SGLT2 Inhibitors Attenuate Neutrophil-Mediated Endothelial Injury in Coronary Artery Disease With Type 2 Diabetes Mellitus Through a GREM1-Associated Smad Signaling Axis

Abstract:

Type 2 diabetes mellitus (T2DM) is a critical risk factor for coronary artery disease (CAD), with neutrophils contributing to CAD through acute inflammatory responses. Sodium–glucose cotransporter 2 (SGLT-2) inhibitors offer cardiovascular benefits beyond glycemic control, but their mechanisms involving innate immunity and neutrophil function in T2DM-related CAD remain unclear. Neutrophils were isolated from healthy controls, patients with CAD and T2DM, and patients receiving SGLT2 inhibitor therapy. Bulk transcriptomic analysis, flow cytometry, ELISA, and western blotting were used to assess gene expression, signaling changes, and neutrophil activation. Co-cultures of neutrophils with luciferase-expressing human aortic endothelial cells (HAEC-Luc) assessed endothelial injury, and a diabetic rat model was used to assess in vivo cardiac injury. Results showed elevated peripheral neutrophil cell-free dsDNA, elastase 2, and GREM1 expression in patients with T2DM-CAD, reversed by SGLT-2 inhibitors. T2DM-CAD neutrophils induced significant endothelial injury, ameliorated by SGLT-2 inhibition, or GREM1 blockade. Mechanistically, SGLT-2 inhibitors suppressed GREM1, restored TGF-β/Smad signaling, and reduced neutrophil-mediated cytotoxicity. In vivo, SGLT-2 inhibitors alleviated cardiac damage in diabetic rat, with suppressed neutrophil activation and upregulated TGF-β signaling. Our findings suggest that increased GREM1 expression in neutrophils is associated with impaired Smad1/5/9-related signaling and enhanced endothelial injury in T2DM-CAD, and SGLT-2 inhibitors exert vascular protection through GREM1 downregulation and TGF-β/Smad pathway modulation, highlighting a novel immunoregulatory mechanism with translational potential for diabetic cardiovascular outcomes.

INTRODUCTION

Cardiovascular diseases (CVDs) are the leading cause of mortality and reduced quality of life worldwide, with coronary artery disease (CAD) being the most prevalent form of CVD.^1,2 The pathogenesis of CAD is complex and closely associated with multiple risk factors, including hyperglycemia, chronic inflammation, and endothelial dysfunction.³ Type 2 diabetes mellitus (T2DM) is a major risk factor for CAD and accelerates atherosclerotic progression.^4,5 Endothelial dysfunction and chronic inflammation are central pathophysiological mechanisms underlying the development and progression of CAD in diabetic populations.^6,7 Recent basic and translational studies have revealed that sodium–glucose cotransporter 2 inhibitors (SGLT2i) ameliorate vascular endothelial dysfunction—a key feature of CAD—through several interrelated pathways. These include the reduction of oxidative stress and reactive oxygen species production, attenuation of proinflammatory cytokine release (such as TNF-α and IL-6), enhancement of nitric oxide (NO) bioavailability to improve endothelium-dependent vasodilation, and inhibition of endothelial cell apoptosis through mitochondrial and metabolic regulation.^8–10 Notably, animal models in which myocardial SGLT2 was knocked out have failed to show direct improvement in myocardial reperfusion injury and in transverse aortic constriction–induced heart failure models, suggesting that the cardioprotective actions of SGLT2i may be mediated through indirect pathways rather than direct actions on cardiac myocytes.^11,12 Growing evidence indicates that SGLT2i confer cardioprotective effects in part by modulating the immune system, particularly through the regulation of monocytes and macrophages, which play pivotal roles in atherosclerotic inflammation and plaque instability.^13,14 Notably, SGLT2i have been shown to attenuate macrophage-driven atherosclerosis by suppressing foam cell formation and reducing proinflammatory phenotypes, thereby stabilizing atherosclerotic plaques.¹⁵

