Raghu Kalluri - US grants

DESCRIPTION (provided by applicant): Glomerular Basement Membrane (GBM) is a key component of the filtration system of the kidney and structural and functional defects in this GBM are associated with many kidney diseases. The rate of GBM turnover and its regenerative potential is unknown. Type IV colagen is the most abundant protein present in the GBM. The basic unit of type IV collagen is a triple helical protomer derived from three 1-chains. With six known isoforms of type IV collagen (11-16), theoretically many combinations of protomers are possible. Type IV collagen in the glomerular basement membrane (GBM) is predominantly composed of 13, 14 and 15 chains, with 13 chain of type IV collagen being most abundant. Mutations in any one of these chains in Alport syndrome (a condition associated with progressive kidney disease, occasional sensorineural hearing loss and anterior lenticonus) leads to an absence or diminished expression of all three chains in the GBM, further highlighting a post-translational assembly requirement between the three chains to create the unique type IV collagen network of the GBM. While recent biochemical and cell biological experiments have provided further support for inter-chain interactions between 13, 14 and 15 chains, molecular and genetic models to define the assembly of GBM type IV collagen are still lacking. Such mouse models are required for understanding the pathogenesis of Alport syndrome and developing new therapy options. In the previous funding cycle, human Alport kidneys and kidneys from mice with 13(IV) collagen deletion were used in biochemical and cell biological experiments to establish the relationship between the structure of GBM type IV collagen and its susceptibility to enhanced GBM degradation in Alport syndrome. New genetic mouse models were generated to address the specificity of type IV collagen assembly in the GBM. New cell-based therapies were tested in Alport mice and new insights into GBM repair and regeneration unraveled. In this competing renewal application, we now propose to continue our studies to understand the molecular drivers that determine the chain specific assembly of type IV collagen, specifically focusing on the 13(IV) chain and its turnover in the GBM. We will study inducible and cell-specific 13(IV) collagen deficient mice to probe turnover rate of the type IV collagen in the GBM. Additionally, new mouse models will be utilized to study the contribution of 13(IV)NC1 domain, a critical domain in 1(IV) chains, in the chain specific organization of type IV colage network in the GBM, and also evaluate the capacity of 13(IV) chain to re-assemble type IV collagen protomers in the GBM. Such studies will provide further insights into possible cell based therapy options for Alport syndrome, and will also provide a basic understanding of type IV collagen assembly in the GBM.

DESCRIPTION (provided by applicant): Renal fibrosis is the hallmark of chronic kidney disease, a prominent problem in clinical nephrology. Unlike physiologic repair of acute kidney injury, progressive renal fibrosis is a self-contained perpetuated process, which fails to cease even when the initial insult has been eliminated. Activated fibroblasts are the main mediators of fibrosis, and fibroblasts isolated from fibrotic kidneys even fail to return to their quiescent state when cultured in vitro. The molecular mechanisms that underlie this persisting fibroblast activation are not yet known. In a genomewide methylation screen we identified the Ras inhibitor RASAL1 as one of 13 genes to be methylated in fibroblasts from fibrotic kidneys, but not in fibroblasts from non-fibrotic kidneys. The central hypothesis of this grant application is that 'hypermethylation of RASAL1, a suppressor of the Ras proto-oncogene, in renal fibroblasts prevents them from returning to their quiescent state and perpetuates progression of fibrogenesis. Based on the preliminary data we will elucidate if hypermethylation of RASAL1 causes perpetuated fibroblast activation via Ras hyperactivity, ultimately leading to renal fibrosis.

