Liraglutide reprograms vascular smooth muscle cell metabolism to suppress extracellular matrix remodeling in diabetic atherosclerosis

ABSTRACT

Background

Metabolic reprogramming contributes to vascular dysfunction in diabetic atherosclerosis, but the mechanisms linking hyperglycemia‐induced metabolic alterations to extracellular matrix remodeling in vascular smooth muscle cells remain incompletely understood. This study investigated whether liraglutide modulates vascular smooth muscle cell metabolism and plaque remodeling under diabetic conditions.

Methods

Primary vascular smooth muscle cells were exposed to normal glucose, high glucose, or high glucose plus liraglutide. Cellular bioenergetics, mitochondrial function, oxidative stress, extracellular matrix remodeling, and AMPK/PGC‐1α, mTOR, and HIF‐1α signaling were assessed. In vivo, diabetic ApoE^−/− mice were treated with liraglutide for 12 weeks, followed by evaluation of metabolic parameters, aortic root plaque burden, lipid deposition, collagen content, and plaque‐associated signaling markers.

Results

High glucose impaired mitochondrial respiration, enhanced glycolysis, reduced mitochondrial membrane potential, increased mitochondrial reactive oxygen species, and promoted mitochondrial fragmentation and extracellular matrix remodeling in vascular smooth muscle cells. Liraglutide restored mitochondrial function, activated AMPK/PGC‐1α signaling, suppressed mTOR activation and HIF‐1α accumulation, reduced collagen I, MMP‐2, and MMP‐9 expression, and partially restored elastin levels. In diabetic ApoE^−/− mice, liraglutide improved systemic metabolic parameters, reduced atherosclerotic plaque burden and lipid accumulation, increased plaque collagen content, restored plaque p‐AMPK expression, and reduced HIF‐1α and MMP‐9 expression.

Conclusions

Liraglutide attenuates hyperglycemia‐induced metabolic reprogramming and extracellular matrix remodeling in vascular smooth muscle cells and improves plaque stability in diabetic atherosclerosis. These effects are associated with restoration of AMPK/PGC‐1α signaling, inhibition of mTOR activation, suppression of HIF‐1α accumulation, and improved mitochondrial homeostasis.

Liraglutide reverses hyperglycemia‐induced metabolic reprogramming in vascular smooth muscle cells, restoring mitochondrial function and AMPK/PGC‐1α signaling while suppressing mTOR activation and HIF‐1α accumulation. These changes attenuate extracellular matrix remodeling, reduce atherosclerotic plaque burden, and improve plaque stability in diabetic mice.

INTRODUCTION

Cardiovascular disease remains the leading cause of morbidity and mortality among patients with diabetes mellitus. Accelerated atherosclerosis is a hallmark of diabetic vascular complications and is characterized by enhanced plaque formation, vascular remodeling, and increased risk of plaque rupture. Among the cellular components of atherosclerotic plaques, vascular smooth muscle cells (VSMCs) play a crucial role in maintaining vascular integrity and extracellular matrix (ECM) homeostasis. Under diabetic conditions, however, VSMCs undergo phenotypic and metabolic alterations that contribute to pathological vascular remodeling and plaque instability ¹ , ² . Understanding the molecular mechanisms linking hyperglycemia to VSMC dysfunction is therefore essential for developing effective therapeutic strategies for diabetic atherosclerosis.

Emerging evidence suggests that metabolic reprogramming is a key driver of vascular dysfunction in metabolic diseases. In response to hyperglycemic stress, vascular cells exhibit profound alterations in cellular metabolism, including impaired mitochondrial oxidative phosphorylation and increased reliance on glycolysis ³ . This metabolic shift is often accompanied by mitochondrial dysfunction, excessive production of reactive oxygen species, and activation of stress‐responsive signaling pathways. Such metabolic disturbances can profoundly influence vascular cell behavior, including proliferation, inflammation, and ECM remodeling. In particular, altered metabolism in VSMCs has been implicated in excessive collagen deposition, elastin degradation, and increased activity of matrix metalloproteinases (MMPs), all of which contribute to plaque progression and instability.

Glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs) are widely used in the treatment of type 2 diabetes and have demonstrated substantial cardiovascular benefits in several large clinical trials ⁴ , ⁵ . Beyond their glucose‐lowering effects, GLP‐1RAs have been shown to exert pleiotropic actions on the cardiovascular system, including anti‐inflammatory, antioxidative, and endothelial‐protective effects ⁶ . Experimental evidence further indicates that liraglutide modulates vascular endothelial cells, monocytes/macrophages, and VSMCs in a context‐dependent manner, involving both AMPK‐dependent and AMPK‐independent mechanisms ⁷ , ⁸ . However, the mechanisms by which GLP‐1RAs regulate vascular cell metabolism and ECM remodeling remain incompletely understood. In particular, whether liraglutide can modulate metabolic reprogramming in VSMCs under hyperglycemic conditions and thereby influence plaque stability has not been fully elucidated.

