METTL3-mediated N6-methyladenosine modification governs pericyte dysfunction during diabetes-induced retinal vascular complication

Abstract

Rationale: Microvascular complication is a major cause of morbidity and mortality among the patients with diabetes. Pericyte dysfunction is the predominant pathological manifestation of microvascular complication. N⁶-methyladenosine (m⁶A) serves as the most prevalent modification in eukaryotic mRNAs. However, the role of m⁶A RNA modification in pericyte dysfunction is still unclear. Methods: Quantitative polymerase chain reactions and western blots were conducted to detect the change of m⁶A RNA modification in pericytes and mouse retinas following diabetic stress. MTT assay, transwell migration assay, caspase 3/7 activity assay, calcein-AM/propidium iodide (PI) staining, and TUNEL staining were conducted to determine the role of METTL3 in pericyte biology in vitro. Retinal trypsin digestion, vascular permeability assay, and IB4-NG2 double immunofluorescent staining were conducted to determine the role of METTL3 in retinal pericyte dysfunction and vascular complication. RNA sequencing, RNA pull-down assays and immunoblots were conducted to clarify the mechanism of METTL3-mediated pericyte dysfunction and vascular complication. Results: The levels of m⁶A RNA methylation were significantly up-regulated in pericytes and mouse retinas following diabetic stress, which were caused by increased expression of METTL3. METTL3 regulated the viability, proliferation, and differentiation of pericytes in vitro. Specific depletion of METTL3 in pericytes suppressed diabetes-induced pericyte dysfunction and vascular complication in vivo. METTL3 overexpression impaired pericyte function by repressing PKC-η, FAT4, and PDGFRA expression, which was mediated by YTHDF2-dependent mRNA decay. Conclusion: METTL3-mediated m⁶A methylation epigenetically regulates diabetes-induced pericyte dysfunction. METTL3-YTHDF2-PKC-η/FAT4/PDGFRA signaling axis could be therapeutically targeted for treating microvascular complications.

Keywords: Diabetic retinopathy; Microvascular complication; Pericyte dysfunction; m6A methylation.

Figure 1

Diabetic stress leads to increased levels of m⁶A RNA modification in vitro and in vivo. (A and B) Pericytes were incubated with 25 mM glucose (High glucose, HG) for 24 h and 48 h. The levels of m⁶A RNAs were detected by the colorimetric quantification (A, n = 4, *P < 0.05 versus NG group, 1-way ANOVA, Bonferroni test) or dot blot assays (B, n = 4, *P < 0.05 versus NG group, 1-way ANOVA, Bonferroni test). (C) The levels of m⁶A RNAs were detected by the colorimetric quantification and dot blot assays in pericytes cultured in the medium containing H₂O₂ (200 µM), VEGF (10 ng/mL), IL-6 (20 ng/mL), TNF-α (10 ng/mL) or left untreated (Ctrl) for 24 h and 48 h (n = 4, *P < 0.05 versus Ctrl group, 1-way ANOVA, Bonferroni test). (D and E) Retinal vessels were isolated from diabetic retinas after 1-month, 2-month, 4-month, and 6-month diabetes induction. The levels of m⁶A RNA modification in retinal vessels were detected by the colorimetric quantification (D, n = 6). After 6-month diabetes induction, dot blot assays were performed to detect the change of m⁶A RNA modification in retinal vessels (E, n = 4). *P < 0.05 versus non-DM group; Mann-Whitney U test, Bonferroni test.

Figure 2

Diabetic stress induces METTL3 expression in pericytes. (A and B) Retinal pericytes were incubated with normal culture medium (normal glucose, NG) or 25 mM glucose (high glucose, HG) for 24 h and 48 h. qRT-PCR assays and western blots were conducted to detect the levels of METTL3, METTL14, WTAP, VIRMA, RBM15, and ZC3H13. n = 4; *P < 0.05 versus NG group; 1-way ANOVA, Bonferroni test. (C and D) qRT-PCR assays and western blots were conducted to detect the levels of Mettl3 in the diabetic retinal vessels and the corresponding controls. The representative blots for Mettl3 were shown. n = 6 retinas per group; *P < 0.05 versus non-DM; Mann-Whitney U test, Bonferroni test.