DESCRIPTION (provided by applicant): This U01 proposal is designed to be part of the network for the study of tumor microenvironment (TMEN) and seeks to improve our understanding of the critical steps involved the initiation and progression of pancreatic ductal adenocarcinoma (PDAC) to a level that will permit the rational development of effective stromal specific therapeutic agents. Pancreatic cancer is the fourth leading cause of cancer death in the US. In 2004, 31,270 deaths were recorded. This disease is devastating and currently specific therapies are lacking and it leads to rapid death of all patients, coupled with pain and despair. The economic toll of pancreatic cancer is estimated at about $3.7 billion dollars to the US healthcare system. The central focus of this project is to elucidate the molecular mechanisms by which fibroblasts/mesenchymal cells, myeloid/immune cells (MIC) and extracellular matrix (ECM) may contribute to the origin and progression of PDAC. While, invasive PDAC is significantly associated with a marked desmoplastic reaction (one of highest of all human tumors), significant recruitment of fibroblasts/myofibroblasts and ensuing fibrosis, the functional contribution of stromal fibroblasts in PDAC pathogenesis is not known. This application tests the central hypothesis that stromal fibroblasts are rate limiting determinants of PDAC progression. The new mouse models described in this project will provide a basic knowledge regarding the role of stromal fibroblasts, immune cells and ECM biochemistry in PDAC and explore potential PDAC microenvironment specific therapeutic opportunities. The proposal brings together the laboratories of Drs. Kalluri, Moses and Weaver, with complementary expertise to the study PDAC microenvironment. The proposal is divided into three specific aims. Specific Aim 1: Determine the functional contribution of stromal fibroblasts in the pathogenesis of PDAC. Specific aim 2. Determine the role of chemokines and host myeloid/immune cells (MICs) in the pathogenesis of PDAC. Specific Aim 3. Determine whether increased tissue tension drives pancreatic adenocarcinoma aggression.

DESCRIPTION (provided by applicant): Tumor growth, angiogenesis, and invasion depends on the recruitment and coordinated activity of different cell types populating the tumor stroma, such as endothelial cells, fibroblasts, pericytes, and infiltrating inflammatory cells. Although tumor associated vessels share similar components to normal vessels, including endothelial cells, basement membranes and pericytes, these vessels are abnormal, tortuous and leaky. While the functional role of endothelial cells in the context of tumor angiogenesis and vasculogenesis is extensively studied, the contribution of pericytes/perivascular cells in cancer progression and metastasis is poorly understood. Several studies suggest that targeting pericytes can control the rate of angiogenesis and growth of late stage tumors; however it is imperative to perform functional studies to evaluate the consequence of anti-pericyte therapy before moving these agents into clinical testing phase. In this regard, NG2, a novel cell surface chondroitin sulfate proteoglycan, is shown to be a reliable marker of pericytes/perivascular cells, along with the PDGF receptor ß (PDGFRß). This grant proposes to perform functional studies utilizing these markers in transgenic mice setting to elucidate the role of pericytes in breast cancer progression and lung metastasis. This application will test the central hypothesis that pericytes associated with primary breast tumors determine the rate of cancer progression and emergence lung metastasis. The proposal will functionally explore therapeutic targeting of pericytes using novel genetic mouse models in combination with novel pharmacological interventions using targeted therapy. The specific aims in this proposal are: 1) To determine the functional contribution of pericytes in primary breast cancer progression and lung metastasis, 2) To investigate the contribution of epithelial to mesenchymal transition in hypoxia induced lung metastasis, and 3) To determine the role of PDGFRß in lung metastasis.

DESCRIPTION (provided by applicant): Cancer accounts for 8-9 million deaths a year, 80% of which are due to systemic spread of cancer to distant organs. It has long been recognized that primary cancers spread to distant organs with 'specific' preference, and skeleton/bone is one of the most common organs to be affected by metastatic cancer. Among primary cancers, prostate cancer is considered to be an orthotropic tumor with specific predilection to form bone metastasis. Over 70% of prostate cancer patients exhibit bone metastasis as detected during the disease course or autopsies. Prostate cancers are amongst the most commonly diagnosed malignancies and the 2nd leading cause of cancer death in men. Bone metastases are the predominant reason for prostate cancer related deaths, and current therapies have very short-term benefits, if any. To achieve the central objective, the laboratories of Kalluri, Pandolfi and Scadden have joined forces to propose a focused and cohesive plan to study the biology of bone metastasis. The three projects in this TMEN network will generate new genetic mouse models of metastatic prostate cancer, offer new insights into the biology of prostate cancer initiating cells, identify possible bone metastasis cell of origin, establish the role of tumor microenvironment in metastatic prostate cancer with specific emphasis on prostate and bone microenvironment, identify new therapeutic targets against metastatic prostate cancer and evaluate a new drugs for metastatic prostate cancer in pre-clinical trials. Successful completion of this proposed network proposal will offer significant new advances in area of metastatic prostate cancer.