AMP‐activated protein kinase (AMPK) is a central metabolic regulator that coordinates cellular energy balance and mitochondrial function. Activation of AMPK promotes mitochondrial biogenesis through peroxisome proliferator‐activated receptor γ coactivator‐1α (PGC‐1α) and can counteract metabolic stress ⁹ . AMPK also negatively regulates mechanistic target of rapamycin (mTOR), a nutrient‐sensitive kinase involved in cell growth, senescence‐associated signaling, and metabolic adaptation. Conversely, hypoxia‐inducible factor‐1α (HIF‐1α) is a key regulator of glycolytic metabolism and is often activated under metabolic or hypoxic conditions, promoting glycolysis and metabolic adaptation. Dysregulation of the AMPK/PGC‐1α/HIF‐1α axis has been implicated in metabolic diseases and vascular remodeling, yet its role in diabetic atherosclerosis remains incompletely defined.

In the present study, we investigated whether liraglutide treatment modulates metabolic reprogramming in VSMCs under hyperglycemic stress and how this process influences ECM remodeling and plaque stability. Using in vitro metabolic assays and in vivo models of diabetic atherosclerosis, we demonstrate that liraglutide restores mitochondrial function, activates AMPK/PGC‐1α signaling, suppresses HIF‐1α accumulation, and attenuates ECM remodeling in VSMCs. We further examined mTOR activation as a downstream component of the AMPK‐related metabolic signaling response. Furthermore, liraglutide reduces plaque burden and improves plaque stability in diabetic mice. These findings provide mechanistic insight into the vascular protective effects of liraglutide and identify metabolic reprogramming as a potential therapeutic target in diabetic atherosclerosis.

MATERIALS AND METHODS

Cell culture and treatments

Primary VSMCs were obtained from ScienCell Research Laboratories (USA) and cultured in Dulbecco's modified Eagle medium (DMEM; Gibco, Thermo Fisher Scientific, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Thermo Fisher Scientific, USA) and 1% penicillin/streptomycin (Gibco, Thermo Fisher Scientific, USA). Cells were maintained at 37°C in a humidified incubator containing 5% CO₂ and used between passages 3–6 for all experiments. To mimic diabetic conditions, cells were exposed to high glucose (HG, 25 mM D‐glucose; Sigma‐Aldrich, USA), while control cells were cultured in normal glucose (NG, 5.5 mM). For treatment experiments, VSMCs cultured under HG conditions were treated with liraglutide (Novo Nordisk, Denmark) at a final concentration of 100 nM for 24 h ¹⁰ .

Seahorse metabolic analysis

Mitochondrial respiration and glycolytic activity were measured using a Seahorse XF Analyzer (Agilent Technologies, USA). VSMCs were seeded in Seahorse XF cell culture microplates and allowed to attach overnight. Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured according to the manufacturer's instructions ¹¹ . For the mitochondrial stress test, oligomycin (1 μM), FCCP (1 μM), and rotenone/antimycin A (0.5 μM each) (Agilent Technologies, USA) were sequentially injected to determine mitochondrial respiratory parameters including basal respiration and maximal respiration. For glycolysis stress testing, glucose (10 mM), oligomycin (1 μM), and 2‐deoxyglucose (2‐DG, 50 mM; Sigma‐Aldrich, USA) were sequentially added to evaluate glycolytic activity.

Measurement of lactate production and ATP levels

Lactate production was determined using a lactate assay kit (BioVision, USA) according to the manufacturer's protocol. Culture medium was collected and incubated with the reaction mixture, and absorbance was measured using a microplate reader (BioTek Instruments, USA). Cellular ATP levels were measured using an ATP determination kit (Beyotime Biotechnology, China). Cells were lysed and ATP‐dependent luminescence was detected using a luminometer. Lactate and ATP levels were normalized to total protein concentration determined using a BCA protein assay kit (Thermo Fisher Scientific, USA).

Mitochondrial function assays

Mitochondrial membrane potential was evaluated using a JC‐1 mitochondrial membrane potential assay kit (Beyotime Biotechnology, China). VSMCs were incubated with JC‐1 dye for 20 min at 37°C and fluorescence images were obtained using a fluorescence microscope. The ratio of red to green fluorescence was calculated to assess mitochondrial membrane potential. Mitochondrial reactive oxygen species (mtROS) were detected using MitoSOX Red (Invitrogen, Thermo Fisher Scientific, USA). Cells were incubated with MitoSOX reagent for 10 min at 37°C, and fluorescence intensity was analyzed by fluorescence microscopy. Mitochondrial morphology and network integrity were visualized using MitoTracker Green FM (Invitrogen, Thermo Fisher Scientific, USA). Images were captured using a fluorescence microscope (Olympus, Japan), and mitochondrial length was quantified using ImageJ software (NIH, USA).

Immunofluorescence staining

For immunofluorescence analysis, VSMCs were fixed with 4% paraformaldehyde (Sigma‐Aldrich, USA) for 15 min and permeabilized with 0.1% Triton X‐100 (Sigma‐Aldrich, USA). After blocking with 5% bovine serum albumin (BSA; Sigma‐Aldrich, USA), cells were incubated overnight at 4°C with primary antibodies against collagen I (Abcam, UK), elastin (Abcam, UK), or HIF‐1α (Cell Signaling Technology, USA). Cells were then incubated with Alexa Fluor/conjugated secondary antibodies (Invitrogen, Thermo Fisher Scientific, USA) for 1 h at room temperature. Nuclei were counterstained with DAPI (Beyotime Biotechnology, China). Fluorescence images were obtained using a fluorescence microscope (Olympus, Japan), and fluorescence intensity was quantified using ImageJ software.

