A radiolabeled methyl group to recombinant histones (Fig. 2a). Reconstituted mammalian
two). We conclude that the Ezl1 protein is capable of catalyzing de novo methylation of histone H3 through its catalytic SET domain. Purified PRC2 is recognized to methylate histone H3 on K27 and K9 in vitro, although the activity towards K9 is low7. To address the lysine specificity of the Ezl1 protein, we performed in vitro methylation reactions on recombinant histone octamers bearing lysine-to-alanine mutations on histone H3 either at K9 (K9A) or K27 (K27A), or each mutations Mirogabalin besylate manufacturer around the same histone H3 tail (K9AK27A). We applied Xenopus histones, in spite of some sequence divergence with Paramecium histones (Supplementary Fig. 2), so that you can straight examine Ezl1 activity with that of PRC2. Reconstituted mammalian PRC2 core complex showed a really sturdy preference for K27 as a methylation substrate, with only a weak activity directed towards K9 (Fig. 2c, lanes 1), as previously described7. In contrast, the methyltransferase activity of GFP-Trap Ezl1 immunoprecipitates was moderately reduced on both K9A and K27A single mutants, when compared with wild-type octamers (Fig. 2c, lanes 6). A stronger reduction was observed on double K9AK27A mutant octamers (Fig. 2c, lane 9), approaching the low level observed for the GFP-Ezl1H526A mutant on wild-type octamers (Fig. 2c, lane ten). The striking difference in activity around the K27A and K9AK27A mutant octamers, which was not observed for PRC2, strongly suggests that H3K9 is one of the principal substrates of Ezl1. The 15 residual activity detected with all the double K9AK27A H3 mutant may perhaps represent a contaminating activity in ourpreparations or, alternatively, a weak activity of Ezl1 toward other residues of H3. To distinguish amongst these possibilities, we purified Ezl1 from soluble nuclear protein extracts using a stringent FLAG-HA double-affinity immunopurification (see Procedures). In vitro methyltransferase activity was strongly reduced on K9AK27A double-mutant H3, confirming that lysines 9 and 27 will be the main targets of Ezl1 (Supplementary Fig. 2). We then performed in vitro methylation reactions with H3-H4 tetramers containing recombinant Paramecium H3, instead of Xenopus H3, either wild-type or bearing lysine-toarginine mutations (K9R and K27R single mutants, and K9R K27R double mutant). Arginine substitution was chosen to keep the constructive charge of the residues. Recapitulating our above results, decreased methyltransferase activity was detected on K9RK27R double-mutant H3 when compared with wild-type H3 or single mutants (K9R and K27R) (Fig. 2d and Supplementary Fig. two), indicating that Ezl1 also targets both lysines 9 and 27 in its conspecific H3 substrate. The H526A mutation severely compromised the methyltransferase activity of Ezl1 in this setting at the same time (Fig. 2d). To investigate which lysine residue of Paramecium H3 is methylated in an Ezl1-dependent manner, we performed quantitati.A radiolabeled methyl group to recombinant histones (Fig. 2a). Reconstituted mammalian PRC2 core complicated was used in parallel as a optimistic handle. We identified that, when incubated with histone octamers or oligonucleosomes, GFP-Ezl1 immunoprecipitates transferred the labeled methyl group exclusively onto histone H3 (Fig. 2a, lanes 3, Fig. 2b). PRC2 exhibited a comparable selectivity toward histone H3, as anticipated (Fig. 2a, lane two). In contrast, sharply reduced signals have been detected with extracts in the Ezl1H526A mutant (Fig. 2a, lanes 6), while no signal might be detected within the mock manage from non-transformed cells (Supplementary Fig.