M6A Modified miR-31-5p Suppresses M1 Macrophage Polarization and Autoimmune Dry Eye by Targeting P2RX7

Abstract

The dysregulation of the M1/M2 macrophage balance plays a pivotal role in autoimmune diseases. However, the interplay between microRNAs (miRNAs) and N6-methyladenosine (m6A) modulation in regulating this balance remains poorly understood. Here, a significant reduction in miR-31-5p levels is observed in the lacrimal glands of rabbit autoimmune dacryoadenitis and the peripheral blood mononuclear cells (PBMCs) of Sjögren's syndrome (SS) dry eye patients. Overexpression of miR-31-5p exhibits preventive and therapeutic effects on rabbit autoimmune dacryoadenitis. Further investigation revealed that miR-31-5p overexpression significantly restored the M1/M2 macrophage balance both in vivo and in vitro. Mechanistically, miR-31-5p directly targets the P2x7 receptor (P2RX7), leading to the inactivation of p38 mitogen-activated protein kinases (MAPK) signaling and reduced expression of M1 markers. Furthermore, methylated RNA immunoprecipitation and luciferase reporter assays demonstrated that fat mass and obesity-associated protein (FTO)-mediated m6A demethylation, which sustains pri-miR-31 stability, is responsible for the decreased miR-31-5p levels in autoimmune dry eye. Notably, PBMC samples from SS dry eye patients further support the link between reduced miR-31-5p levels and M1 macrophage activation observed in rabbits. Overall, these data highlight the critical role of the FTO/miR-31-5p/P2RX7/p38 MAPK axis in autoimmune inflammation, suggesting their potential as therapeutic targets for autoimmune dry eye.

Keywords: m6A; macrophage polarization; miR‐31‐5p; p2x7 receptor; sjogren's syndrome dry eye.

Figure 1

miR‐31‐5p is decreased in LGs of rabbit autoimmune dacryoadenitis and PBMCs of human SS dry eye patients. A) Small RNA sequencing workflow schematic. Autoimmune dacryoadenitis in rabbits was induced by transferring activated PBLs. At 6 weeks post‐induction, LGs of normal and model group rabbits were collected for small RNA sequencing (n = 3/group). B) Small RNA‐sequencing‐based scatter plot showing the differentially expressed miRNAs in LGs isolated from rabbit autoimmune dacryoadenitis relative to those from normal controls. Red dots indicate upregulated miRNAs and blue dots represent downregulated miRNAs C) Dysregulated miRNAs expression level in PBMCs from SS dry eye patients and healthy controls (n = 11 per group). D) Correlation between miR‐31‐5p expression and dry eye severity of SS dry eye patients was calculated by the Pearson correlation test. E) Real‐time qRT‐PCR analysis of miR‐31‐5p expression in LGs and PBMCs from model rabbits and normal controls (n = 3 per group). LG, lacrimal gland; pLGECs, purified lacrimal gland epithelial cells; PBLs: peripheral blood lymphocytes; FC: fold change; HC, healthy controls; SS, Sjogren's syndrome; BUT, tear break‐up time; CFS, corneal fluorescein staining. Data was shown as mean ± SD, and the differences were analyzed by Unpaired Student's t‐test. ^* p < 0.05, ^*** p < 0.001.

Figure 2

miR‐31‐5p overexpression prevents the development of rabbit autoimmune dacryoadenitis. A) Lentiviral vector plasmids used to overexpress miR‐31‐5p. B) Schematic diagram illustrating LV‐miR‐31‐5p administration at the early stage (day 1 post transfer) of rabbit autoimmune dacryoadenitis (2 × 10⁷ transducing units/eye). C) Representative corneal fluorescein staining images. D,E) Tear production, tear break‐up time and corneal fluorescein staining scores of each group of rabbits (n = 6/group). F,G) Representative H&E staining photographs and scores of LGs and conjunctivas in each group of rabbits (n = 6/group). Scale bars, 100 and 50 µm. Arrows indicate infiltrating lymphocytes. PBLs, peripheral blood lymphocytes; The data were shown as mean ± SD. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001. D,E) two‐way ANOVA; G) one‐way ANOVA.

