Prmt1 upregulated by Hdc deficiency aggravates acute myocardial infarction via NETosis.

Neutrophils are mobilized and recruited to the injured heart after myocardial infarction, and neutrophil count has been clinically implicated to be associated with coronary disease severity. Histidine decarboxylase (HDC) has been implicated in regulating reactive oxidative species (ROS) and the differentiation of myeloid cells. However, the effect of HDC on neutrophils after myocardial infarction remains unclear. Here, we found that neutrophils were disorderly recruited into the ischemic injured area of the myocardium of Hdc deficiency (Hdc -/-) mice. Moreover, Hdc deficiency led to attenuated adhesion but enhanced migration and augmented ROS/neutrophil extracellular traps (NETs) production in neutrophils. Hdc -/- mouse-derived NETs promoted cardiomyocyte death and cardiac fibroblast proliferation/migration. Furthermore, protein arginine methyltransferase 1 (PRMT1) was increased in Hdc -/- mouse-derived neutrophils but decreased with exogenous histamine treatment. Its expression could be rescued by blocking histamine receptor 1 (H1R), inhibiting ATP synthesis or reducing SWItch/sucrose non fermentable (SWI/SNF) chromatin remodeling complex. Accordingly, histamine or MS023 treatment could decrease ROS and NETs ex vivo, and ameliorated cardiac function and fibrosis, along with the reduced NETs in plasma in vivo. Together, our findings unveil the role of HDC in NETosis by histamine-H1R-ATP-SWI/SNF-PRMT1-ROS signaling and provide new biomarkers and targets for identifying and tuning the detrimental immune state in cardiovascular disease.

Introduction

Myocardial infarction (MI) is one of the leading causes of death worldwide. After infarction, three sequential processes, inflammatory response, cardiomyocyte necrosis, and scar formation, determine the final ejection fraction and the survival rate of patients with MI,. The inflammatory response, regulated by recruited immunocytes, including neutrophils, monocytes/macrophages, and lymphocytes, profoundly influences heart remodeling.

Neutrophils, as sentinels of inflammation, first infiltrate the infarcted zone in great numbers after MI. Although neutrophils have a relatively short lifespan, they secrete inflammatory mediators, purge debris of apoptotic cardiomyocytes, and activate fibroblasts.

A decade ago, neutrophils were found to form a net-like structure, containing double-stranded DNA, citrullination of histones and neutrophil granule protein, called neutrophil extracellular traps (NETs). Originally, NETs were regarded as the protection of our bodies from fungal and bacterial diseases. Recently, excessive NETs have been reported to be involved in the progression of other diseases, including atherosclerosis, lupus erythematosus, and coronavirus disease 2019 (COVID-19). NETs were also found in patients with MI/acute coronary syndromes, and the level of NETs in plasma was related to infarct size,.

Histidine decarboxylase (HDC) catalyzes l-histidine to histamine and is involved in allergic responses, gastric acid secretion, immune modulation, and the development of hematopoietic stem cells. In the bone marrow of mice, over 50% of CD11b+Ly6G+ granulocytic cells highly express HDC, and HDC-expressing granulocytes derived histamine plays important roles in the differentiation of immune cells. Moreover, it has been reported that serum histamine concentrations significantly increased after MI, indicating the potential roles of HDC/histamine in MI. However, the role of HDC/histamine in the regulation of neutrophils after MI has not been fully clarified.

In this study, we report that HDC plays a key role in regulating the functions of neutrophils. Recruitment, oxidative burst, and NETosis were significantly altered in Hdc−/− mouse-derived neutrophils, which augmented the death of cardiomyocytes and the activation of fibroblasts. Blocking neutrophils to a certain extent could aid reduction of myocardial injury in Hdc−/− mice. Furthermore, by using transcriptomics and proteomics, we found that aggravated cardiac injury in Hdc−/− mice was due to excessive NETs through HDC–protein arginine methyltransferase 1 (PRMT1)–reactive oxygen species (ROS) pathway. The administration of exogenous histamine or the inhibitor of PRMT1 could ameliorate cardiac injury in Hdc−/− mice.

Methods and materials

Animals and MI model

Hdc knockout (Hdc−/−, BALB/c background) mice were generously provided by Professor Timothy C. Wang from Columbia University (New York, USA). The generation of Hdc−/− mice has been described in previous papers. The adopted Hdc mice in the present work is the offspring of Hdc−/− homozygous parents, generating not enough littermates as controls. Thus male wild-type (WT, in BALB/c background) mice at 8 to 12-week-old purchased from CAVENS. LA (Changzhou, China) were used as controls, which were shown the same phenotypes as WT littermates of Hdc−/− mice, as previously described. All procedures in this study were approved by the Institutional Review and Ethics Board of Zhongshan Hospital of Fudan University (Shanghai, China). The myocardial infarction model was performed as described before. Briefly, male mice were anaesthetized with 3% isoflurane inhalation in a chamber. After exposing the hearts by left thoracotomy, a small incision was made at the fourth intercostal space to pop out the heart. The left main descending coronary artery (LCA) was permanently ligated at a site 3 mm from its origin. After the heart was placed back into the intrathoracic space, fast air evacuation and chest wall closure were performed. Histamine (Sigma, St. Louis, MO, USA) was injected with the dose of 4 mg/kg/day intraperitoneally (i.p.) beginning from 3 days before surgery and continuing until euthanasia. MS023 (MedChemExpress, Shanghai, China), the inhibitor of PRMT1 which has been used in in vivo tests, was injected 80 mg/kg i.p. every 3 days beginning from 3 days before surgery. For control groups, the same volume of phosphate buffered saline (PBS) was used. Anti-Ly6G antibodies and IgG isotype antibodies (BioXCell, Lebanon, NH, USA) were injected at 50 μg on Days 1, 3, and 5 post surgery.

Histopathological analysis

Hearts were embedded in optimal cutting temperature compound (O.C.T.; Sakura® Finetek Japan Co., Ltd., Tokyo, Japan) or in paraffin (Leica Biosystems, Wentzler, Germany). 5 μm thick paraffin sections were used for stained with haematoxylin and eosin (H&E) and Masson's trichrome staining according to standard procedures.

Immunofluorescent staining was performed to detect neutrophil infiltration (neutrophil elastase and α-actinin, Abcam, Cambridge, UK) and NETs formation [anti-histone H3 citrulline R2+R8+R17, and myeloperoxidase (MPO), Abcam]. Co-staining was visualized by fluorescence microscope with 488- and 596-conjugated secondary antibodies (Abcam) mounted with fluoroshield mounting medium with 4ʹ,6-diamidino-2-phenylindole (DAPI, Abcam).

