PRMT5-Mediated ALKBH5 Methylation Promotes Colorectal Cancer Immune Evasion via Increasing CD276 Expression

Abstract

Numerous diseases have been connected to protein arginine methylations mediated by protein arginine methyltransferase 5 (PRMT5). Clinical investigations of the PRMT5-specific inhibitor GSK3326595 are currently being conducted, and the results are promising for preventing cancers. However, the detailed mechanism of PRMT5 promoting colorectal cancer (CRC) malignant progression remains unclear. Here, we found that PRMT5 directly catalyzes AlkB homologue 5 (ALKBH5) symmetric dimethylation at the R316 residue (meR316-ALKBH5), which enhances TRIM28-mediated ALKBH5 ubiquitination degradation. Then, an ALKBH5 decrease attenuates ALKBH5-mediated m6A demethylation on the CD276 transcript 3' untranslated region, which increases CD276 messenger RNA stability and its expression in CRC cells. Furthermore, a CD276 expression increase facilitates CRC immune evasion by inhibiting cytotoxic T-cell functions. Moreover, we revealed that PRMT5-mediated meR316-ALKBH5 activates CD276 transcription by increasing its messenger RNA m6A modification to increase CRC immune evasion in vitro and in vivo. Furthermore, we consistently showed a strong association between meR316-ALKBH5 and poor outcomes in patients with CRC. Finally, we demonstrated that combining an anti-PD1 antibody with the PRMT5 inhibitor GSK3326595 markedly halts the progression of CRC. Our findings could serve as a basis for the development of a PRMT5-meR316-ALKBH5-CD276 axis-targeting treatment approach for CRC.

Fig. 1.

Protein arginine methyltransferase 5 (PRMT5) inhibitor or knockdown reduces the global level of m6A methylation modification in colorectal cancer cells. (A) Modified nucleosides of colorectal cancer cells treated with GSK595 or DMSO were quantified using liquid chromatography coupled with tandem mass spectrometry (LC–MS/MS). The relative amounts of modified nucleosides in the comparative group (GSK595/DMSO) are shown in bar plots. log₂(ratio) was taken to symmetrize the ratio; log₂(ratio) > 0 indicates that the nucleoside is up-regulated in the comparison group, and log₂(ratio) < 0 indicates that the nucleoside is down-regulated in the comparison group. The m6A level intensely declined. (B) Scatter plots displaying the change in each modified nucleoside in GSK595/DMSO; m6A modification shows significant down-regulation. (C and D) Dot-blot assay of RNA m6A methylation in HCT116 and DLD1 colorectal cancer cells treated with GSK595 for the dose and duration indicated. (E and F) Dot-blot assay of RNA m6A methylation in HCT116 and DLD1 colorectal cancer cells down-regulating or overexpressing PRMT5. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Fig. 2.

PRMT5 weakens AlkB homologue 5 (ALKBH5) stability by increasing ALKBH5 ubiquitination. (A and B) Cell lysates from HCT116 or MC38 cells were immunoprecipitated with antibodies against indicated proteins followed by immunoblotting with various antibodies indicated. (C and D) Flag-ALKBH5 and Myc-PRMT5, Flag-Alkbh5, and Myc-Prmt5 were coexpressed in HCT116 and MC38 cells, respectively. After immunoprecipitation with appropriate antibodies, bound proteins were examined by Western blotting. (E) Glutathione-S-transferase (GST)-fused ALKBH5 was incubated with HEK293T cell lysates. The binding proteins by GST pull-down assays were examined by Western blotting with the indicated antibodies. The amounts of GST and GST-tagged ALKBH5 were visualized by Coomassie blue staining. (F and G) ALKBH5 protein and messenger RNA (mRNA) levels were analyzed by Western blot and quantitative reverse transcription polymerase chain reaction (qRT-PCR) experiments in HCT116, DLD1, and MC38 cells treated by the PRMT5 inhibitor GSK595. (H and I) Protein and mRNA expression of ALKBH5 were assessed by Western blot and qRT-PCR assays after overexpression PRMT5. (J and K) ALKBH5 proteins and mRNAs were detected by Western blot and qRT-PCR after knockdown of PRMT5. (L and M) The effect of PRMT5 pharmacological (GSK595) or genetic (PRMT5 knockdown) inhibition on ALKBH5 protein expression was detected by Western blot in HCT116 cells following treatment by CHX (50 μg/ml) for the indicated time (left). The relative intensities of ALKBH5 proteins were quantified by the software ImageJ (right). For normalization, GAPDH expression was used as a control. *P < 0.05; ***P < 0.001. (N) Western blot showing the effects of proteasome inhibitor MG132 (10 μM for 8 h) treatment on ALKBH5 protein accumulation in HCT116 cells. (O) Western blot showing effects of proteasome inhibitor MG132 (10 μM for 8 h) treatment on ALKBH5 protein accumulation in MC38 cells. (P) PRMT5 was overexpressed in HCT116 cells, and immunoprecipitation was performed using an anti-ALKBH5 antibody, which suggested that PRMT5 could bind to ALKBH5 and increase the ubiquitination degradation level of ALKBH5. (Q and R) PRMT5 was pharmacologically (GSK595) or genetically inhibited (PRMT5 knockdown) in HCT116 cells, and immunoprecipitation was performed using an anti-ALKBH5 antibody, which suggested that PRMT5 could decrease ALKBH5 ubiquitination degradation. IP, immunoprecipitation; IgG, immunoglobulin G; ns, not significant.

