Lactate-Activated GPR132-Src Signal Induces Macrophage Senescence and Aggravates Atherosclerosis Under Diabetes

Abstract

Diabetes is widely acknowledged as a significant risk factor for atherosclerosis, facilitating plaque formation through various mechanisms. Although both conditions are linked to the aging process, the relationship among cellular senescence, diabetes, and atherosclerosis remains inadequately understood. This study presents evidence that elevated glucose levels expedite the progression of atherosclerosis by promoting macrophage senescence. Increased glucose levels are shown to induce senescence in macrophages, which enhances the uptake of oxidized low-density lipoprotein (ox-LDL) and facilitates the formation of foam cells. This mechanism is driven by lactate production via glycolysis, which activates the lactate receptor GPR132, thereby promoting macrophage senescence. The activation of GPR132 is implicated in mediating senescence and lipid uptake through Src phosphorylation. The deletion of GPR132 markedly reduces macrophage senescence and atherosclerosis in mouse models. Furthermore, saracatinib, a specific Src inhibitor, has been demonstrated to effectively alleviate diabetic atherosclerosis in experimental settings. In clinical samples, elevated plasma lactate levels and the activation of the GPR132-Src pathway in peripheral blood mononuclear cells (PBMCs) are positively associated with coronary stenosis. These findings propose a potential mechanism through which diabetes accelerates atherosclerosis via the lactate-GPR132-Src pathway, underscoring macrophage senescence as a pivotal target in the context of diabetic atherosclerosis.

Keywords: atherosclerosis; diabetes; lactate; macrophage; senescence.

Figure 1

Diabetes aggravates atherosclerosis and reprograms the metabolism of peripheral macrophages. A) A diagram for the clinical data analysis flow. B) Groups of 266 patients according to diabetes and coronary arteriosclerosis. C) Correlation analysis of HbA1c and Atherogenic index of plasma (AIP). The correlations were analyzed with linear regression. D) A diagram for the mouse model of diabetic atherosclerosis. 4 mg kg⁻¹ STZ or citric acid was injected in the first and fifth weeks of the experiment. The mice were totally fed with the Western diet for 8 weeks. Representative images of Oil Red O stained aorta were shown in the right panel and the arterial roots with H&E staining were shown below. (E) The heatmap shows the differently expressed genes of mouse peritoneal macrophages. Samples are ordered by hierarchical clustering similarity. F) The PCA analysis of the RNA‐sequencing data of mouse peritoneal macrophages. G) Bubble charts to show the KEGG analysis of the differently expressed genes. H) Volcano plots to show the differently expressed genes. The marked plots are significantly changed genes belonging to the lipid and glucose homeostasis pathway, according to GO analysis. I–K) GSEA analysis of three pathways: the activation of immune response (I), glycolytic process (J), and lipid homeostasis pathway (K). (L) Relative mRNA levels of genes of macrophage activation, glycolytic process, and lipid homeostasis pathways. n = 3 for each group. The data were analyzed with a student t‐test. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2

High glucose level promotes macrophage senescence in vivo and in vitro. (A) GSEA analysis of the cellular senescence pathway. B) Heatmap to show the genes belonging to the cellular senescence pathway, according to KEGG. C) Analysis of single‐cell RNA‐sequencing data of GSE165816. The volcano plots of differently expressed genes in monocytes/ macrophages, the expression data of some SASP genes, and the KEGG analysis of the genes were shown in the right panel. D) Representative images of β‐Gal stained aorta. E) Representative images of the arterial roots with F4/80 (green) and p21 (Red) staining. F) Western blotting to detect p16, p21, and CD36 in the peritoneal macrophages of the mice as in Figure 1. Quantification of p16, p21, and CD36 relative to tubulin is shown in the below panel. G) Representative whole‐well images of Oil Red O and β‐Gal stained RAW264.7 cells. Low: treated with DMEM with 1.5 mg mL⁻¹ glucose for 48 h. Low+ox‐LDL: pre‐treated with DMEM with 1.5 mg mL⁻¹ glucose for 24 h followed by 40 mg mL⁻¹ ox‐LDL addition for 24 h; Middle/Mid+ox‐LDL: pre‐treated with DMEM with 3 mg mL⁻¹ glucose for 24 h followed by 40 mg mL⁻¹ ox‐LDL addition for 24 hours; High+ox‐LDL: pre‐treated with DMEM with 4.5 mg mL⁻¹ glucose for 24 h followed by 40 mg mL⁻¹ ox‐LDL addition for 24 h. H) Flow cytometry analysis of Dil‐ox‐LDL uptake by the RAW264.7 cells under treatment. The pre‐treatment was the same as in G but the Dil‐ox‐LDL was treated for 4 hours with 10 mg mL⁻¹. I, J) Relative mRNA levels of lipid uptake genes (I) and SASP genes (H) in the RAW264.7 cells as in G. n = 5 for each group. K) Western blotting to detect p21, p16, OLR1, and CD36 in the RAW264.7 cells as in G. The quantitative data of E and F were analyzed with student t‐test, and the quantitative data of I and J were analyzed with one‐way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3

