SETting Up for Epigenetic Regulation of Advanced Prostate Cancer

Abstract

An outgrowth of therapy-resistant prostate cancers (PCa) with enhanced metastatic potential may be triggered by inhibitors of androgen receptor (AR) signalling, often via epigenetic rewiring. In this issue of Cancer Cell, Yuan et al. demonstrate how SETD2 integrates EZH2 and AMPK signalling pathways to keep PCa metastasis in check.

The first-line treatment for metastatic prostate cancer (PCa) is androgen deprivation therapy (ADT). Although initially effective, treatment ultimately fails and progression to castration-resistant PCa (CRPC) occurs. Given that CRPC is still driven by hormonal signaling through aberrant activation of the androgen receptor (AR), more potent AR-targeting therapies have been developed. Yet, CRPC patients ultimately develop secondary resistance paired with enhanced metastatic capability. One form of resistance results in a phenotypic switch leading to AR indifference and progression to an AR negative CRPC phenotype, which no longer respond to AR-targeting therapy and has a mean survival of 12 months (Metzger et al., 2019).

There is mounting evidence that supports epigenetic events as mechanisms for PCa transdifferentiation to an AR-indifferent state (Beltran et al., 2016). Extensive studies have shown that DNA methylation plays a significant role in mediating these mechanisms in PCa among other cancers (Zhao et al., 2020). One example, attributed to the polycomb group protein enhancer of zeste homolog 2 (EZH2), links its enzymatic activity, which tri-methylates histone H3K27 and induces gene silencing, with poor survival in PCa (Beltran et al., 2016; Varambally et al., 2002).

EZH2 is the catalytic core subunit of polycomb repressive complex 2 (PRC2). The activity of EZH2 heavily relies on the functionality of its cystine-rich region and its SET domain. It was shown that to function as a histone methyltransferase (HMT), EZH2 requires at least two other polycomb subunits: EED and SUZ12. The physiological and cellular functions of EZH2 are manifold, including regulation of mammalian development and genomic imprinting. In cancer, however, EZH2 induces anchorage-independent colony formation and cell invasion and it has has been linked to self-renewal in poorly-differentiated cancers and cancer stem cells (Kim and Roberts, 2016). Gain-of-function mutations in the EZH2 SET domain have been reported in a variety of cancers, including Non-Hodgkin lymphoma and melanoma, while in other cancers, such as PCa, EZH2 levels are drastically elevated without traces of direct genomic alterations (Varambally et al., 2002). Particularly, in the small-cell neuroendocrine PCa (NEPC) phenotype, which accounts for 10–20% of all CRPC cases, EZH2 is enriched (Abida et al., 2019). The exact number of NEPC cases harboring elevated levels of EZH2 are still unclear due to the lack of epigenomic data in this context. Therefore, the answer about the underlying mechanisms that orchestrate EZH2 expression in CRPC remains elusive but is needed to pin down the clinical relevance of EZH2 in PCa.

In this issue of Cancer Cell, Yuan et al. demonstrate that the inversely correlated EZH2-mediated H3K27me3 and SETD2-mediated H3K36me3 levels are enzymatically intertwined. Loss-of-function mutations or depletion of the HMT SETD2 have been reported in advanced PCa, emphasizing its role in disease progression (Beltran et al., 2016). Yuan et al. assessed the tumor suppressive function of SETD2 in PCa and discovered EZH2 as one of its direct non-histone substrates, by demonstrating that SETD2 directly monomethylates EZH2 at its K735 lysine residue, which ultimately triggers a Smurf2 E3 ligase-dependent degradation (Figure 1A). These findings were supported in vivo, by engineering SETD2-deficient mice and mice harboring EZH2 mutated at the K735 lysine-residue (K735R) to prevent SETD2-mediated methylation of EZH2. In these mice, they observed accelerated tumor progression compared to control mice, especially when paired with a phenotype monoallelic for PTEN. These results are of particular interest since activating phosphorylation of EZH2 is facilitated by the PI3K/Akt pathway in CRPC (Xu et al., 2012), and, therefore, one might posit an even larger network of pathways feeding the oncogenic role of EZH2. To summarize, a model for PCa progression in which SET2D loss-of-function triggers EZH2 overexpression, while PTEN loss unleashes the chains of the PI3K/Akt pathway to fuel EZH2 activity, could be proposed (Figure 1B).