Figure 3

METTL3 silencing suppresses pericyte dysfunction in vitro. (A) Retinal pericytes were transfected with scrambled (Scr) siRNA, METTL3 siRNA1 (M3 siRNA1), METTL3 siRNA2 (M3 siRNA2), pcDNA 3.1 vector (Vector), pcDNA 3.1-METTL3 (M3 OE), or left untreated (Ctrl) for 48 h. qRT-PCR assays were conducted to detect the levels of METTL3 (A, n = 4, *P < 0.05 versus Ctrl group; Student's t-test). (B-D) Pericytes were transfected with Scr siRNA, M3 siRNA1, M3 siRNA2, or left untreated, and then exposed to 25 mM glucose for 48 h. MTT assays were conducted to detect cell viability (B, n = 4, *P < 0.05 versus HG group;1-way ANOVA, Bonferroni test). Apoptotic cells were determined by caspase 3/7 activity (C, n=4, *P < 0.05 versus HG group; 1-way ANOVA, Bonferroni test), Calcein-AM/PI staining (D, n = 4, Scale bar: 20 µm, *P < 0.05 versus HG group; 1-way ANOVA, Bonferroni test), and TUNEL staining (E, n = 4, Scale bar: 20 µm, *P < 0.05 versus HG group; 1-way ANOVA, Bonferroni test). The expression levels of pericyte markers, including PDGFR-β, NG2, α-SMA, and Desmin were determined by qRT-PCRs in the pericytes after transfection of Scr siRNA, M3 siRNA1, M3 siRNA2, or left untreated (Ctrl) for 48 h (F, n = 4, *P < 0.05 versus Ctrl group;1-way ANOVA, Bonferroni test). The barrier function was analyzed by the measurement of 70-kDa FITC-Dextran leakage. The treated pericytes were co-cultured with endothelial cells in the insert chambers and relative diffusive flux change was used to show the barrier permeability (G, n = 4, *P < 0.05 versus HG group; 1-way ANOVA, Bonferroni test).

Figure 4

Pericyte-specific deletion of Mettl3 alleviates pericyte dysfunction and retinal vascular complication in vivo. (A and B) Mettl3^f/f; Pdgfrβ-Cre (Mettl3^f/f) mice or Mettl3^+/+; Pdgfrβ-Cre (Mettl3^+/+) mice (Ctrl group) were intraperitoneally injected with STZ for diabetes induction. The flat-mounted retinas were stained by IB4 (in green, endothelial cells) and NG2 (in red, pericytes) to detect pericyte coverage on retinal vessels (n = 6 retinas per group, Scale bar: 100 µm). The multiple overlapping images of flat-mounted retinas were captured by a ×10 lens and the individual images were arrayed to obtain the composite images of a leaf of retina vessels. The representative composite images after 6 months of treatment and statistical result was shown. (C and D) Evans blue assay was conducted to detect retinal vascular leakage. Evans blue dye was injected and circulated for 2 h. After fixation, the whole flat-mounted retinas were tiled and observed under a fluorescence microscope at a × 4 lens to take tile-scanning images. The representative composite images of flat-mounted retinas and the quantification results were shown. The red fluorescent is Evans blue signaling (n = 6 retinas per group, Scale bar: 500 µm). (E-H) Retinal trypsin digestion was conducted to detect the number of microaneurysms (F, n = 6 retinas per group, per mm² retina), pericyte ghosts (G, n = 6 retinas per group, per 1000 capillary cells), and acellular capillaries (F, n = 6 retinas per group, per mm² retina). The representative images of acellular capillary, pericyte ghost, and microaneurysm were shown (E, Scale bar: 100 µm). *P < 0.05 versus Ctrl group; ^#P < 0.05 between the marked groups; Kruskal-Wallis test, Bonferroni test.

Figure 5

Identification of the potential targets of METTL3 in pericytes. (A) KEGG pathway analysis was conducted to predict the potential signaling pathways regulated by METTL3 in pericytes. (B and C) Pericytes were transfected with scrambled siRNA (Scr siRNA), METTL3 siRNA1 (M3 siRNA1), METTL3 siRNA2 (M3 siRNA2), or left untreated (Ctrl) for 48 h. qRT-PCRs were conducted to detect the expression of PKC-η, FAT4, and PDGFRA in pericytes (B, n = 4, *P < 0.05 versus Ctrl group). Western blots were conducted to detect the levels of PKC-η, FAT4, and PDGFRA in pericytes. Representative blots and statistical results were shown (C, n = 4, *P < 0.05 versus Ctrl group). (D-F) Retinal pericytes were transfected with pcDNA 3.1 vector (Vector), pcDNA 3.1-METTL3 (M3 OE) without or with PKC-η, FAT4, or PDGFRA overexpression, or left untreated (Ctrl). These cells were exposed to 25 mM glucose for 48 h. MTT assays were conducted to detect cell viability (D, n = 4). The endothelial permeability was measured by FITC-Dextran transwell assay (E, n = 4). Apoptotic cells were determined by Calcein-AM/PI staining. Representative PI staining images at 48 h and statistical result were shown (F, n = 4, Scale bar: 20 µm). *P < 0.05 versus Ctrl group; ^#P < 0.05 versus M3 OE group; 1-way ANOVA, Bonferroni test.