Enzyme-linked immunosorbent assay (ELISA) assay

CK-MB (Jiancheng Bioengineering Institute, Nanjing, China) and MPO (R&D Systems, Minneapolis, MN, USA) were measured with the respective ELISA kits, according to the manufacturer's instructions. Briefly, plasma samples were collected from mice in each group at the indicated time after MI, and centrifuged at 3000 ×g for 10 min, after which the supernatant was collected. The plasma of 100–200 μL was added into the 96-well plates for the measurement of concentrations based on the absorption values.

Neutrophil isolation

Mice neutrophils were harvested and purified as previously described. In brief, bone marrow cells were gathered, added into Percoll gradient consisting of 52%, 65%, 78% Percoll layers (GE, Boston, MA, USA), and centrifuged at 2500 ×g for 30 min at room temperature. The cells between 65% and 78% layers were harvested, and red blood cells were lysed with red blood cells lysis buffer (BD, New York, NY, USA).

Adult mouse cardiomyocyte and fibroblasts isolation

The isolation of adult mouse cardiomyocytes was performed as described previously. In brief, hearts of mice were incised, after injected with EDTA buffer from apical region of left ventricle. Then left ventricle was digested and disassociated to isolate cardiomyocytes and cardiac fibroblasts through gravity sedimentation. For hypoxia, the cardiomyocytes were placed in an anaerobic incubator filled with gas consisting of 1% O2, 5% CO2, and 94% N2, at 37 °C for 24 h.

Adhesion assay

Isolated neutrophils were labeled with Calcein-AM (Thermo Fisher Scientific, Waltham, MA, USA) and plated on human umbilical vein endothelial cells (HUVEC) with the presence of 10 ng/mL tumor necrosis factor-α (TNF-α) as previously described. Then wells were washed with PBS. The number of adhering neutrophils was counted, and the percentage was calculated.

Transwell migration assay

Neutrophil Transwell migration assay was performed as described previously. Briefly, bone marrow derived neutrophils from each genotype were labeled with Calcein-AM, and seeded in serum-free RPMI 1640 medium in the upper chamber of Transwell inserts (BD Falcon, New York, NY, USA). Then recombinant interleukin-8 (IL-8, 100 ng/mL, Protein tech, Rosemont, IL, USA) were added in the lower chambers and incubated for 2 h at 37 °C in 5% CO2 incubator. The migrated cells were analyzed by fluorescent microscopy.

Quantification of ex vivo NETs formation and ROS

The 5 × 105 cells per well were seeded onto coverslips coated with poly-l-lysine (Beyotime, Nantong, China), and treated with phorbol myristate acetate (PMA, 200 nmol/L, MedChemExpress) for 18 h. Neutrophils were stained with DAPI/Sytox green (Thermo Fisher Scientific). Percentage values of NETs were calculated as the area of Sytox green positive cells relative to DAPI positive cells. The parameter was set to filter the Sytox green positive cells with small area to ignore the dead points of neutrophil for more accuracy.

To measure the level of ROS, isolated neutrophils were treated with DCFH-DA (MCE) for 30 min. Then cells were reconstituted with serum-free RPMI 1640 medium and transferred to black at 96-fold intervals, with or without PMA (200 nmol/L) treatment. The plate was measured at an excitation wavelength of 488 nm and an emission length of 530 nm using an automated plate monochrome reader (Molecular Devices, Silicon Valley, CA, USA).

TdT-mediated dUTP nick-end labeling (TUNEL) assay

TUNEL assay using the One Step TUNEL Apoptosis Assay Kit (Beyotime) was performed in cardiac tissue sections or cultured cardiomyocytes following the manufacturer's instructions. Briefly, cardiomyocytes were permeabilized with 0.05% Triton X-100 before labeling the segmented DNA of the apoptotic cells using TUNEL detection buffers. Then the samples were co-stained with anti-α-actinin primary antibody (CST, Danvers, MA, USA) followed by the Alexa Fluor-conjugated secondary antibodies (CST). Each staining in different groups was visualized with identical light exposure parameters under a fluorescence microscope (Leica Biosystems).

5-Ethynyl-2-deoxyuridine (EdU) incorporation assay

EdU incorporation assay was analyzed by using BeyoClick™ EdU Cell Proliferation Kit (Beyotime) according to the manufacturer's instructions. In brief, isolated fibroblasts were incubated in DMEM (Thermo Fisher Scientific) with 10 μmol/L EdU for 2 h at 37 °C/5% CO2. After the incubation, the cells were washed with PBS to remove the DMEM and the free EdU probe. Then fibroblasts were fixed in 4% paraformaldehyde at room temperature and stained with vimentin primary antibody (CST) followed by the Alexa Fluor-conjugated secondary antibodies (CST).

Cell migration assay/wound healing

Fibroblasts labeled with Calcein-AM were seeded in 24-well plates and cultured until cell monolayers formed. Monolayers were wounded by manual scraping with a 10 μL micropipette tip. The cells were then incubated with NETs from neutrophils of each genotype pretreated with PMA for 4 h. Wound repair was analyzed through measuring the injured area recovered from the wounding borders.

DNA accessibility (DNase I sensitivity) assay

DNA accessibility assay were carried out as described previously. In brief, DNA from HL60 cells treated with histamine for 12 h was digested by DNase I (Takara, Kyoto, Japan) and accessed by quantitative real-time PCR (qRT-PCR).

ELISA for MPO–DNA complexes

Capture ELISA was used to measure the MPO/NA complexes as described. Briefly, capture antibody for myeloperoxidase (R&D Systems) was diluted and added into an ELISA plate overnight. After blocking, diluted plasma samples were added to the plate overnight. The supernatant was discarded, and diluted PicoGreen (Thermo Fisher Scientific) with 1× Tris-EDTA (TE) buffer was added to the plate. Fluorescence of the samples was measured using an automated plate monochrome reader (Molecular Devices).

Measurement of DNA concentrations

DNA concentrations were measured using the Quant-iT PicoGreen dsDNA assay kit (Thermo Fisher Scientific) following the manufacturer's instructions. In brief, plasma was mixed with the PicoGreen dye, added to the microplate wells, and then incubated in the dark for 10 min. The fluorescence of the samples was measured using an automated plate monochrome reader (Molecular Devices).

Statistical analysis

Data are shown as mean ± standard error of mean (SEM) of at least three independent experiments for each cellular experimental group and at least five independent experiments for each animal group. The size and count of NETs or cells were accessed by Python using OpenCV (https://github.com/Winston-00/NET-analysis). We evaluated the data with Student's t-test for two-group comparisons, and one or two way analysis of variance for multiple comparisons. The value of P < 0.05 was considered statistically significant.

