EZH2 Inhibition Interferes With the Activation of Type I Interferon Signaling Pathway and Ameliorates Lupus Nephritis in NZB/NZW F1 Mice

Abstract

Enhancer of zeste homolog 2 (EZH2) is a histone-lysine N-methyltransferase mediating trimethylation of H3K27, which represses gene expression and is critical to immune regulation. Inhibition of EZH2 is proved to have the potential of treating many diseases. However, whether inhibition of EZH2 affects type I interferon (IFN-I) signaling pathway, the abnormality of which is an important pathogenic mechanism for SLE, is still elusive. Here, we report, unexpectedly, a positive regulatory function of EZH2 in IFN-I signaling pathway, which contributes to the overactivation of IFN-I signaling pathway in SLE. We show that the expression of EZH2 was upregulated and positively correlated with the overexpression of interferon stimulated genes (ISGs) in both peripheral blood mononuclear cells and renal tissues of SLE patients. In vitro inhibition of EZH2 by either siRNAs or chemical inhibitors reduced the phosphorylation of STAT1 and the induction of ISGs stimulated by IFN-I. Additionally, inhibition of EZH2 interfered with the in vivo and ex vivo activation of IFN-I signaling pathway elicited by intravenous injection of adenovirus vector expressing mouse IFN-α5 and exogeneous stimulation with IFN-α, respectively. We evaluated the therapeutic effects of EZH2 inhibitor in NZB/NZW F1 mice which depend on IFN-I signaling pathway for the lupus-like disease development. Administration of EZH2 inhibitor prolonged the survival, reduced the levels of anti-dsDNA autoantibodies, and improved lupus nephritis of the mice. What's more, EZH2 inhibitor attenuated the expression of ISGs in the kidneys of these mice. In summary, we show that excessive EZH2 contributes to the overactivation of IFN-I signaling pathway in SLE. EZH2 inhibitor has the potential to inhibit IFN-I signaling pathway and alleviate lupus nephritis. Additionally, diverse disease driving pathways exist among systemic lupus erythematosus (SLE) patient, and even in the same patients. Common regulators of different pathogenic pathways can be multivalent therapeutic targets. Together with previous studies showing EZH2 is involved in T-cell and B-cell mediated immune responses, EZH2 could be a potent multivalent therapeutic target for SLE.

Keywords: enhancer of zeste homolog 2; epigenetics; multivalent therapeutic target; systemic lupus erythematosus; type I interferon.

Figure 1

Overexpression of EZH2 associates with the upregulation of ISGs in SLE. (A) The levels of EZH2 mRNA in the PBMCs from SLE patients (n = 30) and healthy controls (n = 30). (B, C) Positive correlation between the expression levels of EZH2 and CXCL10 (B) and IFIT3 (C) in the PBMCs from SLE patients (n = 30). (D–F) RNA-seq analysis was previously performed using kidney biopsies from SLE patients (n = 22) and the carcinoma–adjacent normal kidney tissues (n = 7) (29). (D) The levels of EZH2 transcript in the kidney biopsies from SLE patients and in the carcinoma–adjacent normal kidney tissues. (E, F) Positive correlation between the expression levels of EZH2 and CXCL10 (E) and IFIT3 (F) in the kidney tissues from SLE patients. (A, D) data are presented as mean ± SEM. P values were determined by Mann-Whitney U-test. (B, C, E, F) Each dot represents individual patients. P values were determined by Spearman’s correlation test.