Figure 6

METTL3-PKC-η/FAT4/PDGFRA axis regulates pericyte dysfunction and retinal vascular complication in vivo. (A and B) qRT-PCRs (A, n = 4 retinas per group) and western blots (B, n = 4 retinas per group) were conducted to detect the expression of PKC-η, FAT4, and PDGFRA in pericyte-specific Mettl3 deletion mice (Mettl3^f/f; Pdgfrβ-Cre mice) (Mettl3^f/f) and Mettl3^+/+ mice. The representative dots were shown. (C and D) Diabetic mice (Ctrl) or diabetic pericyte-specific Mettl3 deletion mice received an intravitreous injection of PKC-η shRNA, FAT4 shRNA or PDGFRA shRNA, or were left untreated. Pericyte coverage was quantified by staining the whole-mount retinas with IB4 and NG2 after 1 month, 2 months, 4 months, and 6 months of treatment (C, n = 6 retinas per group, Scale bar: 100 µm). Evans blue assay was conducted to detect retinal vascular leakage. Evans blue dye was injected and circulated for 2 h. After the fixation, the whole flat-mounted retinas were tiled and observed under a fluorescence microscope at a × 4 lens to take tile-scanning images. The representative composite images of flat-mounted retinas and the quantification results were shown. The red fluorescent is Evans blue signaling (n = 6 retinas per group, Scale bar: 500 µm). *P < 0.05 versus Ctrl group, #P < 0.05 among the marked groups. Kruskal-Wallis test, Bonferroni test.

Figure 7

METTL3 regulates the stability of PKC-η, FAT4, and PDGFRA mRNA in YTHDF2-dependent manner. (A) Pericytes were transfected with pcDNA3.1-YTHDF2 (YTHDF2), pcDNA3.1 vector (Vector), or left untreated (Ctrl) for 48 h. Western blots were conducted to detect the expression of PKC-η, FAT4, or PDGFRA. Representative blots and quantification result was shown (n = 4, *P < 0.05 versus Ctrl group, 1-way ANOVA, Bonferroni test). (B) RNA pull-down assay was conducted to confirm the association between YTHDF2 and PKC-η, FAT4, or PDGFRA. The levels of PKC-η, FAT4, and PDGFRA were detected by qRT-PCR assays (n = 4, *P < 0.05 versus IgG group, Student t test). (C) qRT-PCR assays were conducted to detect the levels of PKC-η, FAT4, and PDGFRA mRNA in pericytes overexpressing pcDNA3.1-YTHDF2 (YTHDF2) or pcDNA3.1 vector (Vector) at indicated time after the treatment of transcription inhibitor, actinomycin D (*P < 0.05 versus Vector group, n = 4, 1-way ANOVA, Bonferroni test). (D) The pericytes were treated as shown. Western blots were conducted to detect the expression of PKC-η, FAT4, and PDGFRA. Representative immunoblots and quantification results were shown (n = 4, *P < 0.05 versus Ctrl group, #P < 0.05 M3 siRNA group+YTHDF2 versus M3 siRNA group, 1-way ANOVA, Bonferroni test). (E and F) Pericytes were transfected with Scr siRNA, M3 siRNA, M3 siRNA+YTHDF2, vector, or left untreated (Ctrl), and then exposed to 25 mM glucose for 48 h. FITC-Dextran assay was conducted to evaluate the barrier function (E, n = 4, 1-way ANOVA, Bonferroni test). Apoptotic cells were determined by PI staining (F, n = 4, Scale bar: 20 µm). *P < 0.05 versus Ctrl group; ^#P < 0.05 between the marked groups; 1-way ANOVA, Bonferroni test.