Enzyme-independent role of EZH2 in regulating cell cycle progression via the SKP2-KIP/CIP pathway

Abstract

While EZH2 enzymatic activity is well-known, emerging evidence suggests that EZH2 can exert functions in a methyltransferase-independent manner. In this study, we have uncovered a novel mechanism by which EZH2 positively regulates the expression of SKP2, a critical protein involved in cell cycle progression. We demonstrate that depletion of EZH2 significantly reduces SKP2 protein levels in several cell types, while treatment with EPZ-6438, an EZH2 enzymatic inhibitor, has no effect on SKP2 protein levels. Consistently, EZH2 depletion leads to cell cycle arrest, accompanied by elevated expression of CIP/KIP family proteins, including p21, p27, and p57, whereas EPZ-6438 treatment does not modulate their levels. We also provide evidence that EZH2 knockdown, but not enzymatic inhibition, suppresses SKP2 mRNA expression, underscoring the transcriptional regulation of SKP2 by EZH2 in a methyltransferase-independent manner. Supporting this, analysis of the Cancer Genome Atlas database reveals a close association between EZH2 and SKP2 expression in human malignancies. Moreover, EZH2 depletion but not enzymatic inhibition positively regulates the expression of major epithelial-mesenchymal transition (EMT) regulators, such as ZEB1 and SNAIL1, in transformed cells. Our findings shed light on a novel mechanism by which EZH2 exerts regulatory effects on cell proliferation and differentiation through its methyltransferase-independent function, specifically by modulating SKP2 expression.

Figure 1

EZH2 knockdown induces expression of CIP/KIP family CKIs. (A) HeLa cells synchronized by double-thymidine block were released into the cell cycle. Cell lysates post release were blotted for PRC2 components (e.g., SUZ12, EZH2, BMI1, Ring1A), and p-H3, H3K9me3 and β-actin. Modified EZH2 and BMI1 with slow mobility are marked (in blue). (B) HeLa cells transfected with EZH2 or CNTL (siLuc) siRNAs for 48 h were collected, and cell lysates were blotted for various cell cycle proteins as indicated. (C) HeLa cells were treated with EPZ-6438 (an EZH2 inhibitor) or vehicle (DMSO) for 48 hor transfected with EZH2 (siEZH2) or CNTL (siControl) siRNAs for 48 h. Cells of both treatments were then stained with Click-iT EdU and FxCycle Violet (Invitrogen) for flow cytometry. (D) Cell cycle distributions (based on DNA content only) of various treatments as shown in (B) were summarized. (E) HeLa cells were treated with EPZ-6438 or vehicle (DMSO) for 48 h or transfected with an independent sequence targeting EZH2 (siEZH2) or CNTL (siControl) siRNAs for 48 h were collected, and cell lysates were blotted for various cell cycle proteins as indicated.

Figure 2

EZH2 knockdown induces expression of KIP family proteins in various cell lines. (A) HeLa and A549 cells were treated with EPZ-6438 (1 µM) or DMSO for 48 h or transfected with EZH2, or control (siLuc), siRNAs for 48 h. Cell lysates were then blotted for EZH2, p27, cyclin D3, cyclin E2, cyclin A2, cyclin B1, and H3K27me3. (B) HCT116 (WT or p53-null) cells were treated with EPZ-6438 (1 µM) or vehicle, or transfected with EZH2, or control (siLuc), siRNAs for 48 h. Cell lysates were blotted for various cell cycle markers as shown. (C) HeLa and non-transformed BEAS-2B cells were treated with EPZ-6438 (1 µM) or DMSO for 48 h or transfected with EZH2, SKP2 or control (siLuc), siRNAs for 48 h. Cell lysates were then blotted for various proteins as shown.

Figure 3

EZH2 knockdown but not enzymatic inhibition modulates SKP2 protein expression. (A) HeLa cells were treated with EPZ-6438 (1 µM) or transfected with EZH2 siRNAs (siEZH2) for 48 h, or synchronized with the thymidine block or by nocodazole treatment for 18 h. Cell lysates were then blotted for various cell cycle markers as shown. (B) HeLa cells were transfected with EZH1 siRNAs or EZH2 siRNAs or both for 48 h. Equal amounts of cell lysates were then blotted for polycomb group proteins (EZH1, EZH2, SUZ12 and EED), histone modifications (p-H3 and H3K27me3), and cell cycle markers including SKP2, p27 and p57. (C) HeLa cells were transfected with FLAG-tagged EZH2 or vector control (CNTL) for 24 h. Asynchronized cells or cells synchronized by thymidine block (G1/S) or treatment with nocodazole (M) were collected and lysed. Equal amounts of cell lysates were immuno-precipitated with the anti-FLAG antibody. Immunoprecipitants, along with cell lysate inputs of various treatments, were blotted for FLAG, SKP2, SUZ12, phospho-H3, cyclin D3, and β-actin. (D) HeLa cells transfected with EZH2 or CNTL (siLuc) siRNAs for 24 h. Cells treated with 50 µg/ml of cycloheximide (CHX), a protein-synthesis inhibitor, were collected at the indicated times. Cell lysates were blotted for EZH2 and SKP2 as indicated. (E) The graph reflects EZH2 and SKP2 band density relative to the β-actin level quantified by ImageJ.

Figure 4

EZH2 knockdown downregulates SKP2 mRNA and modulates EMT. (A) HeLa cells were treated EPZ-6438 (1 µM) or vehicle (DMSO) for 48 h or (B) transfected with EZH2 or control siRNA (siLuc) for 48 h. Total RNAs were extracted from cells of both experiments and subjected to quantitative PCR analysis. Genes being analyzed include p16, p21, p27, and p57, as well as EZH2 and SKP2. Significance was analyzed using PRISM software as described in Methods. Error bars indicate triplicates. (C) Transient luciferase assay detection of the activity of the SKP2 promoter. HEK293T cells were co-transfected with the pGL4 construct containing the full‑length SKP2 promoter driving the luciferase gene and EZH2 WT or EZH2^H689A. Following 24 h transfection, luciferase activity was determined. All data shown are the mean ± SEM from experiments performed in triplicate. (D and E) Violin plots reporting the relative expressions of EZH2 and SKP2 in paired adenocarcinomas (designated as C) or matched normal tissues (designated as N) among lung, colon and pancreatic cancer patients’ groups from the “Cancer Genome Atlas” database. P values were computed using the Wilcoxon rank sum test and adjusted using the Benjamini–Hochberg method. ****, adjusted P value < 0.0001; ns represents P > 0.05. (F) HeLa cells were treated EPZ-6438 (1 µM) or vehicle (DMSO) or transfected with EZH2 (siEZH2) or control (siLuc) siRNAs for 48 h. Equal amounts of cell lysates were then blotted with various EMT markers and β-actin.