Deciphering m6A methylation in monocyte-mediated cardiac fibrosis and monocyte-hitchhiked erythrocyte microvesicle biohybrid therapy

Rationale: Device implantation frequently triggers cardiac remodeling and fibrosis, with monocyte-driven inflammatory responses precipitating arrhythmias. This study investigates the role of m6A modification enzymes METTL3 and METTL14 in these responses and explores a novel therapeutic strategy targeting these modifications to mitigate cardiac remodeling and fibrosis. Methods: Peripheral blood mononuclear cells (PBMCs) were collected from patients with ventricular septal defects (VSD) who developed conduction blocks post-occluder implantation. The expression of METTL3 and METTL14 in PBMCs was measured. METTL3 and METTL14 deficiencies were induced to evaluate their effect on angiotensin II (Ang II)-induced myocardial inflammation and fibrosis. m6A modifications were analyzed using methylated RNA immunoprecipitation followed by quantitative PCR. NF-κB pathway activity and levels of monocyte migration and fibrogenesis markers (CXCR2 and TGF-β1) were assessed. An erythrocyte microvesicle-based nanomedicine delivery system was developed to target activated monocytes, utilizing the METTL3 inhibitor STM2457. Cardiac function was evaluated via echocardiography. Results: Significant upregulation of METTL3 and METTL14 was observed in PBMCs from patients with VSD occluder implantation-associated persistent conduction block. Deficiencies in METTL3 and METTL14 significantly reduced Ang II-induced myocardial inflammation and fibrosis by decreasing m6A modification on MyD88 and TGF-β1 mRNAs. This disruption reduced NF-κB pathway activation, lowered CXCR2 and TGF-β1 levels, attenuated monocyte migration and fibrogenesis, and alleviated cardiac remodeling. The erythrocyte microvesicle-based nanomedicine delivery system effectively targeted inflamed cardiac tissue, reducing inflammation and fibrosis and improving cardiac function. Conclusion: Inhibiting METTL3 and METTL14 in monocytes disrupts the NF-κB feedback loop, decreases monocyte migration and fibrogenesis, and improves cardiac function. Targeting m6A modifications of monocytes with STM2457, delivered via erythrocyte microvesicles, reduces inflammation and fibrosis, offering a promising therapeutic strategy for cardiac remodeling associated with device implantation.

Yifei Li, liyfwcsh@scu.edu.cn.§ These authors contributed equally The PDF file includes:   (A) Enrichment prediction score for METTL3 and METTL14 across each cell type profiled in human heart tissue according to The Human Protein Altas.This score calculates the mean correlation between METTL3/METTL14 and three reference transcripts representing each cell type.(B) t-SNE plot of single-cell RNA-seq data from 3-week-old mouse hearts.Analysis using the Mouse Cell Atlas reveals high METTL3 and METTL14 expression in cardiac macrophages, indicating significant transcript abundance in these cells.(C) UMAP mapping of single-cell RNA sequencing data from mouse hearts post-Ang Ⅱ induction, displaying clusters of macrophages, fibroblasts, and other cell types.(D) GO-Term analysis shows significant enrichment of signaling pathways in macrophages before and after Ang Ⅱ induction, including I-κB phosphorylation, macrophage activation, chemotaxis, and the production of IL-1β, IFN-1, and IL-6.(E) GSEA highlights the trends in enriched pathways following Ang Ⅱ induction.(F) AUCell analysis of gene sets in macrophages reveals enhanced activity levels post-Ang Ⅱ induction, particularly in producing pro-inflammatory cytokines.(A-F) Clinical chemistry analysis for constituents in blood and urine, including aminotransferase (ALT and AST), alkaline phosphatase (ALP), creatinine, total bilirubin (BIL), and total proteins.n = 9 per group.(G-L) Hematological analysis for STM@MVnex-u injected into mice's tail veins.The blood cell count levels at 29 days post-injection, including red blood cells (RBCs), white blood cells (WBCs), platelets, neutrophils, lymphocytes, and monocytes.n = 9 per group.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Fig. S25. H&E of mouse organs following different treatments in the Ang Ⅱ-induced cardiac fibrosis and remodeling model
Mice received the same doses of formulations as in the efficacy study (Fig. 8).Organs of mice were analyzed by H&E staining.Representative of n = 6 biologically independent animals per group.

