HMGB2 Promotes Cardiomyocyte Proliferation and Heart Regeneration Through MTA2-Driven Metabolic Reprogramming

Abstract

The neonatal heart possesses the unique ability to regenerate post-injury. Underlying related mechanisms and reactivation of this process are crucial for regeneration medicine. Using quantitative proteomics with tandem mass tag labeling, RNA-sequencing (RNA-seq) and single-nucleus RNA-seq dataset analyses, high mobility group box 2 (HMGB2) is identified as a key regulator of cardiomyocyte proliferation, whose expression declines during postnatal heart development and increases in the high regenerative potential cardiomyocyte populations in hearts post-injury. Cardiomyocyte-specific HMGB2 knockdown curtails cardiomyocyte proliferation and impairs heart regeneration following apical resection in neonatal mice, while cardiomyocyte-specific HMGB2 overexpression enhances cardiomyocyte proliferation and facilitates cardiac regeneration and repair in adult mice post-myocardial infarction. Mechanistically, RNA-seq analysis reveals that HMGB2 promotes cardiomyocyte proliferation via activating hypoxia inducible factor 1ɑ (HIF-1α)-mediated glycolysis. This study further finds HMGB2 can directly interact with metastasis-associated protein 2 (MTA2) and inhibit its ubiquitination degradation to stabilize HIF-1α protein through immunoprecipitation-mass spectrometry (IP-MS) analysis. Finally, overexpression of HIF-1α or MTA2 also promotes cardiomyocyte proliferation and cardiac repair in adult mice following MI. Taken together, these findings highlight that HMGB2 plays a crucial role in promoting heart regeneration through regulating glycolysis. Activating the HMGB2-MTA2-HIF-1α axis might serve as a potential therapeutic option for regenerative therapies post-myocardial injury.

Keywords: HMGB2; MTA2; cardiomyocyte proliferation; glucose metabolism; heart regeneration.

Figure 1

Identification of HMGB2 as a vital protein involved in neonatal heart regeneration. A, Venn diagram showing 10 candidates by cross‐comparing these differentially downregulated genes with proteins. Red represents the differentially downregulated proteins between adult mouse cardiomyocytes (AMCMs) and neonatal mouse cardiomyocytes (NMCMs). Blue represents the differentially downregulated genes between adult mouse hearts and neonatal mouse hearts. Differentially downregulated proteins and genes were identified when adjusted P value < 0.05 and Log₂FC < ‐2.5. B, Single‐nucleus RNA‐sequencing analysis of HMGB2 expression in regenerative hearts 1 and 3 days after MI or sham surgery at P1 mice from the GEO repository (GSE130699). dps: day post‐sham. dpi: day post‐infarction. C, Western blot and quantification of HMGB2 protein expression in the sham mouse myocardium and apical resection (AR) mouse myocardium (n=6 each). D, Quantitative reverse‐transcription polymerase chain reaction (qRT‐PCR) analysis of HMGB2 mRNA expression in the sham mouse myocardium and AR mouse myocardium (n=6 each). E, Immunofluorescence (IF) staining analysis of HMGB2 expression in the sham mouse myocardium and AR mouse myocardium at 1 dpr. cTNT (green), HMGB2 (red), Hoechst (blue). scale bar, 10 µm. F, IF staining analysis of HMGB2 expression in the sham mouse myocardium and AR mouse myocardium at 3 dpr. cTNT (green), HMGB2 (red), Hoechst (blue). scale bar, 10 µm. G, Western blot and quantification of HMGB2 protein expression at different time points in mouse myocardium (n=6 each). H, qRT‐PCR analysis of HMGB2 mRNA expression at different time points in mouse myocardium (n=6 each). Statistical analysis was performed by an unpaired Student's t‐test for C‐D. Statistical analysis was performed by one‐way ANOVA followed by Tukey multiple comparisons for G‐H. Ns, not significant. ^*** P<0.001.

