Overexpression of the SETD2 WW domain inhibits the phosphor-IWS1/SETD2 interaction and the oncogenic AKT/IWS1 RNA splicing program

Abstract

AbstractOur earlier studies had shown that AKT phosphorylates IWS1, and that following phosphorylation, IWS1 recruits the histone methyltransferase SETD2 to an SPT6/IWS1/ALY complex, which assembles on the Ser2-phosphorylated CTD of RNA Pol II. Recruited SETD2 methylates histone H3 at K36, during transcriptional elongation of target genes, and this regulates multiple steps in RNA metabolism. By regulating the RNA splicing of U2AF2, it controls cell proliferation. Importantly, pathway activity correlates with histological grade, clinical stage, and metastatic potential of lung adenocarcinomas, especially those with EGFR mutations. By regulating nucleocytoplasmic mRNA transport of intronless genes, including those encoding type I IFNs, it regulates sensitivity to viral infection. Here, we show that SETD2 interacts with IWS1 via its WW domain, that the interaction is IWS1 phosphorylation-dependent, and that WW domain overexpression blocks the interaction and inhibits the pathway and its biological outcomes. We conclude that blocking the phosphor-IWS1/SETD2 interaction is feasible and has significant therapeutic potential in human cancer.SignificanceIWS1 phosphorylation by AKT, primarily AKT3, promotes tumor cell proliferation and correlates positively with the histological grade, clinical stage, and metastatic potential of lung adenocarcinomas, especially those with EGFR mutations. Therefore, targeting the phosphor-IWS1 pathway may benefit patients with these tumors. Inhibiting AKT3 should block this pathway. However, we do not currently have clinical grade AKT3-specific inhibitors, and even if we did, inhibition of AKT3 would have off target effects. Here, we show that SETD2 binds phosphor-IWS1 via its 33 aa-long WW domain and that overexpression of this domain blocks the phosphor-IWS1/SETD2 interaction, inhibiting the pathway and tumor cell proliferation in culture and in animals. Therefore, blocking the phospho-IWS1/SETD2 interaction is feasible and has significant therapeutic potential in human cancer.