Inhibition of EHMT2/G9a epigenetically increases the transcription of Beclin-1 via an increase in ROS and activation of NF-κB

Abstract

We previously reported that BIX-01294 (BIX), a small molecular inhibitor of euchromatic histone-lysine N-methyltransferase 2 (EHMT2/G9a), induces reactive oxygen species (ROS)-dependent autophagy in MCF-7 cells. Herein, we analyzed the epigenetic mechanism that regulates the transcription of Beclin-1, a tumor suppressor and an autophagy-related gene (ATG). Inhibition of EHMT2 reduced dimethylation of lysine 9 on histone H3 (H3K9me2) and dissociated EHMT2 and H3K9me2 from the promoter of Beclin-1. To this promoter, RNA polymerase II and nuclear factor kappa B (NF-κB) were recruited in a ROS-dependent manner, resulting in transcriptional activation. Moreover, treatment with BIX reversed the suppression of Beclin-1 by the cooperative action of EHMT2 and DNA methyltransferase 1 (DNMT1). Accordingly, a combination treatment with BIX and 5-Aza-2'-deoxycytidine (5-Aza-Cd), a DNMT1 inhibitor, exerted a synergistic effect on Beclin-1 expression. Importantly, high levels of EHMT2 expression showed a significant association with low levels of Beclin-1 expression, which was related to a poor prognosis. These findings suggest that EHMT2 can directly repress Beclin-1 and that the inhibition of EHMT2 may be a useful therapeutic approach for cancer prevention by activating autophagy.

Keywords: Beclin-1; EHMT2/G9a; autophagy; epigenetic regulation; histone methyltransferase.

Figure 1. Inhibition of EHMT2 increases Beclin-1 expression

a. GFP-MCF-7 cells were treated with 10 μM BIX for 2 h. Cells were examined using fluorescence microscopy. Scale bar, 50 μm. b. Two breast cancer cell lines (MCF-7, Beclin-1^+/−; HS578T, Beclin-1^+/+) were treated with 10 μM BIX for the indicated amounts of time. Beclin-1 transcripts and protein expression were determined by RT-PCR (lower panel) and Western blotting (upper panel), respectively. c. RT2 Profiler PCR analysis for ATG expression. This figure includes a “heat map,” which is the part of the figure containing colors (red, green and black), in which the color represents the expression level of the gene; red and green represent high and low expression, respectively. Expression levels are continuously mapped on the color scale provided at the top of the figure. The heat map represents the differential gene expression pattern in MCF-7 cells treated without or with 10 μM BIX for 4 or 24 h. d. Graph represents the fold-change in the level of mRNA transcript in cells treated with BIX for 24 h versus untreated cells. Each circle represents an autophagy gene; those with the greatest fold-changes are indicated in red. The central line represents no changes in expression; above the central line indicates genes with increased expression; below the line are those genes with reduced levels. Grey lines indicate a 2-fold increase or decrease. e. MCF-7 cells were transfected with siCont or EHMT2 siRNA (siEHMT2) for 48 h. Western blot analysis was performed using the indicated antibodies. f. MCF-7 cells were transfected with 2 μg pcDNA3 empty vector (MOCK) or 1 or 2 μg pcDNA3-Flag-EHMT2 (FLAG-EHMT2) for 24 h; pcDNA3 empty vector was used as a negative control. Western blot analyses were performed using the indicated antibodies.

Figure 2. Epigenetic transcriptional activation of Beclin-1 by EHMT2 inhibition

a. A pair of breast cancer cell lines (MCF-7: Beclin-1^+/− and HS578T: Beclin-1^+/+) were treated with 10 μM BIX, 10 μg/ml CHX, 1 μg/ml ACD, or combinations of these compounds as indicated for 4 h. Beclin-1 transcripts and protein expression were determined by RT-PCR (upper panel) and Western blotting (lower panel), respectively. b. MCF-7 cells were treated with 10 μM BIX and/or 1 μg/ml ACD for 4 h. Cell morphology was examined under a light microscope. Scale bar: 50 μm. c, d and e. MCF-7 cells and HS578T cells were treated with 10 μM BIX for 4 h or MCF-7 cells were transfected with 1 μM siEHMT2. Treated cells were analyzed by ChIP. The ChIP analysis performed with anti-H3K9me2, anti-EHMT2, and anti-RNA pol II antibodies was compared with normal rabbit IgG as a negative control. An equal amount (input) of DNA–protein complex was applied. Real-time quantification of the Beclin-1 promoter sequences was carried out with anti-H3K9me2 ChIP, anti-EHMT2 ChIP, and anti-RNA pol II ChIP (right panel). Results are presented relative to the input and are the fold-changes over the control expressed as means ± SEM of three independent experiments. *P < 0.001 compared with control or siCont by one-way ANOVA.

