The functional landscape of patient derived RNF43 mutations predicts Wnt inhibitor sensitivity

Abstract

A subset of Wnt-addicted cancers are sensitive to targeted therapies that block Wnt secretion or receptor engagement. RNF43 loss-of-function mutations that increase cell surface Wnt receptor abundance cause sensitivity to Wnt inhibitors. However, it is not clear which of the clinically identified RNF43 mutations affect its function in vivo. We assayed 90 missense and 45 truncating RNF43 mutations found in human cancers, using a combination of cell-based reporter assays, genome editing, flow cytometry and immunofluorescence microscopy. Patent-derived xenograft (PDX) models with C-terminal truncating RNF43 mutations were tested for Wnt inhibitor sensitivity. We find that five common germline variants of RNF43 have wild-type activity. The majority of cancer-associated missense mutations in the RING and PA domains are either loss of function or hyperactivating. Hyperactivating mutants appear to function through formation of inactive dimers with endogenous RNF43 and/or ZNRF3. C-terminal truncation mutants including the common G659fs mutant, have discordant behavior in in vitro versus in vivo assays. PDXs and cell lines with C-terminal truncations show increased cell surface FZD, Wnt/{beta}-catenin signaling and are responsive to PORCN inhibition in vivo, providing clear evidence of RNF43 loss of function. In conclusion, RNF43 nonsense and frameshift mutations, including those in the C-terminal domain, and specific missense mutations in RING and PA are loss of function and predict response to upstream Wnt inhibitors in microsatellite stable cancers. This study expands the landscape of actionable RNF43 mutations, potentially extending the benefit of these therapies to additional patients. Statement of SignificanceLoss of function RNF43 mutations, first described in pancreatic cancers, drive progression of multiple cancers by increasing cellular sensitivity to Wnt ligands. These cancers are therefore uniquely sensitive to agents such as PORCN inhibitors that block Wnt production. As the PORCN inhibitors and other upstream inhibitors advance into clinical trials it is important to identify the right patients to treat with these upstream Wnt inhibitors. Hence a detailed map of mutations that are actionable is required. Here we systematically examined a spectrum of 135 patient-derived RNF43 mutations from multiple cancers. Using cell-based reporter assays, genome editing and patient-derived xenografts, we identify rules to guide patient selection. MSS cancers with either truncating mutations anywhere in the gene, including C-terminal truncations around the G659 position, or point mutations in well-defined functional domains, are likely to have RNF43 loss of function and hence a response to therapy.