Aim 1 and 2: In 2018, 81,190 new cases of bladder cancer (BCa; urothelial (transitional) cell carcinoma of the bladder) and 17,240 bladder cancer-related deaths were estimated in the U.S. alone. Although 70% of newly diagnosed disease is confined to the mucosa, recurrence and progression are frequent and long-term surveillance by cystoscopy is required. The remaining 30% of new cases are more advanced, with muscle invasion, lymph node involvement or distant metastases. Half of those individuals with muscle-invasive bladder cancer fail definitive therapy (surgery or chemoradiation) within 5 years and succumb to the disease. The 5- and 10-year survival rates for patients with lymph node involvement are 31 and 23%, respectively. Standard of care combination platinum-based chemotherapy for patients with metastatic disease provides a median survival of only 15 months and a 5-year survival rate of 15%. These circumstances underscore the urgent need to identify oncogenic pathways and therapeutic targets for this disease. Interrogating BCa datasets in The Cancer Genome Atlas (TCGA) project using the cBioPortal revealed potentially oncogenic alterations in MET or the gene encoding its cognate ligand, HGF in 15% of cases in the TCGA Provisional dataset. In addition to genetic alterations, overproduction of wild type RTKs and/or their cognate ligands, by tumor cells or by the tumor microenvironment (TME), can contribute profoundly to tumor cell survival, proliferation, immunosuppression and metastasis. In particular, overproduction of HGF in the TME has been found in many cases where genetic alterations in MET are absent: HGF overproduction was found in 50% head and neck cancers and high levels were correlated with metastasis, and with concerning frequency in cases of drug resistance in melanoma, colorectal, pancreatic, breast and lung cancers. Evidence of HGF/Met pathway involvement in BCa has been found in model systems and in vivo. Positive interim results from a phase II NCI clinical trial of the multikinase inhibitor cabozantinib for patients with advanced BCa (see NCT01688999/NCI-12-C-0205-N: A Phase 2 Study of Cabozantinib (XL184) in Patients With Advanced /Metastatic Urothelial Carcinoma, Principal Investigator: AB Apolo, GMB, NCI) also implicate RTKs in disease progression. Known cabozantinib targets include the RTKs encoded by MET, KDR, RET, VEGFR1, VEGFR3, KIT, FLT3, ROS1, and AXL. Our unpublished results (below) identify those encoded by MERTK, TYRO3, CSF1R, DDR1, and DDR2 as additional high affinity targets. Potentially oncogenic gene alterations of these 18 RTKs occur in a combined 68% of cases in the BCa TCGA Provisional database as determined using tools available in the cBioPortal. Moreover, 6 of these BCa-associated RTKs (BCaRTKs) show significant co-occurrence of 2-fold expression for their cognate ligands in 34% of TCGA BCa cases combined, suggestive of oncogenic autocrine activation. Identifying the BCaRTKs that are likely to mediate clinical responses to existing TKIs will provide a basis for improving RTK inhibitor-based therapies by informing drug combination strategies, driving the development of inhibitors with relevant target selectivity, and guiding the development of needed biomarkers for patient selection. This information will also further define BCa molecular pathogenesis and diversity. Quantitative mRNA analysis of RTKs and their cognate ligands in BCa-derived cell lines revealed robust co-expression of MST1R/MST1, AXL/GAS6 and CSF1R/CSF1 in 13/13, 10/13 and 8/13 BCa cell lines, respectively. In the same cell lines we found mRNA overexpression of 11 common downstream transcriptional activators (TAs) known to promote cell cycle progression and/or contribute to epithelial to mesenchymal transition (EMT): ZEB1, ZEB2, RUNX2, SOX2, SOX9, POU5F1, NANOG, MYC, SNAI1, SNAI2 and TWIST1. These genes were also queried for overexpression (mRNA z-score 2) in the BCa Provisional TCGA dataset using the cBioPortal. Overexpression was noted exclusively, i.e. no cases of reduced expression were present, and occurred at individual frequencies from 3 to 8%, yielding a combined frequency of 44%. Statistically significant frequency of co-occurrence was noted among several TAs (ZEB1/2, SOX9/MYC, SNAI2/MYC, SNA1/TWIST1 and SOX9/TWIST1) and 24 unique BCaRTK/TA pairs. Functional evidence of autocrine signaling by BCaRTKs in BCa cells was obtained using siRNA suppression. Suppressing GAS6 in J82 cells significantly decreased basal AXL activation and AKT activation. siRNA suppression of AXL in J82 attenuated the motogenic response to exogenous GAS6 and mimicked inhibition of basal migration by cabozantinib. Similar results were observed in cell proliferation assays, where siRNA suppression of either GAS6 or AXL mimicked the effect of cabozantinib. siRNA suppression of MST1R (RON) in FL3 cells also attenuated the motogenic response to exogenous MST1 and significantly reduced basal motility, and suppression of either ligand or receptor significantly reduced FL3 cell proliferation. These results: (1) further define the group of relevant, druggable RTK targets, (2) reveal patterns of their overexpression in BCa cell models and subpopulations of BCa patients, and (3) reveal frequent co-overexpression of cognate ligands and shared TAs that are known to drive tumor cell proliferation, invasion and EMT. They provide a framework for an efficient, multilevel approach to better define the functional relevance of BCaRTKs in disease progression, to develop biomarkers that indicate their criticality in individual BCa patients, and in turn, to identify those in whom relevant RTK-targeted therapies will be most effective.
Aim 3 : Renal cancer (RCC) is comprised of multiple subtypes of which clear cell renal carcinoma (ccRCC) is the most common. The Cancer Genome Atlas (TCGA) provisional ccRCC data indicates that MET mutation, copy number alteration, and amplification/over-expression occur in 12% of 446 total cases. In prostate cancer, androgen receptor (AR) represses MET expression and hence negatively regulates Met signaling. AR, a steroid nuclear receptor, is activated by androgens testosterone/dihydrotestosterone (T/DHT) leading to AR translocation into the nucleus and regulation of AR dependent genes. AR structure contains a N-terminal activation function (AF1-Ligand independent) domain, DNA binding domain, hinge region, and C-terminal ligand binding domain (LBD/AF2) domain. Deletion of exons 4-8 of AR removes the LBD resulting in a truncated constitutively activated mutant (ARv7/ARv3). ARv7 is expressed in prostatic tissues and associated with androgen independent growth of PCa cells. Intriguingly, AR over-expression occurs in 6% of 534 ccRCC TCGA cases and was associated with significantly improved over-all (OS) and progression-free survival (PFS). Normal kidney has higher AR expression compared to RCC tissues and is negatively associated with stage and grade. We found that reconstitution of AR in ccRCC cell lines with robust MET expression did not significantly reduce MET mRNA levels. However, it did reduce Met protein content, implying that an additional negative regulatory mechanism by AR may exist. To gain a more global view of how AR expression might potentially benefit ccRCC patients and identify potentially novel therapeutic strategies, we performed RNA-Seq on ccRCC cell lines over-expressing wild-type AR or the constitutively active variant ARv7 in the presence or absence of the AR ligand dihydrotestosterone (DHT). Data analysis using Partek Flow identified significant alterations in steroid pathways (e.g. PPAR, thyroid hormone, oxytocin and estrogen receptor signaling) and RCC relevant pathways (e.g. Hippo, TGF-beta, Wnt, FOXO, oxidative phosphorylation pathways), including those of therapeutically targetable proteins.
Showing the most recent 10 out of 38 publications