Figure 2

HMGB2 is essential for the proliferation of neonatal mouse cardiomyocytes in vitro. A, Ki67 immunofluorescence staining and quantification of Ki67‐positive CMs in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). cTNT (red), Ki67 (green), Hoechst (blue). scale bar, 50 µm. B, EdU immunofluorescence staining and quantification of EdU‐positive CMs in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). cTNT (red), EdU (green), Hoechst (blue). scale bar, 20 µm. C, pH3 immunofluorescence staining and quantification of pH3‐positive CMs in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). cTNT (red), pH3 (green), Hoechst (blue). scale bar, 50 µm. D, Aurora B immunofluorescence staining and quantification of Aurora B‐positive CMs in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). cTNT (red), Aurora B (green), Hoechst (blue). scale bar, 50 µm. E, eGFP representative images and quantification of CM number in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). eGFP (green). scale bar, 20 µm. F, Representative images and quantification of CM nucleation in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). cTNT (green), Hoechst (blue). scale bar, 20 µm. G, Flow cytometry detection of cell cycle alterations in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). H, qRT‐PCR analysis of cell cycle‐related genes expression in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). Statistical analysis was performed by an unpaired Student's t‐test for A‐H. ^** P<0.01; ^*** P<0.001.

Figure 3

HMGB2 knockdown inhibits heart regeneration in neonatal mice after apical resection. A, Schematic representation of injecting Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 into the resection border zone of P1 mice after apical resection (Created in https://BioRender.com). B, qRT‐PCR analysis of HMGB2 mRNA expression in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 7 dpr (n=6 each). C, Western blot and quantification of HMGB2 protein expression in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 7 dpr (n=6 each). D, Ki67 immunofluorescence staining and quantification of Ki67‐positive CMs in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 7 dpr (n=6 each). cTNT (red), Ki67 (green), Hoechst (blue). scale bar, 20 µm. E, EdU immunofluorescence staining and quantification of EdU‐positive CMs in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 7 dpr (n=6 each). cTNT (red), EdU (green), Hoechst (blue). scale bar, 20 µm. F, pH3 immunofluorescence staining and quantification of pH3‐positive CMs in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 7 dpr (n=6 each). cTNT (red), pH3 (green), Hoechst (blue). scale bar, 20 µm. G, Aurora B immunofluorescence staining and quantification of Aurora B‐positive CMs in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 7 dpr (n=6 each). cTNT (red), Aurora B (green), Hoechst (blue). scale bar, 20 µm. H, WGA immunofluorescence staining and quantification of CM size in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 28 dpr (n=6 each). WGA (red), Hoechst (blue). scale bar, 20 µm. I, Representative Masson's trichrome staining and quantification of fibrotic area in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 28 dpr (n=6 each). scale bar, 500 µm. J, Echocardiography analysis of the left ventricle ejection fraction (EF) and fractional shortening (FS) in mice injected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 at 1 dpr and 28 dpr (n=8 each). Statistical analysis was performed by an unpaired Student's t‐test for B‐J. Ns, not significant. ^*** P<0.001.

Figure 4

HMGB2 overexpression promotes cardiac regeneration and repair in adult mice after myocardial infarction. A, Schematic representation of the injection of AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 into the infarct border zone of adult mice after MI (Created in https://BioRender.com). B, Representative Masson's trichrome staining and quantification of fibrotic area in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 28 dpi (n=6 each). scale bar, 1000 µm. C, Echocardiography analysis of the left ventricle ejection fraction (EF) and fractional shortening (FS) in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 1 dpi and 28 dpi (n=7 each). D, Ki67 immunofluorescence staining and quantification of Ki67‐positive CMs in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi (n=6 each). cTNT (red), Ki67 (green), Hoechst (blue). scale bar, 20 µm. E, EdU immunofluorescence staining and quantification of EdU‐positive CMs in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi (n=6 each). cTNT (red), EdU (green), Hoechst (blue). scale bar, 20 µm. F, pH3 immunofluorescence staining and quantification of pH3‐positive CMs in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi (n=6 each). cTNT (red), pH3 (green), Hoechst (blue). scale bar, 20 µm. G, Aurora B immunofluorescence staining and quantification of Aurora B‐positive CMs in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi (n=6 each). cTNT (red), Aurora B (green), Hoechst (blue). scale bar, 20 µm. H, WGA immunofluorescence staining and quantification of CM size in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 28 dpi (n=6 each). WGA (red), Hoechst (blue). scale bar, 10 µm. I, Quantification of CM number in CMs isolated from MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 28 dpi (n=6 each). J, Representative images and quantification of CM nucleation and ploidy in CMs isolated from MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 28 dpi (n=6 each). The percentage of mononucleated (mono), binucleated (bi), and multinucleated (multi) CMs in total CMs. cTNT (green), Hoechst (blue). scale bar, 10 µm. Statistical analysis was performed by an unpaired Student's t‐test for I‐J. Statistical analysis was performed by one‐way ANOVA followed by Tukey multiple comparisons for B, D‐H. Statistical analysis was performed by two‐way ANOVA followed by Tukey multiple comparisons for C. Ns, not significant. ^*** P<0.001.

