Inhibition of Macrophage ARID3A Alleviates Myocardial Ischemia-Reperfusion Injury After Heart Transplantation by Reducing THBS1/CD47 Signaling-Mediated Neutrophil Extracellular Traps Formation

Abstract

Mitigating myocardial ischemia-reperfusion (IR) injury is essential for enhancing the success of heart transplantation (HT) and improving patient outcomes. During HT, infiltrating neutrophils are influenced and regulated by various other cell types, contributing to myocardial IR injury through the excessive release of neutrophil extracellular traps (NETs). Nonetheless, the precise mechanisms underlying the interactions between neutrophils and other non-cardiomyocytes remain largely unexplored. Single-cell RNA sequencing is employed to characterize the cellular landscape and to explore the crosstalk between neutrophils and other non-cardiomyocytes. The role of AT-rich interactive domain-containing protein 3A (ARID3A) during HT is further examined using myeloid-specific ARID3A-knockout mice. Molecular docking analyses are conducted to identify the target of 4-octyl itaconate (4-OI). These results reveal that M1 macrophages recruited during the reperfusion of HT promote NETs formation and myocardial IR injury through THBS1/CD47 axis, whereas CD47 induces NETosis by activating the p38 MAPK signaling. Exogenous administration of 4-OI specifically inhibits ARID3A in macrophages, thereby suppressing NETosis and alleviating myocardial IR injury. These findings indicate that THBS1/CD47 signaling is a critical bridge mediating the interaction between M1 macrophages and NETs-associated neutrophils, and identify 4-OI as a promising therapeutic candidate for the treatment of myocardial IR injury following HT.

Keywords: heart transplantation; intercellular crosstalk; ischemia‐reperfusion injury; neutrophil extracellular traps (NETs); single‐cell RNA sequencing.

Figure 1

NETs are involved in the occurrence and development of IR injury during HT. A) Quantitative analysis of MPO‐DNA complex in serum of patients undergoing HT during the perioperative period (n = 7). B) The Spearman correlation analysis of cTnT and MPO‐DNA complex in serum (n = 20). C) Spontaneous NETs formation of human peripheral blood neutrophils was measured using SYTOX Green staining. Scale bar = 20 µm. D, E) Representative immunofluorescence images and quantification of Ly6G⁺ CitH3⁺ double‐positive cells with staining for Ly6G (green) and CitH3 (red) in heart tissues (n = 6). Scale bar = 50 µm. F) Quantitative analysis of MPO‐DNA complex in serum of mice (n = 6). G) Representative image of H&E staining of heart tissues. Scale bar = 20 µm. H, I) Representative peak perfusion images in arbitrary units (AU) and statistical analysis of mean perfusion (n = 5). Pseudo‐colored scale bar represents level of perfusion with higher values in red and lower values in blue. J) Beating score measured by the Stanford Cardiac Surgery Laboratory graft scoring system (n = 5). K) The serum levels of cTnI were evaluated in different groups of mice (n = 5). L) SYTOX Green staining was used to measure the morphology and levels of NETs in human peripheral blood neutrophils co‐cultured with H9c2 and HL‐1 cells derived H/R supernatants (n = 6). Scale bar = 20 µm. M, N) Cell viability of H9c2 and HL‐1 cells in different groups (n = 3). O, P) The levels of Bax and Bcl‐2 proteins in H9c2 and HL‐1 cells were measured by western blotting (n = 5). ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 2

ScRNA‐seq reveals a NETs‐associated neutrophil subset in the hearts of mice undergoing HT. A) Graphical overview of this single‐cell sequencing study. The hearts of mice were dissociated into single‐cell suspensions, and non‐cardiomyocytes were analyzed for scRNA‐seq with 10 × Genomics, followed by comprehensive analysis. B) UMAP analysis of 36835 cells from control hearts and 36704 cells from IR hearts was performed, with 9 clusters shown in each plot (n = 3). C) Barplot presenting the proportion of all cell types in each group. D) UMAP plot of neutrophil sub‐clusters in each group. E) Ro/e analysis assessed the percentage difference in each neutrophil cluster between two groups. F) The left pseudotime trajectory is shown colored in a gradient from dark blue to yellow and the start of pseudotime is indicated. The middle pseudotime trajectory is shown colored by different neutrophil sub‐clusters. The right pseudotime trajectory is shown colored by different cell states. G) RNA velocity of neutrophil sub‐clusters. H) Violin plots showing NETs score for each neutrophil cluster. I) KEGG pathway enrichment analysis of up‐regulated differential genes in Ero1l⁺ neutrophil subpopulation, which was significantly enriched in the NETs pathway. J) Immunofluorescence co‐staining for Ly6G (green), CitH3 (red), and Ero1l (magenta) in the hearts of mice from Con and IR groups. Scale bar = 50 µm. K, L) Quantification analysis of Ly6G⁺ Ero1l⁺ cells and Ly6G⁺ Ero1l⁺ CitH3⁺ cells (n = 5). M) The levels of cell‐free DNA (cfDNA) in the supernatants of HL‐60 cells co‐cultured with H9c2‐derived H/R supernatants (n = 5). N) The levels of PAD4 and CitH3 proteins in HL‐60 cells were measured by western blotting (n = 5). ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 3

