The androgen receptor (AR) responds to androgen hormones, including testosterone, primarily produced in the testis and adrenal glands. In prostate cancer cells, activation of AR promotes growth, survival, and metabolic functions. For this reason, “AR is a key therapeutic target for prostate cancer, with most treatments designed to block AR signaling,” stated Dr. Pete Nelson, Vice President of Precision Oncology and Professor in the Human Biology Division at Fred Hutch. “However, recent work also demonstrates that ‘overdriving’ AR activity, through high, supraphysiological levels of testosterone (SPT), also represses prostate cancer growth. The mechanism underlying this growth suppressive effect has not been established,” but is a direction that the Nelson lab has sought to investigate. From this pursuit, they discovered that heightened testosterone levels restrict prostate cancer proliferation by blocking cell division. These findings were published recently in Cancer Research.
While androgen deprivation or AR inhibition therapies are highly effective as front-line prostate cancer therapies, almost inevitably, over multiple years resistance occurs. The rebound of this cancer is typically accompanied by mutations that enhance AR signaling by overcoming low androgen levels or AR inhibition. This type of prostate cancer that is insensitive to androgen deprivation is termed castration resistant prostate cancer (CRPC). As in primary prostate cancer, the spotlight is on AR regulation to control CRPC progression. Counterintuitively, clinical use of physiologically high levels or ‘supraphysiological’ testosterone or androgen (SPT or SPA) treatment has been shown to inhibit CRPC growth suggesting that the AR may have a goldilocks response when it comes to androgens. To better understand the mechanism at play, the Nelson lab investigated changes in transcript levels, protein interactions, and cis-acting elements to gain a comprehensive view of gene expression changes in CRPC treated with SPA.
Since SPA has variable outcomes across the tested prostate cancer cell lines and in vivo systems, dissecting the genetic differences will help inform on biomarkers of SPA susceptible cancer. Some evidence suggested that the retinoblastoma (RB1) tumor suppressor may participate in SPA resistance. However, the Nelson lab did not observe an effect of RB1 presence or absence on SPA-mediated restriction of tumor growth in an in vivo model of CRPC. Similarly, combined loss of RB1 and TP53, a gene typically mutated in RB1 loss in clinical contexts, did not alter the response to SPA.
The researchers next investigated the role of E2F-related transcriptional signaling. SPA alters gene expression, specifically it represses E2F target genes involved in cell cycle progression. Intriguingly, the absence of RB1 did not restrict the signaling role of SPA. The researchers conducted transcriptome-wide analysis of untreated and SPA treated prostate cancer cell lines and showed that following treatment, there was increased AR signaling and decreased mitotic signaling, which included E2F target genes. Taking a closer look at the E2F transcription factor, they performed CUT&RUN chromatin immunoprecipitation and observed a decrease in E2F:chromatin interaction in prostate cancer cells treated with SPA as compared to untreated cells. The researchers also discovered that 71% of genes whose expression was decreased by SPA treatment were bound by E2F1 under normal conditions but not under SPA treatment conditions. These findings support the role of SPA to reduce E2F1:chromatin contacts and restriction of E2F1-dependent gene expression to promote CRPC progression.
The canonical pathway of E2F1 regulation is via RB1, however, other RB-family members RBL1 and RBL2 may similarly decrease the expression of E2F1 target genes. In fact, the dimerization partner, RB-like, E2F and Multi-vulval class B (DREAM) complex can reduce expression of genes involved in cell cycle progression by preventing E2F:chromatin interactions. While this pathway includes TP53, the researchers proposed that SPA signaling may occur through a non-canonical pathway of E2F regulation involving the DREAM complex since TP53 loss had not altered the effect of SPA treatment. To test this, the researchers compared the growth of RB1/TP53 double knock out prostate cancer cells to cells with additional RBL1/RBL2 deletions. Under SPA treatment, the RB1/TP53 double knockout cells were sensitive to growth inhibition (68% growth inhibition) by SPA but only a minor growth restriction was observed for the four gene knockout cells (21% growth inhibition). These results confirmed that RBL1/RBL2-dependent signaling occurs downstream of AR activation following SPA treatment.
Lastly, the researchers wanted to identify other genes repressed following SPA treatment that were independent of E2F regulation. A subset of genes with the greatest repression under SPA treatment were several cMYC-regulated genes. Strikingly, when the researchers overexpressed MYC in the context of SPA treatment, there was a complete rebound of cell growth, reversing the cell cycle stall caused by SPA treatment. Together, “we found that SPT [SPA treatment using Testosterone] blocks prostate cancer proliferation by activating a group of proteins termed the DREAM complex which regulates the cell cycle to inhibit cell division,” summarized Dr. Nelson. “SPT activation of the DREAM complex was also regulated through the MYC oncogene which has known reciprocal interactions with the AR.”
Dissection of AR signaling activated by SPA revealed factors involved in SPA-dependent repression of CRPC cell growth. “Since not all prostate cancers respond to treatment with SPT, these findings provide a molecular mechanism that can now be studied in patients to potentially determine which patients would be resistant to SPT versus which patients are most likely to respond,” commented Dr. Nelson. “The results also identify strategies to improve responses to SPT, for example by co-targeting MYC or MYC-activated cellular programs.” Altogether, “the project involved collaborations with researchers and staff at both Fred Hutch and the University of Washington, novel prostate cancer models generated through the Consortium’s prostate cancer program including a series of patient derived xenografts, and key shared resources including comparative medicine, and the scientific computing infrastructure.”
The spotlighted research was funded by the National Cancer Institute, the Congressionally Directed Medical Research Program, the Prostate Cancer Foundation, and the IPCR. The FHCRC Scientific Computing Infrastructure was funded by ORIP.
Fred Hutch/University of Washington/Seattle Children's Cancer Consortium members Elahe Mostaghel, Stephen Plymate, Eva Corey, Michael Haffner and Pete Nelson contributed to this work.
Nyquist MD, Coleman IM, Lucas JM, Li D, Hanratty B, Meade H, Mostaghel EA, Plymate SR, Corey E, Haffner MC, Nelson PS. 2023. Supraphysiological Androgens Promote the Tumor Suppressive Activity of the Androgen Receptor Through cMYC Repression and Recruitment of the DREAM Complex. Cancer Res. CAN-22-2613.