Figure 5

HMGB2 stimulates cardiomyocyte proliferation through HIF‐1α‐mediated glycolysis. A, Volcano plots of differentially expressed genes in NMCMs from the HMGB2 group compared with the vector group (adjusted P value < 0.05 and fold change > 1.2 or < 1/1.2). Red represents upregulated genes and blue represents downregulated genes (n=3 each). B, Gene Ontology (GO) term analysis of differentially expressed genes in cardiomyocytes from HMGB2 group and the vector group. C, Kyoto Encyclopedia of Genes and Genomes (KEGG) term analysis of the enriched signaling pathways in cardiomyocytes from HMGB2 group and vector group. D, Gene Set Enrichment Analysis (GSEA) analysis of gene clusters in the regulation of glycolysis /gluconeogenesis. E, Immunofluorescence staining of HK2 expression in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi. cTNT (green), HK2 (red), Hoechst (blue). scale bar, 10 µm. F, Immunofluorescence staining of LDHA expression in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi. cTNT (green), LDHA (red), Hoechst (blue). scale bar, 10 µm. G, Seahorse XF analysis of glycolysis stress tests in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). H, Detection of intracellular lactate concentration in NMCMs transfected with Ad5‐cTNT‐mir155‐HMGB2, Ad5‐cTNT‐mir155‐scramble, Ad5‐cTNT vector, or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). I, Western blot and quantification of HMGB2, LDHA, HK2, and HIF‐1α proteins expression in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2 for 24 h (n=6 each). J, Western blot and quantification of HMGB2, LDHA, HK2, and HIF‐1α proteins expression in sham mice and MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 at 14 dpi (n=6 each). Statistical analysis was performed by an unpaired Student's t‐test for G‐I. Statistical analysis was performed by one‐way ANOVA followed by Tukey multiple comparisons for J. ^** P<0.01; ^*** P<0.001.

Figure 6

The Δ80‐165 domain of HMGB2 binds to MTA2. A, Schematic diagram of the procedure identifying downstream targets of HMGB2 (Created in https://BioRender.com). B, Identification of MTA2 as the downstream target of HMGB2 by immunoprecipitation coupled with mass spectrometry (IP/MS) analysis. C, Identification of MTA2 as a possible protein that interacts with HMGB2. D‐E, CO‐immunoprecipitation (CO‐IP) and Western blot analysis of the interaction of HMGB2 and MTA2 in NMCMs. F‐G, CO‐IP and Western blot analysis of HMGB2 and MTA2 protein interactions in neonatal mouse myocardium at 7 dpr. H, Immunofluorescence staining analysis of HMGB2 and MTA2 protein interactions in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2 for 24 h. HMGB2 (red), MTA2 (green), Hoechst (blue). scale bar, 5 µm. I, Design diagram of HMGB2 full‐length and truncated plasmids (∆1‐165 and ∆1‐79). J, CO‐IP and Western blot analysis of Flag‐HMGB2 and HA‐MTA2 interaction in HEK293T cells transfected with HA‐MTA2‐FL and truncated Flag‐HMGB2 plasmids in I. K, Molecular docking analysis of the interaction between the MTA2 and HMGB2 domains, with HMGB2 in magenta and MTA2 in cyan.

