Androgen receptor (AR) pathway inhibitors (ARPIs) improve outcomes in advanced prostate cancer (PC) in combination with androgen deprivation therapy (ADT). However, PC rapidly develops ARPI resistance, frequently through expression of truncated AR variants (AR-Vs), like AR-V7, highlighting a need for more effective therapies. The sodium-glucose co-transporter 2 inhibitor (SGLT2i) canagliflozin, an approved diabetes drug, also suppresses PC growth and inhibits AR-related gene expression. Therefore, we hypothesized that canagliflozin may directly inhibit AR. Cellular and tumor models of PC were subjected to proliferation, clonogenic, and xenograft studies. RNA-seq and siRNA knockdown approaches defined molecular mechanisms. Molecular docking, thermal shift, and surface plasmon resonance assays assessed drug-target interactions. Stable sh-AR full-length (sh-AR-FL) and sh-AR-V7 cell lines were generated to interrogate the transcriptomic impact of AR and prognostic analysis was performed using clinical datasets. We found that canagliflozin suppresses PC growth through AR. It interacts with the AR ligand binding domain (LBD) with estimated affinity comparable to ARPIs and blocks AR signaling. Canagliflozin reduces the transcript and protein levels of the HSP70 chaperone and suppresses the cytoplasmic and nuclear levels of AR-FL and AR-Vs through proteasomal degradation. It mediates substantial reprogramming of PC transcriptional activity, including inhibition of AR pathway, cell-cycle, E2F and MYC hallmark targets. Its gene expression profile overlaps with silencing AR-FL or AR-V7 is associated with improved prognosis in clinical datasets. The results of this study demonstrate the potential for canagliflozin to function as a clinically useful ARPI and support prospective clinical investigation of this drug in PC.
Keywords: AR variants (AR-Vs); Androgen receptor (AR); Canagliflozin; Castrate-resistant prostate cancer (CRPC); Castrate-sensitive prostate cancer (CSPC); Sodium-glucose cotransporter 2 inhibitor (SGLT2i).
Copyright © 2026 The Authors. Published by Elsevier B.V. All rights reserved.