Figure 1: Setting the Stage for Epigenetic Regulation of Advanced Prostate Cancer.

A. Metformin-activated AMPK phosphorylates (P) FOXO3 transcription factor, which in turn binds to promoter regions containing FOXO-binding motifs, e.g. SETD2 promoter. Thereby, FOXO3 triggers transcription of SETD2. SETD2 monomethylates (Me) EZH2 at the lysine-residue K735, which triggers ubiquitinylation (Ub) by Smurf2 and consequential degradation of EZH2. This results in low PCR2-mediated tri-methylation (Me3) of H3K27 and high SETD2-mediated tri-methylation of H3K36 (top).

Activated AMPK mediates phosphorylation of EZH2 at the threonine-residue T311 residue to force deassembly of the EZH2-associated PRC2 complex. This results in low PCR2-mediated tri-methylation of H3K27 and normal SETD2-mediated tri-methylation of H3K36 (bottom).

B. PCa patients harboring SETD2 loss-of-function mutation at R1523 accumulate high levels of EZH2. In concert with PTEN loss, EZH2 becomes further activated by uncontrolled PI3K-Akt signalling, which phosphorylates EZH2 at the serine-residue S21. Consequentially, SETD2-mutation alone or in combination with PTEN loss results in high tri-methylation of H3K27, as well as, low SETD2-mediated tri-methylation of H3K36, which promotes PCa progression.

This study’s clinical implications were exemplified by the use of cancer-associated missense mutated SETD2 variants, manifested at the R1523 residue, in an in vitro setting. The most interesting observation was that the forced expression of the mutant R1523H led to abrogated EZH2 interaction without affecting the H3K36me3 activity of SETD2 (Figure 1B). This highlights that modification of EZH2 may be the leading oncogenic force rather than decreased H3K36 methylation. The mechanism mediated by SETD2 mutation may be relevant for a small minority of CRPC cases with EZH2 dysregulation, since fewer than 2% of CRPC patients have SETD2 mutations (Armenia et al., 2018; Yuan et al., 2020). This prompts the question of which key factors further contribute to the dysregulation of EZH2 and regulation of SETD2 itself.

To study upstream mechanisms governing SETD2 expression, Yuan et al. applied a chemical screen on ADT-insensitive PCa cells and identified a network driven by the AMP-dependent kinase (AMPK), which in turn activates transcription factor FOXO3, ultimately governing SETD2 expression. The researchers posited that an AMPK agonist could trigger methylation of the EZH2 lysine-residue K735 (Figure 1A), and, in fact, treatment with the AMPK agonist metformin reduced colony formation efficiency of CRPC organoids and impaired tumor growth in patient-derived xenografts (PDXs). SETD2 levels increased while EZH2 levels decreased upon metformin treatment, phenocopying the results of reduced colony formation efficiency. Surprisingly, organoid sizes were affected in a SET2D-KO and EZH2 K735R setting upon metformin treatment. Inspired by studies that indicate that AMPK antagonistically regulates PRC2 assembly by phosphorylating the T311 residue of EZH2 (Wan et al., 2018), the researchers explored short-term effects of metformin treatment. They observed impaired interaction between EZH2 and SUZ12, but neither levels of SETD2 or EZH2 methylation levels were increased. This indicates that both phosphorylation by AMPK and methylation by SETD2 lead to reduced EZH2 levels and activity (Figure 1A), while abrogating tumor formation evidenced by empirical approaches presented throughout the study. Nevertheless, direct EZH2 inhibition remains a viable approach to targeting CRPC, as evident in the on-going execution of clinical trials. Even in this study, the authors provided data that indicates that EZH2 inhibition is still a feasible option in SETD2-mutated PCa. Therefore, it would be necessary to assess the benefits of metformin or other AMPK agonists over EZH2 inhibition in a clinical setting, especially when SETD2 is not mutated.

In summary, the work of Yuan et al. elegantly exposes the antagonistic relationship of a SETD2-AMPK-EZH2 axis in PCa. The authors have untangled a long-standing question in advanced CPRC about how EZH2 is regulated and potentially plays a role in disease progression. This discovery of a SETD2-AMPK-EZH2 pathway opens new pharmaceutical options and provides an opportunity for patients that develop CRPC and AR-indifferent forms of the disease.