Fig. 3.

PRMT5-mediated meR316-ALKBH5 decreases ALKBH5 stability by TRIM28-mediated ALKBH5 degradation. (A) Symmetric dimethylation (SDMA) and MMA of endogenous ALKBH5 in HCT116 cells were assessed by IP assays. (B) In vitro methylation assays. Purified GST-tag fusion proteins of GST–ALKBH5 (292 to 395 AA) were incubated with the PRMT5/MEP50 complex. Methylated proteins were detected via immunoblotting, and the total amounts of proteins were visualized by Coomassie blue staining (arrows: position of GST–ALKBH5 [292 to 395 AA]). (C) PRMT5 and Flag-ALKBH5 were overexpressed in HCT116 cells simultaneously, and IP was performed using an anti-Flag antibody, which suggested that PRMT5 could increase the SDMA level of ALKBH5. (D) Detection of ALKBH5 SDMA modification status after IP Flag-ALKBH5 in PRMT5 down-regulated HCT116 cells. (E) Detection of ALKBH5 SDMA modification status after IP Flag-ALKBH5 in HCT116 cells treated with GSK595. (F) In vivo methylation detection of SDMA-ALKBH5 levels in HEK293T cells overexpression of wild-type (WT) ALKBH5 and R249K/R250K/R269K/R283K/R316K/R357K/R359K mutant ALKBH5, respectively. (G) Purified GST-tag fusion proteins of WT and R316K GST–ALKBH5 (292 to 395 AA) were incubated with the PRMT5/MEP50 complex in in vitro methylation assays. Methylated proteins were detected via immunoblotting, and the total amounts of proteins were visualized by Coomassie blue staining. (H) Mass spectrometry (MS) analysis of ALKBH5 methylation. The fragmentation of the ALKBH5 peptide LSGNNRDPALKPK identified a dimethylated residue at R316. (I) Sequence alignments of ALKBH5 in mammals. The ALKBH5-R316 site is denoted in the protein sequences. (J) The ALKBH5 expression was detected by Western blotting after expression of ALKBH5-WT or ALKBH5-R316K in HCT116 cells treated with CHX (50 μg/ml). The bands of ALKBH5 proteins treated by CHX were quantified by the software ImageJ. *P < 0.05; **P < 0.01. (K) Western blots of Flag-ALKBH5-associated ubiquitination after IP ALKBH5-Ub in HCT116-ALKBH5-WT cells and HCT116-ALKBH5-R316K cells. (L) Western blot analysis of exogenous interaction between ALKBH5 and TRIM28 after IP Tagged-ALKBH5 in HCT116 cells. (M) TRIM28 overexpression decreased ALKBH5 protein half-life. CHX (50 μg/ml) was added in HCT116 and TRIM28-overexpressing HCT116 cells at the indicated time points. Cell lysates were then subjected to immunoblotting analyses. **P < 0.01; ***P < 0.001. (N) Western blot showing effects of proteasome inhibitor MG132 (10 μM for 8 h) treatment on ALKBH5 protein accumulation in HCT116 cells. (O) Western blots of meR316-ALKBH5 and TRIM28 binding with Flag-ALKBH5 after IP Flag-ALKBH5 in Flag-ALKBH5- and PRMT5-coexpressing HCT116 cells compared to HCT116 Flag-ALKBH5 cells. (P) Western blots of meR316-ALKBH5 and TRIM28 interaction with ALKBH5 in shCtrl and shPRMT5 cells after IP Flag-ALKBH5. (Q) Western blots detected meR316-ALKBH5 and TRIM28 interacting with ALKBH5 after coimmunoprecipitation (Co-IP) Flag-ALKBH5 from HCT116 cells treated by PRMT5 inhibitor GSK595. (R) Western blot analysis of meR316-ALKBH5 and TRIM28 binding with ALKBH5 after IP Flag-ALKBH5 in HCT116-ALKBH5-WT cells and HCT116-ALKBH5-R316K cells. AA, amino acids.

