Inseparable RNA binding and chromatin modification activities of a nucleosome-interacting surface in EZH2

Abstract

Polycomb repressive complex 2 (PRC2) interacts with RNA in cells, but there is no consensus on how RNA regulates PRC2 canonical functions, including chromatin modification and the maintenance of transcription programs in lineage-committed cells. We assayed two separation-of-function mutants of the PRC2 catalytic subunit EZH2, defective in RNA binding but functional in methyltransferase activity. We find that part of the RNA-binding surface of EZH2 is required for chromatin modification, yet this activity is independent of RNA. Mechanistically, the RNA-binding surface within EZH2 is required for chromatin modification in vitro and in cells, through interactions with nucleosomal DNA. Contrarily, an RNA-binding-defective mutant exhibited normal chromatin modification activity in vitro and in lineage-committed cells, accompanied by normal gene repression activity. Collectively, we show that part of the RNA-binding surface of EZH2, rather than the RNA-binding activity per se, is required for the histone methylation in vitro and in cells, through interactions with the substrate nucleosome.

Fig. 1. An RNA-binding surface in EZH2 is required for the methylation of naïve chromatin in an RNA-independent manner.

a, EZH2 mutants used in this study. In parentheses are citations for studies where a given activity was assayed, unless not determined (ND). b, Amino acids that mutated in this study are indicated (in blue, red and orange for mt1, mt2 and dEZH2, respectively) on the structure of PRC2 (gray) and a substrate nucleosome (DNA in black, H3 histones in magenta and the other histones in pink). The structure was determined in the presence of AEBP2, JARID2 and a monoubiquitinated nucleosome (PDB 6WKR (ref. )), but visible are only determinants that were assayed herein, including the core PRC2 subunits and the mononucleosome. S-Adenosyl-homocysteine (SAH) is in CPK yellow representation. c, Coomassie blue-stained SDS–PAGE shows the purity of PRC2 complexes, as indicated. d, HMTase assay performed with various concentrations of S-adenosyl-methionine (SAM) (24, 12, 6, 3 and 1.5 µM on the left gel and 6, 3 and 1.5 µM on the right), using 0.6 µM wild-type or mutant PRC2 complexes. 12 µM histone H3 substrates were used in the left panel and chromatinized DNA with the sequence of a polycomb target gene (naïve chromatin) was used in the right panel (0.645 µM chromatin used, determined based on the NCP molar equivalent). Three independent replicates were carried out on different days with similar results. e, HMTase assays at different PRC2 concentrations and time points, as indicated, on 1,200 nM naïve chromatin substrate. The chromatin concentration is defined as an NCP molar equivalent. Produced SAH concentrations were quantified using the MTase-Glo methyltransferase assay (Promega) and are indicated. Presented are the means of three independent replicates that were carried out on three different days and the error bar represents the standard deviation. f, HMTase assay performed using 0.6 µM wild-type or mutant PRC2 complexes and either 1.29 µM H3 histone or 0.645 µM chromatinized DNA substrates, with the reaction buffer supplemented with KCl to a final concentration of 0, 50 and 100 mM. Three independent replicates were carried out on different days with similar results. Uncropped gel images used to generate this figure are in Supplementary Fig. 2. Color key for all the schematic illustrations in this figure: PRC2 in gray, DNA in black, H3 histones in magenta and the other histones in pink. Source data

Fig. 2. An RNA-binding surface of EZH2, but not its RNA-binding activity, is required for maintaining H3K27me3 at repressed genes.

The control cell line (Ctrl) was transduced with the empty rescue vector and an AVVS guide RNA (gRNA) expression vector. Across the figure, the EZH2 mutants are color coded as in Fig. 1. a, Schematic illustration of the experimental strategy used for generating EZH2 knockout with rescue in K562 cell lines. Large circles represent cells, small circles surrounded by dots represent lentiviruses and green color represents the presence of an EGFP selection marker. b, Western blot analysis of whole-cell lysates collected from EZH2 knockout and rescued K562 cell lines, as indicated. The lentivirus EZH2 rescue vector is indicated, with an empty vector used as a control (Ctrl). The lentiviral gRNA vector is indicated, either if EZH2 gRNA (+) or the AVVS gRNA control vector (−) is used. Immunoblotting antibodies are indicated on the left. Shown are from two independent replicates that were carried out on two different days, with an additional four independent replicates shown in Supplementary Fig. 2a. c, Representative CUT&Tag genomic tracks. d, Heat maps representing the H3K27me3 and H3K27ac CUT&Tag signals 5 kb upstream and downstream from all the transcription start sites (TSS) in the human genome. e, Enrichment profiles represent the average distributions of H3K27me3 and H3K27ac CUT&Tag over 5 kb upstream and downstream from TSS, with color code indicated. Three independent biological replicates of CUT&Tag were carried out on three different days. One replicate is presented in this figure and additional replicates are presented in Supplementary Fig. 5.

Fig. 3. An RNA-binding surface in EZH2, but not its RNA-binding activity, is required for maintaining transcription programs in lineage-committed cells.

a, Unsupervised clustering of genome-wide correlation between H3K27me3 CUT&Tag data from Fig. 2. Values in boxes represent Pearson’s correlation coefficients between corresponding samples. Different samples are represented by color codes at the top and the left of the plot, with the color key is indicated at the bottom left of the figure. Three independent replicates were carried out on three different days, with different replicates indicated to the right and the bottom of the plot (rep). b, RNA-seq was carried out on cell lines that were prepared according to the same scheme as in Fig. 2a. PCA of the RNA-seq data is color coded as in panel a. RNA-seq recorded from four independent biological replicates that were carried out on four different days.

Fig. 4. A catalytic defective EZH2 mutation leads to lower PRC2 chromatin occupancy compared to the mutations in the nucleic acid-binding surfaces of EZH2.

a, Representative ChIP-Rx genomic tracks. b, Enrichment profiles represent the average distributions of SUZ12, EZH2 and H3K27me3 ChIP-Rx over 5 kb upstream and downstream from TSS, with color code indicated.

Fig. 5. Model: a surface of EZH2 that can interact with either RNA or nucleosomes, rather than its RNA-binding activity per se, is required for maintaining the H3K27me3 mark at facultative heterochromatin.

a, EZH2 mutants and their identified molecular properties, as determined herein. b, Model: PRC2 uses part of its RNA-binding surface to interact with chromatin during histone methylation (top left). At that point, the RNA-binding surface, not its RNA-binding activity, is being used for interactions with chromatin. When PRC2 is not methylating chromatin, EZH2 can use its RNA-binding surface to interact with RNA (bottom left). The molecular mechanism of how EZH2 mutants affect the canonical functions of PRC2 in H3K27me3 deposition (blue line) and transcriptional regulation (red line) in cells are illustrated in the top right: mt1 is indistinguishable from the wild-type EZH2, while mt2 phenocopy the catalytically defective EZH2 (dEZH2).