Despite advances in elucidating the immunomodulatory functions of SGLT2 inhibitors, their effects on neutrophil-mediated endothelial injury in the context of diabetic coronary vasculature remain inadequately characterized. Neutrophils are key effectors of both acute and chronic vascular injury, primarily through the release of cytotoxic mediators. Therefore, clarifying the role of neutrophils in SGLT2i-mediated vascular protection is of both clinical and translational relevance.^16,17 Neutrophils, the most abundant white blood cells, are key mediators of innate immunity.^18,19 After coronary ischemia, neutrophils rapidly infiltrate the myocardium, releasing proteolytic enzymes and proinflammatory cytokines that aggravate vascular injury and promote plaque instability.²⁰ Critically, neutrophils can form extracellular traps (NETs) composed of DNA, modified histones (CitH3), and enzymes, such as MPO and neutrophil elastase, which amplify inflammation and tissue damage.^21–24 Excessive NET formation is recognized as a major contributor to vascular injury in T2DM and CAD.^23,25

In this study, we aimed to elucidate the effects of SGLT2 inhibitors on neutrophil-induced vascular endothelial injury in patients with CAD and T2DM and to explore the underlying mechanisms through which SGLT2i regulate neutrophil activation and cytotoxicity.

MATERIALS AND METHODS

Transcriptome Sequencing Analysis

Blood samples were collected from healthy individuals, patients with CAD and T2DM, and patients with CAD and T2DM who received SGLT-2 inhibitor treatment. Human peripheral blood neutrophils were isolated from fresh EDTA-anticoagulated whole blood using the EasySep™ Direct Human Neutrophil Isolation Kit (STEMCELL Technologies, Cat. #19666) according to the manufacturer’s instructions. Briefly, non-neutrophil cells were labeled and removed by immunomagnetic negative selection, and the untouched neutrophil-enriched fraction was collected for downstream analysis. The purity of isolated neutrophils was verified by flow cytometry before RNA extraction. Total RNA was then extracted from purified neutrophils, and transcriptome sequencing was performed to identify differentially expressed genes (DEGs). Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis was conducted to determine the enrichment of signaling pathways.

Single-Cell RNA Sequencing

Single-cell RNA sequencing data were downloaded from the CZ CELLxGENE Discover database in .h5ad format, converted to a matrix using Python, and imported into R to construct a Seurat object (v5.0.0). Cells were filtered to retain those with 500–8000 detected genes, <50,000 UMI counts, and <7% mitochondrial reads. Data were normalized using LogNormalize with a scale factor of 10,000. The top 2000 highly variable genes were selected, and PCA was performed on them. The first 15 principal components were used for Louvain clustering (resolution = 0.5), visualized by UMAP. Cell types were annotated based on canonical marker genes. Differentially expressed genes were identified using FindAllMarkers (min.pct = 0.25, logfc.threshold = 0.25). Expression of GREM1 was visualized using feature and violin plots, and average expression per cell type was calculated.

Isolation of Primary Neonatal Rat Cardiomyocytes, Cardiac Fibroblasts, and Neutrophils

Hearts were excised from neonatal Wistar rats (1–3 days old), minced, and digested sequentially with trypsin (overnight at 4°C) and collagenase type II (15–20 minutes at 37°C). The cell suspension was filtered and subjected to differential plating for 90 minutes; the adherent fraction was cultured as cardiac fibroblasts (passages 2–3), while nonadherent cells were collected as cardiomyocytes and cultured in the presence of 0.1 mM Brdu. Neutrophils were isolated from whole blood or bone marrow by density gradient centrifugation using Histopaque, followed by hypotonic lysis of red blood cells. All cells were maintained in DMEM with 10% FBS, and purity was verified by immunofluorescence (cardiomyocytes: cardiac troponin T; fibroblasts: vimentin) or flow cytometry (neutrophils: HIS48).