NETosis is strongly associated with M1 macrophages induced by IR after HT. A) Cell‐cell interaction networks and chord diagram showing the potential interaction magnitude between Ero1l⁺ neutrophil subpopulation and other different cell types in each group. The line thickness is proportional to the number of ligand‐receptor pairs. B) UMAP plot of macrophage sub‐clusters in each group. C) RNA velocity of macrophage sub‐clusters. D) The left pseudotime trajectory is shown colored in a gradient from dark blue to yellow, and the start of pseudotime is indicated. The middle pseudotime trajectory is shown colored by different macrophage sub‐clusters. The right pseudotime trajectory is shown colored by different cell states. E) The left panel shows the difference in the number of reciprocal pairs between the different cell types between the two sets, with the receptor cell type in the horizontal coordinates and the ligand cell type in the vertical coordinates, the colored bars at the top represent the sum of the column values displayed in the incoming signals, and the colored bars on the right represent the sum of the row values of the outgoing signals, and in the color columns, the red (or blue) color indicates that the signals in the second dataset increase (or decrease). F) Heatmap of the number of differential interaction pairs between two groups, with red representing more interaction pairs in the IR group and blue representing more interaction pairs in the Con group. G) Heatmap showing differential genes expression in M1 macrophage subpopulation associated with markers of M1 polarization and M2 polarization. H) Gene expression profiles of M1 macrophage markers and M2 macrophage markers in hearts following myocardial IR after HT (n = 3). I) Representative flow cytometry plots showing the percentages of neutrophils (CD11b⁺ Ly6G⁺) and M1 macrophages (CD11b⁺ F4/80⁺ CD86⁺) within the heart tissues of control and IR mice. J) Quantification of neutrophils and M1 macrophages within the heart tissues of control and IR mice (n = 5). K) KEGG pathway enrichment analysis of up‐regulated differential genes in M1 macrophage subpopulation, which was significantly enriched in the NETs pathway. L) Images and quantitative analysis of NETs formation in human or mouse peripheral blood neutrophils by SYTOX Green staining (n = 5). Scale bar = 20 µm. M) The levels of cfDNA in the supernatants of peripheral blood neutrophils co‐cultured with BMDMs induced by LPS/IFN‐γ and H9c2‐derived H/R supernatants (n = 5). ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 4

M1 macrophages could regulate IR‐mediated NETosis and myocardial injury in HT through the THBS1/CD47 pathway. A) Dot plot of selected ligand‐receptor interaction between Ero1l⁺ neutrophils and macrophage subpopulations in Con and IR hearts. The size of the dots represents the interactions' significant values and the color of the dots represents the intensity of interactions. B) Co‐immunoprecipitation assay to detect the interaction between THBS1 and CD47 in myocardial tissues. C) The PPI network is constructed based on the STRING database. D) Immunofluorescence co‐staining for Ero1l (green), CD47 (red), F4/80 (magenta), and THBS1 (cyan) in the hearts of mice. Scale bar = 50 µm. E) Immunofluorescence co‐staining for Ero1l (green) and CD47 (red) in human peripheral blood neutrophils co‐cultured with BMDMs induced by H9c2‐derived normal and H/R supernatants. Scale bar = 20 µm. F) Immunofluorescence co‐staining for iNOS (green) and THBS1 (red) in BMDMs and THP‐1 cells co‐cultured with H9c2‐derived normal and H/R supernatants. Scale bar = 20 µm. G) The Spearman correlation analysis of THBS1 and MPO‐DNA complex in serum (n = 20). H) Beating score measured by the Stanford Cardiac Surgery Laboratory graft scoring system (n = 5). I) The serum levels of cTnI were evaluated in different groups of mice (n = 5). J) The levels of cfDNA in peripheral blood (n = 5). K) Representative immunofluorescence images and quantification of Ly6G⁺ CitH3⁺ double‐positive cells with staining for Ly6G (green) and CitH3 (red) in heart tissues (n = 5). Scale bar = 50 µm. L) The levels of PAD4 and CitH3 proteins in HL‐60 cells co‐cultured with THP‐1 cells induced by H9c2‐derived H/R supernatants were measured by western blotting (n = 3). M) The levels of cfDNA in the supernatants of HL‐60 cells co‐cultured with THP‐1 cells induced by H9c2‐derived H/R supernatants (n = 3). ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 5

CD47 mediates NETosis through the p38 MAPK signaling pathway. A, B) GSEA showed MAPK signaling pathway was activated in Ero1l⁺ neutrophils after HT. C) The levels of MAPK signaling pathway proteins in HL‐60 cells co‐cultured with THP‐1 cells induced by H9c2‐derived normal and H/R supernatants were measured by western blotting (n = 4). D) The levels of cfDNA in the supernatants of HL‐60 cells co‐cultured with THP‐1 cells induced by H9c2‐derived H/R supernatants (n = 4). E) The levels of PAD4 and CitH3 proteins in HL‐60 cells co‐cultured with THP‐1 cells induced by H9c2‐derived H/R supernatants were measured by western blotting (n = 4). ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 6