Results

Transcriptomic profiling of Hdc−/− mice with deteriorative outcomes after MI

To elucidate the role of histamine deficiency in myocardial infarction, we compared the severity of myocardial infarction injury between Hdc−/− and WT (Hdc+/+) mice. As expected, the levels of CK-MB increased significantly on Day 1 after MI in both Hdc−/− and Hdc+/+ mice (Fig. 1A). However, the level of CK-MB in the plasma of Hdc−/− mice was much higher than that in Hdc+/+ mice (Fig. 1A), indicating that Hdc deficiency aggravated cardiomyocyte death. Furthermore, cardiac function on Day 7 after MI surgery was measured by echocardiography. The results demonstrated that the ejection fraction (EF) and fractional shortening (FS) (Fig. 1B and C, and Supporting Information Fig. S1A for the representative images) in Hdc−/− mice were significantly lower than those in Hdc+/+ mice. These data indicated that Hdc deficiency was associated with poor prognosis after MI.

Figure 1

Hdc deficiency aggravates infarct injury and alters the cardiac transcriptome after MI. (A) Plasma levels of CK-MB from Hdc−/− and Hdc+/+ mice on Day 1 after ligation were determined by ELISA according to the manual instructions. (B and C) Quantitation of left ventricle ejection fraction (EF) and fractional shortening (FS) of Hdc−/− and Hdc+/+ mice on the 7th day after infarction. Data are shown as mean ± SEM (n = 6–8 mice in each group). ∗P < 0.05, ∗∗P < 0.01. (D and E) Gene expression profiling of hearts from Hdc+/+ and Hdc−/− mice on Days 0, 1, and 7 after MI respectively. The heatmaps depicts up/down-regulated differentially expressed genes (DEGs) at the indicated time point (D). The number of DEGs and the overlap genes consistently differentially expressed on Days 1 and 7 after MI were shown in Venn diagram (E). n = 3 per group. (F–H) gene ontology (GO) analysis of DEGs between Hdc−/− and Hdc+/+ mice. The top 15 GO terms of biological process (BP, F), cellular component (CC, G) and molecular function (MF, H) at the indicated time points were shown.

To investigate the detailed mechanisms underlying the deteriorated cardiac functions in Hdc−/− mice after MI, RNA-sequencing was performed to dissect the transcriptomic profile of hearts from Hdc+/+ and Hdc−/− mice on Days 0, 1 and 7 after MI (Fig. 1D). A total of 259 genes and 273 genes were identified as differentially expressed genes (DEGs, P < 0.05, FC > 1.5) in the hearts of Hdc+/+ and Hdc−/− mice on Days 1 and 7 after MI. Among them, 92 genes were consistently abnormally expressed both on Day 1 and on Day 7 after MI (Fig. 1E).

To identify the specific biological process (BP) and networks involved in the progression of MI, we performed gene ontology (GO) enrichment analysis with the Day 1 and Day 7 DEGs, respectively. The top 15 GO terms were enriched based on the DEGs and their changes are shown in the time course. In the GO analysis of BP, more than 50% of the top terms were related to the function of immune cells, including leukocyte chemotaxis, leukocyte migration, and regulation of the inflammatory response (Fig. 1F). The DEGs with leukocyte migration/chemotaxis and regulation of the inflammatory response were diminished at the 7th day of MI, while genes with T cell activation were augmented. In the GO analysis of cellular component (CC), DEGs were related to cell vesicles, including lysosomes, lytic vacuoles, and secretory granules (Fig. 1G). In the GO analysis of molecular function (MF), lists of genes were generated in the categories of amide binding, ATPase activity, coenzyme binding, and cytokine receptor binding (Fig. 1H). In general, the DEGs were mainly involved in immune cell function, including chemotaxis/migration, secretory granule production, and other inflammatory responses. These results suggested that immune processes likely play a vital role in the pathological progression of MI and subsequent heart remodeling.

Infiltration of immune cells especially neutrophils was shifted in Hdc−/− mice

The contribution of immunological phenomena to cardiac diseases is of particular clinical relevance. Great numbers of immunocytes successively infiltrate the infarcted zone after MI, influencing the subsequent process of myocardial healing. The results of H&E staining confirmed that there were many inflammatory cells infiltrating into the infarcted zone (Fig. 2A). Therefore, we estimated the pattern of the infiltrated immunocytes by employing the MCP counter algorithm. As shown in Fig. 2B, the infiltration of immunocytes, including neutrophils, monocytes, dendritic cells (DCs), B cells, and T cells, in the infarcted hearts, was changed after Hdc deletion. As Hdc is expressed in myeloid cells, especially in Ly6G+ granulocytes, we further focused on exploring the consequence of Hdc deficiency in neutrophils responding to ischemia injury. As flow cytometry data (Fig. 2C and Supporting Information Fig. S2A for the gating strategy) revealed, the number of infiltrated neutrophils in the infarcted zone peaked on Day 1 and greatly dropped on Days 3 and 7 after MI. However, there were still many more neutrophils in the cardiac infarcted zone of Hdc−/− mice than in Hdc+/+ mice on Days 3 and 7 after MI. The immunostaining result was consistent with the flow cytometry data (Fig. 2D and E).

Figure 2

Hdc deficiency changed the pattern of the infiltrated immune cells in particular of neutrophil in infarcted zone after MI. (A) Representative pictures of H&E staining of hearts from Hdc+/+ and Hdc−/− mice on Day 1 after MI surgery. (B) The size and the color indicate the MCP count score of the indicated immunocytes infiltrating into hearts on Days 1 and 7 after MI (n = 4–5). (C) Flow cytometry analysis of the numbers of Ly6G+ neutrophils in the hearts of Hdc−/− and Hdc mice on Days 1 and 7 after MI or sham operation. (D) Representative images of immunofluorescence staining of neutrophil elastase (NE) and α-actinin in the infarct border zone on Days 1 and 7 after MI or sham operation. (E) Quantification of NE positive neutrophils in the infarct border zone. n = 3–5 per group. (F) Quantitation of left ventricle ejection fraction (LVEF) and FS of Hdc−/− mice injected with anti-Ly6G antibody or its isotype after infarction. n = 3–5 mice in each group. (G) Representative images and quantitation of Masson staining of hearts from Hdc−/− mice injected with anti-Ly6G antibody or its isotype on Day 7 after infarction. n = 3–5 mice in each group. Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.001. n.s., no significant. Scale bar, 50 μm.

To figure out whether the increased neutrophils contributed to the aggravated cardiac injury of Hdc−/− mice, the anti-Ly6G antibody was applied to neutralize neutrophils after myocardial infarction. The EF and FS in Hdc−/− mice were improved by the administration of anti-Ly6G antibody but not IgG (Fig. 2F). Similarly, compared with IgG, anti-Ly6G antibody reduced the cardiac fibrosis of Hdc−/− mice (Fig. 2G). These findings suggested that Hdc deficiency-aggravated myocardial injury after MI may be due to the effect of Hdc deficiency mainly on the regulation of neutrophils.