In this project, we propose to address the functional contribution of hypoxic microenvironment (TME) to the progression of prostate cancer and emergence of metastatic bone disease. This project use the mouse models generated in the Project 1 to study the influence of hypoxia, EMT and myofibroblasts metabolism and validate these finding in human prostate cancer tissue. Such studies will be performed in collaboration Scadden Laboratory (Project 3), which will determine the cellular components of the bone that likely cooperate with bone metastasis. We also propose to evaluate in detail the acquisifion of the ability of cancer cell to mimic mesenchymal cell types and undergo a program of epithelial to mesenchymal transition (EMT), in hopes of gathering information to identify new therapeutic targets against metastatic prostate cancer.

? DESCRIPTION (provided by applicant): Pancreatic ductal adenocarcinoma (PDAC) is amongst the deadliest and fastest progressing cancers, with a 3 to 6 month expected survival that minimally improves with current therapies. Therefore, we are facing an urgent need for new strategies for effective therapies for this cancer. One reason for our failure to significantly impct patient survival is often due to late stage diagnosis of pancreas cancer. Early diagnosis based on clinical symptoms has been a challenge due to the symptom-free disease course until patients develop aggressive and invasive disease. Surgical and other interventions could have a significant impact on improving the outcome of this disease if pancreatic cancer is diagnosed at an early stage. Unfortunately, potentially life saving non-invasive, sensitive and specific early detection methods are currently not available. Our efforts in this area led to the discovery of glypican-1 (GPC1) as specific biomarker present on serum exosomes of patients with early and late stage pancreatic cancer. Performing blinded analysis of 100 microliters of serum from 197 patients with pancreatic cancer before surgical resection of tumors, 100 normal individuals and 26 patients with non-neoplastic diseases of the pancreas, we were able to detect GPC1+ exosomes with 100% specificity and sensitivity in patients with pancreas cancer, including 7 patients with precursor lesions (early pancreatic intraepithelial neoplastic (PanIN) and intraductal papillary mucinous neoplastic (IPMN) lesions). Importantly, serum assay for CA- 19-9 performed less effectively and could not distinguish between pancreatic lesions due to cancer and non-cancer etiology. Successful validation of putative serum biomarkers requires a thorough pre-clinical characterization to identify specific windows of utility and to explore the benefit of combining non-invasive imaging to increase efficacy. In this proposal, we will employ genetically engineered mouse models of PDAC to perform cross-sectional and longitudinal studies that couple histological grading of tumors, genetic analysis of exosomes derived DNA/RNA and non-invasive MRI with the goal of establishing the utility of GPC1+ serum exosomes for reliable early detection of pancreatic cancer. Additionally, such method could also allow for reliable monitoring of post treatment/follow-up disease assessment/burden. By coupling non-invasive MRI as a validation of GCP1+ serum exosomes-based detection of pancreatic cancer, we hope to identify an informative and clinically useful diagnostic system for early detection of pancreatic cancer. Such efforts may enable early and effective therapy options for our patients.

Abstract ? Project 3 (The Functional Contribution of Tumor Immunity to PDAC) PDAC is marked by an extensive desmoplastic reaction/stroma consisting of myofibroblasts, extracellular matrix (ECM) and immune cells, and oncogenic Kras is a critical driver of PDAC genesis and maintenance. The functional importance of PDAC microenvironment and its various cellular constituent in oncogenic Kras dependent and independent PDAC maintenance remains an area of active investigation, but there is minimal knowledge addressing the role of oncogenic Kras in shaping tumor immunity, in particular the T cell response. The central mission of our P01 program is to collectively identify a range of novel Kras driven vulnerable nodes that can be inhibited as a treatment for this devastating disease. The inducible oncogenic Kras genetically engineered mouse model (GEMM) of PDAC indicated that tumor regression could be achieved via oncogenic Kras extinction, a process accompanied by depletion of oncogenic Kras dependent cancer cells and significant, albeit incompletely characterized, alterations in immune response. Studies from Projects 1 and 2 using this PDAC GEMM also offered mechanistic insights into spontaneous tumor relapse, which includes the presence oncogenic Kras extinction resistant cells (KRC) with distinct metabolic dependencies (OXPHOS), and the emergence of tumors with activation of YAP or other yet unidentified mechanism for PDAC maintenance. Projects 1 and 2 will focus on exploiting the metabolic vulnerabilities of cancer cells and the central hypothesis of Project 3 is that ?oncogenic Kras and its signaling surrogates have a causal effect on tumor immunity to facilitate cancer progression?. Knowledge of this oncogenic Kras driven circuitry may illuminate therapeutic points of intervention in the tumor microenvironment and offer an improved understanding of how to deploy checkpoint blockade therapies in combination with cancer cell-specific therapies targeting metabolic or salvage pathways. The study could inform combination treatment that includes checkpoint blockade therapies ? i.e., whether OXPHOS inhibition or autophagy inhibition can help or impede such immunotherapy strategies. The studies proposed by Project 3 are highly interconnected with studies proposed in Projects 1 and 2, and rely directly on all the Cores of this P01. Specifically, the integrated specific aims of Project 3 are to investigate Kras* dependent and independent immune response in PDAC (Aim 1), to determine the impact of targeting metabolic and salvage pathways on tumor immunity (Aim 2) and to identify novel strategies to enhance efficacy of checkpoint blockade immunotherapy in PDAC (Aim 3). This effort is integral to the overall P01 Program Goal of developing a mechanism-based rational combination strategy that can lead to meaningful therapeutic advances for PDAC patients.