High levels of glucose promoted macrophage ox‐LDL uptake through cell senescence. A) Representative whole‐well images of Oil Red O stained RAW264.7 cells, which were pre‐treated with DMEM with different concentrations of glucose and 50 µmol L⁻¹ SSO for 6 h followed by 24 h stimulation with 40 mg mL⁻¹ ox‐LDL. Low, DMEM with 1.5 mg mL⁻¹ glucose; Mid, DMEM with 3 mg mL⁻¹ glucose; High, DMEM with 4.5 mg mL⁻¹ glucose. B) Quantification of the Oil Red O staining of the cells in A. n = 4 for each group. C) Western blotting to detect p21 and p16 in the RAW264.7 cells as in A. Quantification of p21 relative to β‐actin from two independent experiments was shown in the right panel. D) Relative mRNA levels of SASPs in the RAW264.7 cells as in A. n = 4 for each group. E) Western blotting to detect p16, p21, CD36, and OLR1 in the RAW264.7 cells pre‐treated with 4.5 mg mL⁻¹ glucose DMEM and different concentrations of quercetin (QC) for 12 h followed by 24 h stimulation with 40 mg mL⁻¹ ox‐LDL. F) Representative whole‐well images of Oil Red O stained RAW264.7 cells as in E. Quantification of the Oil Red O staining of the cells in the right panel. n = 4 for each group. G) Flow cytometry analysis of Dil‐ox‐LDL uptake by the RAW264.7 cells under 24‐h QC treatment followed by 4‐h 10 mg mL⁻¹ Dil‐ox‐LDL treatment. H) Western blotting to detect p16, p21, CD36, and OLR1 in the RAW264.7 cells pre‐treated with 1.5 mg mL⁻¹ glucose DMEM and different concentrations of paclitaxel (PTX) for 12 hours followed by 24 h stimulation with 40 mg mL⁻¹ ox‐LDL. I) Representative whole‐well images of Oil Red O stained RAW264.7 cells as in H. Quantification of the Oil Red O staining of the cells in the right panel. n = 4 for each group. J) Flow cytometry analysis of Dil‐ox‐LDL uptake by the RAW264.7 cells under 24‐h PTX treatment followed by 4‐h 10 mg mL⁻¹ Dil‐ox‐LDL treatment. All the quantitative data were analyzed with one‐way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001.