Hdc deficiency dysregulates the characteristic cellular activities of neutrophils

In a previous study, Hdc deficiency was proposed to promote phagocyte infiltration in infectious inflammation. Nevertheless, little is known about its role in sterile inflammation. We first investigated the effect of Hdc deficiency on the adhesion and migration of neutrophils. Compared with Hdc+/+ neutrophils, Hdc−/− neutrophils showed significantly decreased adhesion to HUVECs after stimulation with TNF-α (Fig. 3A). However, Hdc−/− neutrophils have a higher migration ability with the attraction of IL-8 through Transwells (Fig. 3B). These data indicated that Hdc deficiency led to disordered neutrophil priming in a sterile inflammatory environment.

Figure 3

Hdc deficiency dysregulates the characteristic cellular activities of neutrophils. (A) Calcein-AM-labeled neutrophils isolated from the bone marrow of Hdc−/− and Hdc mice were incubated with HUVECs. The percentage values of Calcein-AM+ adherent cells were measured after washing (n = 4). (B) Calcein-AM-labeled neutrophils isolated from the bone marrow of Hdc−/− and Hdc mice were cultured in Transwell plate to assess migration capacity. Green fluorescence showed the migrated cells across Transwell (n = 4). (C) FACS analysis of the generation of ROS in Hdc−/− and Hdc mice derived neutrophils before and after PMA treatment (n = 4–6). (D) Representative pictures of immunofluorescence staining of Hdc−/− and Hdc neutrophils after exposure to PMA for 18 h with Sytox green and Hoechst 33342 (blue). (E) The areas of Sytox green+ NETs derived from Hdc−/− and Hdc neutrophils were quantified (n = 4). (F) Representative images of immunofluorescence co-staining of NE (green), citH3 (red), and DAPI (blue) in cardiac slices of Hdc−/− and Hdc mouse on Days 1 and 7 after MI or shame operation (n = 4–5). (G) Percentage values of neutrophils with evidence of NET formation (NE+/citH3+) in the cardiac slices of Hdc−/− and Hdc mice were analyzed (n = 5). Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01. Scale bar: 50 μm.

The response of neutrophils to stimulation relies on the production of ROS and the formation of NETs under both infectious and sterile inflammatory conditions. Therefore, we investigated whether Hdc affected the generation of ROS in neutrophils by flow cytometry. Regardless of PMA treatment, Hdc−/− neutrophils produced significantly higher levels of ROS than Hdc+/+ neutrophils did (Fig. 3C). Accumulating studies have reported that excessive ROS promote the formation of NETs, which is a unique way of exacerbating peripheral cell death. Therefore, we assessed the areas of NETs formation under PMA treatment. Bone marrow-derived neutrophils from Hdc−/− mice, after treatment with PMA in vitro, were shown to have a larger area of Sytox green-positive cells that represented NETs (Fig. 3D and E).

Furthermore, although there was no significant difference in the number of neutrophils infiltrating into the infarcted zone on Day 1 after MI in Hdc+/+ and Hdc−/− mice as shown in Fig. 2, the co-staining of neutrophil elastase and citrullinated histone H3 revealed that there were much more NETs in the cardiac infarcted zone of Hdc−/− mice even on Day 1 after MI, compare with Hdc+/+ mice (Fig. 3F and G). These results indicated that Hdc played an essential role in regulating the characteristic cellular activities of neutrophils, which may be involved in the more severe pathological progression of MI in Hdc−/− mice.

Hdc deficiency in neutrophils promotes cardiomyocyte death and cardiac fibroblast proliferation/migration through NETosis

Hdc deficiency in neutrophils affected the cellular activities of neutrophils, as observed above. Next, we investigated the effects of neutrophils on cardiomyocytes. Neutrophils isolated from the bone marrow of Hdc+/+ or Hdc−/− mice were co-cultured with cardiomyocytes for 24 h under hypoxia condition after activated by PMA. The TUNEL assay data demonstrated that Hdc−/− neutrophils caused more cardiomyocyte death than Hdc+/+ neutrophils did (Fig. 4A). However, the difference was abolished by treatment with DNase I, which could disrupt the structure of NETs. Moreover, by staining with citH3 and TUNEL in cardiac sections, the in vivo study revealed that there were more NETs accompanied by more cardiomyocyte death on Day 1 after MI in Hdc−/− mice (Fig. 4B) than in Hdc+/+ mice, indicating a positive relationship between NETs and cardiomyocyte death. These findings suggest that Hdc deletion in neutrophils promoted cardiomyocyte death through NETosis.

Figure 4

Hdc deficiency in neutrophils augments cardiomyocytes death and cardiac fibroblasts proliferation/migration through NETosis. (A) Representative images of TUNEL assay of cardiomyocytes treated with NETs from Hdc−/− and Hdc neutrophils under hypoxia condition. DNase I was employed to interrupt the structure of NET. TUNEL positive cells are shown in red. α-Actinin positive cardiomyocytes are displayed as green. Nucleus is stained with Hoechst 33342, in blue. Percentage values of TUNEL+ cells were calculated and are shown as mean ± SEM (n = 4–5). ∗∗P < 0.01; n.s., no significant. (B) Representative images of immunofluorescence co-staining of citH3 (green), TUNEL (red) and DAPI (blue) in cardiac slices of Hdc−/− and Hdc mice on Day 1 after MI or sham operation. The percentage values of citH3+ NETs and TUNEL+ cells were quantified and are shown as mean ± SEM (n = 5). ∗P < 0.05. (C) Representative images of EdU (red), Vimentin (green), DAPI (blue) co-staining for fibroblasts treated with NETs from Hdc−/− and Hdc neutrophils. DNase I was employed to interrupt the formation of NETs. The percentage values of EdU + cells were calculated and are shown as mean ± SEM (n = 4–5). ∗∗P < 0.01; n.s., no significant. (D) Representative images of fibroblast migration assay. A confluent monolayer of fibroblasts was scraped with 200 μL pipette tip and co-cultured with NETs from Hdc−/− and Hdc neutrophils for 24 h. DNase I was employed to interrupt the formation of NETs. The ratios of wound healing were quantified and are shown as mean ± SEM (n = 3). ∗∗P < 0.01 vs. vehicle group; ΔP < 0.05 vs. Hdc+/+ + DNase; ##P < 0.01 vs. Hdc group; $$P < 0.01 vs. Hdc−/− group. Scale bar: 50 μm.

Studies have reported that excessive NETs can induce fibrosis27, 28, 29, which participates in multiple pathological processes. Thus, we explored whether Hdc deficiency altered the effect of neutrophils on cardiac fibroblasts. Cardiac fibroblasts were co-cultured with Hdc+/+ or Hdc−/− mouse-derived NETs. An EdU incorporation assay was employed to determine the proliferation of cardiac fibroblasts. The results revealed that Hdc−/− neutrophils boosted the proliferation of fibroblasts to a higher degree than Hdc+/+ neutrophils did (Fig. 4C). However, when treated with DNase I simultaneously, the ability of NETs to promote the proliferation of fibroblasts was diminished to similar levels in both groups (Fig. 4C). Next, we performed wound healing assays to investigate the influence of Hdc+/+ and Hdc−/− NETs on the migration of fibroblasts. The migration of fibroblasts co-cultured with Hdc−/− NETs was faster than that of fibroblasts co-cultured with Hdc+/+ NETs (Fig. 4D). Similarly, DNase I treatment decreased the migration ability of both groups to the same degree (Fig. 4D). All these results demonstrated that Hdc deficiency in neutrophils exerted its pro-effect on cardiomyocyte death and fibroblast proliferation and migration by NETosis.

PRMT1 is negatively regulated by histamine in myeloid cells

To unveil the mechanism through which Hdc regulates NETosis, we first sorted CD11b positive cells from the bone marrow of Hdc−/− or Hdc+/+ mice for microarray analysis and then screened out the DEGs regulated by Hdc deficiency or histamine, through combining our data and published myeloid cell transcriptome data from other research groups.

Differential gene analysis revealed that there were 3096 DEGs between the Hdc+/+ and Hdc−/− groups, and 1551 DEGs between groups with or without histamine treatment (Fig. 5A). Additionally, according to a previous study, there were 2628 DEGs between the Hdc-high expression group (Hdc-GFPHigh) and the Hdc-low expression group (Hdc-GFPlow) (Fig. 5A). Among them, 64 DEGs were overlapped in the three groups (Fig. 5A). Their relative expression levels were shown in the heatmaps (Fig. 5B).

Figure 5

PRMT1 is negatively regulated by histamine in myeloid cells. (A) The Venn diagram shows the numbers of significantly DEGs in the paired groups of Hdc−/−vs. Hdc myeloid cells (blue), with vs. without histamine treated myeloid cells (yellow), Hdc high expressed (Hdc-GFPhi) vs. following Hdc low expressed (Hdc-GFPlo) myeloid cells (red). 64 DEGs were identified in all three paired groups. (B) The heatmaps of the shared 64 DEGs from the above three paired groups. (C) Triangle plot shows the increased/decreased expression of the 64 DEGs in the indicated groups. (D) The representative images of Western blots of the protein levels of PRMT1 in neutrophils isolated from the bone marrow of Hdc−/− and Hdc mice (n = 6). Data are shown as mean ± SEM. ∗P < 0.05 vs. vehicle group. (E) The representative images of Western blots of the protein levels of PRMT1 in HL60 cells treated with the indicated doses of histamine for 24 h (n = 3). Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01 vs. vehicle group; n.s. no significant. (F) The representative images of Western blots of the protein levels of PRMT1 in HL60 cells treated with 10−4 mol/L histamine for the indicated time (n = 3). Data are shown as mean ± SEM. ∗∗P < 0.01 vs. vehicle group. (G) The relative ROS levels produced in neutrophils from Hdc−/− and Hdc mice were determined after the treatment of PMA along with histamine (HA) or PRMT1 inhibitor (MS023). ∗∗P < 0.01, Hdc−/− + PMA vs. Hdc−/− + PMA/HA; ##P < 0.01, Hdc−/− + PMA vs. Hdc−/− + PMA/MS023. Data are shown as mean ± SEM (n = 4). (H) The areas of NETs formed by neutrophils isolated from Hdc−/− and Hdc mice were calculated after the treatment of PMA along with HA or MS023. ∗∗P < 0.01, Hdc−/− + PMA vs. Hdc−/− + PMA/HA; ##P < 0.01, Hdc−/− + PMA vs. Hdc−/− + PMA/MS023. Data are shown as mean ± SEM (n = 4).

Among the 64 DEGs, we noted that the expression level of Prmt1 was downregulated in the histamine group and Hdc-GFPhigh group but augmented in the Hdc−/− group (Fig. 5C). PRMT1 is a protein arginine methyltransferase that has been reported to methylate a variety of protein substrates, including those involved in gene transcription, DNA damage repair, signal transduction, and protein translocation.

To verify the alteration of Prmt1 level in myeloid cells, we firstly detected its expression in neutrophils. Data showed that the level of Prmt1 was transcriptionally and translationally increased in Hdc−/− mouse-derived neutrophils (Fig. 5D and Supporting Information Fig. S3A).

To test whether the expression of PRMT1 is regulated by histamine in granule cell like HL60 cells, the concentration gradient and time course assay were performed on HL60 cells. Our data show that the expression of PRMT1 was decreased responding to the treatment of 10 and 100 μmol/L histamine (Fig. 5E). And it began to decrease at 6 h after histamine treatment, kept for more than 72 h (Fig. 5F and Fig. S3B).

Next, we investigated the expression of other members of PRMT family, with the treatment of histamine. As shown in Fig. S3C, only PRMT1, with the greatest Ct value, has been downregulated among PRMT1–6, after the treatment with histamine.

To verify whether the enhanced ROS generation and NETosis caused by Hdc deficiency in neutrophils were mediated through PRMT1, we inhibited the activity of PRMT1 using MS023. MS023 is the specific type I inhibitor of PRMTs, which had been reported to be applied in treating acute myeloid leukemia (AML) in a relatively low dose. To minimize its side effects, we had chosen the optimal concentration of MS023 on PRMT1 according to the reported C50 of MS023 on PRMT1. The data showed that the increased production of ROS and the augmented formation of citH3-positive NETs in Hdc−/− neutrophils were reduced after 24 h of treatment with MS023 or histamine (Fig. 5G and H).

Histamine repressed the transcription of PRMT1 by restricting the DNA accessibility of the transcriptional start site (TSS) region of PRMT1

There are four known histamine receptors (H1R, H2R, H3R, and H4R) expressed in myeloid cells, and we wondered which of them would be responsible for the repression of histamine on PRMT1 in neutrophils. Among the inhibitors of four histamine receptors, only the H1R antagonist pyrilamine rescued the decrease of PRMT1 induced by histamine treatment in HL60 cells (Fig. 6A), suggesting that the regulation of PRMT1 by histamine was dependent on H1R.