Western blot analysis

Total proteins were extracted from VSMCs using RIPA lysis buffer (Beyotime Biotechnology, China) supplemented with protease and phosphatase inhibitor cocktails (Roche, Switzerland). Equal amounts of protein were separated by SDS/PAGE and transferred onto PVDF membranes (Millipore, USA). Membranes were blocked with 5% nonfat milk and incubated overnight at 4°C with primary antibodies against p‐AMPK, AMPK, PGC‐1α, and HIF‐1α (Cell Signaling Technology, USA), collagen I, MMP‐2, and MMP‐9 (Abcam, UK), p‐mTOR and total mTOR (Cell Signaling Technology, USA), and β‐actin (Proteintech, USA). After incubation with HRP‐conjugated secondary antibodies (Cell Signaling Technology, USA), protein bands were visualized using enhanced chemiluminescence reagents (Thermo Fisher Scientific, USA). Band intensities were quantified using ImageJ software and normalized to β‐actin.

Animal model and treatment

All animal experiments were approved by the Institutional Animal Care and Use Committee and performed in accordance with institutional guidelines. Male ApoE^−/− mice (8 weeks old; Jackson Laboratory, USA) were used to establish a model of diabetic atherosclerosis. After 1 week of acclimatization, mice were fasted for 6 h and diabetes was induced by intraperitoneal injection of freshly prepared streptozotocin (STZ; Sigma‐Aldrich, USA) at 50 mg/kg/day for five consecutive days. STZ was dissolved immediately before use in sterile 0.1 mol/L citrate buffer (pH 4.5) and protected from light. Control mice received an equal volume of citrate buffer. Diabetes was confirmed when fasting blood glucose was ≥16.7 mmol/L on two consecutive measurements.

After induction of diabetes, mice were maintained on a high‐fat diet. Animals were randomly divided into three groups: control, diabetic, and diabetic treated with liraglutide (Diabetic + Lira). The atherogenic diet contained 21% fat and 0.15% cholesterol and was continued for 12 weeks. Liraglutide (Novo Nordisk, Denmark) was administered daily by subcutaneous injection at 0.3 mg/kg/day for 12 weeks beginning after confirmation of diabetes. Control and diabetic mice received an equal volume of sterile saline. Body weight and fasting blood glucose were monitored weekly. Systolic and diastolic blood pressure were measured by a tail‐cuff system after acclimatization. At sacrifice, blood was collected for plasma lipid profiling and renal function assays, including total cholesterol, triglycerides, LDL‐cholesterol, HDL‐cholesterol, blood urea nitrogen, and serum creatinine.

Histological staining and plaque analysis

Aortic roots were harvested, fixed in 4% paraformaldehyde (Sigma‐Aldrich, USA), embedded in paraffin, and sectioned at 5 μm thickness. Serial cross‐sections of the aortic root were obtained beginning at the level of the aortic valve cusps, and consecutive sections from the same anatomical region were used for histological analyses. Serial sections were stained with hematoxylin and eosin (H&E; Sigma‐Aldrich, USA) to evaluate plaque morphology. Lipid deposition was assessed using Oil Red O staining (Sigma‐Aldrich, USA). Collagen content within plaques was evaluated using a Masson's trichrome staining kit (Sigma‐Aldrich, USA). Histological images were captured using a light microscope (Olympus, Japan). Low‐magnification images were acquired to visualize the overall aortic root lesion burden and vessel wall morphology. Plaque area, lipid‐positive area, and collagen‐positive area were quantified using ImageJ software (NIH, USA). All histological quantification was performed by investigators blinded to group allocation.

Immunohistochemistry

For immunohistochemical staining, paraffin sections were deparaffinized, rehydrated, and subjected to antigen retrieval using citrate buffer. Sections were blocked with BSA and incubated overnight at 4°C with primary antibodies against p‐AMPK (Cell Signaling Technology, USA), HIF‐1α (Cell Signaling Technology, USA), and MMP‐9 (Abcam, UK). After incubation with HRP‐conjugated secondary antibodies (Cell Signaling Technology, USA), signals were visualized using diaminobenzidine (DAB) substrate (Vector Laboratories, USA) and counterstained with hematoxylin. Staining intensity was quantified as mean optical density (MOD) using ImageJ software.

Statistical analysis

All data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism software (GraphPad Software, USA). Comparisons among multiple groups were conducted using one‐way analysis of variance (anova) followed by Tukey's post hoc test. For physiological and biochemical parameters, the same statistical workflow was applied across groups. A value of P < 0.05 was considered statistically significant.

RESULTS

Hyperglycemia induces metabolic reprogramming in VSMCs

To determine whether hyperglycemia alters metabolic profiles in VSMCs, we evaluated mitochondrial respiration and glycolytic activity using Seahorse metabolic analysis. ECAR was significantly increased under HG conditions, indicating enhanced glycolytic activity (Figure 1a). Quantitative analysis revealed significant decreases in both basal respiration and maximal respiratory capacity in HG‐treated VSMCs (Figure 1b), indicating impaired mitochondrial oxidative phosphorylation. Real‐time OCR tracing further confirmed suppressed mitochondrial respiration under HG conditions (Figure 1c). Consistent with enhanced glycolysis, lactate production was markedly elevated in HG‐treated cells (Figure 1d), while intracellular ATP levels were significantly reduced (Figure 1e). Scatter plot analysis of OCR and ECAR further demonstrated a metabolic shift from oxidative phosphorylation toward glycolysis under hyperglycemic conditions (Figure 1f).

