<?xml version="1.0" ?> <tei> <teiHeader> <fileDesc xml:id="0"/> </teiHeader> <text xml:lang="en"> <p>Steroid hormone receptors are transcription<lb/> factors that regulate transcription in an exqui-<lb/>site manner <ref type="biblio">1</ref> . In the absence of ligand, hormone<lb/> receptors either remain inactive in the cyto-<lb/>plasm or repress promoter activity through the<lb/> recruitment of co-repressors, including histone<lb/> deacetylase enzymes (HDACs). Co-repressors<lb/> are dislodged by ligand binding and hormone<lb/> receptors then initiate promoter activation<lb/> through the recruitment of co-activator com-<lb/>plexes that possess histone acetyltransferase<lb/> and chromatin-remodelling activities. Cycles<lb/> of acetylation and deacetylation at histone<lb/> tails are among the many dynamic covalent<lb/> modifications that nucleosomes endure <ref type="biblio">2</ref> .<lb/> Phosphorylation, ubiquitination, methylation,<lb/> sumoylation and ADP-ribosylation also occur<lb/> on several histone tail residues and are linked<lb/> closely with the transcriptional activity of many<lb/> promoters. It has also been suggested that such<lb/> modifications convey epigenetic information.<lb/> Steroid hormone receptors have served as a<lb/> model for the functional interplay between<lb/> transcription factors and their corresponding<lb/> chromatin-modifying transcriptional co-acti-<lb/>vators or co-repressors. Indeed, recent stud-<lb/>ies have documented the interaction of such<lb/> receptors with a number of enzymes capable<lb/> of modifying histone tails in both normal <ref type="biblio" >3</ref> and<lb/> pathological states <ref type="biblio">4</ref> .<lb/></p> <p>Many of the histone modifications have been<lb/> studied in detail; however, very little is known<lb/> about the biological function of histone H3<lb/> Thr 11 (H3T11) phosphorylation. Early stud-<lb/>ies in plant and mammalian cells suggested that<lb/> phosphorylation of H3T11 might be involved<lb/> in chromosome condensation during mitosis<lb/> and meiosis <ref type="biblio">5</ref> . On page 53 of this issue, Metzger<lb/> et al. <ref type="biblio">6</ref> demonstrate that H3T11 phosphoryla-<lb/>tion is linked to transcriptional regulation in<lb/> response to stimulation with androgen receptor<lb/> agonists. The protein-kinase-C-related kinase<lb/> 1 (PRK1) is required for androgen-receptor-<lb/>dependent gene transcription <ref type="biblio">7</ref> . Interestingly,<lb/> the authors show that inhibition of PRK1<lb/> (either with the specific inhibitor Ro318220 or<lb/> by stable RNA interference-mediated silencing)<lb/> prevents androgen-receptor-mediated H3T11<lb/> phosphorylation at the promoter of prostate-<lb/>specific antigen (PSA). Analysis of the PSA<lb/> promoter has important implications, as aber-<lb/>rant regulation of PSA has not only been asso-<lb/>ciated with a large number of human prostate<lb/> tumours but also serves as an early marker for<lb/> diagnosis of prostate cancer. Metzger et al. show<lb/> that in prostate tumour cells, androgen recep-<lb/>tors interact directly with PRK1 and occupy the<lb/> same region on the PSA promoter. Indeed, the<lb/> presence of PRK1 is strictly required for H3T11<lb/> phosphorylation. In addition, PRK1 is able to<lb/> phosphorylate H3T11 in vitro.<lb/></p> <p>In a previous study, the histone demethylases<lb/> LSD1 and JMJD2C were shown to participate<lb/> in androgen-receptor-dependent transcription<lb/> in a coordinated fashion <ref type="biblio">8</ref> . These authors have<lb/> extended their analyses to show that phosphor-<lb/>ylation of H3T11 facilitates Lys 9 (K9) demeth-<lb/>ylation by the demethylase JMJD2C (Jumonji C<lb/> (JmjC) domain-containing protein). One inter-<lb/>pretation of these results is that trimethylated<lb/> K9-containing nucleosomes are better sub-<lb/>strates for JMJD2C when H3T11 is phosphor-<lb/>ylated. Alternatively, PRK1 may phosphorylate<lb/> JMJD2C directly to enhance its demethylation<lb/> activity. As H3K9 trimethylation (a docking<lb/> site for an important component of hetero-<lb/>chromatin, heterochromatin protein 1 (HP1;<lb/> <ref type="biblio">refs 9, 10</ref>)) has been correlated closely with<lb/> transcriptional repression, reduced levels of<lb/> trimethyl K9 caused by H3T11 phosphoryla-<lb/>tion may lead to transcriptional activation.