Figure S1 :
Figure S1: A working model for the interrelationship between monocyte inflammatory phenotypes, migration, and fibrosis with m 6 A modification.m 6 A modification enhances the stability of MyD88 mRNA through recognition by YTHDF3, thereby strengthening the activation of the NF-κB pathway triggered by LPS.In turn, the NF-κB transcriptional activators act on the promoters of METTL3 and METTL14, enhancing their transcription.On another front, METTL3 and METTL14 promote the expression of m 6 A-modified TGF-β1 mRNA directly, and CXCR2 expression indirectly.

Figure
Figure S3 Single-cell RNA sequencing analysis of METTL3 and METTL14 in human and mouse heart cells following Ang Ⅱ induction.(A) Enrichment prediction score for METTL3 and METTL14 across each cell type profiled in human heart tissue according to The Human Protein Altas.This score calculates the mean correlation between METTL3/METTL14 and three reference transcripts representing each cell type.(B) t-SNE plot of single-cell RNA-seq data from 3-week-old mouse hearts.Analysis using the Mouse Cell Atlas reveals high METTL3 and METTL14 expression in cardiac macrophages, indicating significant transcript abundance in these cells.(C) UMAP mapping of single-cell RNA sequencing data from mouse hearts post-Ang Ⅱ induction, displaying clusters of macrophages, fibroblasts, and other cell types.(D) GO-Term analysis shows significant enrichment of signaling pathways in macrophages before and after Ang Ⅱ

Figure
Figure S4 The efficiency of the siRNA-mediated downregulation of METTL3 and METTL14 was assessed at the mRNA and protein levels-Knockdown of siRNA for METTL3 and METTL14 with 3 target sites, respectively.(A-D) Expression of the METTL3 (A and B) and METTL14 (C and D) mRNA and protein measured in RAW264.7 cells transfected with siRNA of METTL3 (siM3) or siRNA of METTL14 (siM14) using qRT-PCR and western blot.n = 3 per group.(E and F) Knockdown of METTL3 and METTL14 of siRNA (siM3-2, siM14-1) chosen according to results in (A-D), with or without treatment of LPS for BMDMs.BMDMs were harvested 24 hours after LPS treatment.mRNA levels (A) and protein contents (B) of METTL3 and METTL14 were examined by qRT-PCR and western blot.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure
Figure S6 Effects of Clodronate Liposome on Mice with Ang Ⅱ-Induced Myocardial Fibrosis were assessed by flow cytometry.Hearts were harvested and dissociated into single-cell suspensions 5 days post-injection.Controls included Ang Ⅱ-treated mice that did not receive clodronate liposome injections and untreated mice.(A) Flow cytometry gating strategy identified macrophages as CD45 + Ly6G -F4/80 + cells.(B) Representative dot plots of cardiac single-cell suspensions highlight total and resident macrophages (CX3CR1 + ).(C) Flow cytometric quantification of macrophage percentages among leukocytes.(D) Quantitative analysis of resident macrophages as a percentage of total macrophages.Bar charts are presented as mean ± SD.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure
Figure S7 Identification of myocardial fibrosis types through the protein expression ratio of Type I to Type Ⅲ collagen fibers.(A) Western blot analysis of Collagen Ⅰ and Collagen Ⅲ following treatment with Ang Ⅱ, clodronate liposomes (Clo-lipo), the injection of in-vitro-cultured BMDMs that are either untreated, treated by LPS with or without transfection of siM3&14.(B) Quantitative analysis depicting the statistical evaluation of the ratio between Collagen I and Collagen III across the experimental groups detailed above.Bar charts are presented as mean ± SD.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure
Figure S8 Knockdown of METTL3 and METTL14 in adoptively transferred BMDMs alters monocyte recruitment and macrophage polarization.Clodronate liposomes were used to deplete monocytes and macrophages after inducing myocardial fibrosis in mice with Ang Ⅱ. Subsequently, exogenous BMDMs were injected to reconstruct the monocyte/macrophage system.Three groups of mice were injected with BMDMs which were intervened with siRNA-NC (NC group), LPS+siRNA-NC (LPS+NC group), and LPS+siM3&14, respectively.(A and B) LSCM images of PKH-26 positive donor monocytes in hearts (A) and quantification of the number of the recruited monocytes (B).n = 12 per group.(C) Flow cytometry analysis of single-cell suspension of hearts 10 days after monocytes (prelabeled with PKH-26) injection.By forward scatter versus side scatter plots we exclude cell doublets and other types of cells with big morphological differences with macrophages, such as cardiomyocyte and endothelial cells.Cells displaying triple positivity for CD11b, F4/80, and PKH-26 were identified as donor cells.Gate shows the double positive population within the CD11b + population.The gating area was chosen based on the sample from untreated mice.(D) Quantifying the percentage of macrophages from donor BMDMs (PKH-26 + ) of total