Flow Cytometry

For purity assessment of isolated human neutrophils, cells were stained with fluorochrome-conjugated antibodies against human CD45 (BD, Cat 561863, San Jose, CA), CD16 (BD, Cat 561313, San Jose, CA), and CD66b (BD, Cat 561927, San Jose, CA). The gating strategy was performed sequentially based on total cells, single cells, CD45+ leukocytes, and CD16+CD66b+ neutrophils. The percentage of CD16+CD66b+ cells within the CD45+ population was used to define the purity of isolated neutrophils.For rat cardiac neutrophil analysis, single-cell suspensions were prepared from rat myocardial tissues and stained with fluorochrome-conjugated antibodies against CD45 (BD, Cat 561586, San Jose, CA), CD11b (BD, Cat 554982, San Jose, CA), and HIS48 (BD, Cat 570874, San Jose, CA). Neutrophils were identified as CD45+CD11b+HIS48+ cells. Flow cytometric acquisition was performed using a BD flow cytometer (BD Biosciences, San Jose, CA, USA), and data were analyzed using FlowJo software. 

Enzyme-Linked Immunosorbent Assay

Blood samples were allowed to clot at room temperature for 1 hour, followed by centrifugation at 1500g for 20 minutes at 4°C. Serum was harvested from the supernatant for subsequent analysis. The concentrations of cell-free dsDNA (cf-dsDNA) and elastase 2 were quantified using the Equalbit dsDNA HS Assay Kit (Vazyme, Cat EQ111-01, Nanjing, China) and a Neutrophil Elastase enzyme-linked immunosorbent assay (ELISA) kit (Abcam, ab270204, Cambridge, United Kingdom), respectively, according to the manufacturers' instructions.

Cell Culture

Neutrophils were isolated from blood samples using the Human Peripheral Blood Neutrophil Isolation Solution Kit (Solarbio, Cat P9040, Beijing, China) according to the manufacturer's protocol. Primary human neutrophils were isolated as described above. Human aortic endothelial cells constitutively expressing firefly luciferase (HAEC-Luc) were generated by lentiviral transduction of the luc2 gene under the CMV promoter and were cultured in RPMI 1640 medium containing 10% FBS. For coculture experiments, neutrophils were added to adherent HAEC Luc monolayers at ratios of 2.5:1, 5:1, 10:1, and 20:1, and incubated for 24 hours. Endothelial injury was quantified by measuring luciferase activity released into the culture supernatant, as increased extracellular luciferase directly reflects loss of plasma membrane integrity and cell death.

Western Blot Assay

Neutrophils were homogenized with RIPA lysis buffer containing protease inhibitors (Roche, Indianapolis, IN) and phosphatase inhibitors (Roche). Protein concentration was determined using a BCA Protein Assay Kit (Beyotime, China). Equal amounts of proteins were separated by SDS–PAGE and transferred to a nitrocellulose (NC) membrane, which was then blocked with 5% nonfat milk. The membrane was incubated overnight at 4°C with specific primary antibodies targeting GREM1 (Proteintech, Cat 18024-1-AP, Rosemont, IL), IL-1β (Proteintech, Cat 26048-1-AP), IL-6 (Proteintech, Cat 30953-1-AP), TNF-α (Proteintech, Cat 29652-1-AP), ICAM1(Proteintech, Cat 10020-1-AP), FAS (Proteintech, Cat 30601-1-AP), neutrophil elastase (Invitrogen, Cat PA5-115648), Smad1/5/9 (Abcam, Cat ab300164), phosphorylated Smad1/5/9 (Affinity, Cat AF8313, Cincinnati, OH), CitH3 (Affinity, Cat AF4365), histone 3 (H3) (Proteintech, Cat 17168-1-AP), and GAPDH (Proteintech, Cat 60004-1-Ig). The membranes were subsequently incubated with secondary antibodies, treated with ECL reagent, and imaged using a Tanon imaging system. The relative intensities of the protein bands were quantified using ImageJ software.

Measurement of Extracellular Luciferase Activity

Extracellular luciferase activity, reflecting endothelial cell membrane damage, was measured in the culture supernatant using a Firefly Luciferase Assay Kit (Beyotime, Cat RG005, China). In brief, coculture supernatants were collected and centrifuged to remove debris, mixed with luciferase assay reagent, and chemiluminescence was immediately measured using a multimode reader (Tanon, Shanghai, China).