Pancreatic Ductal Adenocarcinoma (PDAC) has a dismal prognosis with a median survival of about 6 months for patients with metastatic disease despite the use of current treatments. Therefore, PDAC is in urgent need of effective therapies. About 80% of PDAC patients exhibit mutation in KRAS, with substitution of a glycine residue at codon 12, such as KRASG12D, observed in high frequency. Many studies in mice unequivocally show that direct inhibition oncogenic KRAS significantly suppresses tumor progression. Despite such compelling evidence for the functional role of oncogenic KRAS in the PDAC pathogenesis, direct targeting of oncogenic KRAS has been elusive and it has often been dubbed as ?undruggable?. In this grant application, we propose to explore a novel approach to target oncogenic KRAS using physiological nanoparticles known as exosomes. Exosomes are 40-150 nm extracellular vesicles (EVs) that contain DNA, RNA and proteins, and can deliver their contents efficiently into cells they fuse with and/or enter. Exploiting this unique feature of exosomes, our research team proposes in this grant application to test the central hypothesis that ?exosomes can be engineered to deliver RNA interference (RNAi) molecules to target KRASG12D in pancreatic cancer?. Successful completion of the proposed studies will determine whether exosomes exhibit a superior ability to deliver RNAi molecules for the treatment of pancreatic cancer when compared to liposomes. Additionally, after such increased ability is investigated using multiple experimental systems, the proposal will further identify mechanism/s associated with the enhanced efficacy of exosomes in the delivery of RNAi molecules. Collectively, the proposal is designed to enable in-depth pre-clinical studies coupled with an investigation into the unique mechanism of action of exosomes to potentially facilitate future therapeutic testing in patients with PDAC.

Abstract Exosomes are released by all cells and carry bioactive molecules between diverse cell populations, with significant impact on the biology of target cells and tissues. Our group was the first to identify double- stranded DNA in exosomes and to report that collectively, intraluminal DNA fragments found in exosomes (exoDNA) cover the entire genome and reflect the mutational profiles of the cells of origin. Subsequently, exoDNA from the sera of cancer patients have been used to detect oncogenic mutations. Published studies predict potential use of patient serum/plasma exoDNA for screening for the actionable therapy targets and biomarkers. However, to date, this methodology lacks rigorous, reproducible, and unbiased analytical procedures required for clinical application. Our preliminary studies underscore the need for systematic optimization of the procedures for exosome collection, exoDNA isolation, amplification, sequencing, and computational analysis. Our overarching goal is to generate a rapid, sensitive, and reproducible pipeline for rigorous selection of somatic variants in exoDNA from urine samples, to identify driver mutations, new actionable therapy targets, and biomarkers. Preliminary analysis of exoDNA from the urines of bladder cancer patients using whole exome sequencing (WES) revealed multiple driver mutations in the tumors and in urinary exoDNA and showed that for bladder cancer, urinary exoDNA was superior to serum exoDNA in terms of representation of mutational profiles identified using tumor DNA. Quality analysis of the WES data revealed significant discrepancies that could limit the predictive value of exoDNA and urge the development of more rigorous and reproducible methodologies. We hypothesize that urine exoDNA isolation and analysis could be streamlined by stepwise optimization of processes and procedures used in exosome collection, DNA isolation, whole genome amplification (WGA), and computational analysis. Our investigative team brought together leading experts in exosome biology, bioinformatics, clinical and experimental urology. To attain designated goals, we propose to: (1) Identify procedures for optimal exosome collection and exoDNA extraction from urine; (2) Optimize whole genome amplification procedures, establish quality control panels and determine the minimal exoDNA input for quality analyses; (3) Determine computational procedures for rapid and reproducible identification of mutations and biomarkers using exoDNA; (4) Perform rigorous independent validation of established methodologies internally and by outside collaborators. The proposed studies should establish beyond reasonable doubt the validity of urinary exoDNA for mutational analysis in bladder cancer and provide rigorous and reproducible methodologies for optimal exoDNA isolation and analysis that can be applied for other exoDNA sources and cancer types.

