OR34-4 ATF3 Is A Converging Point For AR Signaling And Fatty Acid Oxidation In Prostate Cancer

Abstract

Prostate cancer (PCa) is the second most prevalent cancer and leading causes of death among men worldwide. Androgen deprivation is the first line of treatment for locally advanced and metastatic PCa. For men who develop metastatic, castration-resistant PCa (CRPC), survival is limited. Thus, there is a need to identify synergistic pathways to either prevent or re-sensitize the castration-resistant phenotype. PCa has a long natural history of development and there is a significant correlation between age, diet, metabolism and progress of PCa. We have previously reported that treating PCa cells with fatty acid oxidation inhibitors (FAOi) sensitizes them to androgen blockade. Here we set out to evaluate how the two disparate pathways: AR signaling and FAO, influence one another hoping to identify new therapeutic vulnerabilities. We used CRPC cell models (LNCaP-MDVres, and LNCaP-C42), as well as the FAOi (etomoxir and ranolazine) and the antiandrogen enzalutamide (MDV3100). Cells were treated for 48 hours, followed by gene expression, androgen signaling and growth analysis. Both FAOi strongly suppressed the androgen response hallmark (FDR q-val < 0.001) in MDVres cells, and this was associated with a significant upregulation of the ATF3 gene (etomoxir; 30-fold, p<0.001 vs vehicle; ranolazine; 2-fold, p< 0.03). Interestingly, knockdown (KD) of ATF3 with shRNAs resulted in differential responses to FAOi in these genetically-related cells. In the MDVres cells, ATF3 KD resulted in increased PSA expression with FAOi treatment (etomoxir: 2-fold, p= 0.002; ranolazine 4-fold, p< 0.001 vs. control shRNA cells). On the other hand, in the C42 cells, ATF3 KD resulted in decreased PSA expression with FAOi treatment (etomoxir: 3-fold, p< 0.001; ranolazine 1.5-fold, p< 0.004 vs. control cells). These results suggested that ATF3 is a convergent point for AR and FAO pathways that is cell-context dependent, and its expression status could be exploited to sensitize these cells to MDV3100 and/or the FAOi. Cell growth assays confirmed that MDVres cells with ATF3 KD were more sensitive to MDV3100 alone and showed a synergistic combination effect with FAOi (p< 0.01, combinatorial index (CI) = 0.6). The C42 cells with ATF3 KD were more sensitive to the FAOi alone and showed a synergistic effect with MDV3100 (p< 0.001, CI= 0.4). Examination of patient databases revealed an inverse correlation between AR and ATF3, a low expression of ATF3 in advanced cancers (primary tumors n=64, metastatic tumors n=24), and better disease-free survival in patients with high ATF3 expression (MSKCC; n=157, Log rank p= 0.01). These data support the role for FAOi to re-sensitize CRPC to endocrine therapies. In conclusion, we have identified ATF3 as a modulator of FAO and AR signaling that has predictive value for clinical outcomes using fat oxidation inhibitors in CRPC. Sources of research support: NIH, ACS-RSG