Figure 6

Histamine repressed the transcription of PRMT1 through restricting DNA accessibility of the TSS region of PRMT1. (A) HL60 cells were stimulated with histamine (100 μmol/L) for 12 h in the presence of specific antagonists for H1 receptor (pyrilamine, 10 μmol/L), H2 receptor (cimetidine, 10 μmol/L), H3 receptor (thioperamide, 10 μmol/L), or H4 receptor (JNJ-7777120, 10 μmol/L). HA, histamine. Data are shown as mean ± SEM (n = 4). ∗∗P < 0.01 vs. vector group; ##P < 0.01 vs. HA treated group. (B) HL60 cells were transfected with siRNA of scramble control (siC or siCon), CEBPB (siCEBPB), or STAT1 (siSTAT1), along with HA treatment. The mRNA levels of PRMT1 were analyzed by qRT-PCR (n = 3). Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01 vs. vector group. (C) Schematic of the PRMT1 locus containing transcription start site (TSS, red) and the later part (blue). Con, control. (D) HL60 cells were treated with or without 10−4 mol/L HA for 12 h. DNase sensitivity assay was performed on the TSS and the intron of prior locus (pl) of PRMT1. The statistical results are shown as mean ± SEM (n = 3). ∗P < 0.05. (E) The mRNA levels of PRMT1 in HL60 treated with HA and/or oligomycin were determined by qRT-PCR. n = 3. ∗∗P < 0.01. (F) HL60 cells were transfected with siC, BRG1/BRM (siBB), along with HA treatment. The mRNA levels of PRMT1 were analyzed by qRT-PCR (n = 3). ∗∗P < 0.01; n.s., no significant.

Next, we explored the mechanism by which histamine inhibited the expression of PRMT1. Since PRMT1 was decreased at the mRNA level, suggesting the regulation of its expression at the transcriptional level, we attempted to identify the key transcription factors of PRMT1. Signal transducer and activator of transcription 1 (STAT1) and CCAAT enhancer binding protein beta (CEBPB) have been reported to bind to the promoter region of PRMT1 in fibroblasts,. However, knocking down STAT1 or CEBPB in HL60 cells did not abolish the repressive effect of histamine on the level of PRMT1 mRNA in HL60 cells (Fig. 6B), which indicated that there was a unique transcriptional mechanism different from those in known cell types during the process of histamine repressing the transcription of PRMT1.

To further characterize the pathway by which histamine regulates PRMT1 expression, DNA accessibility was assessed by DNase digestion assay. The data revealed that TSS of PRMT1 became DNase I resistant after histamine treatment (Fig. 6C and D).

Since the ATP-dependent SWI/SNF complex contributes to chromatin remodeling, we tested whether it was involved in the regulation of PRMT1. Silencing the catalytic subunits of the SWI/SNF complex, Brahma (BRM) and Brahma related gene 1 (BRG1), abolished the repressive effect of histamine on the expression of PRMT1 (Fig. 6E). Similarly, with oligomycin treatment, which inhibited ATP production, the repression of histamine on the mRNA level of PRMT1 was rescued (Fig. 6F). In general, these findings revealed that HDC/histamine downregulated the transcription of PRMT1 by restricting the DNA accessibility of PRMT1 in an ATP–SWI/SNF-dependent manner.

Asymmetric arginine dimethylation profiling of PRMT1 targets are regulated by histamine

To further identify the downstream proteins methylated by PRMT1, the proteins with asymmetric dimethylated arginine (ADMA) were harvested by co-immunoprecipitation (co-IP) with ADMA antibody from the lysate of cells with or without histamine treatment. Coomassie blue staining showed that there was a significant decrease in the level of proteins with ADMA responding to histamine treatment (Fig. 7A and Supporting Information Fig. S4A).

Figure 7

The asymmetric dimethylated arginine (ADMA) profiling of PRMT1 targets were regulated by HA. (A) Coomasie blue staining of proteins with ADMA from HL60 cells with/without HA treatment. (B) The Venn diagram shows the proteins with ADMA in control and HA treated groups. 274 proteins were detected uniquely in the control group and 202 proteins uniquely in the HA treated group. (C) KEGG pathway enrichment analysis for the differentially expressed proteins uniquely detected in control and HA treated groups. The Rich ratio of x-axis refers to the ratio of selected gene numbers annotated in this pathway terms to all gene numbers annotated in this pathway term. The calculating formula is Rich ratio = term candidate gene number/term gene number. The size and color of the bubbles represent the number of differentially expressed proteins enriched in the pathway and enrichment significance, respectively. (D–F) GO of BP (D), CC (E), and MF (F). Enrichment analysis of histamine regulated proteins with ADMA modification were shown. The size and color of the bubbles represent the number of differentially expressed proteins enriched in the pathway and enrichment significance, respectively.

Then, we performed mass spectrometry to identify the profile of the proteins with ADMA before and after histamine treatment. There were 1488 proteins detected in the group without histamine treatment and 1392 proteins detected in the group with histamine treatment. Among them, 339 unique proteins existed in the former, while 243 proteins uniquely existed in the latter (Fig. 7B and Table 136, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49).

Table 1

ADMA proteins of the RAS family related to ROS.

Group	ADMA proteina	ROS
Only in Con	RAP1B36	↑
	RIN3, RAB15, RAB18, RAB5B, RAB6A, RAP2A	–
Both in Con and HA	RAP1A37	↑
	G3BP138	↑
	IQGAP139	↑
	RAC139	↑
	RAC240	↑
	RAB1041	↑
	RAB27A42	↑
	RAB3543	↑
	ARF644	↑
	CDC4245	↓
	RASA346	↓
	RAB3147	↓
	ARF448	↓
	RASSF2, IQGAP2, RAB11A, RAB14, RAB1A, RAB1B, RAB21, RAB2A, RAB32, RAB5A, RAB5C, RAB6B, RAB7A, RAB8A, ARF5, ARF3, RALA	–
Only in HA	NRAS49	↓
	AK3, RAB8B, RAB8B	–

The ADMA proteins of RAS family are divided into 3 groups: only in control (Con), both in Con and histamine (HA) groups, and only in HA group. Their roles in ROS production are indicated as arrows. ↑ represents the gene promotes the production of ROS; ↓ presents the gene decreases the production of ROS.

Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the differentially expressed proteins revealed that proteins with ADMA regulated by histamine were mainly enriched in the cellular process of endocytosis, RNA processing, DNA replication, and amino acid degradation (Fig. 7C). GO analysis indicated that proteins with ADMA regulated by histamine were enriched for autophagy and Golgi vesicle transport (Fig. 7D). Moreover, these proteins were enriched in mitochondrial, nuclear, and cell cortex regions of the cell (Fig. 7E), and their functions were focused on GTP binding, phosphatidylinositol phosphate binding, and nucleoside binding (Fig. 7F).

Furthermore, we noticed that 42 members of the Ras family were enriched in the proteins pulled down by the ADMA antibody (Table 1), among which 15 members were reported to be involved in ROS production. Significantly, 7 out of the 42 members, especially Rap1, the key component of NADPH oxidase, were specifically enriched in the untreated group but disappeared in the histamine-treated group. The AMDA modification of Rap1 was significantly reduced in histamine treated group (Fig. S4B), suggesting Rap1 was likely the downstream target of HDC–PRMT1 signal in regulating the ROS generation and NETs formation.

