Link Between GLP-1 Receptors on Immune-Regulating Cells and Narrowing of Heart Arteries and Inflammation in Heart Disease

This single-center observational study consecutively enrolled 200 patients who presented with chest pain at Dongfeng General Hospital between September 2023 and May 2024. The study protocol was approved by the Institutional Review Board of Sinopharm Dongfeng General Hospital, Hubei University of Medicine, Shiyan, China in compliance with the Declaration of Helsinki. (Approval No. LW-2024-049). Written informed consent was obtained from all participants prior to enrollment.

Patients were classified into two groups based on quantitative coronary angiography (QCA): the CHD group (n = 130), with50% luminal stenosis in1 major epicardial artery (left main, LAD, LCx, RCA), was further stratified into chronic coronary syndrome (CCS) and acute coronary syndrome (ACS) per ESC guidelines. The non-CHD group (n = 70) had50% stenosis in all major arteries and no history of coronary revascularization. The CHD group was further stratified into low-risk and high-risk subgroups based on the median Gensini score. ≥ ≥ <

The exclusion criteria were as follows: (1) metabolic disorders, such as secondary diabetes mellitus or gestational diabetes mellitus; (2) pharmacological interference, such as current or recent use of systemic hormonal therapies, including thyroid hormone replacement or glucocorticoids, which may interfere with immune-metabolic homeostasis; (3) immune dysregulation, such as a history of established autoimmune/inflammatory disorders (including but not limited to rheumatoid arthritis); (4) organ dysfunction, such as severe cardiac insufficiency (left ventricular ejection fraction (LVEF)<40%), decompensated hepatic cirrhosis (Child-Pugh C), or advanced chronic kidney disease (estimated glomerular filtration rate(eGFR) <30 mL/min/1.73 m²); (5) oncological conditions, such as active malignancy; and (6) acute systemic illnesses, such as hospitalization-requiring infections and trauma (excluding index myocardial infarction events).

All participants underwent diagnostic coronary angiography using standard techniques. The severity of coronary artery stenosis was quantified using the Gensini scoring system (based on the 2019 guidelines [25]), which accounts for both the degree of luminal narrowing and the anatomical significance of the affected vessels. The Gensini score was calculated as follows: Gensini score = lesion 1 (severity score × weighting factor) + lesion 2 (severity score × weighting factor) + ⋯ + lesion n (severity score × weighting factor). Severity scores (1, 2, 4, 8, 16, and 32) were assigned based on the relative reduction in coronary luminal diameter (25%, 25% ≤ diameter < 50%, 50% ≤ diameter < 75%, 75% ≤ diameter < 90%, 90% ≤ diameter < 99%, and ≥99%, respectively). Weighting factors were applied according to the anatomical location of the stenosis: left main (LM, ×5); proximal (×2.5), mid (×1.5), and distal (×1) segments of the left anterior descending artery (LAD); first diagonal branch (×1) and second diagonal branch (×0.5); proximal (×2.5), mid (×1), and distal (×1) segments of the left circumflex artery (LCX); proximal (×1), mid (×1), and distal (×1) segments of the right coronary artery (RCA); obtuse marginal branch (×1); posterior descending artery (×1); and posterolateral branch (×0.5). Based on the median Gensini score, the coronary heart disease (CHD) group was further stratified into low-risk and high-risk subgroups.

Standardized Blood Collection: Peripheral venous blood (7 mL) was collected from all participants immediately prior to coronary angiography via sterile venipuncture of the radial vein after an overnight fast (12 hours). Blood samples were drawn into pre-chilled EDTA vacutainers and maintained at 4 °C throughout processing to preserve cellular integrity and cytokine stability. ≥

Sample Processing Workflow: 2 mL of whole blood was centrifuged at 400g for 10 min (room temperature) within 4 hours post-collection. Plasma aliquots were stored at –80 °C until cytokine profiling (avoiding freeze-thaw cycles). The remaining 5 mL of blood was layered onto human lymphocyte separation medium (Solarbio, #P8610) and centrifuged at 800g for 30 min (18–22 °C). The buffy coat layer (1–2 mm thick, opalescent) was aseptically aspirated using sterile polypropylene transfer pipettes (Thermo Fisher) and transferred into pre-labeled 5 mL flow cytometry tubes. Peripheral blood mononuclear cells (PBMCs) were washed twice at 300g for 10 min at 4 °C with phosphate buffered saline (PBS) (Beyotime). The final cell pellets were resuspended in PBS and immediately subjected to flow cytometry analysis (2 hours post-isolation) to preserve surface epitope integrity. × × ×