Figure 7

HMGB2 inhibits MTA2 degradation through the ubiquitin‐proteasome pathway and stabilizes HIF‐1α level through the deacetylation pathway. A, Western blot and quantification of MTA2 protein stability in NMCMs transfected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 after treating with CHX (200 µM) for 0, 4, 8, 12h (n=6 each). CHX, cycloheximide. B, Western blot and quantification of MTA2 protein stability in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2 after treating with CHX (200 µM) for 0, 4, 8, 12h (n=6 each). C, Western blot and quantification of MTA2 protein expression in NMCMs transfected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 after treating with or without MG132 (20 µM). D, Western blot and quantification of MTA2 protein expression in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2 after treating with or without MG132 (20 µM). E, CO‐IP and Western blot analysis of the ubiquitination level of MTA2 in NMCMs transfected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2. F, CO‐IP, and Western blot analysis of the ubiquitination level of MTA2 in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2. G, CO‐IP, and Western blot analysis of HIF‐1α and MTA2 protein interactions in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2. H, CO‐IP, and Western blot analysis of the acetylation level of HIF‐1α and the interaction between MTA2 and HIF‐1α proteins in NMCMs transfected with Ad5‐cTNT vector or Ad5‐cTNT‐HMGB2. I, CO‐IP and Western blot analysis of the acetylation of HIF‐1α and the interaction between MTA2 and HIF‐1α proteins in NMCMs transfected with Ad5‐cTNT‐mir155‐scramble or Ad5‐cTNT‐mir155‐HMGB2 and control plasmid or MTA2 plasmid. Statistical analysis was performed by an unpaired Student's t‐test for A‐B. Statistical analysis was performed by two‐way ANOVA followed by Tukey multiple comparisons for C‐D. Ns, not significant. ^*** P<0.001.

Figure 8

MTA2 knockdown diminishes the protective effects of HMGB2 on cardiac regeneration and repair in adult mice after myocardial infarction. A, Schematic representation of the injection of AAV9‐cTNT‐HMGB2 or AAV9‐cTNT vector and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 into the infarct border zone of adult mice after myocardial infarction (Created in https://BioRender.com). B, Representative Masson's trichrome staining and quantification of fibrotic area in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 28 dpi (n=6 each). scale bar, 1000 µm. C, Echocardiography analysis of the left ventricle ejection fraction (EF) and fractional shortening (FS) in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 1 dpi and 28 dpi (n=6 each). D, Western blot and quantification of HMGB2, MTA2, LDHA, HK2, and HIF‐1α proteins expression in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 14 dpi (n=6 each). E, Ki67 immunofluorescence staining and quantification of Ki67‐positive CMs in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 14 dpi (n=6 each). cTNT (red), Ki67 (green), Hoechst (blue). scale bar, 20 µm. F, EdU immunofluorescence staining and quantification of EdU‐positive CMs in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 14 dpi (n=6 each). cTNT (red), EdU (green), Hoechst (blue). scale bar, 20 µm. G, pH3 immunofluorescence staining and quantification of pH3‐positive CMs in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 14 dpi (n=6 each). cTNT (red), pH3 (green), Hoechst (blue). scale bar, 20 µm. H, Aurora B immunofluorescence staining and quantification of Aurora B‐positive CMs in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 14 dpi (n=6 each). cTNT (red), Aurora B (green), Hoechst (blue). scale bar, 20 µm. I, WGA immunofluorescence staining and quantification of CM size in MI mice injected with AAV9‐cTNT vector or AAV9‐cTNT‐HMGB2 and AAV9‐cTNT‐mir155‐scramble or AAV9‐cTNT‐mir155‐MTA2 at 28 dpi (n=6 each). WGA (red), Hoechst (blue). scale bar, 10 µm. Statistical analysis was performed by two‐way ANOVA followed by Tukey multiple comparisons for B‐I. Ns, not significant. ^** P<0.01; ^*** P<0.001.