Atherosclerosis is a smooth muscle cell-driven tumor-like disease

Abstract

Atherosclerosis, the leading cause of cardiovascular disease, is a chronic inflammatory disease involving pathological activation of multiple cell types, such as immunocytes (e.g., macrophage, T cells), smooth muscle cells (SMCs), and endothelial cells. Multiple lines of evidence have suggested that SMC "phenotypic switching" plays a central role in atherosclerosis development and complications. Yet, SMC roles and mechanisms underlying the disease pathogenesis are poorly understood. Here, employing SMC lineage tracing mice, comprehensive molecular, cellular, histological, and computational profiling, coupled to genetic and pharmacological studies, we reveal that atherosclerosis, in terms of SMC behaviors, share extensive commonalities with tumors. SMC-derived cells in the disease show multiple characteristics of tumor cell biology, including genomic instability, replicative immortality, malignant proliferation, resistance to cell death, invasiveness, and activation of comprehensive cancer-associated gene regulatory networks. SMC-specific expression of oncogenic Kras^G12D accelerates SMC phenotypic switching and exacerbates atherosclerosis. Moreover, we present a proof of concept showing that niraparib, an anti-cancer drug targeting DNA damage repair, attenuates atherosclerosis progression and induces regression of lesions in advanced disease in mouse models. Our work provides systematic evidence that atherosclerosis is a tumor-like disease, deepening the understanding of its pathogenesis and opening prospects for novel precision molecular strategies to prevent and treat atherosclerotic cardiovascular disease.

Fig. 1.. Extensive genomic instability exists in SMC lineage cells in atherosclerosis.

(A) Immunohistochemistry (IHC) of oxidative DNA damage marker, 8-hydroxy-2’-deoxyguanosine (8-OHdG), during progression of atherosclerosis. ROSA26^{LSL-ZsGreen1/+}; Ldlr^−/−; Myh11-CreER^T2 mice on western diet (WD) for 0, 8, 10, 12, or 16 weeks were sacrificed for immunostaining. Representative images of brachiocephalic artery (BCA) sections from each time point are shown. (B and C) Analysis of 8-OHdG⁺ZsGreen1⁺ area in BCA sections (B) and ZsGreen1⁺ area in intima (C) suggests that oxidative DNA damage occurs at early stages of the SMC phenotypic switching process in atherosclerotic lesions. Values were indicated. N=6 mice at each time point. Scale bars, 50 μm. Significance was determined by unpaired two-tailed t test. N=3, ***P<0.001. (D) Comet assay indicates tailed nuclei with single/double-strand DNA breaks. SMCs and SDCs were isolated through fluorescence-activated cell sorting (FACS) from aortas of ROSA26^ZsGreen1/+; Ldlr^−/−; Myh11-CreER^T2 mice on 0-week and 26-week WD, respectively. Statistical analysis showed the percentage of SMCs and SDCs with tailed nuclei. N=3. (E) Clustered heat map showing copy number profiles estimated by CopyKAT in ZsGreen1⁺ SMC lineage cell types (including SMC, SEM cell, and fibrochondrocyte (FC)) from scRNA-seq database of 16-week WD fed mice. SMCs from 0-week WD fed mice were as reference (SMC-ref.). (F) Line plot indicates the consensus of mouse scRNA-seq copy number profiles of each cell cluster estimated by CopyKAT in (E).

Fig. 2.. SMC-derived cells formed in atherosclerosis show multiple tumor cell-like characteristics.

(A) Staining of senescence-associated beta-galactosidase (SA-β-Gal), a biomarker of cellular senescence, in ex vivo SMCs (at passage 5 (P5)) and SDCs (at P5 and P45). Proportion of SA-β-Gal⁺ SMCs or SDCs at each passage was analyzed. (B) Immunoblotting of cellular senescence markers, p21 and p16, indicates that senescence marker proteins are reduced in SDCs at early passage (P5) and late passage (P45) compared to SMCs at P5. (C) Calculation of cell population doubling levels of SMCs and SDCs starting from passages 3 indicates that SDCs harbor replicative immortality while SMCs encounter growth arrest within 10 passages. (D) SMCs and SDCs were incubated with 5-ethynyl-2’-deoxyuridine (EdU) for 2 hours. Click-iT EdU assay indicated that SDCs had much higher proliferative rate (proportion of EdU⁺ nuclei) than SMCs. (E) Ex vivo SDCs form colonies at low seeding density (200/well, 6-well plate). Cell colonies per well were counted. (F) Schematic and representative images of cell invasion assay with SMCs and SDCs. Number of cells invading through Matrigel layer were shown. (G) Schematic and representative images of 3D spheroid formation assay with SMCs and SDCs. SMCs and SDCs were seeded into ultra-low attachment plate and cultured with 3D Tumorsphere Medium for 10 days. 3D spheroids were counted. Scale bars, 50 μm. Significance was determined by unpaired two-tailed t test. N=3, **P<0.01, ***P<0.001.