Figure 1.

Hyperglycemia induces metabolic reprogramming in vascular smooth muscle cells. (a) Extracellular acidification rate (ECAR) measured in VSMCs under NG, HG, or HG + Lira conditions following sequential injection of glucose, oligomycin, and 2‐deoxyglucose (2‐DG). (b) Quantification of basal respiration and maximal respiration derived from the OCR assay. (c) Real‐time measurement of oxygen consumption rate (OCR) in VSMCs cultured under normal glucose (NG), high glucose (HG), or high glucose with liraglutide treatment (HG + Lira) using a Seahorse XF analyzer. Oligomycin, FCCP, and rotenone/antimycin A (Rot/AA) were sequentially injected at the indicated time points to evaluate mitochondrial respiratory function. (d) Lactate production in VSMCs cultured under the indicated conditions. (e) Cellular ATP levels measured in VSMCs exposed to NG, HG, or HG + Lira. (f) Scatter plot showing the relationship between OCR and ECAR in individual samples, illustrating the metabolic shift from oxidative phosphorylation toward glycolysis under hyperglycemic conditions. Data are presented as mean ± SD from six independent experiments. ***P < 0.001 vs NG or HG as indicated by brackets.

Liraglutide restores mitochondrial function and reduces oxidative stress in VSMCs

We next investigated whether the GLP‐1RA liraglutide could restore mitochondrial function in VSMCs exposed to hyperglycemic stress. JC‐1 staining demonstrated a significant reduction in mitochondrial membrane potential in HG‐treated cells, as evidenced by a decreased red/green fluorescence ratio (Figure 2a,b). Liraglutide treatment partially restored mitochondrial membrane potential. Mitochondrial oxidative stress was assessed using MitoSOX staining. HG exposure markedly increased mtROS production compared with NG controls (Figure 2c,d). Liraglutide significantly attenuated mtROS accumulation. In addition, MitoTracker staining revealed profound alterations in mitochondrial morphology under hyperglycemic conditions, characterized by fragmented mitochondrial networks (Figure 2e). Quantitative analysis showed a significant reduction in average mitochondrial length in HG‐treated cells (Figure 2f). Liraglutide treatment partially restored mitochondrial network integrity.

Figure 2.

Liraglutide restores mitochondrial function and reduces oxidative stress in VSMCs under hyperglycemic conditions. (a) Representative JC‐1 staining images showing mitochondrial membrane potential in VSMCs cultured under NG, HG, or HG + Lira. Green fluorescence indicates JC‐1 monomers, while red fluorescence represents JC‐1 aggregates. Merged images are shown below. Scale bar, 20 μm. (b) Quantification of mitochondrial membrane potential expressed as the JC‐1 red/green fluorescence ratio. (c) Representative fluorescence images of mtROS detected using MitoSOX staining in VSMCs under the indicated conditions. Scale bar, 20 μm. (d) Quantification of MitoSOX mean fluorescence intensity. (e) Representative MitoTracker staining showing mitochondrial morphology and network structure. Scale bar, 20 μm. (f) Quantification of average mitochondrial length. Data are presented as mean ± SD from six independent experiments. ***P < 0.001 between indicated groups.

Liraglutide activates the AMPK/PGC‐1α signaling axis and suppresses HIF‐1α accumulation

To explore the molecular mechanisms underlying liraglutide‐mediated metabolic regulation, we examined key signaling pathways involved in mitochondrial biogenesis and metabolic adaptation. Western blot analysis showed that phosphorylation of AMPK was significantly reduced in HG‐treated VSMCs compared with NG controls (Figure 3a,b). Liraglutide treatment restored AMPK activation. Consistent with reduced AMPK signaling, expression of the mitochondrial biogenesis regulator PGC‐1α was markedly decreased under HG conditions, whereas liraglutide significantly increased PGC‐1α levels (Figure 3a,c). In contrast, HIF‐1α expression was substantially elevated in HG‐treated cells (Figure 3a,d). Immunofluorescence analysis further demonstrated increased nuclear accumulation of HIF‐1α under hyperglycemic conditions (Figure 3e,f). Liraglutide treatment significantly suppressed HIF‐1α expression and nuclear localization. Because AMPK is an upstream negative regulator of mTOR signaling, we further examined mTOR activation in VSMCs. High glucose markedly increased the p‐mTOR/mTOR ratio, whereas liraglutide significantly reduced p‐mTOR/mTOR under HG conditions (Figure S1A,B). Total mTOR expression normalized to β‐actin was not significantly altered among NG, HG, and HG + Lira groups (Figure S1A,C). These data indicate that liraglutide primarily suppresses high‐glucose‐induced mTOR activation rather than total mTOR abundance.

Figure 3.