Figure
Figure S10.m 6 A modification compared WT macrophages with METTL3-KO macrophages with or without LPS intervention by MeRIP-seq.(A and B) Pie chart presenting fractions of m 6 A peaks in different transcript segments of WT and METTL3-KO macrophages with LPS intervention.(C and D) RNA-seq volcano plots of WT and METTL3-KO macrophages with or without LPS intervention.(E) Quadrant_plots of differential transcripts.(F, G) Gene Ontology (GO) and Kyoto Encyclopedia of Genes and

Figure
Figure S12.METTL3/METTL14 overexpression promotes CXCR2 expression in monocytes, which regulates chemotaxis and migration in vitro.(A) mRNA and protein expression of CXCR2 measured by qRT-PCR and immunoblot in the above cells.RAW264.7 cells were transfected with NC, siM3&14, or OE-M3&14 before treatment with LPS (100 ng/ml).n = 4 per group.(B) Immunofluorescence of above cells.(C) Typical flow cytometry results of CXCR2 surface expression and percentage of CXCR2 + cells (D) Mean fluorescence values of CXCR2 in cells of (C).n = 3 per group.(E and F) Typical transwell images (E) and the chemotactic number (F) of the above cells toward CXCL1 in the lower chamber after 6 h.n = 6 per group.(G and H) Cell migration (G) and quantification (H) detected by wound healing assay.n = 4 per group.(I) Prediction sites of CXCR2 mRNA m 6 A peaks as produced from SRAMP.(J) Schematic of CXCR2 mRNA segments.(K) MeRIP-PCR was performed to detect the m 6 A abundance in LPS-treated cells with or without overexpression of METTL3 and METTL14.n = 3 per group.(L and M) qPCR and western blot were used to detect the mRNA expression and CXCR2 protein content after treatment with NC, OE-M3&14, OE-M3&14+LPS, OE-M3&14+JSH23, OE-

Figure S15 .
Figure S15.METTL3 and METTL14 genes are activated by NF-κB transcriptional activator p65.(A and B) qPCR (A) and western blot (B) were used to detect the METTL3 and METTL14 mRNA and protein expression after treatment with LPS and NF-κB inhibitor JSH23.n = 8 per group.(C) Sequence logo representing consensus motif of DNA binding sites of NF-κB transcription factors according to JASPAR.(D) Schematic representation of NF-κB transcription factors binding sites of METTL3 promoter region predicted by EPD and JASPAR.(E) ChIP-PCR assay was used to measure the binding sites (S1, S2, S3, S4) of NF-kB on METTL3 with or without treatment of LPS.n = 3 per group.(F) Schematic representation of NF-κB transcription factors binding sites of METTL14 promoter region.(G) ChIP-PCR assay was used to measure the binding sites of NF-kB on METTL14.n = 3 per group.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure S16 .
Figure S16.Flow cytometry analysis and Annexin V expression quantification in MVs and RBCs.(A) Flow cytometry of MVs and RBCs for annexin V expression.MVs were inducted in RBCs treated with PBS, A23187, and t-BOOH for 24 hours or stored at 4°C for 10 days.MVs were separated from RBCs using 4 μm diameter PE-labeled microspheres.This method distinguishes PE-negative RBCs and MVs by their forward scatter height (FSH): the former have a larger FSH, while the latter has a smaller FSH, compared to microspheres.(B) Quantification of the percentage of annexin V positive RBCs and MVs, highlighting elevated annexin V expression in t-BOOH treated MVs.n = 3 per group.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure S17 .
Figure S17.Characterization of the distinctive features of MVs compared to erythrocytes and RAW264.7 cells.Western blot analysis illustrates the detection of MV markers ALIX and TSG101 alongside the marker for erythrocytes-derived MVs, Stomatin (STOM), in comparison to GAPDH, serving as a loading control.The samples include MVs from erythrocytes, cell lysates from erythrocytes, or RAW264.7 cells.