The ARID3A of macrophages could regulate THBS1 at the transcriptional level. A) The levels of ARID3A protein in BMDMs and THP‐1 cells induced by LPS/IFN‐γ and H9c2‐derived H/R supernatants were measured by western blotting (n = 4). B) Immunofluorescence co‐staining for ARID3A (green) and THBS1 (red) in RAW264.7 cells co‐cultured with H9c2‐derived normal and H/R supernatants. Scale bar = 20 µm. C) Schematic diagram of ARID3A binding motif predicted based on the JASPAR database. D) Dual‐luciferase reporter assay to detect the THBS1 promoter activity. E) ChIP assay was performed to analyze ARID3A binding to the THBS1 promoter using an anti‐ARID3A antibody. F) Agarose gel electrophoresis for ChIP assay to validate ARID3A binding to the promoter region of the Thbs1 gene. G, I) The levels of THBS1 protein and mRNA in THP‐1 cells co‐cultured with H9c2‐derived normal or H/R supernatants (n = 5). H, K) Quantification of THBS1 secreted by THP‐1 cells and BMDMs. J, L) The levels of THBS1 protein and mRNA in ARID3A^cKO mice‐derived BMDMs co‐cultured with H9c2‐derived normal or H/R supernatants (n = 4). ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 7

Macrophage ARID3A deficiency ameliorates IR injury in HT by modulating THBS1/CD47‐mediated NETosis. A) Representative image of H&E staining of heart tissues. Scale bar = 20 µm. B) The serum levels of cTnI were evaluated in different groups of mice (n = 5). C) Beating score measured by the Stanford Cardiac Surgery Laboratory graft scoring system (n = 5). D) Quantitative analysis of MPO‐DNA complex in serum of mice (n = 5). E, F) Representative immunofluorescence images and quantification of Ly6G⁺ CitH3⁺ double‐positive cells with staining for Ly6G (green) and CitH3 (red) in heart tissues (n = 5). Scale bar = 50 µm. G) The levels of cfDNA in the supernatants of human peripheral blood neutrophils co‐cultured with THP‐1 cells (n = 6). H) Images and quantitative analysis of NETs formation in human peripheral blood neutrophils co‐cultured with THP‐1 cells by SYTOX Green staining (n = 6). Scale bar = 20 µm. ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 8

Exogenous 4‐OI administration ameliorates IR injury and inhibits NETosis during HT. A) The chemical structure of 4‐OI. B) Representative image of H&E staining of heart tissues. Scale bar = 20 µm. C, D) Representative peak perfusion images in AU and statistical analysis of mean perfusion (n = 4). Pseudo‐colored scale bar represents level of perfusion with higher values in red and lower values in blue. E) The serum levels of cTnI were evaluated in different groups of mice (n = 4). F) Beating score measured by the Stanford Cardiac Surgery Laboratory graft scoring system (n = 4). G) ELISA analysis for MPO‐DNA complex levels in serum (n = 4). H, I) Representative immunofluorescence images and quantification of Ly6G⁺ CitH3⁺ double‐positive cells with staining for Ly6G (green) and CitH3 (red) in heart tissues (n = 4). Scale bar = 50 µm. J) The schematic of the RNA‐seq between IR mice and IR + 4‐OI mice (n = 3). K) Results of DEGs for GO enrichment research (Biological Process). L) GSEA showed that the NETs pathway in the IR + 4‐OI group is enriched. M) KEGG pathway enrichment analysis showed that DEGs are enriched in the NETs pathway. ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.

Figure 9

The 4‐OI may target macrophage ARID3A to exert therapeutic effects. A) A molecular docking simulation was performed to investigate the interaction between 4‐OI and ARID3A. B) The SPR assays showing the binding of 4‐OI to ARID3A protein. C) The levels of ARID3A protein in BMDMs and THP‐1 cells induced by 4‐OI and H9c2‐derived H/R supernatants were measured by western blotting (n = 5). D) Myocardial infarct size was measured by TTC staining and expressed as the percentage of infarct relative to the total area (n = 5). E) The serum levels of cTnI were evaluated in different groups of mice (n = 5). F) Beating score measured by the Stanford Cardiac Surgery Laboratory graft scoring system (n = 5). G) Representative image of H&E staining of heart tissues. Scale bar = 20 µm. H) Representative immunofluorescence images and quantification of Ly6G⁺ CitH3⁺ double‐positive cells with staining for Ly6G (green) and CitH3 (red) in heart tissues (n = 5). Scale bar = 50 µm. I) The levels of NRF2, HO‐1, and THBS1 proteins in THP‐1 cells were measured by western blotting (n = 5). J) Nuclear translocation of NRF2 (red) in THP‐1 cells was detected by immunofluorescence assay. Scale bar = 20 µm. K) Schematic illustration of the mechanism of effect of the 4‐OI targeting macrophage ARID3A to regulate NETosis through the THBS1/CD47 axis in IR injury after HT. ns: No significant, ^* p <0.05, ^** p <0.01, ^*** p <0.001, ^**** p <0.0001.