Detection of H₂O₂, NO, and TNFα

The levels of H₂O₂, NO, and TNF-α in the cell culture medium were assessed using the Hydrogen Peroxide Assay Kit (Beyotime, Cat S0038, China), NO Detection Kit (Beyotime, S0021S, China), and Human TNF-α ELISA Kit (Beyotime, Cat PT518, China), respectively.

Diabetic Rat Models

Male Zucker diabetic fatty rats and Zucker lean rats were obtained from the Beijing Vital River Laboratory Animal Technology Co, Ltd All rats were kept at SPF animal care facility of Zhongshan Hospital, Fudan University. Male Zucker diabetic fatty rats aged 8 wk were fed with a high-fat diet. After 2 wk, those with a random blood glucose level >301 mg/dL were selected and subcutaneously injected with isoprenaline (ISO) at a dose of 1 mg/kg body weight for 10 consecutive days. These rats were then randomly divided into 2 groups: the model group and the SGLT-2 inhibitor treatment group. The SGLT-2 inhibitor treatment group was administered empagliflozin at a dose of 10 mg/kg/day, while the model group was given an equal volume of PBS. Male Zucker lean rats aged 8 wk, fed with a normal diet, were used as the normal control group. After 10 days, all rats were sacrificed, and myocardial tissues and blood samples were collected for subsequent detection. The experimental protocols were approved by the Animal Care and Use Committee of Zhongshan Hospital, Fudan University. The levels of blood glucose in plasma were detected by Blood Glucose Content Detection Kit (Solarbio, Cat BC2495, China), the activities of CK and CK-MB were detected by Creatine Kinase Assay Kit (Nanjing Jiancheng, Cat A032-1-1, Nanjing, China) and Rat CKMB ELISA Kit (Elabscience, Cat E-EL-R1327, Houston, TX).

HE Staining, Masson's Trichrome Staining, and Immunofluorescence Staining

Myocardial tissues were fixed in 4% paraformaldehyde (PFA) for 48 hours. Then the tissues were embedded in paraffin and cut into 5-μm slices with HE and Masson's trichrome staining. For immunofluorescence staining, mouse anti-MPO (Proteintech, Cat 66177-1-Ig) or anti-α-actin (CST, Cat 19245, Danvers, MA) and anti-RP-1 (BD Pharmingen, Cat 550000) were used as primary antibodies. After 3 washes with PBS, the slides were incubated with the corresponding fluorophore-conjugated secondary antibodies (CST) at room temperature for 30 minutes in the dark. For nuclear visualization, the sections were stained with DAPI solution for 10 minutes at room temperature. Finally, the slides were sealed with an antifluorescence quenching mounting medium. Slices were observed and analyzed by a microscope (Nikon, Tokyo, Japan) and a fluorescence microscope (Nikon, Japan).

Statistics

Statistical analysis was performed using SPSS software. In this study, continuous variables are reported as the mean ± standard error of the mean (SEM), with categorical variables expressed as percentages. Parametric data were analyzed using Student's t test and one-way analysis of variance with Bonferroni correction. For nonparametric data, the Kruskal-Wallis test with Dunn's post hoc test for multiple comparisons was applied. Fisher's Exact test and Fisher Freeman-Halton test were used to determine significant differences for categorical variables. A P-value of <0.05 was considered to be statistically significant.

RESULTS

SGLT2 Inhibitor Treatment is Associated With Changes in GREM1 Expression and Smad Signaling