Project Abstract/Summary The goal of our Training Grant in Cancer Biology is to educate predoctoral trainees and postdoctoral fellows to be the next generation of cancer researchers and overcome the greatest challenges in cancer research. This Program will provide trainees with educational and training opportunities that will allow them to assimilate multiple cutting-edge methodologies and technologies. Our program has highly directed foci on signaling networks in cancer cells, tumor microenvironment, metastasis and interactions between cancer cells and host cells. Training includes state-of-the-art topics and methodologies including extracellular vesicles, metabolomics, and microRNA biology. Our Program mentors have outstanding track records in both funding and publication in high-impact journals. All of them have peer-reviewed research support directly related to cancer; thus this program fulfills and exceeds the NCI Programmatic Requirement minimum of 50% of mentors having peer- reviewed cancer research support. Program mentors hold about $23,629,683 in total extramural research funding, of which about $15,401,301 (~65%) is provided by the NIH. A seven-member Internal Executive Committee (IEC) will evaluate potential mentors interested in joining the program, as well as current mentors, during the funding period. In addition, four subcommittees (Education, Science Conduct, Retreat/Symposium and Diversity) will assist the IEC and Program Director/Co-director in facilitating program-related functions and workshops. Based on applications to the training program, a five-member Education Subcommittee will select potential candidates (two predoctoral trainees and six postdoctoral fellows) and, after interview by Program Faculty, will submit the final trainee list to the IEC for approval, along with two additional predoctoral trainees supported by the Provost at The University of Texas MD Anderson Cancer Center. Each trainee will complete 2 years of training in the Program and will have an Individual Development Plan within 1 month of entering the Program. The predoctoral trainees follow the guidelines of The University of Texas Graduate School of Biomedical Sciences, whereas the postdoctoral trainees are required to complete two core courses in the first year and three seminar courses within two years. All trainees are required to 1) enroll in program-specific workshops and activities outlined herein; 2) have four standing advisory/mentoring committee meetings within the 2-year training period; 3) submit an annual progress report evaluated by advisory/mentoring committee. All trainees will have at least four one-hour RCR workshops each year, and follow-up for senior trainees will be done to ensure retraining every 4 years. The Program will be evaluated by external advisory board/internal advisory board members, trainees, mentors, and alumni regarding current approaches and consideration of innovative suggestions. Upon completion of the training, Program's trainees will have the ability to integrate multiple, diverse disciplines that encompass all relevant areas in cancer science.