Freshly isolated PBMCs were stained with the following antibody cocktail for surface markers (30 min, room temperature): PerCP-conjugated anti-human CD4 (Clone RPA-T4, BioLegend, #300528), PE-conjugated anti-human CD25 (Clone BC96, BioLegend, #302606), and Alexa Fluor 488-conjugated anti-human GLP-1R (Clone 358903, R&D Systems, #FAB6292G). Intracellular FoxP3-stained cells were fixed/permeabilized using the FoxP3/Transcription Factor Staining Buffer Set (BioLegend, #420801) and stained with Alexa Fluor 647-conjugated anti-FoxP3 (Clone 236A/E7, BioLegend, #320214) for 45 min at room temperature. Cytokine levels interleukin-2 (IL-2), interleukin-4 (IL-4), interleukin-6 (IL-6), interleukin-10 (IL-10), interleukin-35 (IL-35), and tumor necrosis factor-alpha (TNF-)) were analyzed using a flow cytometric multiplex cytokine kit (Bio-Techne, #IS135; 96 tests). α

Using NovoExpress software (version1.3.4, Santa, Clara, CA, USA), cells or cytokines are automatically collected and analyzed. Treg cells are defined as CD4⁺CD25⁺Foxp3⁺ T cells. Lymphocyte populations (lymphocytes) are gated based on FSC-SSC, followed by gating for CD4⁺ T cells. From the CD4⁺ T-cell population, CD25⁺Foxp3⁺ Treg cells were identified, and the number of CD4⁺CD25⁺Foxp3⁺ Treg cells relative to that of CD4⁺ T cells was determined. From the CD4⁺CD25⁺Foxp3⁺ Treg cell population, CD4⁺CD25⁺Foxp3⁺ Treg cells expressing GLP-1R were gated, and their percentage was calculated (see Supplementary Fig. 1). For the detection of Treg cells and their GLP-1R expression levels, each sample was collected until 10,000 Treg cells or 500 µL of PBMC suspension was obtained to ensure sufficient statistical power for rare population analysis. The termination condition for collecting cytokines was set at capturing 150 microspheres or reaching a sample volume of 100 µL.

For all variables, normality assumptions were checked with the Shapiro-Wilk test. Normally distributed continuous variables are expressed as the means ± standard deviations (SDs); between-group differences are analyzed via the Student's t test (two groups); nonnormally distributed continuous variables are reported as the medians (interquartile ranges, IQRs) and analyzed using the Mann-Whitney U test (two groups); categorical variables are described as frequencies (percentages) and compared via Pearson's χ² test or Fisher's exact test (expected cell counts <5). Univariate and multivariate binary logistic regression analyses were used to evaluate odds ratios (ORs) with 95% confidence intervals (95% CIs). The predictive performance of relevant indicators was assessed using receiver operating characteristic (ROC) curves. Spearman correlation analysis was used to examine the relationship between GLP-1R expression levels and cytokines. All statistical analyses were performed using SPSS software (version 23.0, Armonk, NY, USA), with p < 0.05 considered statistically significant.

Demographic and clinical characteristics of the study cohorts are presented in Table 1. Compared with controls, patients with CHD presented a significantly greater prevalence of hypertension (CHD vs. non-CHD: 59.23% vs. 38.57%, χ² = 7.780, p = 0.005) and smoking history (43.08% vs. 27.14%, χ² = 4.929, p = 0.026). The CHD group was older [median age: 64 (IQR: 57–70) vs. 59 (IQR: 53–65) years, Z = –3.037, p = 0.002] and had a larger waist circumference (93.34 ± 9.77 cm vs. 90.13 ± 10.03 cm, t = –2.196, p = 0.029). In contrast, no significant differences were observed in sex distribution (male: 66.92% vs. 57.14%, p = 0.171), alcohol consumption (22.31% vs. 20.00%, p = 0.705), or lipid profiles, including total cholesterol (4.52 ± 0.96 vs. 4.47 ± 1.12 mmol/L, p = 0.745), low-density lipoprotein cholesterol (LDL-C) (2.62 ± 0.90 vs. 2.48 ± 0.89 mmol/L, p = 0.267), high-density lipoprotein cholesterol (HDL-C) (1.12 ± 0.28 vs. 1.16 ± 0.37 mmol/L, p = 0.402), and triglycerides [median: 1.62 (IQR: 1.20–2.19) vs. 1.72 (IQR: 1.01–2.21) mmol/L, p = 0.438]. Body mass index (BMI) also showed no group differences (24.94 ± 3.53 vs. 24.47 ± 3.45 kg/m², p = 0.363). Notably, high-sensitivity c-reactive protein (hs-CRP) levels tended to be higher in the CHD group than in the control group but did not reach statistical significance [median: 1.33 (IQR: 0.62–3.71) vs. 0.97 (0.53–2.49) mg/L, p = 0.218]. These results underscore hypertension, smoking, advanced age, and central adiposity as key distinguishing factors in CHD patients, whereas traditional lipid markers and systemic inflammation showed limited discriminative capacity in this cohort.