In this study, CAD was diagnosed by coronary angiography (≥50% stenosis in at least one major epicardial artery or its branches), with images independently reviewed by 2 experienced cardiologists blinded to clinical data. Transcriptome sequencing analysis revealed that the use of SGLT-2 inhibitors reduced the expression of GREM1 in neutrophils compared with the control group, with enrichment analysis indicating involvement in the TGF-β signaling pathway (Figs. 1A, B). In neutrophils, TGF-β1 inhibits their immune killing ability, mediating the immunosuppressive function of myeloid-derived suppressor cells. GREM1 encodes Gremlin 1, a member of the bone morphogenetic protein antagonist protein family, which can directly bind to bone morphogenetic protein and block the TGF-β signaling pathway. Based on this, we hypothesize that in neutrophils of patients with CAD, upregulated GREM1 expression leads to the blockade of the TGF-β signaling pathway, enhancing neutrophil immune killing capacity and damaging the vascular endothelium. In patients with CAD with T2DM, the use of SGLT-2 inhibitors may inhibit GREM1 expression, upregulate the TGF-β signaling pathway, reduce neutrophil immune killing capacity, and exert a protective effect. To further clarify the cellular source of cardiac GREM1, we performed single-cell sequencing analysis on human heart samples, which revealed that GREM1 expression in cardiomyocytes is very low (Fig. 1C). In addition, we isolated primary cardiomyocytes and cardiac fibroblasts from rat cardiac tissue and neutrophils from rat whole bone marrow, followed by GREM1 expression analysis. As shown in Figure 1D, the most pronounced alterations in GREM1 expression were observed in neutrophils. Therefore, we focused our investigation on the role of GREM1 in neutrophils.

FIGURE 1.

Transcriptome sequencing analysis. A, Heatmap of DEGs enriched from patients with CAD with type 2 diabetes treated with or without SGLT-2 inhibitors. B, Kyoto Encyclopedia of genes and genomes pathway analysis of DEGs to identify enriched signaling pathways. C, Single-cell sequencing analysis was performed to examine Grem1 expression in human heart tissue. t-SNE dimensionality reduction revealed 6 major cell types in the heart, including cardiomyocytes, fibroblasts, smooth muscle, endothelial cells, pericytes, and T cells. Grem1 expression is represented by red dots. D, GREM1 expression in primary cardiomyocytes, fibroblast and neutrophils, n = 3. ***P < 0.001.

SGLT-2 Inhibitor Treatment Improved Circulating Neutrophil Activation Markers

Peripheral blood was collected from healthy individuals, patients with CAD combined with T2DM, and patients with CAD with T2DM treated with SGLT-2 inhibitors. Human neutrophils were isolated using an immunomagnetic negative-selection approach, and flow cytometry was performed to verify the purity of isolated neutrophils before transcriptomic and functional analyses. As shown in Figure 2A, cells were sequentially gated on total events, single cells, CD45+ leukocytes, and CD16+CD66b+ neutrophils. The isolated neutrophil fraction showed high purity, with CD16+CD66b+ neutrophils accounting for 99.2% of CD45+ cells. Compared with healthy individuals, the levels of cell-free dsDNA and elastase 2 were upregulated (Figs. 2B, C), and the expression level of GREM1 mRNA was increased (Fig. 2D). After treatment with SGLT-2 inhibitors, these indices were significantly reduced in the peripheral blood of patients.

FIGURE 2.

Flow cytometric validation of isolated human neutrophils and SGLT-2 inhibitor-associated changes in circulating neutrophil activation markers. A, Representative flow cytometry gating strategy for purity assessment of isolated human neutrophils. Cells were sequentially gated on total events, single cells, CD45+ leukocytes, and CD16+CD66b+ neutrophils. B, C, Serum levels of cell-free dsDNA and elastase 2 were measured by ELISA. D, qRT-PCR was used to detect the expression level of GREM1. ***P < 0.001.

SGLT-2 Inhibitor Therapy Affects Peripheral Blood Neutrophil Function

Next, we measured the expression of GREM1, the TGF-β/Smad signaling cascade, ICAM1, FAS, neutrophil elastase, and CitH3. The results revealed that in neutrophils from the peripheral blood of patients with CAD with T2DM, the expression levels of GREM1, TNF-α, ICAM1, FAS, IL-1β, IL-6, and neutrophil elastase as well as CitH3 were increased, suggesting enhanced neutrophil functions. Meanwhile, the expression level of phosphorylated Smad1/5/9 was significantly decreased (Figs. 3A–L). In patients treated with SGLT-2 inhibitors, the expression levels of these markers were reversed.

FIGURE 3.

SGLT-2 inhibitor therapy affects peripheral blood neutrophil function. (A–C), The protein levels of GREM1, TNF-α, ICAM1, FAS, IL-1β, IL-6, Smad1/5/9, phosphorylated Smad1/5/9, neutrophil elastase, and CitH3 were measured by western blotting. D–L, Quantification of relative protein expression from the western blot assay. n = 3, **P < 0.01, ***P < 0.001.

