DUB regulators of YAP/TAZ activity in mesothelioma and head and neck cancer

The Hippo pathway is evolutionarily conserved to regulate multiple biological processes including development, organ size, apoptosis, and tissue regeneration, thereby maintaining tissue homeostasis. This tumour-suppressive signalling cascade acts inactivates the downstream effectors and oncogenes YAP/TAZ, preventing their binding to the TEAD transcription factors. Disruption of the Hippo pathway and aberrant YAP/TAZ activity are observed in many cancers such as pleural mesothelioma (PM) and head and neck squamous cell carcinoma (HNSCC), where there are known genetic drivers. Deubiquitinating enzymes (DUBs) have been identified as direct or indirect regulators of the pathway. However, questions remain unresolved, notably around the specificity of their interaction with the Hippo components which may depend on the cancer type and/or genetic background. To uncover novel roles for DUBs in malignancies with YAP/TAZ aberrant activation, the regulation of these oncogenes in PM and HNSCC was investigated by two complementary, unbiased DUB siRNA library screens in cells with hyper-active YAP/TAZ. Among a panel of 24 cell lines, the PM NCI-H2052 and HNSCC PE/CA-PJ41 exhibited highly expressed or hyper-activated YAP coupled with robust TEAD transcriptional activity and were selected as appropriate cell models for DUB siRNA screen. Method development included two main readouts: (i) a TEAD-specific dual luciferase assay to measure the output of the Hippo pathway upon silencing of 94 human DUBs; (ii) real-time monitoring of cell confluency to examine cell viability upon DUB silencing. Data revealed different top DUB hits in each cell line, whose depletion strongly reduced TEAD activity, representing potential therapeutic targets. Majority of the six best performing DUB hits in each cancer model were not previously associated with the Hippo pathway or these malignancies. DUBs regulators of YAP/TAZ in HNSCC were further explored in a complementary project in our group. In PM, USP42 emerged as the strongest candidate compared to the others when followed up across multiple analyses. USP42 over-expression is associated with lower survival in mesothelioma. Its depletion significantly decreased TEAD activity, and it also affected cell proliferation and YAP(Ser127)/TAZ(Ser89) phosphorylation in NCI-H2052. Moreover, an unreported USP42 interaction with Hippo components was highlighted in HEK293, where USP42 levels also influenced their protein expression, suggesting its role in modulating the Hippo signalling, and hence YAP/TAZ-TEAD activity in cancer. To explore the underlying molecular mechanism, USP42 knockout MCF 10A clones were generated but proved impossible to grow, indicating USP42 essentiality. In future, an inducible shUSP42 system, that was developed in this work, will allow a time-controlled depletion for more in-depth analysis. Overall, the findings uncovered cell/cancer-specific DUB candidates linked with YAP/TAZ oncogenic behaviour in HNSCC or PM, strengthening the hypothesis that roles for different DUBs in cancer-associated Hippo signalling may emerge in specific cancer types. USP42 was identified as the most promising DUB candidate to regulate YAP/TAZ activity in PM. Additional validation may unveil its therapeutic potential in cancers with disrupted Hippo signalling and YAP/TAZ aberrant activation.