation of the prostatic epithelium and the occurrence of high-grade PIN and/or cancer lesions following castration. Importantly, androgen-independent lesions of the Nkx3.1;Pten mutant mice have wild-type AR and require AR signaling, as is frequently the case for hormone refractory prostate cancer in humans. We propose to utilize these Nkx3.1;Pten mutant mice and their derivative cell lines to pursue the following Aims: Aim 1: Based on our previous findings showing that constitutive activation of Akt and Erk MAP kinase signaling act additively in culture and synergistically in vivo to promote androgen-independence, we will investigate the individual and collaborative roles of the Akt and B-Raf/Erk signaling pathways for androgen-independence by integrating studies in established cell lines and primary prostate tissues in culture and in vivo, and with validation to human prostate cancer. Aim 2: Based on our gene expression profiling analyses of cancer progression in Nkx3.1;Pten mutant mice, we will identify novel genes that are de-regulated in the mutant mice and investigate their relevance for hormone refractory prostate cancer in humans. Aim 3: Based on our recent generation of an inducible model system to achieve stochastic gene deletion specifically in the prostate epithelium, we will develop a "next-generation" mutant mouse model to investigate the evolution of androgen-independence and relevant signaling pathways in vivo. Relevance: Currently, patients with advanced prostate cancer have limited treatment options, the most effective being androgen-ablation therapy. However, this is only a temporary solution since hormone deprivation inevitably leads to the emergence of androgen-independent disease, which is highly aggressive, metastatic, and ultimately lethal. Therefore, it is of the utmost clinical significance to understand the evolution of androgen-independent prostate cancer and to delineate the underlying molecular pathways.

DESCRIPTION (provided by applicant): This Leadership Application in Chemoprevention complements my U01 Research application that has been submitted for its third renewal to the MMHCC. My proposed U01 research will integrate analyses of our unique mouse models of prostate cancer with analyses of human prostate cancer using state-of-the-art systems biology approaches to define an interactome of mouse and human prostate cancer; the ultimate goal will be to identify new druggable targets for advanced prostate cancer and to evaluate the efficacy of such targets for therapeutic intervention. Although my U01 MMHCC research application is focused on cancer progression and advanced disease, I am highly committed to the effective utilization of mouse models for studying cancer initiation, early detection, and prevention. Indeed, despite many recent advances in the design and demonstrated relevance of GEM models for human cancer, GEM models are vastly underutilized in cancer prevention research. In my view, this reflects the need for improved communication among investigators across many disciplines, and particularly in the areas of mouse modeling and cancer prevention, as well as greater clarity in defining prevention research and establishing realistic goals, expectations, and predicted outcomes for using GEM models in chemoprevention. Accordingly, I along with Dr. Bill Nelson have assembled an outstanding team of investigators (Drs. Brown, Chodosh, Colburn, Hanahan, Hawk, Kohl, Neugut, Pandolfi, Tuveson, Van Dylce) with diverse interests and expertise from within the MMHCC and the broader community who will serve as a Scientific advisory team. Through ongoing discussions, the team will to create a dialogue among leaders from relevant disciplines with the ultimate goal of establishing a common language and clearly elucidating research objective and outcomes. In the course of such discussion, we will define specific applications for GEM models in prevention research, taking into consideration their potential benefits as well as a realistic understanding of their limitations. Co-chaired by Bill Nelson and myself, the Advisory team will: (i) establish a common language to apply prevention research to GEM models, (ii) define state-of-the-art strategies for cancer prevention and early intervention that can be informed by GEM models; (iii) explicate potential applications of GEM mice for prevention and intervention; and (iv) validate the use GEM models for cancer prevention in well-defined demonstration trials. Finally, I believe that I have demonstrated the necessary leadership skills to coalesce multi-disciplinary teams of investigators to effectively address these complex issues and I am confident that the plan I have outlined will have a high impact on our ability to effectively engage mouse models research for chemoprevention. GEM models provide a valuable but largely untapped resource for informing on cancer prevention; however their efficacy is entirely dependent on the appropriate design of experimental paradigms, the choice of relevant GEM models to address specific experimental questions, and the effective validation of studies in GEM models to human cancer. I understand these problems and I am committed to resolving them to ensure the effective use of mouse models for chemoprevention.