Liraglutide activates the AMP‐activated protein kinase (AMPK)/peroxisome proliferator‐activated receptor γ coactivator‐1α (PGC‐1α) signaling axis and suppresses hypoxia‐inducible factor‐1α (HIF‐1α) accumulation in VSMCs under hyperglycemic stress. (a) Representative Western blot analysis of p‐AMPK, total AMPK, PGC‐1α, and HIF‐1α in VSMCs cultured under NG, HG, or HG + Lira. β‐Actin was used as a loading control. (b) Quantification of p‐AMPK normalized to total AMPK. (c) Quantification of PGC‐1α normalized to β‐Actin. (d) Quantification of HIF‐1α normalized to β‐Actin. (e) Representative immunofluorescence images showing HIF‐1α localization (green) and DAPI‐stained nuclei (blue). Scale bar, 20 μm. (f) Quantification of nuclear HIF‐1α fluorescence intensity. Data are presented as mean ± SD from six independent experiments. ***P < 0.001 between indicated groups.

Liraglutide attenuates extracellular matrix remodeling in VSMCs

Because metabolic reprogramming is closely linked to vascular remodeling, we next evaluated ECM alterations in VSMCs. Immunofluorescence analysis revealed that collagen I expression was markedly increased under HG conditions compared with NG controls (Figure 4a,b). Liraglutide significantly reduced collagen I accumulation. Conversely, elastin expression was markedly decreased in HG‐treated VSMCs (Figure 4c,d). Treatment with liraglutide partially restored elastin levels. Western blot analysis further demonstrated that HG exposure significantly increased expression of MMP‐2 and MMP‐9 (Figure 4e,f), key enzymes involved in ECM degradation and plaque instability. Liraglutide treatment significantly reduced the expression of both MMP‐2 and MMP‐9.

Figure 4.

Liraglutide attenuates extracellular matrix remodeling in VSMCs under hyperglycemic conditions. (a) Representative immunofluorescence images of collagen I expression (red) in VSMCs cultured under NG, HG, or HG + Lira. Nuclei were stained with DAPI (blue). Scale bar, 20 μm. (b) Quantification of collagen I mean fluorescence intensity. (c) Representative immunofluorescence images showing elastin expression (green) with DAPI counterstaining (blue). Scale bar, 20 μm. (d) Quantification of elastin mean fluorescence intensity. (e) Representative Western blot analysis of collagen I, MMP‐2, and MMP‐9 protein expression. β‐Actin was used as a loading control. (f) Quantification of collagen I, MMP‐2, and MMP‐9 normalized to β‐Actin. Data are presented as mean ± SD from six independent experiments. ***P < 0.001 between indicated groups.

Liraglutide improves systemic metabolic parameters, reduces plaque burden, and improves plaque stability in diabetic mice

To validate the vascular protective effects of liraglutide in vivo, we examined atherosclerotic lesions in diabetic mice. Physiological and biochemical assessment confirmed successful diabetes induction and systemic metabolic deterioration in diabetic ApoE^−/− mice. Compared with control mice, diabetic mice showed lower body weight, higher fasting blood glucose, increased blood pressure, worsened plasma lipid profiles, and impaired renal function. Liraglutide improved fasting blood glucose, blood pressure, plasma lipid levels, and renal function parameters compared with untreated diabetic mice (Table 1).

Table 1.

Physiological and biochemical characteristics of experimental mice

Parameter	Control	Diabetic	Diabetic + Lira
Body weight (g)	28.4 ± 2.1	24.1 ± 1.8***	25.9 ± 1.6 #
Systolic blood pressure (mmHg)	118 ± 8	131 ± 9**	122 ± 7 #
Diastolic blood pressure (mmHg)	76 ± 6	86 ± 7**	80 ± 6 #
Fasting blood glucose (mmol/L)	7.1 ± 0.8	25.8 ± 2.7***	18.6 ± 2.4 ###
Total cholesterol (mmol/L)	13.5 ± 1.8	20.9 ± 2.4***	16.5 ± 2.2 ##
Triglycerides (mmol/L)	1.3 ± 0.3	2.2 ± 0.4***	1.6 ± 0.3 ##
LDL‐cholesterol (mmol/L)	7.8 ± 1.2	13.6 ± 1.6***	10.1 ± 1.4 ##
HDL‐cholesterol (mmol/L)	1.4 ± 0.3	0.9 ± 0.2**	1.2 ± 0.2 #
Blood urea nitrogen (mmol/L)	7.8 ± 1.1	12.6 ± 1.8***	9.7 ± 1.3 ##
Serum creatinine (μmol/L)	21.5 ± 3.1	34.2 ± 5.3**	27.0 ± 4.1 #

Data are mean ± SD (n = 6 per group). **P < 0.01; ***P < 0.001 vs Control; ^# P < 0.05; ^## P < 0.01; ^### P < 0.001 vs Diabetic. HDL, high‐density lipoprotein; LDL, low‐density lipoprotein.

H&E staining revealed substantial plaque formation in diabetic mice compared with control animals (Figure 5a). Liraglutide treatment significantly reduced plaque size. Oil Red O staining demonstrated marked lipid accumulation within plaques of diabetic mice (Figure 5b). Quantitative analysis confirmed a significant increase in lipid‐positive area in the diabetic group (Figure 5e), which was significantly reduced following liraglutide treatment. Masson's trichrome staining showed decreased collagen content in plaques from diabetic mice, indicating reduced plaque stability (Figure 5c). Liraglutide treatment significantly increased collagen deposition within plaques (Figure 5f).