Figure 3

Overexpression of miR‐31‐5p restores the M1/M2 macrophage balance in vivo. LGs were collected from miR‐31‐5p overexpressing‐rabbits or control rabbits at week 8 after adoptive transfer of activated PBLs, and then subjected to real‐time qRT‐PCR or Western blot analysis. A) The relative expression of M1‐related genes. B) The mRNA expression of M2‐associated genes. C) The protein levels of Arg1 and NOS2. LV‐miR‐31‐5p, miR‐31‐5p‐overexpressing rabbits; LV‐Ctrl, control rabbits. Data was from at least three independent experiments and presented as mean ± SD. Data was analyzed by one‐way ANOVA. ^* p < 0.05, ^** p < 0.01.

Figure 4

miR‐31‐5p inhibits M1 macrophage activation and facilitates macrophages into M2 phenotype in vitro. A) Scheme of experimental procedures to evaluate the role of miR‐31‐5p on macrophage polarization. B,C) PBMCs isolated from model rabbits were transfected with miR‐31‐5p mimics of control mimics (300 nmol) and cocultured with irradiated pLGECs for 48h. Relative expression of M1 markers and M2‐related genes were analyzed by real‐time qRT‐PCR. D–I) Macrophage derived from THP‐1 cells were transfected with miR‐31‐5p mimics, miR‐31‐5p inhibitors or their negative controls for 24 h, and then stimulated with LPS and IFN‐γ to induce M1 macrophage polarization. D–G) Real‐time qRT‐PCR analysis of M1 and M2 macrophage‐related gene expression. H,I) Western blot analysis of Arg1 and NOS2 protein level. J,K) Representative confocal images of Arg1 (red) and NOS2 (green) immunofluorescence staining in macrophages transfected with miR‐31‐5p mimics or control mimics. Scale bars, 20 µm. Data were representative of at least three independent experiments and were analyzed by Unpaired Student's t‐test or Mann–Whitney U test. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ns, not significant.

Figure 5

P2RX7 is a functional target of miR‐31‐5p. A) Sequence alignment between miR‐31‐5p and its potential binding sites (in red letters) in the 3^′UTR of R2RX7 mRNA. The mutation of the miR‐31‐5p binding sites is shown in bule. B) Luciferase activity analysis of reporter carrying the wild‐type or mutant P2RX7 3^′UTR co‐transfected into HEK293T cells with miR‐31‐5p mimics or control mimics. C–E) Macrophage derived from THP‐1 cells were transfected with indicated mimics or inhibitors, and the expression of P2RX7 was measured by real‐time qRT‐PCR and western blot. F,G) Real‐time qRT‐PCR analysis of P2RX7 expression in LGs of each group of rabbits. H–N) Macrophage derived from THP‐1 cells were transfected with indicated siRNA or inhibitors for 24 h, and then stimulated with LPS+IFN‐γ to induce M1 macrophage polarization. H) Real‐time qRT‐PCR analysis of P2RX7 expression in each group of cells. I–N) The relative expression of M1 and M2 macrophage associated genes, detected by real‐time qRT‐PCR or western blot, is shown. Data were shown as mean ± SD from at least three independent experiments. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ns, not significant. (B‐K) Unpaired Student's I‐test; L–N) one‐way ANOVA.

Figure 6

miR‐31‐5p/P2RX7 inhibits M1 macrophage polarization by suppressing p38 MAPK signaling. A) KEGG analysis on target genes of miR‐31‐5p. The top 10 pathways are summarized. B–F) THP‐1 derived macrophages were transfected with indicated mimics, siRNA or inhibitors, and then stimulated with LPS + IFN‐γ to induce M1 macrophage polarization. The phosphorylation level of p38, JNK and ERK were detected by western blot 48 h later. G) THP‐1‐derived macrophages were pretreated with 10 µm SB203580 p38 inhibitor or DMSO for 1 h prior to stimulation with LPS and IFN‐γ to induce M1 macrophage polarization. The relative mRNA expression of M1‐related genes was then examined by real‐time qRT‐PCR. I) THP‐1‐derived macrophages were transfected with miR‐31‐5p inhibitor or Ctrl inhibitor following pretreatment with 10 µm SB203580 p38 inhibitor or DMSO for 1h. The levels of M1‐associated genes were analyzed. Data were shown as mean ± SD from at least three independent experiments. ^* p < 0.05, ^** p < 0.01, ns, not significant. (C‐D and G) Unpaired Student's t‐test or Mann–Whitney U test; (F and H) one‐way ANOVA.