<lb/></p> <p>Enhancement of H3K9 demethylation<lb/> accounts for only part of the effects of PRK1/<lb/> H3T11 phosphorylation in regulating pro-<lb/>moter activity. Metzger et al. provide additional<lb/> data correlating PRK1 occupancy at target pro-<lb/>moters with the activation of the transcription<lb/> complex from a pre-initiation to an initiation<lb/> state, as measured by phosphorylation of the<lb/> RNA polymerase II at Ser 5. Impairment of<lb/> PRK1 activity prevents Ser 5 phosphorylation<lb/> but not the recruitment of RNA polymerase<lb/> II to the promoter, suggesting a provocative<lb/> crosstalk between H3T11 phosphorylation,<lb/> H3K9 demethylation and the critical switch<lb/> from a pre-initiation to initiation complex.<lb/></p> <p>Although it seems that androgen receptors,<lb/> together with PRK1, fine-tune gene regulation,<lb/> it is not clear how ligands of androgen receptors<lb/> activate the Rho signalling pathways known to<lb/> control PRK1 activation or whether there is<lb/> cross-talk between ligands and Rho GTPases<lb/> at the plasma membrane. In a recent study<lb/> demonstrating that progestins trigger the cyto-<lb/>plasmic Src/Ras/Erk/Msk1 signalling cascade,<lb/> Vicent et al. describe a model for progesterone-<lb/>receptor-mediated mouse mammary tumour<lb/> virus (MMTV) promoter activation <ref type="biblio">11</ref> . When a<lb/> ligand binds to the androgen receptor, Erk and<lb/> Msk1 navigate to promoters together with pro-<lb/>gesterone receptors. At the promoters, binding<lb/> of the Msk1–progesterone receptor complex<lb/> causes Msk1-dependent phosphorylation of<lb/> H3K10, HP1 displacement, RNA polymerase<lb/> II recruitment and activation of transcription.<lb/> Both Metzger et al. and Vicent et al. analysed<lb/> histone phosphorylation in response to hor-<lb/>mone stimulation; however, their findings<lb/> suggest that the consequence of such phospho-<lb/>rylation on transcriptional regulation differs<lb/></p> <figure>a<lb/> b<lb/> c<lb/> d<lb/> ON<lb/> PRK1<lb/> ?<lb/> KDM<lb/> PSA<lb/> AR<lb/> AR<lb/> Ligand<lb/> Signal<lb/> cascade<lb/> OFF<lb/> PSA<lb/> OFF<lb/> PRK1<lb/> AR<lb/> PSA<lb/> OFF<lb/> PRK1<lb/> KDM<lb/> AR<lb/> PRK1<lb/> RNA Polymerase II<lb/> = H3K9 tri-methylation<lb/> = CTD-S5 phosphorylation<lb/> = H3T11 phosphorylation<lb/> = Histone core<lb/></figure> <p>for the two hormone receptors. Thus, proges-<lb/>terone-receptor-mediated H3S10 phosphoryla-<lb/>tion leads to release of HP1; in contrast, Metzger<lb/> et al. suggest that phosphorylation of H3T11<lb/> enhances demethylation. Even so, could PRK1<lb/> also be recruited in the cytoplasm and trans-<lb/>located to the nucleus by androgen receptors?<lb/> Moreover, given the cyclic activation of genes<lb/> regulated by hormones <ref type="biblio">12</ref> , how is phosphoryla-<lb/>tion of H3T11 reversed? One could envision a<lb/> direct recruitment of a protein phosphatase, as<lb/> has been suggested by studies in plants show-<lb/>ing that PP2A and PP1 phosphatases might be<lb/> involved in such a process <ref type="biblio">5</ref> .<lb/></p> <p>Although uncovering a new mode of regu-<lb/>lation for the PSA gene has potential clinical<lb/> implications, it is unclear at present whether<lb/> other genes targeted by androgen receptors are<lb/> phosphorylated at H3T11 after ligand admin-<lb/>istration and whether other steroid hormone<lb/> receptors would regulate transcription in a<lb/> similar manner.<lb/></p> <p>Finally, Metzger et al. show that high levels<lb/> of PRK1 and H3T11 phosphorylation exist<lb/> in early-stage prostate carcinomas, where<lb/> androgen receptors are known to be crucial<lb/> in controlling cell growth. The authors also<lb/> show that knockdown of PRK1 in prostate<lb/> tumour cells markedly reduces cellular<lb/> proliferation. They thus identify PRK1 as a<lb/> promising target for therapeutic interven-<lb/>tion in tumours where androgen receptors,<lb/> and possibly other steroid hormone recep-<lb/>tors, may be aberrantly regulating gene<lb/> expression.<lb/></p> <figure>Figure 1 Schematic representation depicting the sequence of events following association of androgen<lb/> receptors (AR) with PRK1. (a) Ligand-binding stimulates the association of AR with PRK1. (b) AR–PRK1<lb/> interaction causes phosphorylation of H3T11.(c) Demethylation of H3K9 by a lysine demethylase<lb/> (KDM) is enhanced. (d) Phosphorylation of RNA polymerase II on Ser 5 activates transcription.</figure> </text> </tei>