DESCRIPTION (provided by applicant): This application represents a continuation of our efforts to utilize mouse models to understand the molecular mechanisms of human prostate cancer, as part of the NCI Mouse Models of Human Cancer Consortium. Our team of investigators includes long-standing members of our previous group (Drs. Cory Abate-Shen and Michael Shen), as well as new members (Drs. Andrea Califano and Carlos Cordon-Cardo) who will bring essential expertise in two key areas, systems/computational biology and systems pathology. At the heart of our program is a series of genetically-engineered mouse models of prostate cancer that have been generated and characterized to identify new biomarkers for biochemical recurrence of human prostate cancer, to characterize key signaling pathways responsible for hormone-refractory prostate cancer, and to pursue pre-clinical studies. In the current application, our objective is to use reverse-engineering of regulatory programs in mouse and human prostate cancer to develop accurate molecular interaction maps for human (human Prostate Cancer interactome, hPCi) and mouse (mPCi). We will employ innovative bioinformatics tools that will identify the drivers that are causally responsible for tumor progression, as distinct from other discovery approaches that identify passenger genes whose expression is simply correlated with a particular cancer phenotype. Thus, the human and mouse PCis will be interrogated using systems biology tools to elucidate key regulators of the genetic pathways that are dysregulated in cancer (Specific Aim 1) and to identify druggable targets that can affect such altered pathways (Specific Aim 2). Candidate key regulatory genes identified by these analyses will be biochemically and functionally validated for their causal role in tumorigenesis and tumor progression, and their relevance for human prostate cancer will be validated using systems pathology approaches. Finally, lead candidate druggable targets will be evaluated as potential therapeutic targets in pre-clinical trial programs, using our in vivo mouse models.

The Nkx3. 1 homeobox gene is a key regulator of prostate epithelial differentiation whose loss-of-function represents an enabling event in prostate cancer initiation. Our proposed studies are based on the hypothesis that the molecular processes by which NkxS. 1 regulates prostate epithelial differentiation are causally linked to its role in cancer initiation, while its consequences for malignancy are limited by cellular senescence. We will utilize state-of-the-art molecular approaches applied to in vivo mouse models, organ transplant models, and cell-based approaches, and combined with validation to human prostate cancer to Identify key molecular pathways involved in prostate cancer initiation. Specifically: In Aim 1, we will investigate the relationship between cellular senescence and cancer initiation. Based on our observation that Nkx3.1 mutant mice display senescence coincident with the occurrence of PIN (Preliminary Data), we will investigate the relationship of the senescence phenotype to cell type, and prostate stem cells. We will evaluate the functions of senescence modulators in renal graft assays and validate their relevance to human prostate cancer with assistance of Core A. In Aim 2, we will investigate the relationship between cellular specification and cancer initiation. Based on our finding that gain-of-function of Nkx3.1 in nonprostatic epithelium is sufficient to induce prostate growth in vivo (Preliminary Data), we will now investigate the mechanisms by which Nkx3.1 induces prostate specification and their relationship to cancer initiation. To do so, we will use gene expression profiling followed by functional analyses of selected genes, which will be prioritized in conjunction with the bioinformatic component of Core B and validated to human prostate cancer with Core A. In Aim 3, we will elucidate a transcriptional program of cancer initiation. Having identified putative target genes that are relevant for cancer initiation and are both regulated by (i.e., from gene expression profiling) and bound (i.e., from chromatin immunoprecipitation, ChIP) by Nkx3.1 (Preliminary data), we will now pursue a comprehensive analysis of Nkx3.1 target genes using gene expression profiling combined with high-throughput genome-wide sequencing of ChlP-enriched DNA fragments (ChlP-Seq). Candidate target genes will be prioritized with assistance from the bioinformatic component of Core B and based on their expression in human prostate cancer with Core A. In addition to essential interactions with Cores A and B, our studies will benefit from interactions with Michael Shen (Project 1) to evaluate the molecular properties of senescence in the context of prostate stem cells, and with Edward Gelmann (Project 3), who will provide an in-depth evaluation of the consequences of Nkx3.1 for DNA damage.

