Access Type

Open Access Dissertation

Date of Award

January 2017

Degree Type


Degree Name



Cancer Biology

First Advisor

Manohar Ratnam


Prostate cancer (PC) is generally dependent on the androgen signaling axis for tumor growth. PC is managed by androgen deprivation therapy (ADT). The tumors then frequently progress by restoring ADT-resistant AR signaling through mechanisms such as intratumoral androgen synthesis, overexpression of AR, expression of splice variants of AR and alteration in the balance of AR co-regulators. This stage of progression is termed castrate recurrent prostate cancer (CRPC). Moreover, ADT has many major undesirable acute and chronic side effects on various normal tissues. Therefore a more strategic therapy approach is one that would disrupt a functional arm of AR signaling critical for PC/CRPC growth but not for the essential physiological roles of AR in normal adult tissues. This thesis describes two different mechanism-based approaches to develop small molecule drugs that address the above problems. The transcription factor ELK1 tethers the androgen receptor (AR) to chromatin, enabling sustained activation of genes critical for growth in prostate cancer cell lines. The N-terminal A/B domain of AR [AR(A/B)], which excludes the ligand binding pocket of AR, is adequate for interaction with ELK1. This is significant because the major splice variants of AR (AR-V7) that lack the ligand binding domain, as well as overexpressed full length AR, are known to strongly support growth of castration resistant prostate cancer (CRPC). In our first approach to develop small molecule drugs for prostate cancer, we showed that both wtAR and AR-V7 synergize with ELK1 by coopting the two ERK docking sites on ELK1, independent of the classical mechanism of (transient) activation of ELK1 via phosphorylation by ERK. As the association of ELK1 and AR is only required for prostate tumor growth, disrupting this interaction should selectively inhibit the growth of CRPC cells without interfering with the physiological role of androgen in normal tissues. Therefore, small molecules that disrupt binding of AR to ELK1 should inhibit the growth of a broader spectrum of advanced prostate tumors than androgen ablation or conventional anti-androgen therapies without the many acute and chronic side effects associated with those treatments. We have developed and conducted a stringent in situ high throughput screen for small molecules that selectively disrupt the ELK1-AR synergy. We initially screened over 18,000 compounds from diversity sets that follow the Lipinski guidelines for “drug-likeliness”. Our top hit from the screen inhibited ELK1-dependent promoter activation by androgen in a dose-dependent manner but did not inhibit promoter activation via canonical androgen response elements. Follow up structure-activity studies identified a lead compound that was much more stable than the initial hit. We report discovery of this small molecule (KCI807) that selectively disrupts ELK1-dependent promoter activation by wild-type and variant forms of AR without interfering with ELK1 activation by ERK. KCI807 has an obligatory flavone scaffold and functional hydroxyl groups on C5 and C3'. KCI807 binds to purified AR, blocking ELK1 binding, and selectively blocks recruitment of AR to chromatin by ELK1. KCI807 narrowly affects a subset AR target growth genes and selectively inhibits AR-dependent growth of PCa cell lines and Enzalutamide-resistant PCa tumor xenografts. The results offer a mechanism-based therapeutic paradigm for disrupting the AR growth signaling axis in the spectrum of prostate tumors while avoiding global attenuation of testosterone actions.

The second approach to developing new small molecule drugs against prostate cancer involved the development of hybrid molecules. Histone deacetylase inhibitors (HDACis) can disrupt androgen signaling through the down regulation of heat shock protein 90 (HSP90). However despite their ineffectiveness in prostate cancer (PCa) cells non-selective toxicities are associated with these molecules. We designed hybrid molecules containing partial scaffolds of the AR targeted drug, enzalutamide, and the HDAC inhibitor, suberoylanilide hydroxamic acid (SAHA), to weaken the intrinsic pan-HDACi activity of the molecule and to selectively target the cytosolic AR-HSP90 complex in AR overexpressing and enzalutamide-resistant PCa cells. These new molecules, 2-75 and 1005, showed reduced potency in intrinsic HDAC inhibitor activity, degraded the HSP90 chaperone protein, induced hyper acetylation of the HDAC6 substrate α tubulin, induced p21, and caused loss of viability of the enz-resistant C4-2 cells all to a greater extent compared to either parent compound alone. These results suggested that these new molecules could be used as prototypes for the development of hybrid HDAC inhibiting drugs with reduced pan HDAC inhibitor activity and increased selectivity for AR overexpressing PC cells.

Included in

Oncology Commons