Figure 5.

Liraglutide attenuates atherosclerotic plaque burden and improves plaque stability in diabetic mice. (a) Low‐magnification hematoxylin and eosin (H&E) staining of aortic root sections from control, diabetic, and diabetic mice treated with liraglutide (Diabetic + Lira). Scale bar, 500 μm. (b) Low‐magnification Oil Red O staining showing lipid accumulation in aortic root lesions from the indicated groups. Scale bar, 500 μm. (c) Low‐magnification Masson's trichrome staining illustrating collagen deposition within aortic root plaques. Collagen fibers are stained blue. Scale bar, 500 μm. (d) Quantification of plaque area expressed as percentage of total vessel area. (e) Quantification of lipid deposition expressed as Oil Red O/positive area (%). (f) Quantification of collagen content within plaques expressed as collagen volume fraction. Data are presented as mean ± SD (n = 6 mice per group). ***P < 0.001 between indicated groups.

Liraglutide restores AMPK signaling and suppresses HIF‐1α‐mediated matrix remodeling in vivo

Finally, we examined whether the AMPK/HIF‐1α signaling pathway observed in vitro was also altered in vivo. Immunohistochemical staining revealed reduced p‐AMPK expression in atherosclerotic plaques of diabetic mice compared with controls (Figure 6a,d). Liraglutide treatment restored p‐AMPK expression. Conversely, HIF‐1α expression was markedly elevated in plaques from diabetic mice (Figure 6b,e). Liraglutide significantly suppressed HIF‐1α accumulation. Consistent with increased ECM remodeling in diabetic plaques, MMP‐9 expression was significantly elevated in diabetic mice (Figure 6c,f). Liraglutide treatment markedly reduced MMP‐9 expression.

Figure 6.

Liraglutide restores AMPK signaling and suppresses HIF‐1α‐mediated matrix remodeling in atherosclerotic plaques of diabetic mice. (a) Low‐magnification immunohistochemical staining of phosphorylated AMPK (p‐AMPK) in aortic root plaque regions from control, diabetic, and Diabetic + Lira mice. Scale bar, 500 μm. (b) Low‐magnification immunohistochemical staining of HIF‐1α in aortic root plaque regions. Scale bar, 500 μm. (c) Low‐magnification immunohistochemical staining of MMP‐9 in aortic root plaques. Scale bar, 500 μm. Images were obtained from serial adjacent aortic root sections corresponding to the histological analyses shown in Figure 5. (d) Quantification of p‐AMPK staining expressed as optical density. (e) Quantification of HIF‐1α staining intensity expressed as optical density. (f) Quantification of MMP‐9 staining intensity expressed as optical density. Data are presented as mean ± SD (n = 6 mice per group). ***P < 0.001 between indicated groups.

DISCUSSION

Diabetic atherosclerosis is characterized by profound metabolic disturbances and vascular remodeling that contribute to plaque instability and cardiovascular events ¹² . In the present study, we demonstrate that hyperglycemia induces metabolic reprogramming in VSMCs, characterized by impaired mitochondrial respiration and enhanced glycolysis. Importantly, the GLP‐1RA liraglutide restored mitochondrial metabolic function, suppressed oxidative stress, and attenuated ECM remodeling. Mechanistically, liraglutide treatment was associated with activation of the AMPK/PGC‐1α signaling axis and suppression of HIF‐1α accumulation, ultimately reducing matrix degradation and improving plaque stability in diabetic atherosclerosis. These data further show that high glucose activates mTOR signaling and that liraglutide suppresses the p‐mTOR/mTOR ratio without markedly changing total mTOR protein expression, supporting involvement of AMPK‐linked mTOR pathway regulation in the metabolic response to liraglutide.

Metabolic reprogramming has emerged as a critical mechanism linking hyperglycemia to vascular dysfunction. Under diabetic conditions, VSMCs undergo a metabolic shift from oxidative phosphorylation toward glycolysis, a phenomenon resembling the ‘Warburg‐like’ metabolic phenotype observed in other proliferative vascular diseases ¹³ . Consistent with previous reports, we observed that high glucose markedly reduced mitochondrial respiration and increased glycolytic activity in VSMCs, accompanied by elevated lactate production and reduced ATP levels. These metabolic alterations were associated with mitochondrial dysfunction, including decreased mitochondrial membrane potential, increased mtROS, and fragmentation of the mitochondrial network. Mitochondrial dysfunction is increasingly recognized as a key contributor to vascular inflammation and remodeling in diabetes, suggesting that restoring mitochondrial homeostasis may represent an important therapeutic strategy.