Project Summary/Abstract Most prostate cancer deaths are due to failed treatment response or to metastasis, particularly to bone. We have been studying the mechanisms responsible for treatment failure using genetically-engineered mouse models (GEMMs) that recapitulate key features of advanced prostate cancer, including castration-resistant prostate cancer (CRPC) and aggressive-variant CRPC with neuroendocrine differentiation (CRPC-NE), and that develop highly penetrant metastatic prostate cancer including to bone, which is the primary site of metastasis in humans. Using novel cross-species computational approaches, we have shown that the mechanisms underlying disease progression in these GEMMs are conserved with human prostate cancer, while drug response in GEMMs can be predictive of drug response in humans. Among our major findings, we have shown that GEMMs based on combined loss-of-function of Pten and p53 (NPp53), which are frequently co-mutated in human CRPC, model key phenotypic and molecular features of treatment-induced aggressive variant CRPC. Not only do these NPp53 mice fail to respond to treatment with anti-androgens, treatment actually accelerates disease progression, which we have called exceptional non-responders. Furthermore, we have shown that the treatment-induced neuroendocrine phenotype (CRPC-NE) of these exceptional non-responders is related to lineage plasticity. Thus, our proposed studies will test the hypotheses that: (i) co-clinical analyses of GEMMs and human prostate cancer can elucidate biological and molecular mechanisms of drug response, and (ii) lineage plasticity is an important novel mechanism of drug resistance. In Aim 1 we will perform co-clinical investigations capitalizing on our GEMMs that model key aspects of advanced prostate cancer, and complemented with analyses of human prostate cancer organoid models. We will evaluate the efficacy of clinically-relevant drugs and drug combinations, focusing on those that: (a) counteract cellular plasticity associated with treatment resistance in CRPC; and (b) target bone metastasis. In Aim 2, we will investigate molecular mechanisms of drug response leveraging our genome-wide regulatory networks that enable cross-species integration between data from GEMMs and human prostate cancer. We will focus on identifying ?treatment response regulators? that inform on response to: (i) lineage plasticity and/or (ii) bone metastasis. In Aim 3 we will identify novel drivers of advanced prostate cancer using a forward genetic screening approach. Toward this end, we have undertaken a genetic screen utilizing the Sleeping Beauty (SB) murine transposon-based system, and have shown that mice harboring the activated transposon display accelerated lethal prostate cancer phenotypes. Taken together, the successful implementation of these Aims will identify tumor contexts that are responsive to drug treatment and molecular mechanisms of drug response, and will identify new targets for intervention.

? DESCRIPTION (provided by applicant): Muscle invasive bladder cancer (MIBC) affects approximately 20,000 patients each year in the United States alone, and is associated with significant morbidity and mortality. Yet, despite its significance, it has been relatively neglecte in basic cancer research, and as a consequence, many salient issues regarding its molecular and cellular origins have not been fully elucidated. In particular, the cell(s) of origin of bladde cancer and consequences of specific oncogene/tumor suppressor alterations in these cells of origin have not been systematically investigated. To a large extent, this reflects the limitations f current model systems for pursuing such studies. In the current proposal, we describe innovative genetically-engineered mouse models for investigating the cell type(s) of origin of bladder cancer. In particular, we have established an approach to achieve inducible gene targeting in specific cell types in the bladder urothelium, which will enable our studies of the cel of origin in vivo as well as their isolation for tumor engraftment assays and explant approaches. In addition, we have established novel culture methods to derive patient-derived organoids to model human bladder cancer in explant culture, which will provide a vital resource for extrapolating our findings from mouse to man using integrative systems biology approaches. Thus, we are ideally poised to investigate the hypothesis that the cell type of origin for bladder cancer, combined with specific oncogene/tumor suppressor alterations, can generate tumor subtypes that are associated with distinct treatment responses and/or patient outcomes. Toward this end, we will pursue the following Specific Aims: In Aim 1, we will systematically analyze the tumorigenic potential of urothelial cell types in GEM models in vivo, and will perform cross-species systems analyses of these findings compared with patient-derived human organoids. In Aim 2, we will investigate whether tumorigenic properties of potential cell type(s) of origin are defined by specific oncogene/tumor suppressor alterations. In particular, we will evaluate the oncogenic activities of ERBB2 (HER2/Neu) and the tumor suppressor activity of TSC1, both of which are altered in MIBC, in order to investigate their functional relevance in orthotopic models and ultimately in GEM models. In Aim 3, we will perform preclinical studies to determine whether distinct cell types of origin help define treatment response, by focusing on chemotherapy agents that are widely used in the clinic, as well as on agents that target PI3 kinase, which is highly relevant for bladder cancer. Thus, the studies outlined in this proposal will provide a comprehensive analysis of the cell types of origin of bladder cancer, their roles in defining bladder cancer subtypes, and their relevance for treatment response.

DESCRIPTION (provided by applicant): Although most prostate cancer deaths are due to metastasis, our understanding of the biological and molecular mechanisms underlying metastatic prostate cancer is still limited. Our proposed research is focused on elucidating molecular mechanisms involved in tumor progression, dissemination, and metastasis of prostate cancer using analyses of in vivo mouse models to inform on human prostate cancer. Thus, we have generated genetically-engineered mouse (GEM) models of metastatic prostate cancer. Using lineage-tracing of these models, we have defined a temporal progression from tumorigenesis to metastasis, and have isolated primary tumor, disseminated, and metastatic cells for molecular analyses. In parallel, we have established state-of-the- art systems biology approaches to accurately integrate molecular data from mouse models to human cancer. By using these advanced systems approaches to interrogate our GEM models, we have identified transcriptional regulatory and epigenetic control mechanisms associated with metastasis progression. In particular, we have shown that a transcription factor pair, namely FOXM1 and CENPF, are master regulators of prostate cancer malignancy. We further demonstrate that the ETS transcription factor, ETV4, is required for metastasis. Moreover, we have found that among regulatory pathways that distinguish metastases from primary tumors are those that are associated with epigenetic modifications, such as the histone methyltransferase, SUV39H1. Thus, we have hypothesized that the molecular cascade associated with metastasis progression includes transcriptional regulatory mechanisms, which drive tumor aggressiveness and promote their metastasis, as well as epigenetic regulators, which are associated with formation of metastases. We will investigate these mechanisms as follows: In Aim 1, we will investigate transcriptional mechanisms associated with metastasis progression, focusing on FOXM1-CENPF and ETV4, to delineate their individual functions and potential cross-talk, as well as to elucidate their mechanism(s) of action. In Aim 2, we will investigate epigenetic regulatory mechanisms that promote metastasis, focusing on histone modifications and particularly the histone-modifying enzyme SUV39H1, to delineate its functions and mechanism(s) of action. In Aim 3, we will pursue discovery-based systems analyses to investigate key molecular aspects of metastasis progression, in particular, to identify mechanisms associated with organ specificity of metastases and molecular mechanisms associated with clonal relationship of tumors and metastases.

Project Summary Our proposed studies will address major clinical challenges associated with differential drug treatment response, focusing on advanced prostate cancer. Our overall approach is based upon the co-clinical trial paradigm, in which genetically-engineered mouse (GEM) models are assayed for drug response to provide information that can be incorporated into patient treatment regimens. Here, we propose a novel augmented co- clinical paradigm that uses analyses of tissue-specific organoids together with GEM models to expedite investigation of differential drug response and drug synergy, as well as sophisticated computational systems biology approaches to identify molecular regulators of drug response that are conserved from mouse models to human cancer. In preliminary studies, we have established methods for novel three-dimensional culture of tumor organoids, which display drug responses characteristic of the GEM models from which they were established. Furthermore, we have used computational systems methods to generate gene regulatory networks (interactomes) for both mouse and human prostate cancer, and have demonstrated their utility for cross-species identification of candidate master regulators of tumor aggressiveness. We have expanded these systems biology approaches for prediction of drug response in preclinical studies and to extrapolate these data to human prostate cancer. Based on these preliminary studies, we hypothesize that systematic analysis of drug response in genetically-engineered mouse (GEM) models followed by cross-species systems analyses can inform human cancer treatment by enabling the systematic evaluation of optimal drug treatments in distinct tumor contexts as well as by identifying patients who are most likely to respond to treatment. Thus, our proposed research will pursue the broad objective of identifying the underlying mechanisms of differential drug response in distinct prostate tumor contexts. Our specific plans are: Aim 1: To identify optimal drug treatments for specific tumor contexts, we will use organoid lines derived from a series of GEM models of prostate cancer to assay response to a range of drugs, including agents currently used for treatment of advanced prostate cancer. These findings from organoid models will be experimentally validated in GEM models in vivo. Aim 2: To analyze molecular mechanisms of drug response and drug synergy, we will use cross-species computational systems approaches to identify genes and pathways that regulate drug response in human prostate cancer. These results will be experimentally validated in organoid, GEM, and xenograft models. Impact: Our proposed studies directly address the broad goals of the Oncology Models Forum, since they introduce new experimental paradigms for effective translation of mouse models to achieve unmet translational needs. Our newly developed organoid models and cross-species computational analyses and validation will be of broad value to members of the Oncology Forum and are sharable through the NCIP Hub.

Project Summary/Abstract: This new Program Project will investigate the molecular mechanisms underlying the pathogenesis of bladder cancer. Our objectives are: (i) to study clonal evolution from non-muscle invasive to invasive to metastatic disease, and to elucidate molecular drivers for each stage of evolution; (ii) to study mechanisms of disease pathogenesis, with a major focus on the functional role of mutations that affect the epigenome; (iii) to elucidate mechanisms of disease response and resistance, with a major focus on understanding those that affect response to chemotherapy; and (iv) to generate novel human patient-derived and genetically-engineered mouse models (GEMMs) to study disease pathogenesis and to pursue co-clinical investigations. To achieve these objectives, we have assembled a highly accomplished multi-disciplinary team at Columbia University Medical Center (Cory Abate-Shen, Michael Shen, James McKiernan, Tian Zheng) and Memorial Sloan Kettering Cancer Center (David B. Solit, Hikmat Al-Ahmadie, Barry Taylor) with complementary expertise in genomic analyses (DBS, HA, BT), molecular investigations (CAS, MS, DBS), treatment response (CAS, MS, DBS, JM, BT), biostatistics and bioinformatics (TZ, BT), molecular pathology (HAA), generation of human and mouse cancer models (CAS, MS), co-clinical and clinical investigations (DBS, JM, CAS, JM, MS). We will pursue three interrelated Projects that are supported by three Cores. Project 1, led by David Solit, will seek to define the biologic functions of the histone demethylase, KDM6A, in bladder cancer pathogenesis, as well as the temporal occurrence of mutations in this gene during tumor progression. Project 2, led by Cory Abate-Shen, will analyze the functions of the chromatin remodeling gene, ARID1A, in muscle- invasive bladder cancer, focusing on its role in disease pathogenesis, its effect on treatment response and its molecular mechanisms of action. Project 3, led by Michael Shen, will investigate tumor evolution and drug response in human patient-derived bladder cancer organoids, focusing on genetic determinants of genomic instability, alterations of the epigenome, and the role of heterogeneity in drug sensitivity and resistance. The Molecular Pathology Core (Core A), led by Hikmat Al-Ahmadie, will maintain a biorepository of urothelial cancer tumors, including the tissues used for organoid generation, and will provide histopathological analyses for all three projects. The Bladder Cancer Models Core (Core B), led by Michael Shen will serve as a central hub for the generation and analysis of cancer models for all three projects, including human patient-derived organoid and xenograft models, and GEMMs. The Administrative Core (Core C), led by Cory Abate-Shen, will provide support for bioinformatic and biostatistical analyses and organizational support for Program Project investigators, and will coordinate with the Internal and External Advisory Boards. In summary, this Program Project is led by a highly collaborative team of investigators with diverse but complementary expertise who have substantial track records in cancer research, and are located in close proximity in New York City.

Project Summary/Abstract: The Administrative Core will provide coordination and oversight of all activities of the Projects and Cores in this Program, and will maximize interactions and potential synergies among Project Leaders, Core Leaders, and laboratory personnel at the two participating institutions, namely, Columbia University Medical Center (CUMC) and Memorial Sloan Kettering Cancer Center (MSKCC). The major scientific functions of the Core will be to provide expert bioinformatic and biostatistical support for the Projects and other Cores. In addition, to facilitate program integration and effective communication among its various members, the Core will organize monthly meetings composed of two parts: (i) Research presentations, which will include all program participants, as well as relevant Internal Advisory Board members; and (ii) Executive meetings, which will be limited to the Executive Committee (Drs. Abate-Shen, Shen, Solit, and Al-Ahmadie) and members of the Internal Advisory Board. The latter meetings will provide a forum for candid assessment of research progress, oversight of Core utilization, prioritization of resources and objectives, and resolution of potential conflicts. An important function of the Core will be to provide logistical support through management of data and resources, fiscal review and budget analysis, preparation of progress reports, and regulatory compliance. Finally, the Core will organize an annual full-day review of the Program Project by members of the Internal and External Advisory Boards, who will provide detailed feedback on research progress and future directions, as well as advice on program management. This Core will have the following specific aims: (1) To provide bioinformatic and (2) biostatistical support for the Projects, which will be led by co-Investigators, Drs. Barry Taylor (MSKCC) and Tian Zheng (CUMC), respectively. (3) Program management, which will include: (a) Coordination of the overall program by maintaining the organization of the program and establishing mechanisms for effective interaction among its personnel; (b) Data and resource management by promoting effective utilization of Cores A and B, supporting access to secure storage of research presentations and databases, and facilitating data and resource sharing with external laboratories; (c) Review of the Program Project by Internal and External Advisory Boards to evaluate ongoing progress of each Project, review the functionality of the Cores, and provide advice on future research directions. Taken together, the operations of the Administrative Core will ensure the smooth and efficient functioning of the overall Program, as well as planning and oversight for its operations, and will promote the sharing of data and resources within the Program Project as well as with external investigators.

Project Summary/Abstract Although the chromatin remodeling gene ARID1A is mutated in ~27% of muscle invasive bladder cancer (MIBC), the functional consequences have not been extensively studied. We propose to investigate the role of ARID1A in bladder cancer pathogenesis and treatment response using cell-based models, human patient-derived organoids, and genetically-engineered mouse models (GEMMs) that we have generated and characterized in collaboration with colleagues on this Program Project. Our studies are based on significant preliminary data showing that ARID1A has tumor suppressor functions in human bladder cancer cells as well as in mouse models. In particular, loss-of-function of Arid1a in a GEMM of bladder cancer accelerates tumorigenesis, in part through activation of PI-3 kinase signaling. Additionally, we have generated human patient-derived bladder cancer organoids that have ARID1A mutations as occur in human bladder cancer, and have shown that these organoids have reduced response to chemotherapy in cell culture. Lastly, using cell- based models, we have shown that ARID1A interacts with components of the SWI/SNF chromatin complexes in bladder cells and is necessary for formation of these complexes, providing a foundation for molecular investigations of ARID1A in bladder cancer contexts. We will now investigate the hypothesis that ARID1A deficiency promotes bladder cancer pathogenesis by altering chromatin structure and global gene expression, thereby affecting treatment response. In Aim 1, we will investigate the functions of ARID1A in bladder tumorigenesis, and its collaboration with other relevant epigenetic regulators and tumor suppressors, by studying the consequences of its loss-of-function in GEMMs, as well as in human bladder cancer cell-based models. We will augment these studies with analyses of human patient-derived organoids that harbor ARID1A mutations. In Aim 2, we will perform co-clinical analyses to investigate the consequences of ARID1A inactivation for response to chemotherapy, and to test whether such response can be improved by combining chemotherapy with targeted agents or immunotherapy. In Aim 3, we will study the molecular mechanisms by which ARID1A deficiency promotes bladder cancer, by performing biochemical analyses in bladder cancer cells and investigating the consequences of ARID1A deficiency for gene expression and chromatin structure, which will guide mechanism-based translational efforts. Integration: Our proposed studies are highly complementary with Project 1, which is focused on the epigenetic regulator KDM6A, and Project 3, which will study resistance to cisplatin in patient-derived bladder cancer organoids. Additionally, the success of this project will require input from Core A, which will provide human tumors for correlation of molecular findings and will define the timing at which ARID1A mutations arise in bladder cancer; Core B, which will assist with the generation of human organoid and GEMMs; and Core C, which will provide essential bioinformatic and statistical support, as well as program integration.

PROJECT SUMMARY The Herbert Irving Comprehensive Cancer Center (HICCC) was designated as an NCI Cancer Center in 1972 and gained comprehensive status in 1979. The HICCC is a component of Columbia University Medical Center is associated with New York-Presbyterian Hospital (NYPH). During the period 2008-2013, CUMC and NYPH committed over $350 million for i) new research initiatives in basic, clinical and population science; ii) new and expanded facilities for laboratory research and clinical activities; iii) recruitment and program restructuring; and iv) support of the Center's administrative office. A new Director, Stephen G. Emerson MD, PhD was appointed in 2012 and given: i) authority over new facilities, including the Irving Cancer Research Center (ICRC), a 10-story building dedicated to cancer research, with state-of-the-art laboratories and clinical space in the Herbert Irving Pavilion (HIP); ii) a broad-based recruitment plan, which has already attracted 40 new faculty members to the HICCC; and iii) restructuring and expansion of the shared resources. The structure of the HICCC has been organized to increase cancer focus and interdisciplinary collaboration, and now includes 245 members from 22 Departments and 6 Schools, assigned to two Basic Research programs (Cancer Regulatory Networks and Cancer Genetics & Epigenetics), four Disease- Specific programs (Breast Cancer, Prostate Cancer, Neuro-Oncology and Lymphoid Development & Malignancy), and two Population Science programs (Cancer Epidemiology and Prevention, Control & Disparities). The HICCC administers and supports a total of 15 Shared Resources. During the period 2008-2013, the HICCC research activities have been documented in a total of 3057 publications of which 17% are inter-programmatic and 15% are intra-programmatic. The current NCI funding base is $26M (total direct costs), a 13% increase over the 2007 NCI funding base of $23M (total direct costs). The Center is requesting CCSG support of $2,607,495 (total direct costs) for the initial budget period.