Fig. 3.. Analysis of cancer-associated signaling pathways activated in SDCs versus SMCs in atherosclerosis.

(A) Heat map showing median scores of 12 cancer-associated signaling pathways in SMCs, SEM cells, and FCs estimated by Pathway RespOnsive GENes (PROGENy). (B) Violin plot shows PROGENy score for NFκB pathway. (C) Immunoblotting results indicate that phospho-NFκB p65 (S536) was increased in SDCs versus SMCs. (D) Violin plot shows PROGENy score for PI3K pathway. (E) Immunoblotting results indicate that phospho-AKT (S473) was increased in SDCs versus SMCs. (F) Violin plot shows PROGENy score for MAPK pathway. (G) Immunoblotting results indicate that phospho-ERK (T202/Y204) and phospho-MEK1/2 (S217/221) were increased in SDCs versus SMCs. (H) Summary of 12 cancer-prone signaling pathways that were activated in SDCs. Key transducers of each signaling pathway that were validated via immunoblotting were marked by red (activated) or blue (repressed) dots. Cancer-related functions of each signaling pathway were indicated. P values are shown.

Fig. 4.. SMC-specific expression of Kras^G12D accelerates SMC phenotypic switching during atherosclerosis progression.

(A) Kras+/+; ROSA26LSL-ZsGreen1/+; Ldlr^−/−; Myh11-CreER^T2 (SMC-Kras+/+) and Kras^LSL-G12D/+; ROSA26^{LSL-ZsGreen1/+} Ldlr^−/−; Myh11-CreER^T2 (SMC-Kras^G12D/+) mice were sacrificed for IHC staining after 0, 8, 10, or 12 weeks of WD. Representative images of mouse BCA sections stained with oxidative DNA damage marker, 8-OHdG, at each time point are shown. (B-D) Statistical analysis of 8-OHdG⁺ZsGreen1⁺ area (B) and ZsGreen1⁺ area (C) in neointima and total atherosclerotic lesion area (D) in BCA sections. (E) SMC phenotypic switching was divided into four stages: (i) contractile SMC, (ii) early remodeled SMC/SDC, (iii) fibrous cap SMC/SDC, and (iv) neointimal SDC. Proportion and number of SMC-Kras^+/+ and SMC-Kras^G12D/+ mice at each stage of SMC phenotypic switching at each time point are indicated. N=8 mice/group at each time point. Scale bars, 50 μm. Significance was determined by unpaired two-tailed t test. *P<0.05, **P<0.01, ***P<0.001.

Fig. 5.. Niraparib has both preventive and therapeutic effects on atherosclerosis in mouse models.

(A) Schematic of administration of niraparib to ROSA26^{LSL-ZsGreen1/+}; Ldlr^−/−; Myh11-CreER^T2 mice during initiation and progression of atherosclerosis. The mice were induced with tamoxifen (TAM) for one week and then fed one-week chow diet, followed by WD. After 4 weeks of WD, mice were treated with niraparib (Nira, 10 mg/kg mice, 3 times/week) or vehicle (Veh, corn oil, 3 times/week) and sacrificed after 16-week WD. Aortic sinus sections were subjected to hematoxylin and eosin (H&E) stain. Representative images from Veh and Nira-treated mice were shown. (B-D) Statistical analysis of lesion area (B), necrotic core area (C), and ratio of fibrous cap/lesion area (D) in H&E-stained sections from Veh and Nira-treated mice in (A). (E) Schematic of niraparib treatment to ROSA26^{LSL-ZsGreen1/+}; Ldlr^−/−; Myh11-CreER^T2 mice with established atherosclerosis. The mice were induced with TAM and fed chow diet followed by WD as the same in (A). After 16 weeks of WD, mice were treated with Nira (10 mg/kg mice, 3 times/week) or Veh (corn oil, 3 times/week) and sacrificed at total of 24-week WD. En face aortas were stained with Oil Red O and captured under microscope. Representative images from mice of two groups were shown. (F) Proportions of plaque area in aortas from Veh and Nira-treated mice in (E). (G) Representative images of H&E-stained aortic sinus sections from Veh and Nira-treated mice with established atherosclerosis were shown. (H-J) Statistical analysis of lesion area (H), necrotic core area (I), and ratio of fibrous cap/lesion area (J) in H&E-stained sections from Veh and Nira-treated mice in (G). Scale bars, 500 μm. Significance was determined by unpaired two-tailed t test. N=8 mice/group. P values are indicated.