Figure 4

Lactate produced from glycolysis promotes senescence in an extracellular way. A) A diagram to show the key steps of glycolysis and the inhibitors. B) Western blotting to detect p21 and CD36 in the RAW264.7 cells pre‐treated with 1.5 mg mL⁻¹ glucose (Low), 4.5 mg mL⁻¹ glucose (High), 4.5 mg mL⁻¹ glucose with 2 mg mL⁻¹ 2‐DG (2‐DG), 4.5 mg mL⁻¹ glucose with 1 µmol L⁻¹ aldometanib (Al‐i), 4.5 mg mL⁻¹ glucose with 10 mmol L⁻¹ oxamic acid sodium (Oxa) for 12 h followed by 24 h stimulation with 40 mg mL⁻¹ ox‐LDL. Quantification of p21 relative to β‐actin from two independent experiments is shown below, n = 4 for each group. (C) Representative whole‐well images of β‐Gal stained RAW264.7 cells as in B. (D) Western blotting to detect p21, p16, CD36 and OLR1 in the RAW264.7 cells pre‐treated with 1.5 mg mL⁻¹ glucose (Low), 4.5 mg mL⁻¹ glucose (High), 1.5 mg mL⁻¹ glucose with 5 mmol L⁻¹ lactate (5LA), 1.5 mg mL⁻¹ glucose with 15 mmol L⁻¹ lactate (15LA) for 12 h, followed by 24 h stimulation with 40 mg mL⁻¹ ox‐LDL. (E) Representative whole‐well images of β‐Gal and Oil Red O stained RAW264.7 cells as in D. (F) The levels of lactate in the plasma of the mice in Figure 1. n = 7 for each group. (G) A diagram to show lactate transport and recognition in cells. AZD3965 specific inhibits MCT1; 7ACC1 targets both MCT1 and MCT4; AR‐C155858 targets MCT1 and MCT2. (H) Western blotting to detect p21, p16, and CD36 in the RAW264.7 cells cultured with 1.5 mg mL⁻¹ glucose, respectively pre‐treated with 10 mmol L⁻¹ lactate, 1 µmol L⁻¹ AZD3965, 0.5 mmol L⁻¹ 7ACC1, 0.1 µmol L⁻¹ AR‐C155858 for 12 h, followed by 24 h stimulation with 40 mg/mL ox‐LDL. I) Relative mRNA levels of Cdkn1a and Vegfa in the RAW264.7 cells as in H. J) Representative whole‐well images of β‐Gal stained RAW264.7 cells as in H. The quantitative data of B and I were analyzed with one‐way ANOVA, while the data of F were analyzed with student t‐tests. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 5

Lactate‐mediated macrophage senescence through GPR132. A) Bubble charts to show the GO analysis of the differently expressed genes, with the RNA‐seq data shown in Figure 2. B) Heatmap to show the expression of GPCR genes. C) Western blotting to detect GPR132 and HCAR1 in mouse peritoneal macrophages as in Figure 2. Quantification of GPR132 was shown in the right panel. D) Western blotting to detect GPR132 and HCAR1 in the RAW264.7 cells pre‐treated with DMEM with 4.5 mg/mL glucose (High), 1.5 mg mL⁻¹ glucose (Low), 1.5 mg mL⁻¹ glucose with 10 mmol L⁻¹ lactate (LA) for 24 h. E) Western blotting to detect p21 and p16 in the RAW264.7 cells cultured with 1.5 mg mL⁻¹ glucose DMEM, respectively pre‐treated with DMSO (DMSO), 10 mmol L⁻¹ lactate (LA), different concentrations of 3,5‐DHBA or ONC212 for 24 h, followed by 24 h stimulation with 40 mg mL⁻¹ ox‐LDL. (F) Representative images of β‐Gal stained RAW264.7 cells in E. G) Representative images of GPR132 immunohistochemistry of the aortic roots in the mice as in Figure 1. H) Western blotting to detect the deletion of GPR132 in BMDMs. I) Representative images of Oil Red O stained aorta. J) Representative images of the arterial roots with H&E staining, Oil Red O staining, F4/80 (green), and p21 (red) staining. K) Quantification of the plaques in I‐J. n = 9 for each group. L) Western blotting to detect p21, p16, OLR1 and CD36 in the BMDMs from Apoe^−/− (WT) and Gpr132^−/−Apoe^−/− (KO) mice, pre‐treated with 1.5 mg/mL glucose DMEM (low), 4.5 mg mL⁻¹ glucose DMEM (high), and 1.5 mg/mL glucose DMEM with10mmol/L lactate (LA) for 24 h, followed by 24 h stimulation with 40 mg/mL ox‐LDL. M) Representative whole‐well images of Oil Red O and β‐gal stained RAW264.7 cells as in L. Quantification of Oil Red O were shown in the right panel. n = 4 for each group. The quantitative data in C and K were analyzed with student t‐tests, and the data in M were analyzed with one‐way ANOVA. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 6