Comparative immunophenotypic analysis demonstrated significant alterations in Treg cell dynamics among CHD patients. The frequency of circulating regulatory T cells (Tregs, CD4⁺CD25⁺FoxP3⁺) was substantially reduced in the CHD cohort compared with that in the non-CHD control group (Fig. 1A). In the non-CHD group, the proportion of Treg cells was 5.93%, whereas in the CHD group, it was 3.50%. The difference was highly statistically significant (Fig. 1B). This depletion was further accentuated in CHD patients with high-risk coronary stenosis (Gensini score ≥ median), who exhibited a significant reduction relative to controls (p < 0.001).

Notably, the immunoregulatory defect extended to the subset of Tregs expressing glucagon-like peptide-1 receptor (GLP-1R⁺ Tregs). In the non-CHD group, the proportion of GLP-1R⁺ Treg double-positive cells was 5.67% compared with 3.15% in the CHD group (Fig. 1C). The proportion of GLP-1R⁺ Tregs was significantly greater in the non-CHD group than in the CHD group (Fig. 1D). Stratification by stenosis severity revealed a progressive decline in GLP-1R⁺ Treg representation, with high-risk patients displaying significantly lower levels than their low-risk counterparts (p < 0.05).

These findings collectively suggest a dual defect in both Treg abundance and functional GLP-1R-mediated immunomodulatory capacity, potentially contributing to sustained inflammatory activation in coronary atherosclerosis.

Univariate logistic regression revealed a significant inverse correlation between Treg frequency and CHD risk (OR = 0.749, 95% CI: 0.657–0.854; p < 0.001). This association persisted after multivariate adjustment for sex, smoking history, and lipid profiles (adjusted OR = 0.752, 95% CI: 0.645–0.877, p < 0.001; Supplementary Table 1), indicating that Treg depletion is an independent protective factor against CHD. ROC curve analysis demonstrated a certain degree of discriminative capacity of Treg frequency for CHD risk, with an AUC of 0.663 (95% CI: 0.581–0.746) (Supplementary Fig. 2A). A restricted cubic spline (RCS) model further delineated a nonlinear dose-response relationship, wherein CHD risk decreased progressively with increasing Treg frequency (Supplementary Fig. 2B).

Univariate logistic regression revealed that a reduced GLP-1R⁺ Treg frequency (% of total Tregs) was significantly associated with elevated CHD risk (OR = 0.852, 95% CI: 0.765–0.948, p = 0.003). Additional risk factors included smoking history (OR = 2.031, 95% CI: 1.081–3.817, p = 0.028), hypertension (OR = 2.314, 95% CI: 1.276–4.195, p = 0.006), advanced age (OR = 1.050, 95% CI: 1.017–1.084, p = 0.003), and increased waist circumference (OR = 1.035, 95% CI: 1.003–1.069, p = 0.031). After adjusting for sex, lipid profiles, and the aforementioned covariates, the inverse correlation between GLP-1R⁺ Treg frequency and CHD risk remained robust (adjusted OR = 0.859, 95% CI: 0.762–0.969, p = 0.013) (Table 2), confirming its role as an independent protective factor against coronary atherosclerosis.

ROC curve analysis demonstrated that GLP-1R⁺ Treg frequency (% of total Tregs) exhibited modest discriminative capacity for differentiating CHD patients from non-CHD controls, with an AUC of 0.600 (95% CI: 0.514–0.685, p = 0.032) (Fig. 2A). While the AUC value suggests limited standalone diagnostic utility, its statistical significance (p < 0.05) implies potential relevance in composite risk models. Furthermore, the RCS model revealed a nonlinear inverse relationship between GLP-1R⁺ Treg frequency and CHD risk (Fig. 2B). Elevated GLP-1R+ Treg levels were associated with a reduction in CHD risk, reinforcing the protective role of GLP-1R signaling in Treg-mediated immunomodulation.

