Atherosclerotic plaque vulnerability is increased in mouse model of lupus

Anti-apolipoprotein A-1 (anti-apoA-1 IgG) and anti-double stranded DNA (anti-dsDNA IgG) autoantibodies have been described as mediators of atherogenesis in mice and humans. In the present study, we aim to investigate the association between atherosclerotic parameters, autoantibodies and plaque vulnerability in the context of systemic lupus erythematosus (SLE). We therefore bred a lupus prone-mouse model (Nba2.Yaa mice) with Apoe−/− mice resulting in Apoe−/−Nba2.Yaa mice spontaneously producing anti-apoA-1 IgG antibodies. Although Apoe−/−Nba2.Yaa and Apoe−/− mice subject to a high cholesterol diet displayed similar atherosclerosis lesions size in aortic roots and abdominal aorta, the levels of macrophage and neutrophil infiltration, collagen, MMP-8 and MMP-9 and pro-MMP-9 expression in Apoe−/−Nba2.Yaa mice indicated features of atherosclerotic plaque vulnerability. Even though Apoe−/−Nba2.Yaa mice and Apoe−/− mice had similar lipid levels, Apoe−/−Nba2.Yaa mice showed higher anti-apoA-1 and anti-dsDNA IgG levels. Apoe−/−Nba2.Yaa mice displayed a reduction of the size of the kidney, splenomegaly and lymph nodes (LN) hypertrophy. In addition, anti-apoA-1 and anti-dsDNA IgG increased also in relation with mRNA levels of GATA3, IL-4, Bcl-6 and CD20 in the spleen and aortic arch of Apoe−/−Nba2.Yaa mice. Our data show that although atherosclerosis-lupus-prone Apoe−/−Nba2.Yaa mice did not exhibit exacerbated atherosclerotic lesion size, they did show features of atherosclerotic plaque destabilization in correlation with the increase of pro-atherogenic autoantibodies.

Apoe −/− Nba2.Yaa mice have elevated autoantibodies, small kidney, splenomegaly and lymph node hypertrophy. To examine whether the Nba2.Yaa mutation in Apoe −/− mice affects autoantibody production, 11-week-old Apoe −/− and Apoe −/− Nba2.Yaa mice were subjected to HCD for 11 weeks and serum was collected and tested for antibodies against dsDNA and ApoA-1 by ELISA. Levels of anti-dsDNA and anti-ApoA-1 IgG were significantly higher in Apoe −/− Nba2.Yaa mice compared with Apoe −/− mice (Fig. 3a,b). Furthermore, the level of increase for both autoantibodies was correlated (Fig. 3c). Consistently with the vulnerable parameters observed in Apoe −/− Nba2.Yaa mice, PLT and fibrous cap thickness were inversely correlated with the level of anti-dsDNA IgG (Fig. 3d,e). As anti-dsDNA and anti-ApoA-1 IgG production are associated, we observed the same results with PLT, fibrous cap thickness and anti-ApoA-1 IgG level (Fig. 3f,g). Interestingly, the weight of the spleen and lymph nodes increased significantly, while that of the kidney decreased in Apoe −/− Nba2.Yaa mice versus Apoe −/− mice (Fig. 4a,b and Table 1). A correlation between autoantibody titers and the weight of the spleen and lymph nodes was also observed (Fig. 4c,d). However, assessment of kidney function did not suggest the existence of a significant renal dysfunction in Apoe −/− Nba2.Yaa mice compared with Apoe −/− mice, as indicated by the measurement of blood urea nitrogen (BUN) (Fig. 4g). This result is consistent with the negative staining for IgG and IgM deposition in frozen kidney sections (data not shown) and the fact that the Nba2.Yaa mouse model required at least 12 months to fully develop glomerulonephritis.
Apoe −/− Nba2.Yaa mice exhibit Th2, T FH and B cell mRNA markers. