Sohail Tavazoie, MD, PhD - US grants

DESCRIPTION (Provided by the applicant) Abstract: The vast majority of cancer deaths result from the metastatic spread of cancer cells to distal organs. Systemic chemotherapy can prevent metastasis in some patients by killing microscopic tumor cells throughout the body. Chemotherapy can also dramatically reduce the size of metastases in some advanced stage patients. Interestingly, these standard chemotherapeutic regimens are administered to hundreds of thousands of patients without prior knowledge of the sensitivity of individual patients. cancer cells to such treatments. If we could identify the chemotherapeutic responsive and resistant subsets of patients at diagnosis, innumerable patients would be spared from the risks, side effects, and expense of ineffective chemotherapy and instead offered alternative and experimental therapies in the upfront setting. Furthermore, the identification of such biomarkers could provide mechanistic insights into the molecular underpinnings of chemotherapeutic resistance. Working in breast cancer, we recently discovered a set of human microRNAs that strongly suppress metastasis in a robust mouse model of breast cancer. These microRNAs act as biomarkers since their expression levels in primary tumors predict future metastatic relapse, thus guiding clinical decision-making. Colorectal cancer is a highly prevalent and aggressive disease entity with significantly fewer treatment options than breast cancer. We propose to apply a conceptually and technically innovative, systematic, and multi-disciplinary approach to discover .chemotherapeutic-response predictive microRNAs. through an experimental approach that integrates molecular, in vitro, in vivo, and human clinical insights. We will validate the power of these microRNA biomarkers through prospective in vivo human studies. If successful, we envision this powerful approach applied to other common cancers. The identification of such microRNAs will not only be of tremendous clinical value now, it will also lay the foundation for future mechanistic and synthetic efforts aimed at generation of novel microRNA-based therapeutic agents for the prevention and treatment of cancer metastasis. Public Health Relevance: Colorectal cancer is a highly prevalent and aggressive disease entity. Despite surgical resection of primary tumors, tens of thousands of patients will develop metastatic spread of their cancers to organs such as the liver or lung and will consequently die of their disease. Through a multi-disciplinary, collaborative, and highly innovative approach that integrates molecular, in vitro, in vivo, and human clinical analyses, we propose to discover specific microRNA biomarkers that will empower clinicians to identify patients whose tumors will display resistance to standard chemotherapy for colorectal cancer. The identification of such biomarkers will save thousands of patients annually from ineffective treatment and make them eligible for alternative and experimental therapies. The identification of such miRNA biomarkers will not only be of tremendous clinical diagnostic value now, but will lay the foundation for future efforts aimed at generating novel microRNA-based therapeutic agents for the prevention and treatment of colorectal cancer metastasis.

DESCRIPTION (provided by applicant): The proposed inter-disciplinary research plan combines molecular, biochemical, genetic, human pathologic, and animal functional studies to characterize a pro-metastatic regulatory network that has been found to regulate metastatic progression by a broad range of melanomas representing diverse mutational subtypes. Our published work has revealed that in the majority of human melanomas, over-expression of three pro-metastatic microRNAs (miR-1908, miR-199a5p, and miR-199a3p) drives metastatic colonization through direct targeting of ApoE gene and the heat-shock factor DNAJA4 (Pencheva et al., Cell, 2012; Pencheva, Buss et al., Cell, 2014). ApoE and DNAJA4 were found to robustly suppress metastatic colonization and angiogenesis. Our extensive previously published and unpublished work supports a robust and clinico-pathologically validated role for these genes and their network in human metastatic progression. Since our original publications, independent investigators have provided further support of a role for ApoE in cancer progression by demonstrating that the ApoE gene suppresses breast cancer metastasis and that the human ApoE4 polymorphism increases the likelihood of death from malignancy in men and women. These findings as a whole establish ApoE as a potent and dual regulator of metastatic progression and tumor angiogenesis. Thus, an enhanced understanding of the upstream mechanisms that regulate ApoE in cancer, its downstream mechanism(s) of action on tumor angiogenesis, its therapeutic potential, and the impact of ApoE polymorphic variants on cancer is required. The current proposal aims to answer four major questions of great significance to cancer and metastasis biology: (i) what is the mechanism of upstream regulation of ApoE by the DNAJA4 heat-shock factor? (ii) How does the binding of ApoE to the endothelial LRP8 receptor suppress angiogenesis? (iii) Could we develop a genetic model of melanoma metastatic progression driven by genetic inactivation of ApoE and use this model to test the anti-metastatic efficacy of an ApoE peptide mimetic? and (iv) does the prevalent human ApoE4 polymorphism drive mouse melanoma metastatic progression. The development of an ApoE inactivation driven genetically initiated model of melanoma metastatic progression is innovative and could beneficially impact the scientific, medical, and pharmaceutical communities-allowing us to test and develop ApoE peptide therapy in an immunoproficient model. Discovery of an unexplored post- transcriptional mechanism that regulates ApoE, a gene implicated in dementia and cardiovascular disease could also have broad scientific and clinical impact and reveal another avenue for therapeutic activation of ApoE. Testing the impact of angiogenic suppression of breast and lung cancer by ApoE- LRP8 could broaden the scope and impact of ApoE therapy to these prevalent cancers as well.?

In order for cancers to progress, metastasize, and become treatment refractory, gene expression programs must be modulated. Transcriptional programs generate transcripts within the cells, while post-transcriptional programs regulate the half-lives, localization, and translational efficiency of cellular transcripts. Such post-transcriptional deregulation has been implicated as a major means by which ?aggressive? gene expression progression are established. We recently showed that RNA-fragments (tRF?s) generated from the processing of specific tRNA molecules suppress breast cancer metastasis through their binding and repression of an oncogenic RNA-binding protein (Goodarzi et al., Cell, 2015). A search for a tRF that could promote cancer progression has led to our identification of a specific tRNA-fragment (tRF) that is increased in highly metastatic breast cancer cells, promotes metastasis, and suppresses expression of transcripts containing its recognition motif. This mode of regulation contrasts RNAi-based mechanisms, since the modulated transcripts contain sense (rather than complementary) sequences relative to the tRF. We hypothesize that this tRF drives metastatic progression by binding and inhibiting an RNA-binding protein (RBP) that otherwise promotes the stability of pro-metastatic transcripts. We aim to investigate the role of this tRF in metastasis formation and progression; to identify the trans-factor (RBP) that this tRF regulates and the downstream regulon impacted; to uncover the upstream mechanism of tRF generation; to investigate its diagnostic potential as a predictive biomarker; and to provide proof-of-concept support for tRF therapeutic inhibition through oligonucleotide anti-sense administration. We will utilize cutting-edge, complementary, and mutli-disciplinary methods to achieve these goals. Successful completion of this study will generate new basic insights into post-transcriptional regulation by a tRNA-derived fragment, reveal how specificity is achieved in tRNA-fragment generation, achieve proof-of-concept for oligonucleotide-based therapeutic inhibition of a tRF in cancer, identify a potentially druggable tRF-producing ribonuclease, and establish diagnostic and prognostic potential for a tRF. As such, this work has significant potential for impacting human health. Our lab has made major mechanistic inroads into the non-canonical roles of tRNAs in gene-expression control in cancer (Goodarzi, Hoang et al., Cell, 2016; Goodarzi et al., Cell, 2015). Moreover, our interdisciplinary approach that integrates molecular, biochemical, genetic, computational, animal modeling, and clinical association methods to investigate basic mechanistic questions of relevance to human cancer position us well for success in these efforts.

We have previously identified sets of tissue-specific microRNAs that regulate metastatic progression by distinct cancer types. In both breast cancer and melanoma, such metastasis regulating miRNA pathways enable cancer cells to avidly recruit endothelial cells into the primary tumor site (Png et al., Nature, 2011; Pencheva et al., Cell, 2012). A key unknown in the field has been the signal(s) that are provided by such recruited endothelial cells that impact metastatic progression. We have used an innovative unbiased approach wherein ribosomes of tumor endothelial cells are genetically marked with an affinity tag. This enables purification of tumor endothelial ribosomes along with their associated transcripts, which then undergo next-generation sequencing. This allowed us to identify Slit2 as a gene significantly induced in endothelial cells by highly metastatic cells. Slit2 is an axon guidance molecule required for the proper establishment of nervous system connectivity. Our preliminary evidence in syngeneic models reveals that genetic inactivation of Slit2 in the endothelial compartment significantly impairs cancer metastasis from the primary tumor site. We propose a model whereby metastatic cells induce Slit2 in endothelial cells, which serves as a signal that promotes migration of cancer cells within the tumor (low Slit2) towards the vasculature (high Slit2), enabling intravasation and metastasis. This model is supported by preliminary clinical association evidence that reveals that increased Slit2 in endothelial compartment relative to the tumoral compartment associates with higher stage tumors that exhibit higher rates of metastatic relapse. In this application, we propose a series of complementary approaches for rigorously confirming this surprising model and further mechanistically dissecting it. We will modulate Slit2 signal sensing by cancer cells through genetic inactivation of endothelial or tumoral Slit2 using cell-type specific genetic inactivation in a genetically initiated model of cancer progression. We will employ live animal multi-photon microscopy to visualize Slit2-driven tumoral trans-endothelial migration and intravasation. We aim to identify the tumoral receptor that senses Slit2, to use immunohistochemical methods to investigate an association between endothelial Slit2 and human cancer progression and metastatic relapse, and to discover the tumor-derived signal that induces endothelial Slit2. Finally, we will apply these insights by determining if a clinically used therapeutic, which we find induces Slit2 promotes cancer metastasis. This work has the potential for major impact on our understanding of mechanisms of cancer progression by establishing endothelial cells as major orchestrators of metastasis. It could have important impact on human disease given that this pathway governs progression of highly prevalent cancer types and associates with human relapse. Moreover, the cell-type specific ribosomal profiling method we have employed could be applied more broadly to study endless cell-types within the tumor microenvironment.