Figure 2

EZH2 regulates the activation of IFN-I signaling pathway. (A) THP-1 cells were transfected with 200 nM of EZH2 targeting siRNAs (siEZH2) or negative control siRNAs (NC). Twenty-four hours after transfection, THP-1 cells were stimulated with 1,000 U/ml of universal type I IFN. (A) The levels of CXCL10 mRNA and IFIT3 mRNA were measured at the indicate time points. (B) Primary monocytes were electroporated with 200 nM of EZH2 targeting siRNAs (siEZH2) or negative control siRNAs (NC). Twenty-four hours after electroporation, the cells were stimulated with 1,000 U/ml of universal type I IFN for 6 h. (B) The levels of CXCL10 mRNA and IFIT3 mRNA were measured. (C, D) RNA-seq analysis of the THP-1 cells stimulated with 1,000 U/ml of universal type I IFN with the perturbation of EZH2 or not. (C) Top 20 most affected ISGs. (D) The 21 ISGs constitute IFN signature scoring panel (33, 34). (E) IFN-I induced phosphorylation of STAT1 was determined in THP-1 cells transfected with 200 nM of siEZH2 or NC by Western bolt. Representative pictures of at least three independent experiments are shown. Band densities quantified by Image J (A) data are presented as mean ± SEM from at least three independent experiments. P values were determined by two-way ANOVA. (B) Each pair of dots represents an individual donor (n = 3). P values were determined by Student’s paired t-test. *P < 0.05, **P < 0.01. (C, D) Each row represents individual genes. Each column is an individual treatment, NC: negative control siRNAs without IFN-I stimulation; NC (+): negative control siRNAs with IFN-I stimulation; siEZH2 (+): EZH2 targeting siRNAs with IFN-I stimulation.

Figure 3

GSK126 attenuates the activation of IFN-I signaling pathway. (A, B) THP-1 cells were pretreated with different concentrations of GSK126 for 30 min, then stimulated with universal type I IFN (1,000 U/ml) for 4 h. The levels of CXCL10 mRNA (A) and IFIT3 mRNA (B) were measured. (C, D) IFN-I induced phosphorylation of STAT1 was determined in THP-1 cells pretreated with 5 μM of GSK126 by Western bolt. (C) Representative pictures of at least three independent experiments. (D) Band densities quantified by Image J (A, B, D) Data are presented as mean ± SEM. (A, B) P values were determined by Mann-Whitney U-test. *P < 0.05, ns, not significant. (D) P value was determined by 2-way ANOVA.

Figure 4

GSK126 inhibits ex vivo activation of IFN-I signaling pathway. (A) Experimental setting for the ex vivo tests. The mice were treated with a single dose of GSK126 (50 mg/kg, i.p.) and then sacrificed after 2 days. Splenocytes were isolated and stimulated with universal type I IFN (1,000 U/ml) for indicated time. (B, C) The levels of ISG15 mRNA (B) and IFIT3 mRNA (C) were measured. (A) Created with BioRender.com. (B, C) Data of three independent experiments are plotted and presented as mean ± SEM. P values were determined by two-way ANOVA.

Figure 5

GSK126 inhibits in vivo activation of IFN-I signaling pathway. (A) Experimental setting for the in vivo tests. Each mouse received 10⁹ pfu of IFNα5-encoding adenovirus particles, or PBS (as control), for in vivo activation of IFN-I signaling pathways. Three days later, the mice received adenoviral particles were given GSK126 (50 mg/kg, i.p.), or vehicle, for 3 consecutive days. (B) The levels of ISG15 mRNA and IFIT3 mRNA in the spleens were measured. (C) The levels of ISG15 mRNA and IFIT3 mRNA in the kidneys were measured. (A) Created with BioRender.com. (B, C) Each dot represents individual mice (n ≥ 5 for each group). P values were determined by Mann-Whitney U-test.

Figure 6

GSK126 alleviates lupus nephritis in NZW/NZB F1 mice. (A) Experimental setting. Briefly, 24–28 weeks old female NZB/NZW F1 mice were randomized and grouped based on their urinary protein levels. The mice were treated with GSK126 (50 mg/kg, i.p.), or vehicles, every other day for 4 weeks. Peripheral blood and 24-h urine were collected at indicated time points. Kidneys were harvested 2 days after the last treatment for histopathological analysis. (B) Kaplan-Meier survival curve (n ≥ 10 for each group). (C) Anti-dsDNA autoantibody in the plasma. (D) Twenty-four-hour urinary protein. (E) Representative pictures of renal H&E staining and immunofluorescent staining of glomerular deposition of immunocomplex. Renal histopathological scores were evaluated for the mice in GSK126 group (n = 10) and vehicle group (n = 5). GN, glomerulonephritis; TI, tubulointerstitial nephritis. (F) The levels of ISG15 mRNA and IFIT3 mRNA in the kidneys from the mice treated with GSK126 or vehicle. (A) Created with BioRender.com. (B) P value was determined by Log-rank test. (C–E right panel, F) Each dot represents individual mice. Data are presented as mean ± SEM. P values were determined by Mann-Whitney U-test.