Neutrophils From Patients With CAD Induce Injury in Human Aortic Endothelial Cells

Results from a series of immunofluorescence double-labeling experiment revealed colocalization of GREM1 and neutrophil elastase in neutrophils (Fig. 4A). Next, neutrophils were mixed with HAEC-Luc at ratios of 2.5:1, 5:1, 10:1, and 20:1, respectively, and cultured for 24 hours. Endothelial injury was quantified by measuring luciferase released into the supernatant, as HAEC Luc cells constitutively express cytoplasmic luciferase and its appearance in the medium indicates loss of membrane integrity. Luciferase release increased in a neutrophil dose–dependent manner, paralleled by elevated H₂O₂ and TNF α and reduced NO levels (Figs. 4B–E), consistent with neutrophil-mediated endothelial cytotoxicity. Conversely, the level of NO decreased with the increasing ratio of neutrophils (Fig. 4D), suggesting that neutrophils from patients with CAD have a damaging effect on HAEC-Luc.

FIGURE 4.

Effects of patient-derived neutrophils on HAEC-Luc cells. A, Expression and cellular localization of GREM1 and neutrophil elastase in neutrophils were observed by indirect immunofluorescence double labeling. B–E, Levels of luciferase, H₂O₂, NO, and TNF-α in the cell culture supernatant were measured. n = 3, ***P < 0.001.

SGLT2 Inhibition Attenuates Neutrophil-Induced Endothelial Injury through Modulation of GREM1-Associated Signaling

When neutrophils from patients with CAD were cocultured with HAEC-Luc cells at a ratio of 20:1 and treated with SGLT-2 inhibitors or GREM1 neutralizing antibodies, both treatments were observed to inhibit neutrophil-induced damage to vascular endothelial cells (Fig. 5A). In addition, treatment with SGLT-2 inhibitors or GREM1 neutralizing antibodies reduced the expression of GREM1, TNF-α, ICAM1, FAS, neutrophil elastase, IL-1β, IL-6, and CitH3 in neutrophils, while promoting the expression of p-Smad1/5/9 (Figs. 5B–M). These data suggest that SGLT-2 inhibitors alleviate endothelial cell damage caused by neutrophils through the regulation of the GREM1/TGF-β signaling cascade.

FIGURE 5.

SGLT-2 inhibitor reduces endothelial cell damage induced by neutrophils through modulation of GREM1 signaling. A, Isolated neutrophils were cocultured with HAEC-Luc at a ratio of 20:1 for 24 hours, and cell damage was evaluated by measuring luciferase intensity. B–D, Western blot analysis was conducted to assess the expression levels of GREM1, TNFα, ICAM1, FAS, IL-1β, IL-6, Smad1/5/9, phosphorylated Smad1/5/9, neutrophil elastase, and CitH3. E–M, Quantification of relative protein expression levels based on the western blot results. n = 3, ***P < 0.001.

Empagliflozin Attenuates Diabetic Cardiac Injury and Reduces Myocardial Neutrophil Accumulation in Diabetic Rats

Next, a diabetic rat model was established to determine the in vivo effects of SGLT-2 inhibitors. As shown in Figures 6A–C, treatment with SGLT-2 inhibitors significantly reduced blood glucose levels and alleviated cardiac damage biomarkers, including serum CK and CK-MB, in diabetic rat. Furthermore, using flow cytometry and immunofluorescence colocalization, we found that SGLT2 inhibitor treatment significantly alleviated neutrophil infiltration in myocardial tissue (Figs. 6D–F). Diabetic rat exhibited elevated serum levels of elastase 2 and cell-free dsDNA (Figs. 7A, B), along with increased GREM1 mRNA expression in cardiac tissue (Fig. 7C). SGLT-2 inhibitor treatment significantly reduced all these parameters. Notably, neutrophils are the predominant cell type expressing high levels of GREM1, and SGLT-2 inhibitor markedly attenuated cardiac MPO-positive neutrophil infiltration (Fig. 7M). Thus, the reduction in total cardiac GREM1 after empagliflozin treatment is consistent with decreased neutrophil-associated inflammatory burden, although direct effects on resident cardiac cells cannot be excluded.