GLP‐1RAs have demonstrated significant cardiovascular benefits in large clinical trials, yet the cellular mechanisms underlying their vascular protective effects remain incompletely understood. Our results reveal that liraglutide effectively restores mitochondrial function in VSMCs exposed to hyperglycemic stress. Liraglutide improved mitochondrial membrane potential, reduced mtROS accumulation, and preserved mitochondrial network integrity, indicating a direct protective effect on mitochondrial homeostasis. These findings are consistent with previous studies showing that GLP‐1 signaling can enhance mitochondrial function and reduce oxidative stress in cardiovascular tissues ¹⁴ . The presence and functional relevance of glucagon‐like peptide‐1 receptor (GLP‐1R) signaling in vascular cells, particularly VSMCs, remain an actively discussed issue. GLP‐1 receptor localization studies have detected receptor expression in selected vascular smooth muscle compartments, whereas other vascular effects of GLP‐1RAs may occur indirectly through endothelial, inflammatory, hemodynamic, or metabolic pathways ¹⁵ , ¹⁶ . In the present study, the protective effects of liraglutide were evaluated in VSMCs; however, receptor dependence was not directly tested using GLP‐1R knockdown, receptor antagonism, neutralizing antibodies, cAMP measurement, or calcium‐mobilization assays. Therefore, the present findings support liraglutide‐associated vascular protection rather than definitive proof of VSMC‐autonomous GLP‐1R dependence.

Mechanistically, our data suggest that activation of the AMPK/PGC‐1α signaling pathway plays a central role in mediating the metabolic effects of liraglutide. AMPK is a key cellular energy sensor that regulates mitochondrial biogenesis and metabolic homeostasis ¹⁷ . Under hyperglycemic conditions, AMPK activity was significantly reduced, accompanied by decreased expression of the mitochondrial regulator PGC‐1α. Liraglutide restored AMPK phosphorylation and upregulated PGC‐1α expression, indicating improved mitochondrial metabolic signaling. AMPK is also a major upstream inhibitor of mTOR, and aberrant mTOR activation contributes to cellular growth, stress adaptation, and senescence‐associated signaling. Moreover, our data show that high glucose increased p‐mTOR/mTOR, whereas liraglutide suppressed mTOR activation while leaving total mTOR abundance relatively unchanged. These results are consistent with a model in which liraglutide restores AMPK‐linked metabolic signaling and limits high‐glucose‐induced mTOR activation in VSMCs. In contrast, hyperglycemia induced marked accumulation of HIF‐1α, a transcription factor that promotes glycolytic metabolism and metabolic reprogramming under hypoxic or metabolic stress conditions ¹⁸ . Liraglutide suppressed HIF‐1α expression and nuclear localization, suggesting that restoration of AMPK signaling may counteract HIF‐1α‐mediated metabolic adaptation in VSMCs.

The relationship between HIF‐1α and plaque biology is context‐dependent. HIF‐1α can promote cell survival and adaptation during acute ischemia; however, in chronically stressed atherosclerotic plaques, persistent HIF‐1α activation can enhance glycolysis, inflammatory activation, oxidative stress, and matrix‐remodeling programs ¹⁹ . In diabetic plaques, hyperglycemia, lipid accumulation, mitochondrial dysfunction, local hypoxia, and increased mtROS provide convergent stimuli for HIF‐1α stabilization. Liraglutide reduced HIF‐1α accumulation in both VSMCs and aortic root plaques, which is mechanistically consistent with improved mitochondrial function, reduced mtROS, restored AMPK signaling, and suppression of mTOR activation. Thus, the decrease in HIF‐1α observed after liraglutide treatment reflects attenuation of chronic maladaptive plaque stress rather than inhibition of an acute protective hypoxic response.

ECM remodeling is a key pathological feature of diabetic atherosclerosis and plays a critical role in plaque stability. Hyperglycemia promotes excessive collagen deposition, elastin degradation, and increased expression of MMPs, which collectively destabilize the vascular wall. In the present study, high glucose significantly increased collagen I expression while reducing elastin levels in VSMCs, accompanied by increased expression of MMP‐2 and MMP‐9. Liraglutide markedly attenuated these ECM alterations, suggesting that metabolic restoration directly influences ECM homeostasis. Importantly, our in vivo experiments further confirmed that liraglutide reduces plaque burden, decreases lipid accumulation, and increases collagen content within atherosclerotic plaques in diabetic mice, indicating improved plaque stability.

The mechanistic relevance of AMPK signaling was further validated in vivo. Diabetic plaques exhibited reduced p‐AMPK expression and increased HIF‐1α and MMP‐9 levels, consistent with enhanced ECM degradation and plaque instability ²⁰ . Liraglutide treatment restored AMPK activation while suppressing HIF‐1α and MMP‐9 expression in plaque regions. The coordinated increase in p‐AMPK, reduction in HIF‐1α, and reduction in MMP‐9 provide a mechanistic link between metabolic restoration and plaque stabilization: AMPK activation supports mitochondrial homeostasis and restrains mTOR/HIF‐1α‐driven metabolic stress; reduced HIF‐1α limits glycolytic and inflammatory adaptation within plaques; and reduced MMP‐9 decreases proteolytic degradation of the ECM. Together with reduced lipid deposition and increased collagen volume fraction, these changes support a more stable plaque phenotype. Our study provides mechanistic insight into the vascular protective effects of liraglutide beyond glucose lowering. In addition to their metabolic benefits, GLP‐1RA may directly regulate vascular cell metabolism and matrix remodeling, thereby stabilizing atherosclerotic plaques. These findings are consistent with clinical evidence demonstrating reduced cardiovascular events in patients with type 2 diabetes treated with GLP‐1RA ²¹ .