Figure 3. Transcription of Beclin-1 driven by NF-κB

a. The ChIP assay reveals the binding of endogenous NF-κB to the Beclin-1 promoter. Non-specific rabbit IgG was used as an antibody control. MCF-7 and HS578T cells were treated with 10 μM BIX for 4 h and then analyzed. An equal amount (input) of DNA–protein complexes was applied. Real-time quantification of Beclin-1 promoter sequences was carried out using anti-NF-κB/p65 ChIP in MCF-7 and HS578T cells (right panel). Data are presented relative to the input and are fold-changes over the control value expressed as means ± SEM of three independent experiments. *P < 0.001 compared with the control by one-way ANOVA. b. NF-κB was detected using an Alexa488 (green)-conjugated antibody against the 65kDa subunit. Nuclei were visualized with Hoechst 33258 (blue). In control cells, NF-κB/p65 immunoreactivity was preferentially observed in the cytoplasm. At 4 h after the addition of 10 μM BIX, NF-κB/p65 was mostly localized to the nucleus. Scale bar, 50 μm. Right panels show a quantification of NF-κB/p65 translocation. The percentage of cells with nuclear NF-κB/p65 is indicated, and error bars represent the S.D. (four experiments with at least 50 cells analyzed per condition and experiment). Data were analyzed by analySIS TS Auto (mean ± S.D., n = 3; paired t-test, *P < 0.001 compared with BIX). c. Western blot of BIX induced the nuclear translocation of NF-κB/p65. Nuclear and cytosolic fractions were separately isolated from each group and blotted with anti-NF-κB/p65, anti-GAPDH (cytoplasmic marker), and anti-Lamin B (nuclear marker) antibodies, respectively. d. Western blotting was performed with the specified antibodies (related to the region upstream of NF-κB) using lysates from MCF-7 cells exposed to 10 μM BIX for the indicated amounts of time. e. The effect of an NF-κB inhibitor, CAPE, on Beclin-1 expression. MCF-7 cells were incubated with 10 mM CAPE combined with or without 10 μM BIX for 4 h. Western blotting was performed with anti-Beclin-1 antibody.

Figure 4. Intracellular ROS-mediated activation of autophagy in response to EHMT2 inhibition

a. NF-κB nuclear translocation was inhibited by a ROS scavenger, NAC. MCF-7 cells were treated with 10 μM BIX for 4 h in the presence of 4 mM NAC. Cells were then examined by fluorescence microscopy. NF-κB was detected using a Alexa488 (green)-conjugated antibody against the 65 kDa subunit. Nuclei were visualized with Hoechst 33258 (blue). Scale bar, 50 μm. Lower panels show the quantification of translocated NF-κB/p65. The percentage of cells with nuclear NF-κB/p65 is indicated, error bars represent the S.D. (four experiments with at least 50 cells analyzed per condition and experiment). Data are expressed as the means ± SEM of three independent experiments. *P < 0.05 compared with BIX by one-way ANOVA. b. Inhibition of ROS by NAC resulted in the reduction of Beclin-1 expression. MCF-7 cells were treated with 10 μM BIX for 4 h in the presence or absence of 4 mM NAC. Western blotting was carried out using the specified antibodies.

Figure 5. Coordination of DNMT1 and EHMT2 in the suppression of Beclin-1 expression

a. Western blotting after IP showed that BIX treatment reduced the physical binding between EHMT2 and DNMT1. b. Representative Western blots for the detection of Beclin-1 expression before and after MCF7 cells were treated with 10 μM 5-Aza-Cd, 10 μM BIX, or a combination of both treatments. c. Methylation-specific PCR (MSP) analysis of DNA methylation in the Beclin-1 promoter region in MCF7 cells treated with 10 μM 5-Aza-Cd for 48 h, 10 μM BIX for 1 h, or a combination of both treatments.

Figure 6. Correlation of high EHMT2 and low Beclin-1 expression levels with a poor prognosis in breast cancer patients

a. Comparison of the expression of EHMT2 and Beclin-1 in lysates from normal and tumor breast tissue by Western blot. b. Western blots of EHMT2 and Beclin-1 in cell lysates from breast cancer patient tumor tissues. c, d. Breast cancer patients in the NKI (n = 295, left panel) and UNC (n = 223, right panel) cohorts were dichotomized based on the expression of Beclin-1 and EHMT2 and patients with relatively high expression levels of both Beclin-1 and EHMT2 or relatively low expression levels of both Beclin-1 and EHMT2 were identified for analysis. e. The correlation between Beclin-1 and EHMT2 expression in TCGA breast cancer patients was estimated using a Pearson's correlation test. f. Schematic diagram delineating the reactivation of Beclin-1. Treatment with BIX reduced the levels of H3K9me2, leading to an open chromatin structure. NF-κB can then be recruited to the promoter in ROS-dependent manner, resulting in increased transcription of Beclin-1 and the activation of autophagy.