FIGURE 6.

Empagliflozin attenuates diabetic cardiac injury and reduces myocardial neutrophil accumulation. A, Blood glucose levels in rat. B, C, Serum concentrations of CK and CK-MB. D, Double immunofluorescence staining for the myocyte marker α-actinin (red) and the neutrophils marker RP-1 (green) in the rat heart tissues. Merged images (orange) showing colocalization; scale bar, 50 μm. E, F Flow cytometric analysis of the proportion of CD45+CD11b+HIS48+ neutrophils in rat myocardial single-cell suspensions. n = 6, ***P < 0.001.

FIGURE 7.

SGLT-2 inhibitor mitigates endothelial cell injury caused by neutrophils in diabetic rat through modulation of GREM1 signaling. A, B, Serum levels of elastase 2 and cell-free dsDNA, as measured by ELISA. C, Relative expression of GREM1 mRNA in cardiac tissue. D–I, Western blot analysis of GREM1, TNFα, ICAM1, neutrophil elastase, FAS, IL-1β, IL-6, p-Smad1/5/9, and CiH3, along with corresponding quantification. J, Hematoxylin and eosin (H&E) staining. K, L, Masson's trichrome staining and quantification of collagen deposition based on Masson staining. M, Immunofluorescence staining for MPO in cardiac tissues. n = 6, ***P < 0.001.

In addition, consistent with the in vitro results, treatment with SGLT-2 inhibitors downregulated the expression levels of GREM1, TNF-α, ICAM1, FAS, neutrophil elastase, IL-1β, IL-6, and CitH3, which had been upregulated in diabetic rat, suggesting suppression of neutrophil functions. Furthermore, the expression level of phosphorylated Smad1/5/9 was significantly elevated by SGLT-2 inhibitor treatment (Figs. 7D–I). Histological staining results showed notable cardiac tissue damage and fibrosis in diabetic rat, accompanied by elevated MPO expression, all of which were notably alleviated by SGLT-2 inhibitor treatment (Figs. 7J–M).

DISCUSSION

T2DM is a recognized independent risk factor for CAD and it significantly exacerbates the burden of atherosclerosis and increases the risk of adverse cardiovascular outcomes.^26,27 Accumulating evidence highlights that, beyond hyperglycemia, chronic systemic inflammation plays a critical role in both the development and prognosis of CAD in DM.²⁸ SGLT-2 inhibitors are well-established oral antidiabetic agents that not only effectively control glycemia in T2DM but also consistently reduce cardiovascular and renal events in large-scale randomized trials.^29–31 Notably, recent studies have demonstrated that SGLT-2 inhibitors exert cardioprotective and anti-inflammatory effects independent of glucose lowering,^32–34 suggesting immunomodulation as an important mechanism of their benefit. Empagliflozin reduced MPO-positive inflammatory cell accumulation in diabetic hearts, consistent with decreased neutrophil infiltration; however, this study was not designed to define the upstream recruitment mechanisms.

In our current study, we conducted transcriptome sequencing analysis on blood samples collected from healthy individuals, patients with CAD and T2DM, as well as those treated with SGLT-2 inhibitors. Our data revealed that the use of SGLT-2 inhibitors decreased the expression of GREM1 in neutrophils compared with the control group, and enrichment analysis indicated involvement in the TGF-β signaling pathway. TGF-β1 can inhibit the immune killing ability of neutrophils, and GREM1 encodes Gremlin 1, which interrupts TGF-β1 signaling transduction. Therefore, we hypothesized that SGLT-2 inhibitors exert protective effects in CAD with T2DM by modulating the immune cytotoxicity of neutrophils.