Project 3 -­ Abstract/Summary  The desmoplastic  stroma,  consisting of fibroblasts,  extracellular  matrix  (ECM)  and  immune  cells,  is  a defining  feature  of  human  pancreatic  ductal  adenocarcinoma  (PDAC),  however  its  precise  contribution  to  cancer  progression is still evolving. Type I collagen (Col1), a dominant component of PDAC desmoplasia, is uniquely  regulated in PDAC. Our preliminary studies uncovered that cancer cells produce a novel Col1 variant consisting  of alpha 1 chain (Col1a1) homotrimers. Genetic deletion of the oncogenic Col1 in cancer cells extends the overall  lifespan of PDAC genetically engineered mice (GEMs). Oncogenic Kras (Kras*), a central driver for PDAC, plays  a key role in determining the host response, including tumor immunity and metabolic adaptation. We discovered  that Kras* drives cancer cells? production of oncogenic Col1 via epigenetic silencing of Col1a2.  In contrast with  the classical Col1 heterotrimer produced by fibroblasts, oncogenic Col1 uniquely promotes PDAC initiation and  progression by influencing tumor metabolism and immunity. Col1 homotrimers bind to a?3b?1 integrin on PDAC  cancer cells, a collagen receptor distinctly upregulated in mouse and human PDAC.  This induces a sustained  activation of FAK, Akt and Erk1/2 when compared with Col1 heterotrimers, promoting cancer cells growth and  survival. Oncogenic Col1 also suppresses intratumoral T cells, and oncogenic Col1 deletion synergizes with anti-­ PD-­1  immune  checkpoint  blockade  to  suppress  tumor  growth  and  increase  overall  survival  of  PDAC  GEMs.  Cancer  cell-­directed  oncogenic  Col1  homotrimer  signaling  also  enhances  glycolysis  and  mTOR  signaling,  supporting the metabolic reprogramming driven by Kras*.  In Project 3, in close collaboration with Projects 1 and  2, we will define the role and regulation of Kras*-­induced oncogenic Col1 signaling in impacting tumor immunity  and metabolism. The proposed aims investigate Col1 homotrimers-­ a?3b?1 integrin targeting, immunotherapy, and  anti-­metabolism  strategies,  as  an  informed  combination  treatment  modality.  Our  preliminary  data  support  the  feasibility  and  significance  of  this  novel  approach,  and  exploit  a  new  vulnerability  in  PDAC  signaling,  with  a  translational potential to inform new therapies for PDAC.        

ABSTRACT Most cancer patients with solid tumors die of metastatic disease and organotropic spread is an understudied aspect of metastasis, for which new insights are urgently required. The prevailing ?seed and soil? concept posits that permissive environment in a distant organ (soil) is necessary to support the survival and growth of lurking tumor cells (seeds). Induction of permissive soil in a non-metastatic organ is proposed to re-route metastasis; however, rigorous experimental evidence and mechanistic analyses are lacking. Fibrotic tissue alterations, including inflammation, provide cues for metastasis, together with the signals from bone marrow-derived cells (BMDCs), the tumor secretome, and circulating extracellular vesicles. Our preliminary data suggest organotropic metastasis is not solely dependent on permissive matrix remodeling and BMDCs in the secondary organs, but is also contingent on the disruption of vascular endothelial barrier function imposed by organ-specific vascular junction proteins. Our findings lead to a central hypothesis that ?vascular heterogeneity functionally contributes to organotropism of metastasis?. We propose studies to unravel the mechanisms, by which distinct fibrotic niches effect organ-specific changes in the vascular beds, leading to organotropic metastasis. Preliminary studies identified angiopoetin-2 (Ang-2) as a putative mediator of lung metastasis. We aim to unravel the mechanisms of Ang-2 dependent tropism to the lung but not the kidney or liver, which also generate high Ang-2 levels in the fibrotic setting. Using single-cell RNAseq and CyTOF, we will determine the cellular and molecular targets of Ang-2 in the pre-metastatic milieu. Preliminary studies show Ang-2 induces vascular leakage in the lung vasculature without impacting kidney or liver vessels, thus directing metastasis to the lung. Exosomes released by the fibrotic organs also increase vascular permeability and metastatic colonization in the lung, without affecting kidney or liver vasculature. Single-cell RNAseq of fibrotic organs, as well as genetically engineered mice (GEMs), will be used to unravel the rate limiting effect of tissue-specific disruption of Ang-2 in breast cancer metastasis. Using novel GEMs generated in the lab, we will trace lineage-specific production of metastasis- inducing exosomes and identify the determinants of organotropism via proteomic analysis. Our preliminary studies show Ang-2 disrupts vascular barriers through repression of claudin-5 that is found exclusively in the lung vasculature, in contrast with the kidney and liver vessels, which present with multiple, redundant endothelial claudins. Integrating mouse models with endothelial-specific deletion of claudin-5 and claudin-5 reporter mice, and with molecular profiling of organ-specific endothelial cells in loss- and gain-of-function experiments, we will elucidate functions of specific claudins in organotropic metastasis. Molecular studies will be performed to identify putative mechanism of Ang-2 mediated suppression of claudin-5 and test whether manipulation of vascular permeability can re-route metastasis regardless of cancer-specific organ predilection. Successful completion of the proposed studies will provide new insights into mechanism of metastasis and therapeutic implications.