Inhibiting the activity of PRMT1 can rescue the aggravation of myocardial injury caused by Hdc deficiency after MI

The above findings implied that Hdc deficiency was deleterious to heart function after MI, possibly through upregulating PRMT1, leading to promoted ROS generation and excessive NETosis. Thus, to evaluate the role of PRMT1 in cardiac function after MI, Hdc−/− mice were intravenously injected with histamine or the inhibitor of PRMT1 (MS023) before and after MI operation. Echocardiography showed that the decreased EF and FS in Hdc−/− mice were partially rescued by treatment with MS023 and histamine on Day 7 after surgery (Fig. 8A, B and C). Meanwhile, Masson trichrome staining revealed that pretreatment with histamine and MS023 significantly slowed the process of fibrosis in Hdc−/− mice after MI (Fig. 8D and Supporting Information Fig. S5A).

Figure 8

Inhibiting the activity of PRMT1 can rescue the aggravation of myocardial injury caused by Hdc deficiency after MI. (A) Representative echocardiographic images of Hdc−/− and Hdc mice on Day 7 after ligation. Hdc−/− and Hdc mice were intravenous injected with HA or MS023 prior to surgery. (B and C) Quantitation of LVEF and FS of Hdc−/− and Hdc mice 7 days after infarction. n = 6–8 mice in each group. Data are shown as mean ± SEM. ∗∗P < 0.01. (D) Representative images of Masson staining of hearts from Hdc−/− and Hdc mice. Hdc−/− and Hdc mice were intravenous injected with HA or MS023 prior to surgery. Scale bar, 50 μm. (E) MPO/DNA and ADMA levels in plasma from Hdc−/− and Hdc mice on Day 7 after ligation were determined by ELISA according to the manual instructions. Hdc−/− and Hdc mice were intravenous injected with HA or MS023 prior to surgery. Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, n = 6–8 mice in each group. (F) Correlation of MPO/DNA and ADMA levels in plasma from Hdc−/− and Hdc mice on Day 7 after ligation. Hdc−/− and Hdc mice were intravenous injected with HA or MS023 prior to surgery. R2 = 0.34, P < 0.01. Data are shown as mean ± SEM, n = 30.

Next, the levels of ADMA and the MPO/DNA complex, which is a measure of NETosis, were detected by ELISA in the plasma of Hdc−/− mice with or without surgery. The levels of ADMA and NETs were greatly augmented in Hdc−/− mice after MI and could be reduced by pretreatment with histamine and MS023 to a level similar to that in Hdc+/+ mice after MI (Fig. 8E, Fig. S5B and C). Moreover, the levels of NETs were positively related to the levels of ADMA (Fig. 8F), suggesting that the pathway of asymmetric arginine dimethylation of proteins, which was the main function of PRMT1, indeed contributed to the increased NETs observed in Hdc−/− mice.

Discussion

Despite the development of percutaneous coronary intervention (PCI), the prognosis of patients with ST-segment elevation myocardial infarction (STEMI) varies greatly. The different immune responses of patients to the ischemic injury during the early phase contribute to the diversity of clinical pathological progress. However, little is known about the intrinsic regulators of immune cells that influence inflammatory responses to myocardial injury. This study revealed that Hdc/histamine regulated neutrophil mobilization and limited additional reactive oxidative species and NET generation after myocardial infarction by the H1R–SWI/SNF–PRMT1 pathway (Fig. 9). Our findings identify a new mechanism for neutrophil dependent inflammation and provide a rationale for neutrophils as a therapeutic target for acute myocardial infarction.

Figure 9

Schematic illustration of the effect of HDC on neutrophils after MI. In neutrophils, autocrine, paracrine, or exogenous of histamine, by activating histamine H1 receptor, repress PRMT1 expression, ADMA production, and ROS generation by ATP–SWI/SNF dependent chromatin remodeling. After MI, the deletion of Hdc, without endogenous or exogenous histamine, aggravates cardiomyocyte death and fibrosis, by enhanced neutrophil infiltration and ROS dependent NETosis.

The expression of Hdc in the bone marrow has been shown to decrease across the lifespan, implicating its immunological role in aging related diseases (Supporting Information Fig. S6). Previous data have shown that Hdc/histamine participates in regulating the progression of acute heart injury and heart failure, but the underlying mechanisms have not been fully unveiled. Here, we illustrated the transcriptional profile of the injured hearts of Hdc−/− mice on Days 0, 1, and 7 after ligation of the left anterior descending coronary artery. Time course analysis of GO showed that immunological phenomena was closely related to the progression of MI and that the immune state in the heart changed along with the development of myocardial remodeling. Immune recruitment analysis indicated that immunocytes, neutrophils especially, infiltrated infarcted hearts within a different time frame in Hdc−/− mice, compared with Hdc+/+ mice. Hdc deficiency greatly increased the counts of neutrophils in the myocardium on Days 3 and 7 after MI.

Although Hdc was globally knockout in our experiments, our previous data had suggested that the expression of HDC in myocardium is extremely low to be detected (data not shown) and there were no abnormalities in the cardiac structure and function of Hdc−/− mice under normal conditions. Moreover, neutralizing Ly6G+ neutrophils ameliorated the prognosis of MI in Hdc−/− mice in the present work, which suggested the contributing role of neutrophils in the aggravated cardiac function loss in Hdc−/− mice with MI.

Neutrophils are vital for clearing pathogens or debris in acute inflammation. However, in the context of MI, abnormal neutrophils are detrimental to cardiac function. Clinically, the neutrophil-to-lymphocyte ratio has been employed to predict the survival of patients with STEMI and the improvement of coronary artery disease events is found to be correlated with neutrophil reduction. These findings suggest that neutrophils to be a potential target for the treatment of myocardial infarction. Here, we first reveal that Hdc/histamine is an endogenous factor involved in regulating the infiltration of neutrophils into the heart after MI.

Our data showed that the migration of neutrophils was reinforced in Hdc−/− mice after MI, in accordance with the phenomenon of infectious inflammation described previously. Surprisingly, the adhesion of Hdc−/− neutrophils to endothelial cells was decreased. Histamine has been reported to activate endothelial cells by promoting the expression of adhesion related genes, but the role of histamine in the adhesion of neutrophils is still unknown. In Hdc−/− mice, with the increased migration but the decreased neutrophil adhesion, why the infiltrated neutrophils were significantly increased on Days 3 and 7, but not on Day 1 after injury is the next scientific question we are eager to answer. Moreover, it will be interesting to analyze the effect of Hdc on neutrophils at the earlier stage such as 1, 3, 6, or 12 h post MI, as the adhesion of neutrophils affected the recruitment at the start of inflammation.

