Cyclin-dependent kinases regulate epigenetic gene silencing through phosphorylation of EZH2

Abstract

The Polycomb group (PcG) protein, enhancer of zeste homologue 2 (EZH2), has an essential role in promoting histone H3 lysine 27 trimethylation (H3K27me3) and epigenetic gene silencing. This function of EZH2 is important for cell proliferation and inhibition of cell differentiation, and is implicated in cancer progression. Here, we demonstrate that under physiological conditions, cyclin-dependent kinase 1 (CDK1) and cyclin-dependent kinase 2 (CDK2) phosphorylate EZH2 at Thr 350 in an evolutionarily conserved motif. Phosphorylation of Thr 350 is important for recruitment of EZH2 and maintenance of H3K27me3 levels at EZH2-target loci. Blockage of Thr 350 phosphorylation not only diminishes the global effect of EZH2 on gene silencing, it also mitigates EZH2-mediated cell proliferation and migration. These results demonstrate that CDK-mediated phosphorylation is a key mechanism governing EZH2 function and that there is a link between the cell-cycle machinery and epigenetic gene silencing.

Figure 1

CDK1 and CDK2 phosphorylate EZH2 at Thr 350 in vitro. (a) Left: in vitro kinase assay. Recombinant CDK1–cyclin B1 protein complex was incubated with [γ-³²P]ATP and the indicated substrates. Reaction samples were resolved by SDS–PAGE and autoradiography. Right: protein substrates indicated by Coomassie blue staining. (b) Left: in vitro CDK1 kinase assay using a wild-type (WT) EZH2 GST-fusion protein fragment (amino-acid residues 1–559), and T350A, T421A, and T492A mutants of EZH2 (amino-acid residues 1–559), as substrates. Right: protein substrates indicated by Coomassie blue staining. (c) Top: in vitro CDK2 and CDK6 kinase assays using GST, a wild-type EZH2 GST-fusion protein segment (residues 1–559) and a EZH2 (residues 1–599) T350A mutant, as substrates. Bottom: protein substrates indicated by Coomassie blue staining. (d) Comparison of the amino-acid sequence of EZH2 homologues near the CDK phosphorylation site (Thr 350 indicated by asterisk). Uncropped images of blot are shown in Supplementary Information, Fig. S6a.

Figure 2

CDK1 and CDK2 phosphorylate EZH2 at Thr 350 in vivo. (a) 293T cells were transfected with plasmids to express Myc-tagged wild-type EZH2 or a EZH2^T350A mutant. Ectopically expressed EZH2 proteins were immunoprecipitated with anti-Myc, and resolved by western blot using antibody raised against phosphorylated Thr 350 (anti-T350–P) or anti-Myc. (b) EZH2 was immunoprecipitated from 293T cells expressing Myc–EZH2. Immunoprecipitated EZH2 proteins were subjected to λ protein phosphatase treatment and resolved by western-blot analysis with anti-T350–P or anti-Myc antibodies. (c) LNCaP cells were transfected with plasmids expressing v5–CDK1 and v5–cyclin B1, v5–CDK2 and v5–cyclin E, or Flag–p21^WAF1. Thr 350 phosphorylation of endogenous EZH2 was detected by the anti-T350–P antibody. Immunoblotting of extracellular signal-regulated kinase 2 (Erk2) was included as a loading control. (d) LNCaP cells were transfected with siRNAs against CDK1 and CDK2 for 48 h and then treated with or without the CDK inhibitor roscovitine, as indicated. Endogenous EZH2 Thr 350 phosphorylation was detected by the anti-T350–P antibody. (e) Lysates from 293T cells were immunoprecipitated with the anti-T350–P antibody and resolved by SDS–PAGE gel analysis. The EZH2 band was excised and analysed by LC–MS/MS mass spectrometry. The MS/MS spectrum of the double-charged ion (m/z 382.2) shows that the Thr 350 residue is phosphorylated (low case p in red) in the peptide 348-IKTPPK-353. The b ions (b1–b5) are the fragmentation ions containing the N terminus of the peptide, whereas the y ions (y1–y5) are the fragmentation ions containing the C terminus of the peptide. (f) Representative immunofluorescence microscopy images of primary human prostate tumours using anti-T350–P, anti-Ki-67, and DAPI to visualize nuclei (n = 12). Scale bar, 10 μm. Uncropped images of blots are shown in Supplementary Information, Fig. S6a.