Although this study focuses on VSMCs, liraglutide‐mediated atheroprotection involves multiple vascular and immune cell types. Previous studies have shown that liraglutide can affect endothelial oxidative stress, endothelial nitric oxide (NO)‐related pathways, and monocyte/macrophage inflammatory responses ⁷ , ⁸ . Therefore, the in vivo reduction in plaque burden likely reflects both VSMC‐intrinsic remodeling effects and systemic or multicellular vascular effects. This differs from metformin, which activates AMPK mainly through cellular energy stress and mitochondrial complex I‐related mechanisms, whereas liraglutide acts as a GLP‐1RA and engages incretin‐dependent systemic metabolic effects together with vascular cell signaling pathways. Both agents can converge on oxidative stress and AMPK‐related vascular protection, but they are pharmacologically distinct and may influence different upstream signaling modules ²² .

The in vivo physiological and biochemical data further support the histological findings. Diabetic ApoE^−/− mice exhibited severe hyperglycemia, dyslipidemia, increased blood pressure, and renal function impairment, all of which promote plaque lipid accumulation, inflammatory activation, endothelial dysfunction, and ECM remodeling ²³ , ²⁴ . Liraglutide improved these systemic parameters while also restoring plaque p‐AMPK and suppressing HIF‐1α and MMP‐9 expression. These findings indicate that liraglutide improves plaque stability through combined systemic metabolic improvement and local vascular signaling modulation.

Several limitations should be acknowledged. First, although our results suggest that AMPK signaling plays a central role in mediating the protective effects of liraglutide, additional experiments using genetic or pharmacological inhibition of AMPK would further clarify the causal relationship between AMPK activation and ECM remodeling. Second, while our study focused on VSMCs, other vascular cell types such as endothelial cells and macrophages also contribute to plaque progression and may respond differently to GLP‐1 signaling. Third, GLP‐1R dependence in VSMCs was not directly established in this study. Future experiments using GLP‐1R knockdown, receptor antagonists such as exendin (9–39), neutralizing antibodies, cAMP assays, and calcium‐mobilization assays will be required to define whether the observed VSMC responses are mediated through canonical GLP‐1R signaling. Fourth, although mTOR activation was assessed, downstream mTOR effectors and senescence markers were not comprehensively evaluated ²⁵ . Finally, translation of these findings to human vascular disease will require further investigation in clinical samples.

In conclusion, our study demonstrates that hyperglycemia induces metabolic reprogramming in VSMCs, leading to mitochondrial dysfunction and ECM remodeling that contribute to diabetic atherosclerosis. Liraglutide restores mitochondrial metabolic homeostasis in association with activation of the AMPK/PGC‐1α signaling axis and suppression of HIF‐1α, thereby attenuating ECM degradation and improving plaque stability. These findings further identify inhibition of high‐glucose‐induced mTOR activation as part of the AMPK‐linked metabolic response to liraglutide. These results support a model in which liraglutide improves diabetic plaque stability by integrating systemic metabolic improvement with local restoration of mitochondrial, AMPK/PGC‐1α/mTOR, HIF‐1α, and matrix‐remodeling pathways. This schematic model is summarized in Figure 7.

Figure 7.

Schematic summary of liraglutide‐mediated metabolic reprogramming and plaque stabilization in diabetic atherosclerosis. Under diabetic or high‐glucose conditions, suppressed AMPK/PGC‐1α signaling and increased HIF‐1α accumulation promote mitochondrial dysfunction, glycolytic shift, mitochondrial ROS production, and extracellular matrix remodeling in vascular smooth muscle cells, leading to plaque progression and instability. Liraglutide restores AMPK/PGC‐1α signaling, suppresses HIF‐1α accumulation, improves mitochondrial metabolic function, attenuates collagen I/MMP‐2/MMP‐9 upregulation and elastin loss, and thereby reduces plaque burden and improves plaque stability.

FUNDING

This work was supported by the Science and Technology Innovation Team Project of Shaanxi Provincial Department of Science and Technology (Grant No. 2025RS‐CXTD‐053) and the Traditional Chinese Medicine Scientific Research and Innovation Talent Project of Shaanxi Provincial Administration of Traditional Chinese Medicine (Grant No. TZKN‐CXRC‐15).

DISCLOSURE

The authors declare that they have no competing interests.

Approval of the research protocol: All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Shaanxi Provincial People's Hospital (Approval No. 2025K‐196) and performed in accordance with the guidelines for the care and use of laboratory animals.

Informed consent: N/A.

Registry and the registration no. of the study/trial: N/A.

Animal studies: N/A.

AUTHOR CONTRIBUTIONS

KZ and HL performed the majority of the experiments and analyzed the data. NH and HW contributed to cell culture experiments and mitochondrial functional assays. JL participated in the animal experiments and histological analysis. KZ drafted the manuscript. QZ and ZL conceived and designed the study, supervised the project, and revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript.

Supporting information

Figure S1. Liraglutide suppresses high‐glucose‐induced mechanistic target of rapamycin (mTOR) activation in VSMCs. (A) Representative Western blot analysis of p‐mTOR, total mTOR, and β‐actin in VSMCs cultured under NG, HG, or HG + Lira conditions. (B) Quantification of p‐mTOR/mTOR ratio. (C) Quantification of total mTOR normalized to β‐actin. Data are presented as mean ± SD from six independent experiments. ***P < 0.001 between indicated groups; ns, not significant.

DATA AVAILABILITY STATEMENT

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.