PRMT5 Promotes Cyclin E1 and Cell Cycle Progression in CD4 Th1 Cells and Correlates With EAE Severity

Abstract

Multiple Sclerosis (MS) is a debilitating central nervous system disorder associated with inflammatory T cells. Activation and expansion of inflammatory T cells is thought to be behind MS relapses and influence disease severity. Protein arginine N-methyltransferase 5 (PRMT5) is a T cell activation-induced enzyme that symmetrically dimethylates proteins and promotes T cell proliferation. However, the mechanism behind PRMT5-mediated control of T cell proliferation and whether PRMT5 contributes to diseases severity is unclear. Here, we evaluated the role of PRMT5 on cyclin/cdk pairs and cell cycle progression, as well as PRMT5's link to disease severity in an animal model of relapsing-remitting MS. Treatment of T helper 1 (mTh1) cells with the selective PRMT5 inhibitor, HLCL65, arrested activation-induced T cell proliferation at the G1 stage of the cell cycle, suggesting PRMT5 promotes cell cycle progression in CD4⁺ T cells. The Cyclin E1/Cdk2 pair promoting G1/S progression was also decreased after PRMT5 inhibition, as was the phosphorylation of retinoblastoma. In the SJL mouse relapsing-remitting model of MS, the highest PRMT5 expression in central nervous system-infiltrating cells corresponded to peak and relapse timepoints. PRMT5 expression also positively correlated with increasing CD4 Th cell composition, disease severity and Cyclin E1 expression. These data indicate that PRMT5 promotes G1/S cell cycle progression and suggest that this effect influences disease severity and/or progression in the animal model of MS. Modulating PRMT5 levels may be useful for controlling T cell expansion in T cell-mediated diseases including MS.

Keywords: PRMT5; T cell; cell cycle; multiple sclerosis; relapsing-remitting experimental autoimmune encephalomyelitis.

Figure 1

PRMT5 inhibition arrests TCR-induced cell cycle progression at the G1/S checkpoint. (A, B) mTh1 cells were activated on anti-CD3/CD28 coated plates, treated with 5 µM HLCL65 or vehicle control (DMSO) and harvested at the time of plating (resting) or 24, 48 or 72h post-plating. They were analyzed by immunoblot with PRMT5 (A) or H4R3 (B). Data plotted represent relative protein expression at indicated timepoints. Plots are combined 4-6 experiments, n=8-13/timepoint/condition. (C) mTh1 cells were plated on anti-CD3/CD28 coated plates, treated with vehicle (DMSO), 5 μMHLCL65 or 5 μM EPZ, and pulsed with tritiated ³H-thymidine at 24, 48 or 72h post-plating with vehicle or HLCL65. Cells were harvested 16-18h later. This is a representative experiment with 6 technical replicates/group, analyzed by two-way ANOVA with Tukey’s multiple comparisons test. (D) Supernatants from mTh1 cells plated on anti-CD3/CD28 coated plates were collected at 4h, 8h, 18h, 24h and 48h ± HLCL65, EPZ or vehicle and analyzed for IL-2 cytokine expression by ELISA. Data pooled from 3 independent experiments. Data analyzed by two-way ANOVA with Tukey’s multiple comparisons test. (E) Diagram of the cell cycle stages including Gap 1 (G1), Synthesis (S), Gap 2 (G2) and Mitosis (M). Cells represent the stage the cell is in at each phase of the cell cycle. Red lines on the cell cycle represent cell cycle checkpoints. Created with BioRender.com, modified from a template. (F) Flow cytometry plots of activated mTh1 cells demonstrating the gating strategy and representative results in the negative control (no BrdU), vehicle (DMSO) or PRMT5 inhibitor (2.5 µM HLCL65). (G) Quantification of the results from analyses in (F), corresponding to four independent experiments. Graphs for Gap 0/Gap 1 (G0/G1), Synthesis (S) and Gap 2/mitosis (G2/M). Results analyzed by paired t-test, n=7 pairs. Graphs are line or bar graphs representing mean ± SD with individual data points plotted (A–D) or before-after (plot individual points) (G). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

PRMT5 inhibition suppresses Cyclin E1/CDK2 expression and Rb hyperphosphorylation in inflammatory T cells. (A) Diagram of the expression of cyclins and CDKs at the different cell cycle stages. Created with BioRender.com, modified from a template. (B–G) mTh1 cells were activated on anti-CD3/CD28 coated plates, treated with 5 µM HLCL65 or vehicle control (DMSO) and harvested at various timepoints for protein analyses. (B) Immunoblots of mTh1 cells probed with cell cycle proteins at resting, 24h and 48h post-plating. (C–G) Protein quantification of (C) Cyclin D1, (D) Cdk4, (E) Cyclin E1, (F) Cdk2, (G) ratio of phosphorylated Rb at S608 to total Rb. DMSO values are relative to resting. Data pooled from 4-6 independent experiments, n=6-10, student’s paired t-test. *p < 0.05, **p < 0.01.

Figure 3

PRMT5, SDM and Cyclin E1 expression in CNS infiltrates correlate with RR EAE disease severity and/or progression. (A) Experimental design diagram: graphical representation of the typical EAE course in SJL mice injected with PLP_(139-151). Created with BioRender.com. For actual EAE data in mice used for data in B through I, please see Figure S3 . (B) Mean scores calculated from mice injected with PBS instead of PLP (sham), mice collected at peak disease [day (D)14], in remission 1 (considered in remission if score dropped ≥1.5, D19-D25), in relapse (considered in relapse if score dropped by 1.5 or more and then rose by 1 or more, D28-33), in remission 2 (considered rem2 if mouse went through relapse and back to remission). (C) CNS infiltrating cells were collected at the times described in (B) and analyzed by immunoblot for PRMT5. (D–I) CNS infiltrating cell protein from individual mice or pooled from multiple mice (for conditions with low infiltrating CNS cell numbers) were used as samples for protein analyses of PRMT5, H4R3 and Cyclin E1 immunoblot and one aliquot was analyzed by flow cytometry for immune cell composition (and correlation analyses were performed. CNS infiltrating cell protein from individual mice or pooled from multiple mice (for conditions with low infiltrating CNS cell numbers) were used for protein analyses. A minimum of ~400,000 cells per sample were used and equal amounts of protein (5-15 μg) loaded and data were normalized to sham (C–I). (D, E) Spearman correlation of score vs. (D) PRMT5 or (E) H4R3 expression. (F, G) Pearson correlations were done between CD3⁺CD4⁺ T cell % and (F) PRMT5 or (G) H4R3 SDM expression in infiltrating CNS cells. (H, I) Correlation between Cyclin E1 expression and (H) PRMT5 expression or (I) score in infiltrating CNS cells. Pearson correlations were done for CD3⁺CD4⁺ T cells vs. protein expression or protein vs. protein. Spearman correlations were done for scores vs. protein expression. For (D–I), n=15-36 with technical replicates from n=59-83 total mice from 2-4 independent experiments. Graphs (B, C) quantify relative expression normalized to actin loading control and combined from 4 independent experiments, (9-18 sham, 13-16 peak, 21-35 rem1, 11-12 relapse, 4 rem2), biological samples were kept separate if possible, but pooled for each timepoint as necessary for each experiment, 1-2 technical replicates/experiment, graphs represent mean ± SD, *p < 0.05, ***p < 0.001, ****p < 0.0001.