Cellular communication network factor 2 regulates smooth muscle cell transdifferentiation and lipid accumulation in atherosclerosis

Abstract

Aims: Accruing evidence illustrates an emerging paradigm of dynamic vascular smooth muscle cell (SMC) transdifferentiation during atherosclerosis progression. However, the molecular regulators that govern SMC phenotype diversification remain poorly defined. This study aims to elucidate the functional role and underlying mechanisms of cellular communication network factor 2 (CCN2), a matricellular protein, in regulating SMC plasticity in the context of atherosclerosis.

Methods and results: In both human and murine atherosclerosis, an up-regulation of CCN2 is observed in transdifferentiated SMCs. Using an inducible murine SMC CCN2 deletion model, we demonstrate that SMC-specific CCN2 knockout mice are hypersusceptible to atherosclerosis development as evidenced by a profound increase in lipid-rich plaques along the entire aorta. Single-cell RNA sequencing studies reveal that SMC deficiency of CCN2 positively regulates machinery involved in endoplasmic reticulum stress, endocytosis, and lipid accumulation in transdifferentiated macrophage-like SMCs during the progression of atherosclerosis, findings recapitulated in CCN2-deficient human aortic SMCs.

Conclusion: Our studies illuminate an unanticipated protective role of SMC-CCN2 against atherosclerosis. Disruption of vascular wall homeostasis resulting from vascular SMC CCN2 deficiency predisposes mice to atherosclerosis development and progression.

Keywords: Atherosclerosis; Cellular communication network (CCN) family; Endoplasmic reticulum stress; Smooth muscle cells.

Graphical Abstract

Figure 1

CCN2 protein and macrophage-like SMCs are increased in the advanced atherosclerotic lesions of human atherosclerosis disease. (A) Representative images of H&E staining of LCA sections from different stages of human atherosclerosis patients. (B and C) Quantification of intima/media thickness ratio (B) and media thickness (C). (D) Representative images of Masson Trichrome staining of LCA sections from different stages of human atherosclerosis patients. (E and F) Quantification of plaque area (E) and collagen fraction (F) normalized to the lesion area. (G) Representative images of immunofluorescent staining of LCA sections for CCN2, MYH11, and DAPI (4′,6-diamidino-2-phenylindole) from different stages of human atherosclerosis patients. (H) Quantification of immunofluorescence intensity of CCN2⁺ area normalized to the intima (Int), media (Med) area. (I) CCN2 protein level in serum samples from normal and CAD patients. (J) Representative images of immunofluorescent staining of LCA sections for MYH11, CD68, and DAPI from different stages of human atherosclerosis patients. (K) Quantification of colocalization of MYH11⁺ and CD68⁺ cells normalized to total number of SMCs. LCA, left coronary artery. (A–G, J) n = 4–7 patients per group; (I) n = 25–32 patients per group. (A and D) Scale bar = 1 mm. (G and I) Scale bar = 50 µm. Stages I-II, filled circle; Stage IV, filled square; Stage VI, filled triangle. Fille circle, filled square and filled triangle indicate biological replicates. Data represented as mean ± SEM. Unpaired t-tests (B, C, E, F), Mann–Whitney test (I) and ordinary one-way ANOVA (K). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2

SMC deficiency of CCN2 predisposes mice to AAV-PCSK9 + HFD-induced atherosclerosis. (A) Representative images for the aorta morphology of control and SMC-CCN2-KO treated with HFD. (B) Representative transthoracic echocardiography images of the ascending aorta of control and SMC-CCN2-KO fed HFD. (C) Statistical analysis of ascending aortic diameter measurements in control and SMC-CCN2-KO mice prior to and at the end of HFD. (D) Representative images of ORO staining of an face preparations of the aortas in control and SMC-CCN2-KO mice fed HFD. (E) Quantitative analysis of the atherosclerotic lesion area normalized to the entire aortic area. (F) Representative images of cryosections of the aortic sinus stained with ORO, with haematoxylin used as a counterstain in control and SMC-CCN2-KO mice fed HFD. (G) Quantification of atherosclerotic lesion area of six individual cross sections from control and SMC-CCN2-KO mice fed HFD, 100–600 µm sections from the aortic sinus displayed as lesion fraction of total vessel surface. (H) Quantification of ORO-positive area of aortic root sections in control and SMC-CCN2-KO mice fed HFD. (I) Representative images of immunofluorescent staining of aortic sinus sections for MYH11, CD68, and DAPI from control and SMC-CCN2-KO mice fed HFD. (J) Quantification of MYH11 and CD68 double positive cells normalized to the MYH11-positive cell count. (K) Representative electron microscopy pictures of SMC alteration from aortic root in control and SMC-CCN2-KO mice fed HFD. EM, electron microscopy; N, nuclear; ER, endoplasmic reticulum. n = 4–13 mice per group. F, Scale bar = 100 µm. I, Scale bar = 50 µm, K, Scale bar = 50 nm. Filled circle and filled square indicate biological replicates. Data represented as mean ± SEM. Unpaired t-tests (E, C, G, H, J). *P < 0.05, **P < 0.01, *** < 0.001, ****P < 0.0001.

