H3 Lysine 4 Methylation Is Required for Full Activation of Genes Involved in α-Ketoglutarate Availability in the Nucleus of Yeast Cells after Diauxic Shift.

We show that in S. cerevisiae the metabolic diauxic shift is associated with a H3 lysine 4 tri-methylation (H3K4me3) increase which involves a significant fraction of transcriptionally induced genes which are required for the metabolic changes, suggesting a role for histone methylation in their transcriptional regulation. We show that histone H3K4me3 around the start site correlates with transcriptional induction in some of these genes. Among the methylation-induced genes are IDP2 and ODC1, which regulate the nuclear availability of α-ketoglutarate, which, as a cofactor for Jhd2 demethylase, regulates H3K4 tri-methylation. We propose that this feedback circuit could be used to regulate the nuclear α-ketoglutarate pool concentration. We also show that yeast cells adapt to the absence of Jhd2 by decreasing Set1 methylation activity.

An high-throughput in vivo screening system to select H3K4-specific histone demethylase inhibitors.

BACKGROUND: Histone demethylases (HDMs) have a prominent role in epigenetic regulation and are emerging as potential therapeutic cancer targets. The search for small molecules able to inhibit HDMs in vivo is very active but at the present few compounds were found to be specific for defined classes of these enzymes. METHODOLOGY/PRINCIPAL FINDINGS: In order to discover inhibitors specific for H3K4 histone demethylation we set up a screening system which tests the effects of candidate small molecule inhibitors on a S.cerevisiae strain which requires Jhd2 demethylase activity to efficiently grow in the presence of rapamycin. In order to validate the system we screened a library of 45 structurally different compounds designed as competitive inhibitors of α -ketoglutarate (α-KG) cofactor of the enzyme, and found that one of them inhibited Jhd2 activity in vitro and in vivo. The same compound effectively inhibits human Jumonji AT-Rich Interactive Domain (JARID) 1B and 1D in vitro and increases H3K4 tri-methylation in HeLa cell nuclear extracts (NEs). When added in vivo to HeLa cells, the compound leads to an increase of tri-methyl-H3K4 (H3K4me3) but does not affect H3K9 tri-methylation. We describe the cytostatic and toxic effects of the compound on HeLa cells at concentrations compatible with its inhibitory activity. CONCLUSIONS/SIGNIFICANCE: Our screening system is proved to be very useful in testing putative H3K4-specific HDM inhibitors for the capacity of acting in vivo without significantly altering the activity of other important 2-oxoglutarate oxygenases.