STEM-04 - Connexin 43 drives glioblastoma cancer stem cell phenotypes through a WNK1 signaling axis

Tumors represent a highly dynamic system that relies on coordinated signaling to drive growth and adapt to selective pressures, including those induced by anti-cancer therapies. Given the need for tumors to engage in rapid and coordinated cell-cell communication, mechanisms including gap junction intracellular communication would be essential. Surprisingly, the connexin proteins that make up gap junctions have traditionally been considered tumor suppressors. This hypothesis was based on the loss of GJIC in cancer cells and, frequently, lower connexin expression in tumor cells compared to non-neoplastic tissue, including in glioblastoma (GBM), where connexin 43 (Cx43) was shown to be decreased. However, recent data suggest that this tumor-suppressive effect is context dependent and that connexins can function in a tumor-promoting role. We previously found that GBM cancer stem cells (CSCs) rely on gap junctional intercellular communication through connexin 46 (Cx46) for survival. We now identify a subset of GBM patient-derived xenograft (PDX) CSC models that also rely on Cx43 in addition to Cx46. This dependence on Cx43 is independent of Cx43 expression level. Knockdown of Cx43 using shRNA in these PDX models compromises their survival and self-renewal while inducing apoptosis. We leveraged a protein kinase phospho array to identify signaling changes associated with Cx43 knockdown and observed reductions in phosphorylation of WNK lysine-deficient protein kinase 1 (WNK1) in addition to ERK1/2, AKT, and p38MAPK. Immunoblotting validated that WNK1 phosphorylation is lost in the absence of Cx43. Treatment with a WNK1 inhibitor was toxic to CSCs. Current work is focused on interrogating the link among Cx43, WNK1 phosphorylation, and downstream signaling events and the therapeutic window of WNK1 inhibition. This work identifies a potential new signaling axis downstream of Cx43 in glioblastoma CSCs and further strengthens the hypothesis that connexins may act as both tumor suppressors and tumor drivers depending on context.