Figure S18 .
Figure S18.Purity assessment of erythrocyte-derived MVs via flow cytometry.MVs isolated from platelets were used as a control.Identification markers included the erythrocytespecific antigen CD235a and the platelet-specific antigen CD41.Flow cytometric analysis demonstrated that the isolated erythrocyte MVs were virtually free of contamination from platelet-derived MVs.

Figure S19 .
Figure S19.Western blot analysis of HBA relative to GAPDH in BMDMs untreated or incubated with 8 × 10 11 erythrocyte-derived MVs for 24 h.

Figure S20 :
Figure S20: Flow cytometric analysis of bone marrow-derived macrophages (BMDMs) after co-incubation with erythrocyte microvesicles (MVs).(A) Flow cytometry gating strategy was employed to identify BMDMs.(B) Representative peak profiles of fluorescently positive BMDMs uptaking DIR-prelabeled erythrocyte MVs.The numbers in the figure represent the proportion of DIR fluorescence-positive cells.(C) Quantification of fluorescence-positive cells (indicating successful uptake of erythrocyte MVs) by: (1) Trypsin treatment of cells post-MV-DiR incubation to remove any surfacebound MVs; (2) MV-cell co-incubation at 4°C or under energy depletion (NNA: NaF, NaN3, and antimycin A) to differentiate between active and passive uptake; (3) Inhibition of cytoskeleton rearrangement (Cyt-D: cytochalasin D) and dynein (Dyn: dynasore) to inhibit endocytic processes; (4) Heparin treatment for 24 hours to investigate the role of heparan sulfate proteoglycans.We compared the statistical differences between intervention conditions and the control PBS group.Bar charts are presented as mean ± SD.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure S21 :
Figure S21: Flow cytometry and quantification of Ter-119a positive nanoparticles in drug-loaded STM@MVnex-u.(A) Flow cytometry showing the proportion of Ter-119a positive nanoparticles in response to varying concentrations of EGCG.(B) Quantitative analysis of the percentage of Ter-119a positive nanoparticles, indicating the formation of MPN nanoparticles with MVs.n = 3 per group.(C) Linear correlation in serially diluted STM@MVnex-u between coumarin 6 fluorescence units (FU), MVs counts and total protein content of MVs as determined by BCA method, highlighting the measurable range of coumarin 6 fluorescence and consistency in drug-loading.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure S22 .
Figure S22.Quantification of MVs isolated from whole blood collected using different anticoagulants.The quantification was conducted utilizing DLS technique.Two hours postinjection of 2×10 11 MVs, the proportion of endogenous MVs in the plasma was nearly negligible.

Figure S23 .
Figure S23.MVs dynamics in serum and cardiac tissue post-treatment (A) Flow cytometric analysis of serum exosomes captured on gel beads after treatment of MVs and STM@MVnex-u.(B) Quantification of DIR fluorescence in gel beads post tail vein injection of MVs or STM@MVnex-u.n = 3 per group.(C) Flow cytometry of cardiac cells assessing DIR-positive cells.(D) Quantification of DIR-positive non-myocardial cells.n = 3 per group.P-values were determined by one-way ANOVA with Fisher's LSD post-hoc test.