GPR132‐induced Src phosphorylation mediates macrophage senescence and saracatinib is a potential medicine for diabetic atherosclerosis. A) Western blotting to detect the phosphorylation of Src, AMPK, ERK1/2 and AKT1 in the RAW264.7 cells pre‐treated with DMEM with 1.5 mg mL⁻¹ glucose (Low), 4.5 mg mL⁻¹ glucose (High), 1.5 mg mL⁻¹ glucose with 10 mmol/L lactate (LA), 1.5 mg mL⁻¹ glucose with different concentration of ONC212 for 24 h pre‐treatment and then 24 h stimulation with 40 mg/mL ox‐LDL. (B) Western blotting to detect the phosphorylation of Src, p16, p21, and CD36 in the RAW264.7 cells with DMEM with 1.5 mg mL⁻¹ glucose (Low), 4.5 mg mL⁻¹ glucose (High), 1.5 mg mL⁻¹ glucose with 10 mmol L⁻¹ lactate (LA), 1.5 mg mL⁻¹ glucose with 1 µmol L⁻¹ of ONC212 (ONC), lactate or ONC212 with different concentration of saracatinib for 24 h pre‐treatment and then 24 hours stimulation with 40 mg mL⁻¹ ox‐LDL. S10, S50, S250: saracatinib treatment at 10, 50, and 250 nM concentrations respectively. (C‐D) Representative images of β‐Gal (C) and Oil Red O (D) stained RAW264.7 cells in B. (E) ChIP‐qPCR performed with anti‐STAT3 antibody in RAW264.7 treated by 4.5 mg mL⁻¹ glucose with DMSO (Ctrl) or 250 nM saracatinib (S250) for 12 h and then stimulation with 40 mg mL⁻¹ ox‐LDL for 12 h. n = 3 for each group. (F) A diagram to show the mouse model of diabetic atherosclerosis. The mice were totally fed with the western diet for 12 weeks. 4 mg/kg STZ or citric acid (CA) was injected in the first and seventh weeks of the experiment. From the 9th week, the mice of the experimental groups were treated with DMSO or 20 mg kg⁻¹ saracatinib (Sarac) every 3 days for 4 weeks. (G) Representative images of Oil Red O stained aorta. The plaque area is shown in the right panel. n = 6 for each group. (H) Representative images of the arterial roots with Oil Red O staining (up), F4/80 (green), and p21 (red) staining (below). (I) Quantification of the plaque area (Oil Red O positive), the macrophage area (F4/80 positive), and the ratio of p21 positive macrophage of the roots. n = 8 for each group. The quantitative data in E were analyzed with student t‐tests and the data in G and I were analyzed with one‐way ANOVA. *p < 0.05; **p <0.01; ***p <0.001.

Figure 7

Lactate levels in plasma and GPR132‐Src activation in PBMC are associated with the severity of coronary stenosis in patients. A) Lactate concentration in the plasma of the non‐stenosis patients (Non), stenosis below 50% (<50%), and stenosis over 50% (>50%). Non, n = 23; <50%, n = 44; >50%, n = 48. B, C) The correlation of lactate concentration with Lp (a) level (B) and AIP (C). n = 115 for each analysis. D) Relative mRNA levels of Gpr132 and Cd68 in the PBMCs of the patients as in A. Non, n = 16; <50%, n = 30; >50%, n = 36. E) The correlation of lactate concentration with Gpr132 expression. n = 82. F) Western blotting to detect GPR132 and the phosphorylation of Src in the PBMCs of the patients as in A. G) Quantification of F. H) A proposed model depicts that lactate‐GPR132‐Src pathway promotes macrophage senescence and foam cell formation to induce atherosclerosis under diabetes. The quantitative data of G were analyzed with student t‐tests, and the quantitative data of A and D were analyzed with one‐way ANOVA. The quantitative data of B, C, E were analyzed with Pearson Correlation. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.