Figure 3

The effect of Thr 350 phosphorylation on EZH2-mediated repression of its target-genes. (a) Hierarchical clustering of 6,450 genes (represented by 10,276 probe-sets) that exhibited expression differences in LNCaP cells. Lanes 1–4; cells were transfected with EZH2-specific siRNA (EZ4) and vectors expressing siRNA-resistant EZH2 or a siRNA-resistant EZH2^T350A mutant, as indicated. Lanes 5–7; cells were treated with a CDK inhibitor or HDAC inhibitor. Gene profiling data from cells transfected with the indicated siRNA and vectors (lanes 2, 3 and 4) were normalized to that in cells transfected with control siRNA (lane 1), and the data from drug-treatment experiments (lanes 6 and 7) were normalized to vehicle (DMSO) treatment (lane 5). Red and green represent upregulation and downregulation, respectively, as indicated in the scale at the top. Wild-type EZH2 and EZH2^T350A mutant proteins were expressed at comparable levels (Supplementary Information, Fig. S3b). Experiments were performed in duplicate (n = 2). (b) LNCaP cells were transfected with EZ4 siRNA or plasmids expressing Myc-tagged, siRNA-resistant EZH2 or Myc-tagged, siRNA-resistant EZH2^T350A, as indicated (empty vectors were used as a control). At 60 h after transfection, expression of HOXA9 (left) and DAB2IP (right) were analysed by real-time RT–PCR. Asterisks indicate P < 0.01. Experiments were performed in triplicate (n = 3). (c) Prostate cancer cells were transfected with plasmids expressing v5–CDK2 and v5–cyclin E (DU145 cells, left), v5–CDK1 and v5–cyclin B1 (PC-3 cells, right), or an empty control vector, in combination with control or EZH2 siRNA. At 72 h after transfection, expression of HOXA9 was evaluated by real-time RT-PCR. Asterisk indicates P < 0.05, and double asterisks indicate P < 0.01. Experiments were done in triplicate (n = 3). Western blots (bottom) were used to identify expression of the indicated proteins. (d) DU145 cells were transfected with siRNAs against CDK1, CDK2 and EZH2 as indicated. At 72 h after transfection, expression of HOXA9 was evaluated by real-time RT-PCR. Asterisks indicate P < 0.01. Experiments were carried out in triplicate (n = 3). Western blots (bottom) were used to identify expression of the indicated proteins. Uncropped images of blots are shown in Supplementary Information, Fig. S6a.

Figure 4

The effect of Thr 350 phosphorylation on H3K27me3 levels and EZH2 recruitment at EZH2-target-gene promoters. (a, b) LNCaP cells were transfected with control or EZH2-specific siRNA and plasmids expressing Myc-tagged, siRNA-resistant EZH2 or Myc-tagged, siRNA-resistant EZH2^T350A (or empty control plasmids), as indicated. (a) At 60 h after transfection, expression of EZH2 was examined by western blot. (b) H3K27me3 levels in promoters of the EZH2-target-genes, HOXA9 (left) and DAB2IP (right), were assessed by ChIP assays using anti-H3K27me3 antibodies. Data are means ± s.d. from three individual experiments (n = 3). Asterisks indicate P < 0.01. (c, d) LNCaP cells were transfected with control or EZH2-specific siRNA and plasmids expressing Myc-tagged, siRNA-resistant EZH2 or Myc-tagged, siRNA-resistant EZH2^T350A (empty plasmids were used as a control), as indicated. (c) At 60 h after transfection, expression of endogenous and restored EZH2 was examined by western blot. (d) The binding of Myc–EZH2 and Myc–EZH2^T350A to HOXA9 (left) and DAB2IP (right) promoters was examined by ChIP assays with anti-Myc antibody. Data are means ± s.d. from three individual experiments (n = 3). Asterisk indicates P < 0.05, double asterisks indicate P < 0.01. IgG; immunoglobulin G used as control antibody in ChIP assay. Uncropped images of blots are shown in Supplementary Information, Fig. S6b.

Figure 5

Phosphorylation of EZH2 Thr 350 is crucial for its function in promoting cell proliferation and migration. (a) RWPE-1 cells were transfected with empty control plasmids, or plasmids expressing wild-type EZH2 or the EZH2^T350A mutant. Graph shows cell proliferation, as monitored by MTS assay, at the indicated times after transfection. The levels of endogenous and ectopically expressed wild-type and EZH2^T350A-mutant proteins were detected with an anti-EZH2 antibody at 48 h after transfection by western blot (inset). Erk2 was included as a loading control. Asterisks indicate P < 0.01 when comparing cells transfected with plasmids expressing wild-type EZH2 to those transfected with plasmids expressing EZH2^T350A. Data are means ± s.d. from experiments with six replicates (n = 6) (b) Effects of EZH2 Thr 350 phosphorylation on anchorage-independent growth of 22Rv1 cells. Representative images of colonies formed by cells infected with lentiviral vectors expressing GFP (control), EZH2 or EZH2^T350A, and cultured in medium with agar for two weeks. Scale bar, 300 μm. Clones with the diameter larger than 300 μm in ten randomly selected fields were counted (graph, inset). Asterisk indicates P < 0.01. Data are means ± s.d. from three individual experiments (n = 3). (c, d) Cell migration evaluated by wound healing assay. RWPE-1 cells were transfected with empty control plasmids, or plasmids expressing EZH2 or EZH2^T350A. Artificial wounds were created on cells in confluence. Images were taken at 0, 24, 48 and 72 h after wound (c), and the wound widths were measured and quantified (d). Asterisk indicates P < 0.01. Data are means ± s.d. from five individual experiments (n = 5). Uncropped images of blot are shown in Supplementary Information, Fig. S6b.