DESCRIPTION (provided by applicant): The Herbert Irving Comprehensive Cancer Center (HICCC) was designated as an NCI Cancer Center in 1972 and gained comprehensive status in 1979. The HICCC is a component of Columbia University Medical Center is associated with New York-Presbyterian Hospital (NYPH). During the period 2008-2013, CUMC and NYPH committed over $350 million for i) new research initiatives in basic, clinical and population science; ii) new and expanded facilities for laboratory research and clinical activities; iii) recruitment and program restructuring; and iv) support of the Center's administrative office. A new Director, Stephen G. Emerson M.D., Ph.D. was appointed in 2012 and given: i) authority over new facilities, including the Irving Cancer Research Center (ICRC), a 10-story building dedicated to cancer research, with state-of-the-art laboratories and clinical space in the Herbert Irving Pavilion (HIP); ii) a broad-based recruitment plan, which has already attracted 40 new faculty members to the HICCC; and iii) restructuring and expansion of the shared resources. The structure of the HICCC has been organized to increase cancer focus and interdisciplinary collaboration, and now includes 245 members from 22 Departments and 6 Schools, assigned to two Basic Research programs (Cancer Regulatory Networks and Cancer Genetics & Epigenetics), four Disease-Specific programs (Breast Cancer, Prostate Cancer, Neuro-Oncology and Lymphoid Development & Malignancy), and two Population Science programs (Cancer Epidemiology and Prevention, Control & Disparities). The HICCC administers and supports a total of 15 Shared Resources. During the period 2008-2013, the HICCC research activities have been documented in a total of 3057 publications of which 17% are inter-programmatic and 15% are intra-programmatic. The current NCI funding base is $26M (total direct costs), a 13% increase over the 2007 NCI funding base of $23M (total direct costs). The Center is requesting CCSG support of $2,607,495 (total direct costs) for the initial budget period.

Project Summary/Abstract We have been studying the processes associated with prostate differentiation and their relationship to prostate cancer through our investigations of the NKX3.1 homeobox gene, which is a master regulator of prostate epithelial specification that protects the prostatic epithelium from assaults associated with cancer initiation, including oxidative stress. Our investigations have now revealed that NKX3.1 defends prostate cells from oxidative stress by regulating gene expression in both the nucleus and mitochondria. We find that, in addition to its expected functions as a transcriptional factor in the nucleus, NKX3.1 also localizes to mitochondria in response to oxidative stress, where it regulates the expression of mitochondrial-encoded genes that control reactive oxygen species (ROS). Thus, we hypothesize that NKX3.1 regulates oxidative stress via its coordinated functions in nuclei and mitochondria, and that these functions are necessary to maintain prostate epithelial differentiation and suppress cancer initiation. Since relatively few nuclear transcriptional regulatory proteins have been shown to function in mitochondria, our studies provide a unique opportunity to understand how a tissue-specific transcription factor can control oxidative stress in different sub- cellular compartments, and the relevance of these activities for cancer. In Aim 1, we will investigate the functions of NKX3.1 in the nucleus for protection from oxidative stress and promotion of differentiation. We will investigate: (i) nuclear transcriptional regulatory functions of NKX3.1 for protection against oxidative stress; (ii) their relevance for prostate epithelial differentiation and cancer; and (iii) whether and if so how these functions impact mitochondrial function. In Aim 2, we will investigate novel functions of NKX3.1 in mitochondria. Based on our preliminary data showing that, in response to oxidative stress, NKX3.1 becomes localized to mitochondria where it regulates the expression of mitochondrial-encoded genes, we will investigate: (i) the mechanisms associated with localization of NKX3.1 to mitochondria; (ii) the mechanisms by which NKX3.1 regulates mitochondrial-encoded genes, particularly in comparison with its regulation of nuclear genes; and (iii) the importance of these mitochondrial-specific functions of NKX3.1 for regulation of oxidative stress and cellular differentiation. In Aim 3, we will complement these mechanistic studies by performing co-clinical studies to evaluate the relevance of regulation of oxidative stress by NKX3.1 for suppression of prostate cancer, and whether these activities can be targeted for cancer prevention using genetically-engineered mouse models and a human prostate tissue organotypic model. Relevance for PAR-17-203: Our proposed studies provide a unique opportunity to elucidate molecular mechanisms that govern the balance between oxidative stress and differentiation and cancer initiation and how these are coordinated between the nucleus and mitochondria.