Several studies have emphasized the importance of innate immune cells, particularly neutrophils, in the progression and destabilization of atherosclerotic plaques, especially in the context of DM and CAD.^25,35,36 While earlier research highlighted the anti-inflammatory and antioxidant properties of SGLT-2 inhibitors, our findings extend this understanding by demonstrating that SGLT-2 inhibitors downregulate GREM1 expression and consequently modulate the TGF-β/Smad signaling pathway in neutrophils. Supporting this mechanism, Karin A. L. Müller et al³⁷ revealed that elevated plasma Gremlin-1 correlates with increased CAD instability in diabetic patients, identifying the GREM1 axis as a critical link between metabolic stress and vascular inflammation. Our data suggest that SGLT2 inhibitor treatment is associated with reduced neutrophil GREM1 expression and attenuated inflammatory activation.

Increasing studies have indicated that higher systemic immune inflammation levels at admission correlate with an increased risk of all-cause and cause-specific mortality in patients with CAD, particularly those with T2DM.^38,39 Low-grade inflammatory stress may contribute to the pathogenesis of atherosclerosis.³⁵ Neutrophils, the most abundant subtype of white blood cells in human circulation, are key mediators of acute inflammatory responses and contribute to atherosclerosis in a stage-dependent manner.³⁶ Furthermore, excessive neutrophil activation and NET formation have emerged as key drivers of vascular injury and thrombotic risk.^38–40 Our results demonstrate that SGLT-2 inhibitor therapy significantly decreases neutrophil activation markers, such as cell-free dsDNA and elastase 2, suppresses GREM1 expression, restores Smad1/5/9-related signaling, and reduces pro-inflammatory factors, collectively indicating attenuation of neutrophil-mediated inflammation and vascular injury.

Functionally, we observed that neutrophils from patients with CAD induced notable injury to human aortic endothelial cells (HAECs), an effect that was significantly ameliorated by SGLT-2 inhibitor or anti-GREM1 antibody treatment. Mechanistically, our study supports a model in which elevated GREM1 in patient-derived neutrophils is associated with suppression of Smad1/5/9-related signaling, increased neutrophil activation, thereby promoting cytotoxicity and endothelial injury. This conclusion is substantiated by the consistent inverse correlation between GREM1 expression and Smad1/5/9 phosphorylation in both human samples and animal models, and by the functional rescue achieved with a GREM1-neutralizing antibody, which recapitulated the effects of SGLT-2 inhibition on Smad phosphorylation, neutrophil activation, and endothelial damage. These findings align with the established function of GREM1 as a direct antagonist of TGF-β superfamily signaling. Nevertheless, definitive proof of causality would require genetic manipulation of GREM1 in neutrophils, such as conditional knockout or overexpression systems, which were not feasible in the current study due to technical constraints in primary human neutrophils. Future studies using neutrophil-specific GREM1-deficient mice will be important to validate the causative role of this axis in diabetic vascular complications. These findings have important implications. We have identified neutrophil GREM1 as a novel molecular target for cardiovascular protection in DM, thus providing a mechanistic rationale for the clinical cardiovascular benefits observed with SGLT-2 inhibitors in recent trials, such as EMPEROR-Preserved and DAPA-heart failure.^30,31,41 Importantly, our study extends the current understanding of SGLT-2 inhibitor immunomodulation from macrophages to neutrophils, highlighting a previously unappreciated axis of TGF-β signaling regulation.

Nonetheless, some limitations warrant consideration. The cohort size, although representative, was moderate, and validation in larger and more diverse populations is needed. The interplay of the GREM1/TGF-β pathway with other inflammatory mediators in diabetic vascular disease deserves further exploration.

In summary, our findings demonstrate that SGLT-2 inhibitors reduce neutrophil-mediated endothelial injury in CAD with T2DM by downregulating GREM1 and restoring TGF-β signaling. Notably, the reduction in cardiac GREM1 expression observed in diabetic rat primarily reflects the diminished contribution of infiltrating neutrophils, which are highly enriched for GREM1, rather than direct modulation of cardiomyocyte GREM1. Whether SGLT 2 inhibitors directly affect GREM1 transcription in cardiac resident cells remains an intriguing question for future studies. This adds a nuanced understanding of the cardiovascular benefits of SGLT-2 inhibitors and identifies the neutrophil GREM1/TGF-β axis as a promising therapeutic target in diabetic vascular complications.