Androgen-Induced MIG6 Regulates Phosphorylation of Retinoblastoma Protein and AKT to Counteract Non-Genomic AR Signaling in Prostate Cancer Cells

Abstract

The bipolar androgen therapy (BAT) includes the treatment of prostate cancer (PCa) patients with supraphysiological androgen level (SAL). Interestingly, SAL induces cell senescence in PCa cell lines as well as ex vivo in tumor samples of patients. The SAL-mediated cell senescence was shown to be androgen receptor (AR)-dependent and mediated in part by non-genomic AKT signaling. RNA-seq analyses compared with and without SAL treatment as well as by AKT inhibition (AKTi) revealed a specific transcriptome landscape. Comparing the top 100 genes similarly regulated by SAL in two human PCa cell lines that undergo cell senescence and being counteracted by AKTi revealed 33 commonly regulated genes. One gene, ERBB receptor feedback inhibitor 1 (ERRFI1), encodes the mitogen-inducible gene 6 (MIG6) that is potently upregulated by SAL, whereas the combinatory treatment of SAL with AKTi reverses the SAL-mediated upregulation. Functionally, knockdown of ERRFI1 enhances the pro-survival AKT pathway by enhancing phosphorylation of AKT and the downstream AKT target S6, whereas the phospho-retinoblastoma (pRb) protein levels were decreased. Further, the expression of the cell cycle inhibitor p15INK4b is enhanced by SAL and ERRFI1 knockdown. In line with this, cell senescence is induced by ERRFI1 knockdown and is enhanced slightly further by SAL. Treatment of SAL in the ERRFI1 knockdown background enhances phosphorylation of both AKT and S6 whereas pRb becomes hypophosphorylated. Interestingly, the ERRFI1 knockdown does not reduce AR protein levels or AR target gene expression, suggesting that MIG6 does not interfere with genomic signaling of AR but represses androgen-induced cell senescence and might therefore counteract SAL-induced signaling. The findings indicate that SAL treatment, used in BAT, upregulates MIG6, which inactivates both pRb and the pro-survival AKT signaling. This indicates a novel negative feedback loop integrating genomic and non-genomic AR signaling.