Patients with CHD were stratified into low-risk (n = 68) and high-risk (n = 62) subgroups based on the median Gensini score. Baseline characteristics of both groups are presented in Table 3. No statistically significant differences were observed between the low-risk and high-risk groups across all evaluated parameters (all p > 0.05). Specifically, demographic and clinical factors such as male sex (61.76% vs. 72.58%; p = 0.190), smoking history (39.71% vs. 46.77%; p = 0.416), and prevalence of hypertension (61.76% vs. 56.45%; p = 0.538) showed comparable distributions. Similarly, metabolic markers, including lipid profiles (total cholesterol: 4.25 vs. 4.54 mmol/L, p = 0.366; LDL-C: 2.52 vs. 2.73 mmol/L, p = 0.188) and inflammatory indicators (hs-CRP: 1.13 vs. 1.51 mg/L, p = 0.185), did not differ significantly between groups. Although alcohol consumption (16.18% vs. 29.03%; p = 0.079) and waist circumference (p = 0.073) demonstrated trends toward higher values in the high-risk group, these differences did not reach statistical significance. These findings suggest that baseline characteristics were well balanced between the two subgroups, supporting the validity of risk stratification based on the Gensini score in this cohort.

Flow cytometry was used to analyze the proportions of regulatory T cells (Tregs) and GLP-1R⁺ Tregs in the low-risk and high-risk CHD groups. As shown in Fig. 3A, the proportion of Treg cells, determined by the coexpression of CD25 and Foxp3, was 4.47% in the low-risk CHD group and 3.32% in the high-risk CHD group. As shown in Fig. 3C, based on the expression of SSC and GLP-1R, the proportion of GLP-1R⁺ Treg cells was 5.41% in the low-risk CHD group and 1.83% in the high-risk CHD group. Statistical analysis was conducted to evaluate the differences in the proportions of Treg and GLP-1R⁺ Treg cells between the two groups. The results revealed a significant difference in the proportion of Treg cells between the low-risk CHD group and high-risk CHD group (p < 0.05; Fig. 3B), and the proportion of GLP-1R⁺ Treg cells also exhibited a significant inter-group difference (p < 0.05; Fig. 3D). These results indicate that obvious differences exist in the immune cell populations related to regulatory T cells between the low-risk CHD group and high-risk CHD group, which may have an impact on the pathophysiology of CHD.

Univariate analysis identified that Treg cell depletion was a significant predictor of advanced stenosis (OR: 0.762, p = 0.009). Multivariate adjustment for demographic and metabolic confounders preserved this association (adjusted OR: 0.760, p = 0.021) (Supplementary Table 2), highlighting the independent protective role of Treg cells. ROC analysis supported their discriminatory capacity (AUC: 0.635, 95% CI: 0.539–0.731) (Supplementary Fig. 3A), whereas RCS modeling revealed a nonlinear inverse relationship between Treg proportions and stenosis severity (Supplementary Fig. 3B), underscoring their therapeutic potential in halting atherosclerosis.

Univariate logistic regression revealed a protective role of GLP-1R⁺ Treg cells against coronary stenosis (OR: 0.840, p = 0.023). Multivariate adjustment for demographic and clinical confounders preserved this association (adjusted OR: 0.840, p = 0.040), underscoring GLP-1R signaling as a potential modulator of plaque stability. For detailed regression coefficients, refer to Table 4. ROC analysis demonstrated that the proportion of GLP-1R⁺ Tregs moderately predicted stenosis severity (AUC: 0.619, 95% CI: 0.521–0.717; Fig. 4A). RCS modeling further revealed a dose-dependent protective effect: each incremental increase in GLP-1R⁺ Treg proportions was correlated with attenuated stenosis severity (Fig. 4B), implicating GLP-1R signaling as a therapeutic target to decelerate coronary atherosclerosis.

To investigate the regulatory role of GLP-1R⁺ Treg cells in cytokine networks, correlation analyses were performed between their proportions and key cytokine levels in the study cohort. A statistically significant negative correlation was observed between GLP-1R⁺ Treg cells and interleukin-4 (IL-4) levels (r = –0.150, p = 0.047; Fig. 5B), whereas a significant positive correlation was detected with interleukin-35 (IL-35) levels (r = 0.185, p = 0.016; Fig. 5E). No significant associations were found with other cytokines, including interleukin-2 (IL-2; r = –0.003, p = 0.967; Fig. 5A), interleukin-6 (IL-6; r = 0.023, p = 0.759; Fig. 5C), interleukin-10 (IL-10; r = 0.030, p = 0.695; Fig. 5D), or tumor necrosis factor-alpha (TNF-α; r = 0.118, p = 0.116; Fig. 5F). These results indicate that GLP-1R⁺ Treg cells selectively modulate IL-4 and IL-35, potentially suppressing IL-4 while enhancing IL-35 without broad effects on other cytokines. This differential regulation may contribute to immune homeostasis in CHD, underscoring the nuanced role of GLP-1R signaling in Treg-mediated immune responses.