Figure 7

m6A methylation regulates the expression of miR‐31‐5p in autoimmune dry eye. A) The sequences of pri‐miR‐31, miR‐31‐5p, and the potential m6A motif (GAACU) were highlighted with different colors. B) Relative expression of m6A enzymes (METTL3, METTL14, WTAP and FTO) in LGs of normal and model rabbits were detected by real‐time qRT‐PCR. C,D) PBMCs from model rabbits were transfected with indicated FTO siRNA. The relative expression of FTO, pri‐miR‐31‐5p and miR‐31‐5p was detected by real‐time qRT‐PCR. E) Pri‐miR‐31 levels in FTO knockdown and negative control PBMCs after actinomycin D treatment at the indicated times. F) Flow diagram of MeRIP‐qPCR assays. G) MeRIP‐qPCR analysis of the m6A levels of pri‐miR‐31 in FTO silenced PBMCs and control cells. H) Luciferase activity analysis was performed on a reporter carrying the wild‐type or mutant sequence of pri‐miR‐31, which was co‐transfected into HEK293T cells with either FTO siRNA or control siRNA. I) Total protein from HEK293T and THP‐1 cells was immunoprecipitated with an anti‐DGCR8 antibody. Western blots for FTO and DGCR8 are shown. J) Detection of the abundance of pri‐miR‐31 binding to DGCR8 in HEK293T cells by immunoprecipitation with an anti‐DGCR8 antibody. Data were shown as mean ± SD, and the differences were analyzed by one‐way ANOVA, Unpaired Student's t‐test or Mann–Whitney U test. ^* p < 0.05, ^** p < 0.01, ns, not significant.

Figure 8

Overexpression of miR‐31‐5p at the developed stage efficiently alleviates autoimmune dacryoadenitis in Rabbits. A) Schematic diagram illustrating LV‐miR‐31‐5p administration (2 × 10⁷ transducing units/eye) at the developed stage (day 15 post transfer) of rabbit autoimmune dacryoadenitis. B) Representative corneal fluorescein staining images. C) Tear production, tear break‐up time and corneal fluorescein staining scores of each group of rabbits (n = 5/group). D) Representative specimens of H&E staining in LGs and conjunctivas. Scale bars, 100 and 50 µm. Arrows indicate infiltrating lymphocytes. PBLs, peripheral blood lymphocytes. Data was shown as mean ± SD and analyzed by two‐way ANOVA. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001.

Figure 9

Downregulation of miR‐31‐5p is associated with increased M1 macrophage activation in SS dry eye patients. A) Schematic outline of experimental procedures designed to assess gene expression in patients with SS dry eye patients. B) Expression levels of macrophage‐related genes in PBMCs from SS dry eye patients and controls subjects. C) Correlation analysis between miR‐31‐5p expression and NOS2 levels in PBMCs of SS dry eye patients. D,E) Quantification of FTO mRNA levels in PBMCs of SS dry eye patients and its correlation with miR‐31‐5p expression. F,G) Expression of P2RX7 in PBMCs of SS dry eye patients and its association with miR‐31‐5p level. H,I) PBMC‐derived macrophages from SS dry eye patients were transfected with miR‐31‐5p mimics or negative controls and subsequently polarized into M1 macrophages. The expression of M1 and M2 macrophage‐related gene was measured by real‐time qRT‐PCR. Data was shown as mean ± SD and analyzed by Unpaired Student's t‐test or Mann–Whitney U test. ^* p < 0.05, ^** p < 0.01.

Figure 10

Schematic depicting the mechanisms by which miR‐31‐5p regulates M1/M2 balance in autoimmune dry eye. The upregulation of FTO in autoimmune dry eye suppresses the recognition of pri‐miR‐31 by DGCR8, thereby inhibiting the maturation of miR‐31‐5p and reducing its inhibitory effect on the target gene P2RX7. The activation of P2RX7 subsequently triggers the p38 MAPK signaling pathway and increases the level of M1‐related genes, thus exacerbating the development of autoimmune dry eye. Conversely, overexpression of miR‐31‐5p converts M1 macrophages into an M2 phenotype, which secretes anti‐inflammatory mediators and alleviates the symptoms of autoimmune dry eye.