The emerging role of histone lysine demethylases in prostate cancer

Abstract

Early prostate cancer (PCa) is generally treatable and associated with good prognosis. After a variable time, PCa evolves into a highly metastatic and treatment-refractory disease: castration-resistant PCa (CRPC). Currently, few prognostic factors are available to predict the emergence of CRPC, and no curative option is available. Epigenetic gene regulation has been shown to trigger PCa metastasis and androgen-independence. Most epigenetic studies have focused on DNA and histone methyltransferases. While DNA methylation leads to gene silencing, histone methylation can trigger gene activation or inactivation, depending on the target amino acid residues and the extent of methylation (me1, me2, or me3). Interestingly, some histone modifiers are essential for PCa tumor-initiating cell (TIC) self-renewal. TICs are considered the seeds responsible for metastatic spreading and androgen-independence. Histone Lysine Demethylases (KDMs) are a novel class of epigenetic enzymes which can remove both repressive and activating histone marks. KDMs are currently grouped into 7 major classes, each one targeting a specific methylation site. Since their discovery, KDM expression has been found to be deregulated in several neoplasms. In PCa, KDMs may act as either tumor suppressors or oncogenes, depending on their gene regulatory function. For example, KDM1A and KDM4C are essential for PCa androgen-dependent proliferation, while PHF8 is involved in PCa migration and invasion. Interestingly, the possibility of pharmacologically targeting KDMs has been demonstrated. In the present paper, we summarize the emerging role of KDMs in regulating the metastatic potential and androgen-dependence of PCa. In addition, we speculate on the possible interaction between KDMs and other epigenetic effectors relevant for PCa TICs. Finally, we explore the role of KDMs as novel prognostic factors and therapeutic targets. We believe that studies on histone demethylation may add a novel perspective in our efforts to prevent and cure advanced PCa.

Figure 1

Epigenetic mechanisms of gene silencing and activation. Nucleosomes are composed of DNA (black strip) wrapped around 4 couples of core histones. Histone tails and DNA cytosines may be modified by several enzymes (orange circles), which add or remove epigenetic marks. Local gene function results from the combined interaction of all epigenetic enzymes. Me, methylation; K, lysine; ub, ubiquitination; DNMT, DNA methyltransferase; PRC, Polycomb repressive complex; G9a, histone H3K9 methyltransferase; HAT, histone acetyltransferase; HDAC, histone deacetylase; KDM, lysine-specific demethylases.

Figure 2

KDM genes significantly modulated in PCa (Oncomine database). Genes listed in Table 3 were queried in the Oncomine database. Significant associations are presented in this figure. PIN, prostatic intra-epithelial neoplasia. KDM1B, p < 0.001, fold-change, 2.746; KDM3B, p < 0.001, fold-change, 2.516; KDM4C, p < 0.001, fold-change, 5.89; KDM5D, p < 0.001, fold-change, -2.178. With the exception of KDM4C, we found one significant correlation per gene. For KDM4C, we report the most significant correlation, while the other one (tumor grade) is only quoted in the main text.

Figure 3

Chemical structures of KDM inhibitors. The compounds cited in the text are indicated in bold.