Fig. 4.

meR316-ALKBH5 plays a pivotal role in regulating colorectal cancer cells’ immune evasion in vitro and in vivo. (A) ALKBH5 expression positively associated with 3 kinds of tumor-infiltrating immune cells (TIICs) in TIMER2.0. (B) Effect of ALKBH5-WT or ALKBH5-R316K on HCT116 cells’ proliferation with T-cell killing as assessed by Real-Time Cell Analyzer (RTCA, xCELLigence) assays. (C) MC38 cells were injected into C57BL mice after the transfection of Alkbh5-WT, Alkbh5-R317K, and the vector respectively. (D and E) The growth curves and weights of tumors from subcutaneously injected C57BL mice treated with Alkbh5-WT, Alkbh5-R317K, and the vector separately. (F) Representative images of the immunohistochemistry (IHC) staining of methylated ALKBH5 (meALKBH5), Ki-67, CD3, CD8a, and GZMB in the tumor xenografts. (G) Quantification of meALKBH5, Ki-67, CD3, CD8a, and GZMB staining intensity in xenograft tumors. *P < 0.05; ***P < 0.001.

Fig. 5.

PRMT5 and ALKBH5 regulate CD276 expression in an N6-methyladenosine (m6A)-dependent manner. (A) Distribution of genes with a significant change in both the m6A level and the gene expression level in HCT116 cells treated with GSK595 compared with that in control cells. (B) Schematic of downstream analysis that was mediated by the PRMT5 inhibitor GSK595 in colorectal cancer cells using methylated RNA immunoprecipitation sequencing (MeRIP-seq) and RIP-seq. (C) The expression of PRMT5 is positively correlated with CD276 in colorectal cancer cells verified by the TCGA database. (D) Distribution of m6A peaks across the CD276 mRNA transcript. The 3′ untranslated region (3′UTR) of the CD276 mRNA is highlighted in the red box. (E and F) CD276 protein levels were analyzed by Western blot in HCT116 cells treated by the PRMT5 inhibitor GSK595 or overexpressing PRMT5. (G and H) CD276 mRNA expression were assessed by qRT-PCR assays after HCT116 cells were treated by the PRMT5 inhibitor GSK595 or overexpressed PRMT5. (I and J) The effect of GSK595 or PRMT5 on the stability of CD276 mRNA in HCT116 cells. (K) Schematic diagram of the mutation of the m6A site in CD276 3′UTR. For the CD276 3′UTR-Mut reporter, A–T substitutions (shown in red) were made within m6A consensus. (L and M) Relative luciferase activity of CD276 3′UTR-WT/CD276 3′UTR-Mut reporter when cotransfected with ALKBH5-WT/ALKBH5-R316K expression vector respectively. *P < 0.05; **P < 0.01; ***P < 0.001. FC, fold change; Luc, luciferase.

Fig. 6.

PRMT5-mediated ALKBH5-R316 methylation is associated with poor prognosis in colorectal cancer patients. (A) Representative immunohistochemistry images of PRMT5, meR316-ALKBH5, and CD276 expressions in PRMT5-low case and PRMT5-high case were presented in colorectal cancer patients. (B to D) Correlations between PRMT5 and meR316-ALKBH5 expression (B), PRMT5 and CD276 expression (C), and meR316-ALKBH5 and CD276 expression (D) were examined by Pearson correlation coefficient tests, respectively. (E to G) Correlations between PRMT5 and meR316-ALKBH5 expression (E), PRMT5 and CD276 expression (F), and meR316-ALKBH5 and CD276 expression (G) were examined by Fisher’s exact test, respectively. (H and I) Percentages of a high level of meR316-ALKBH5 expression correlated with tumor sizes (H) and lymph node metastasis (I) as examined by χ² tests. (J) Representative images of weak and strong meR316-ALKBH5 staining in colorectal cancer tissues. (K) High meR316-ALKBH5 expression correlated with a poorer 5-year overall survival for 212 colorectal cancer patients (P < 0.0001, log-rank test). (L) High meR316-ALKBH5 expression correlated with a poorer 5-year disease-free survival for 136 colorectal cancer patients (P < 0.0001, log-rank test). **P < 0.01; ***P < 0.001. CI, confidence interval; HR, hazard ratio.

Fig. 7.

GSK3326595 treatment eradicates colorectal cancer cells’ progress when anti-PD-1 is concurrently applied. (A) Down-regulated PRMT5 in HCT116 cells evaluates the specific cytotoxicity of T cells. (B) PRMT5 inhibitor GSK595 promotes the cytotoxicity of T cells targeting colorectal cancer cells. (C) Schematic diagram of C57BL mouse implanted with MC38 cells under different treatments. (D and F) Representative images (D), tumor growth curves (E), and tumor volume (F) of different treatments in each group. (G) Representative images of the IHC staining of meALKBH5, Ki-67, CD3, CD8a, GZMB, and CD276 in the tumor xenografts. (H) Quantification of meALKBH5, Ki-67, CD3, CD8a, GZMB, and CD276 staining intensity in xenograft tumors. (I) Proposed model to describe the role of PRMT5-mediated ALKBH5-R316 methylation in promoting m6A-dependent CD276 activation.*P < 0.05; **P < 0.01; ***P < 0.001. Ab, antibody; ip, intraperitoneal; PBS, phosphate-buffered saline.