Introduction: AR/AR-variants and AKR1C3 play critical roles in prostate cancer progression and driving resistance to antiandrogens, and are attractive targets for therapeutic intervention for advanced prostate cancer. However, there are currently no clinically available therapies that simutanously target both AR/AR variants and AKR1C3. We have developed a number of novel small molecules (LX) that can concurrently inhibit both AR/AR-variants and AKR1C3. Methods: A library of novel compounds (LX) was designed and synthesized according to structure based computer modeling. Cell growth assays were used to assess their efficacy at inhibiting CRPC growth. The effects of the LXs on AR/AR-variants and AKR1C3 expression were evaluated by Western blot. PSA-luciferase assays were used to determine effects on AR signaling activity. RNA-seq was performed on lead LXs (LX-1 and LX-5). Resistant cell sublines generated from C4-2B cells resistant to enzalutamide (MDVR), apalutamide (ApalR), darolutamide (DaroR), or abiraterone (AbiR) were treated with LX-1 or their respective antiandrogen and cell number was determined. Mice bearing VCaP xenograft tumors and LuCaP35CR PDX tumors were treated with LX-1 and effects on tumor growth were assessed. Results: The lead compound, LX-1, showed the greatest effect at reducing cell number, AR/AR variant expression, and AKR1C3 activity. PSA-luciferase activity was greatly reduced by LX-1. RNA-seq analysis showed a robust reduction in AR and AR-V7 signaling gene expression by LX-1. LX-1 treatment significantly inhibited the growth of resistant cells including enzalutamide resistant MDVR, apalutamide resistant ApalR, darolutamide resistant DaroR, and abiraterone resistant AbiR. Co-treatment of LX-1 and anti-androgens in resistant lines improved treatment response. LX-1 inhibited conversion of the testosterone precursor androstenedione into testosterone in AKR1C3 overexpressing C4-2B and LNCaP cells. Additionally, LX-1 treatment reduced testosterone production by LUCaP35CR tumor cells which express high levels of AKR1C3 in the presence of androstenedione in a dose-dependent manner ex vivo. Furthermore, treatment with LX-1 reduced tumor growth in both VCaP and LuCaP35CR PDX models and reduced intratumoral testosterone. Conclusions: We generated novel small molecule inhibitors that concurrent target AR/AR-variant and AKR1C3. These compounds, specifically LX-1, effectively reduce CRPC growth and synergize with antiandrogens including enzalutamide in vitroand in vivo, suggesting potential for treating advanced prostate cancer. SOURCE OF Funding: NIH CA253605, CA 225836, CA271327