Figure 3

scRNA-seq reveals that SMC deficiency of CCN2 promotes endocytosis and lipid transport signalling pathways in different cell populations in the context of atherosclerosis. (A) Workflow of Sc-seq for control and SMC-CCN2-KO mice fed HFD. (B) UMAP plot showing all cells coloured according to the five major cell types in merged Sc-seq dataset. (C) Significantly expressed marker genes according to respective cell types within the merged Sc-seq dataset. The size of the points indicates the percentage of cells that expressed the particular marker within the specific cell type. Colour depicts the average log normalized expression. (D and E) UMAP plots revealing cell clustering from control and SMC-CCN2-KO groups based on gene expression. (F) Proportions of cell populations in control and SMC-CCN2-KO groups fed HFD. (G–I) Top biological process enrichment analysis from GO database of down-regulated and up-regulated DEGs from control vs. SMC-CCN2-KO groups in SMC (G), FB (H), and MAC (I) subsets. (J) Heatmap of DEGs in positive regulation of endocytosis and regulation of lipid transport pathways between control and SMC-CCN2-KO groups treated with HFD in SMC, FB, and MAC subsets. DEGs, differentially expressed genes; SMC, smooth muscle cell; FB, fibroblast; MAC, macrophage.

Figure 4

scRNA-seq identifies CCN2 deficiency promotes SMC phenotype transition to macrophage-like SMCs in the context of atherosclerosis. (A) UMAP plot showing all cells coloured according to the three SMC cell subpopulations of merged Sc-seq dataset. (B) Dot plot showing specific marker gene expression according to the three SMC cell subpopulations. (C) UMAP plots revealing new cell clustering of SMC subsets from control and SMC-CCN2-KO groups based on gene expression. (D) Proportions of SMC subpopulations in control and SMC-CCN2-KO groups fed HFD. (E) UMAP plots revealing Acta2⁺ and Cd68⁺ cells in SMC subsets from control and SMC-CCN2-KO groups fed HFD. (F) Proportions of Acta2⁺ and Cd68⁺ cells in macrophage-like SMC subpopulation from control and SMC-CCN2-KO groups treated with HFD. (G) Vlnplots showing the expression levels of smooth muscle contractile markers (Myh11, Acta2, Cnn1, and Tagln) and macrophage markers (Cd68, Lgals3, Abca1, and Abcg1) in each SMC subpopulations from control and SMC-CCN2-KO groups fed HFD. (H) Heatmap of DEGs in response to ER stress and response to oxidative stress pathways between control and SMC-CCN2-KO groups fed HFD in SMC subpopulations. (I) Transcription factor target enrichment analysis of DEGs in control vs. SMC-CCN2-KO in SMC subset based on GO database. DEGs, differentially expressed genes.

Figure 5

Knockdown of CCN2 in SMCs enhances cholesterol-induced SMC phenotype transition to macrophage-like cells and ER stress. (A and B) Representative images (A) and statistical analysis (B) of the CCN2 protein expression in human aortic SMCs treated with MBD-cholesterol at different time points. (C) Statistical analysis of mRNA expression of CCN2 in human aortic SMCs treated with MBD-cholesterol at different time points. (D and E) Representative images (D) and statistical analysis (E) of ORO staining in human aortic SMCs treated with either shCCN2 or MBD-cholesterol. (F and G) Statistical analysis of percentage of cholesterol efflux (F) and total cholesterol uptake in human aortic SMCs treated with either shCCN2 or 22-nitrobenzoxadiazole-cholesterol. (H and I) Representative images (H) and statistical analysis (I) of phagocytosis assay in human aortic SMCs treated with either shCCN2 or MBD-cholesterol. n = 3–8 biological experiments per group. Scale bar = 100 µm. The filled circle indicates biological replicate. Data represented as mean ± SEM. ns = not significant. Unpaired t-tests (B) and ordinary one-way ANOVA (C, E–G, I). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6

Knockdown of CCN2 in SMCs enhances protein and gene expression of cholesterol-induced SMC phenotype transition to macrophage-like cells and ER stress. (A and B) Representative pictures (A) and statistical analysis (B) of protein level of smooth muscle contractile markers (ACTA2, Calponin1, and SM22α), macrophage markers (CD68 and Galectin3), ER stress markers (p-PERK, ATF4, and CHOP), KLF4, and HMGCR in human aortic SMCs treated with either shCCN2 or MBD-cholesterol. (C) Statistical analysis of mRNA expression of smooth muscle contractile markers (MYH11, ACTA2, CNN1, and SM22α), macrophage markers (CD68, LGASL3, and ABCA1), ER stress markers (ATF6, PERK, ATF4, and CHOP), KLF4, and HMGCR in human aortic SMCs treated with either shCCN2 or MBD-cholesterol. (D) Representative electron microscopy images of human aortic SMCs treated with either shCCN2 or MBD-cholesterol. EM, electron microscopy; N, nuclear; ER, endoplasmic reticulum. Scale bar = 0.2 µm. n = 3–5 biological experiments per group. shCtrl, filled circle; shCtrl+MBD-Chol, filled square; shCCN2, filled triangle; shCCN2+MBD-Chol, filled inverted triangle. These symbols indicate the biological replicates. Data represented as mean ± SEM. ns, not significant. Unpaired t-tests (B, C). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 7

Treatment of recombinant CCN2 in SMCs suppreses cholesterol-induced SMC phenotype transition to macrophage-like cells and ER stress. (A and B) Representative images (A) and statistical analysis (B) of ORO staining in human aortic SMCs treated with either rbCCN2 or MBD-cholesterol. (C and D) Representative images (C) and statistical analysis (D) of protein level of smooth muscle contractile markers (ACTA2, Calponin1, and SM22α), macrophage markers (CD68 and Galectin3), ER stress markers (p-PERK, ATF4, and CHOP), KLF4, and HMGCR in human aortic SMCs treated with either rbCCN2 or MBD-cholesterol. n = 4–5 biological experiments per group. Data represented as mean ± SEM. Ordinary one-way ANOVA (B) and Unpaired t-tests (D). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.