Prostate cancer can be controlled by androgen-hormone treatment until the cancer becomes refractory. It is believed that hormone sensitivity is largely dependent on androgen receptor (AR) activity. Here, we found the histone demethylase KDM7A which demethylates histone H3K27 to be overexpressed in enzalutamide resistant castration-resistant prostate cancer cell line C4-2b, and investigated the molecular mechanism whereby androgen receptor activity is regulated by KDM7A. We engineered AR-positive LNCaP cells to stably express a short-hairpin RNA against KDM7A mRNA from a lentiviral vector. By measuring AR downstream gene expression after androgen stimulation, we found that a KDM7A-deficient cell line showed lower AR downstream gene expression compared to a control cell. KDM7A knock-down in LNCaP cell line caused decreased cell proliferation. Western blot analysis with modified-histone antibody revealed that the KDM7A-knock-down LNCaP cell line had increased H3K27 di-methylation. We confirmed KDM7A binding on AR target-gene promoters after hormone stimulation in chromatin-immunoprecipitation experiments. And increased H3K27 di-methylation was observed in KDM7A knock-down LNCaP stable cell. Treatment with KDM7A inhibitor, TC-E 5002, reduced proliferation and induced apoptosis of prostate cancer cells. Finally, we observed that the KDM7A protein was significantly upregulated in prostate cancer tissue, and that this difference correlated with the Gleason score. These data suggested that KDM7A is potentially a good therapeutic target for prostate cancer drugs and can be used as potentially a good prognostic indicator for prostate cancer and related treatment strategies.

No abstract

PHF8 is a JmjC domain-containing histone demethylase, defects in which are associated with X-linked mental retardation. In this study, we examined the roles of two PHF8 homologs, JMJD-1.1 and JMJD-1.2, in the model organism C. elegans in response to DNA damage. A deletion mutation in either of the genes led to hypersensitivity to interstrand DNA crosslinks (ICLs), while only mutation of jmjd-1.1 resulted in hypersensitivity to double-strand DNA breaks (DSBs). In response to ICLs, JMJD-1.1 did not affect the focus formation of FCD-2, a homolog of FANCD2, a key protein in the Fanconi anemia pathway. However, the dynamic behavior of RPA-1 and RAD-51 was affected by the mutation: the accumulations of both proteins at ICLs appeared normal, but their subsequent disappearance was retarded, suggesting that later steps of homologous recombination were defective. Similar changes in the dynamic behavior of RPA-1 and RAD-51 were seen in response to DSBs, supporting a role of JMJD-1.1 in homologous recombination. Such a role was also supported by our finding that the hypersensitivity of jmjd-1.1 worms to ICLs was rescued by knockdown of lig-4, a homolog of Ligase 4 active in nonhomologous end-joining. The hypersensitivity of jmjd-1.1 worms to ICLs was increased by rad-54 knockdown, suggesting that JMJD-1.1 acts in parallel with RAD-54 in modulating chromatin structure. Indeed, the level of histone H3 Lys9 tri-methylation, a marker of heterochromatin, was higher in jmjd-1.1 cells than in wild-type cells. We conclude that the histone demethylase JMJD-1.1 influences homologous recombination either by relaxing heterochromatin structure or by indirectly regulating the expression of multiple genes affecting DNA repair.