Jeremy Jones, Ph.D.

Development of a noncompetitive AR inhibitor to treat metastatic prostate cancer

Despite incredible advances in detection and treatment, prostate cancer (PC) remains the second leading cause of cancer death in American men. Signaling by the AR controls normal prostate growth and homeostasis but also drives the proliferation of malignant cells. AR is a classic hormone nuclear receptor that contains an N-terminal transactivation domain (NTD), a DNA binding domain (DBD), a short hinge region and a C-terminal ligand binding domain (LBD). The binding of testosterone or dihydrotestosterone (DHT) to the AR LBD results in AR translocation to the nucleus and transcriptional regulation of AR target genes.

In the setting of metastatic PC, removal of testicular androgens by surgical or chemical castration initially leads to cancer regression; however, new tumors almost universally recur and most of these castration-resistant PCs (CRPCs) remain dependent on AR function for growth. Several molecular mechanisms have been described to account for continued AR signaling in CRPC, including AR amplification, AR gain-of-function mutations that confer greater sensitivity to androgens or increased recruitment of coactivator proteins, LBD-independent N-terminal activation by growth factors, expression of constitutively active AR splice variants, tumoral conversion of adrenal androgens and intratumoral androgen production. Discoveries of these mechanisms have prompted the development of several new drugs that block the AR/androgen signaling axis. Despite these advances, recent data indicate that resistance is still a major problem and that it most likely occurs via renewed AR signaling, suggesting the emergence of androgen-independent, AR-dependent cancers. Therefore, orthogonal approaches that inhibit AR activity by means other than preventing ligand binding may offer an important therapeutic complement to existing treatments.

Recently, we reported the discovery of pyrvinium pamoate (PP), the first non-competitive, androgen-independent AR inhibitor. PP inhibits AR with nanomolar potency in vivo and functions synergistically with competitive antagonists, leading to CRPC growth inhibition in mice. Our data indicate that PP functions by binding to the DBD and preventing the recruitment of RNA polymerase II (Fig 1). Since our discovery, several other LBD competitive antagonists and NTD targeted inhibitors have been developed, including EPI-001. Both LBD and NTD targeted agents are likely to face resistance as AR is frequently mutated at their binding sites in clinical prostate cancer specimens. In contrast, PP is predicted to bind to a highly conserved element of the DBD, in which there are no reported mutations, suggesting that resistance to PP is unlikely to arise. We are working to advance a molecule with superior drug-like properties clinical trial and to further elucidate PP’s mechanism of action.

Warfarin-dependent gamma-carboxylation of the AR and prostate cancer risk

Prostate cancer (PC) is the second leading cause of cancer death in men. Retrospective studies have shown that prolonged use of the anticoagulant warfarin reduces the incidence of PC, an effect that is not observed for other cancers. However, there has been little investigation into the molecular mechanism behind this clinical observation. Warfarin is a vitamin K antagonist known to inhibit the activity of vitamin K epoxide reductase (VKOR), a key enzyme in the vitamin K cycle (Fig 2). Vitamin KH2 is used by γ-glutamyl carboxylase (GGCX) as a co-factor for γ-carboxylation, an important post-translational modification for several proteins involved in coagulation. Coincidentally, we identified several vitamin K antagonists (including warfarin) in a high -throughput screen for drugs that inhibited the conformational change and transcriptional activity of the androgen receptor (AR), which controls the growth of both benign prostate and PC cells. Our preliminary data show that AR is γ-carboxylated in a warfarin-dependent manner at aaE2, a residue known to be mutated in partial androgen insensitivity (PAIS) patients. Because PC is unique in its dependence on AR signaling and because warfarin reduces the incidence of PC, but not other cancers, we hypothesize that warfarin regulates AR activity in prostate cells via γ-carboxylation, which results in decreased tumorigenesis and lowered PC risk. We propose that defining the mechanism by which warfarin controls AR activity will allow us to separate its anti-coagulant activity from its anti-cancer effect and identify new, safe drug targets for PC prevention.

Understanding the role of androgens and androgen receptor in prostate tumorigenesis

The risk of prostate cancer (PC), an androgen-dependent disease, rises coincident with age; however, serum androgens decline with age. Our long term goal is to understand the relationship between this age-related androgen decline and PC. We hypothesize that declining systemic androgens contribute to both PC incidence and aggressiveness. In support of the former, it has been shown in meta-analyses that low, not high, levels of serum androgens correlate with increased PC risk. Furthermore, low androgen levels accelerate PC in rodent models compared to intact or castrate androgen levels. We specifically hypothesize that declining androgen levels increase the frequency of regenerative intermediate or stem cells (SCs) which can become cancer-initiating cells (CICs) if they encounter genetic damage. The increased opportunity to create CICs leads to increased PC incidence. Declining androgens are also related to PC aggressiveness, as men with lower systemic testosterone (T) are more likely to present with higher grade disease and to fail initial treatment. We hypothesize that declining androgen levels select for cells that can better proliferate in a low androgen environment by altering the androgen/androgen receptor (AR) signaling axis (AARSA) in fully differentiated and regenerative prostate luminal cells. These changes can be passed on to cancerous progeny which can then serve as a pre-configured reservoir of refractory cells that can expand during hormone therapy and lead to castration-resistant PC (CRPC). Rodents provide robust models in which to study these phenomena. We have shown that both rat and mouse benign prostate tissue can adapt to low systemic T by altering genes in the AARSA to maintain functional levels of tissue androgens. We now aim to demonstrate a causative relationship between low T and PC incidence and aggressiveness in several rodent models and to provide mechanisms that explain these relationships, namely through increasing the frequency of cells that have the potential to become CICs and the alteration of the AARSA in those cells and their progeny.

Understanding resistance to androgen receptor targeted therapy in castration resistant prostate cancer

Two new androgen receptor (AR)-targeted therapies (enzalutamide and abiraterone) have been approved for treatment of metastatic prostate cancer and more are in clinical trials. Many patients respond to these agents, but both de novo and acquired resistance are common. We urgently need to understand the mechanisms of resistance to these drugs and identify new targets to treat resistant cancers. We have initiated a clinical trial to serially sample blood from 36 patients from initiation of abiraterone or enzalutamide therapy to time of resistance. We are enumerating CTCs and the percentage of CTCs that express synaptophysin, a marker of neuroendocrine cells. Neuroendocrine differentiation is a leading hypothesis to account for resistance to AR targeted therapies, and our study will be the first to determine if expression of synaptophysin on CTCs can function as a biomarker for resistance to abiraterone and/or enzalutamide. We are RNA-sequencing single or small pools of CTCs from drug-sensitive vs. drug-resistant patient samples in an attempt to identify additional biomarkers for drug sensitivity and novel mechanisms of resistance.