Oxidative burst is the way of neutrophils that eliminate pathogens or respond to inflammatory stimuli. Histamine is thought to inhibit ROS driven by opsonized zymosan particles or fMLP in a PKC-independent way,. Our data showed that deletion of Hdc led to neutrophils in a hyperoxidative state not only under the treatment of PMA, an agonist of PKC, but also in the baseline state.

NETosis is a unique mechanism of cell death in neutrophils driven by ROS with various stimulations. It has been reported that NETs lead to the death of cells by histones,,. In addition, NETs have recently been reported to promote fibrosis and cell activation/proliferation by neutrophil elastase (NE) dependent matrix remodeling. Herein, in vivo experiments revealed that excessive NETs accompanied with augmented cardiac injury after MI in Hdc−/− mice. Our data showed that cardiomyocyte death and fibroblast proliferation/migration were aggravated by Hdc−/− NETs. These findings indicate that aggregated neutrophils, which migrate into the heart, may promote cell death of the myocardium, expand the injured zone, and pave the way for fibroblasts rather than purging cell debris or reducing inflammation.

To clarify the underlying mechanisms through which Hdc/histamine regulated ROS generation and the formation of NETs, we analyzed the DEGs regulated by Hdc, by combining our transcriptomic data and the published transcriptomic data of myeloid cells in the GEO database. After screening, we found that the expression of PRMT1 was tightly repressed by HDC/histamine at the mRNA and protein levels, while rescued by the antagonist of H1R. PRMT1 is an important enzyme responsible for asymmetric arginine dimethylation. It has been reported that there is a negative correlation between histamine and methylation in the serum of patients with behavioral disorders. In the current study, histamine treatment, which repressed the transcriptional level of PRMT1, reduced the level of proteins with ADMA modification, which is consistent with the phenomenon observed in behavioral disorders patients. Moreover, inhibiting the catalytic activity of PRMT1 by MS023 could limit the excessive ROS generation and NETosis, and improve the cardiac function of Hdc−/− mice post MI, suggesting that the effect of PRMT1 on NETosis and MI mediated by HDC/histamine–PRMT1 signaling is dependent on its catalytic activity of arginine methylation.

STAT1 and CEBPB have been reported to be responsible for the regulation of PRMT1 transcription in fibroblasts,,. However, in our work, the data showed that neither transcription factor participated in the regulation of histamine on PRMT1 expression. This may be due to the different kinds of cell types. Instead, histamine repressed PRMT1 transcription by restricting DNA accessibility to its TSS region through ATP-dependent chromatin remodeling. Chromatin remodeling occurs in the process of cell differentiation, while PRMT1 has been reported to participate in the maintenance of short-term hematopoietic stem cells, which implies that PRMT1 may participate in histamine-induced myeloid cell maturation. This suggested that ATP analogues or inhibitors of enzymes of chromatin remodeling are the potential targets to modulate PRMT1 expression and NETosis. Beside that histamine repressed the transcription of PRMT1 through restricting the DNA accessibility to the TSS region of PRMT1, our results of co-IP revealed that PRMT1 and HDC could interact with each other slightly, which was not affected by histamine or MS023 treatment as shown in Fig. S6B and C. The phenomenon aroused an interesting question that besides for the catalytic activity of arginine methylation, would PRMT1 act on HDC in a novel way as a negative feedback regulation of histamine inhibiting PRMT1 transcription? Or may PRMT1 and HDC form a complex with other proteins to function in cells? These questions need more efforts to figure out in the future work.

Type I PRMTs, including PRMT1–4, PRMT6, and PRMT8, promote ADMA, reducing the positive charge and hydrophilicity of proteins. Many published studies certified that this type of modification involved in multiple cell activities, including gene transcription, cell signaling, mRNA translation, DNA damage signaling, protein trafficking, protein stability, and pre-mRNA splicing,64, 65, 66. Previous study implicated that the deficiency of Ddah1, which is the key enzyme for ADMA degradation, raised ROS levels in cardiomyocytes. Moreover, it has also been reported that ADMA promotes ROS generation in macrophages and MPO release in polymorphonuclear neutrophils in the context of cardiovascular disease. In our work, the levels of ADMA were shown to be positively related to the levels of NETs in a mouse model of MI, implying that the ADMA playing a role in regulating ROS generation and the formation of NETs.

Although Pyun et al. reported that mice null for cardiac PRMT1 could exhibit dilated cardiomyopathy after 2 months, there is no other work exploring the role of PRMT1 during acute MI up to now. Our work revealed that the expression of PRMT1 in neutrophils would be increased but showed no significant change in myocardium after Hdc deletion. Inhibiting PRMT1 could limit the excessive ROS generation and NETosis in Hdc-deficient neutrophils, improving the cardiac function of Hdc−/− mice post MI. Similarly, directly reducing neutrophils could also attenuate cardiac injury post MI in Hdc−/− mice. Based on these, we consider that PRMT1 mainly in neutrophils played a potential role in regulating the aggravated cardiac injury post MI in Hdc−/− mice. Our results and the work of Pyun et al. suggested that PRMT1 likely possess different functions in different cells and different diseases models.

To determine the downstream factor of PRMT1 in regulating the formation of NETs, the ADMA modified profiling of proteins were determined by mass spectrometry in our work. GTP binding and GTPase activity were specifically enriched in the histamine untreated group. In neutrophils, NADPH oxidase is the main complex responsible for ROS generation. It contains catalytic gp91phox (also known as NOX2), regulatory components, p22phox, p40phox, p47phox, p67phox, and Rac/Rap1,. Rac/Rap1, belongs to the Ras family, many members of which are involved in the regulation of ROS (Table 1). The hydrophilicity of Rap1 would be reduced by its ADMA modification, which may facilitate its localization to the NADPH-oxidase complex on the membrane of mitochondria. The reduction of the ADMA modification of Rap1 caused by histamine suggested the downstream role of Rap1 in HDC–PRMT1 signaling. Taken together, our findings strongly suggested that the activity of NADPH oxidase could be regulated by PRMT1 to promote ROS production and NET formation, indicating that NADPH oxidase or Rap1 could also be the potential target for the immune modulation for the cardiovascular disease.

Conclusions

Our study demonstrates that the lack of Hdc promotes NETosis by the H1R–SWI/SNF–PRMT1–ROS pathway, and hence leads to aggravated cardiomyocyte death and fibroblast activation. Targeting any one of them may provide insights into new treatments for myocardial infarction and consequent cardiac remodeling. These findings improve our understanding of the significant role of neutrophil and NET formation in cardiovascular disease and provide new biomarkers and pharmaceutical targets for identifying and tuning the detrimental immune state in cardiovascular disease.