Dr. Michael Deel, Duke University

Background: Rhabdomyosarcoma is the most common soft-tissue sarcoma in childhood and young adults and is associated with skeletal muscle lineage. Although more than 20 years have passed since the discovery that FP-RMS is driven by the fusion gene PAX3-FOXO1 (P3F), therapeutically tractable components of the P3F tumorigenic program have yet to be uncovered. While the P3F fusion is known to be the principal mutation responsible for otherwise karyotypically simple FP-RMS tumors, survival rates remain dismal (5-yr overall survival <50%). Today, most children with FP-RMS will become chemorefractory or relapse1 , underscoring an urgent need for novel ways to target P3F transcriptional programming, resistance mechanisms and cancer stem cell renewal. My data suggests the transcriptional co-activator TAZ interacts with and regulates P3F, and also potentiates FP-RMS stemness and resistance to anti-tubulins. Hypothesis: TAZ is a critical mediator of pro-growth transcription factors TEADs and AP-1 2 , which, similar to P3F3 , primarily bind to distal enhancers2 . Interestingly, TEAD1 and AP-1 are among the top enriched transcription factor motifs at P3F binding sites3 and are also among the top candidate interactors of P3F4 (Fig.1A). DNA-binding sites for P3F are enriched for PAX3 motifs3 , and the biologic function of P3F is believed to be due to potently activating aberrant expression of wild-type PAX3 transcriptional targets5-7 . In melanocytes, TAZ is essential for the transcriptional activity of wild-type PAX38 through direct binding interactions between the WW domain of TAZ and the PPxY motif of PAX38 (Fig.1B). Key Results: Importantly, we have found via coimmunoprecipitation that P3F and TAZ physically interact (Fig.1C). Furthermore, using a P3F (6xPRS9) luciferase reporter, we show that manipulation of TAZ (KO, knockout) or (S89A, constitutively active mutant), modulates P3F transcriptional output (Fig.1D). In addition, we demonstrate that TAZ knockout decreases P3F protein expression, as well as expression of P3F targets FGFR4 and RASSF4 (Fig.1E). These data show that targeting TAZ may be a novel means of attenuating P3F expression and P3F-mediated transcriptional activity. Serial passaging of FPRMS cells grown as 3D-spheres resulted in a 3-12 fold increase in TAZ expression9 , suggesting that TAZ may be important for stem cell enrichment. We then examined whether the expression of stem cell markers SOX2, OCT4, and NANOG were TAZdependent in FP-RMS cells. Indeed, TAZ suppression led to decreased expression of each markers10. Conversely, propagating spheres expressing TAZS89A led to a 15-150 fold increase in expression of these stem cell markers (unpublished) (Fig.1F). We performed limiting dilution assays to determine the functional role of TAZ in FP-RMS cancer cell stemness, and found that the sphereforming frequency of RP-RMS cells is dramatically reduced in TAZ-deficient cells9 . These data suggest that TAZ is functionally required for FP-RMS stemness and cancer cell self-renewal. From a cell cycle analysis, we had observed that TAZ suppression leads to a G2/M arrest in FP-RMS cells9 . Because of this, we hypothesized that TAZ may mediate resistance to antitubulin drugs, which has been shown in other malignancies11-15 in vitro. We treated FP-RMS cells expressing vector or TAZS89A with vincristine (VCR), an antitubulin used in the backbone of RMS therapy1 . The IC50 was shifted from 2.2 to 4.7 in cells expressing TAZS89A, suggesting that TAZ does promote vincristine chemoresistance9 (Fig.1G). We also saw that combining TAZ/TEAD inhibition using porphyrin compounds with vincristine was better than either agent alone in attenuating FP-RMS cell and xenograft tumor growth (Fig.1H). Together, these data show that inhibiting TAZ may be a novel means of targeting the FPRMS cancer stem cell population and attenuating chemoresistance. Future Plans: The progress to date has identified TAZ (and likely YAP, the better-known paralog of TAZ) as a potential vulnerability for PAX3-FOXO1 (P3F). As we initially hypothesized, TAZ physically interacts with P3F (Fig.1C) and also regulates P3F-mediated transcriptional output (Fig.1D-E). However, to our surprise, suppression of TAZ also appears to regulate the protein expression of PAX3-FOXO1 itself (Fig.1E). While preliminary, these are very exciting findings and offers a lot of promise that pharmacologically targeting TAZ is a way of attenuating the main driver of fusion-positive rhabdomyosarcoma. The data generated from our work thus far p 


Children's Cancer Therapy Development Institute


To address the pressing & unmet clinical need to identify treatments for rhabdomyosarcoma, cc-TDI have generated multiple genetically-engineered mouse (GEM) models of the alveolar and embryonal subtypes of RMS and the undifferentiated pleomorphic sarcoma (UPS) subtype of NRSTS. Development of these GEM models has resulted in both the mechanistic understanding and the therapeutic approaches to RMS. Soft tissue sarcoma models were diligently characterized to be representative of the human diseases by histopathology, gene expression and other features, yet ccTDI’s 3-11 allele Cre/LoxP-mediated conditional models have the special features of knowing the cell-of-origin and mutational profile of a given tumor.

Through cc-TDI’s academic-pharma partnership with a major Swiss pharmaceutical company, 640,000-compounds have been screened across a range of cells-of-origin and mutational profiles for primary tumor cell cultures taken from these murine sarcomas. Four of the most potent hits share near-identical chemical structures. All compounds have cell growth inhibition activity against alveolar RMS (ARMS) and embryonal RMS (ERMS), only mixed activity against UPS, but no activity against normal fibroblasts. One of these compound hits is an FDA-approved cardiovascular medicine with a favorable long-term side effect profile, and completely unstudied with respect to rhabdomyosarcoma.

Our in vitro studies in 3 human rhabdomyosarcoma cell lines also affirm the activity of 4 other drugs in this class of FDA-approved medicines. Through defining the target of these drugs and validating their single-agent activity in vivo for a wide range of ARMS and ERMS models, cc-TDI will pre-clinically validate a compound/class of compounds (and their biomarkers) for RMS that translate quickly to cooperative group clinical trials as an upfront targeted therapy. Given the tolerability of this class of compounds, the potential for use as a maintenance therapy for RMS is also very high.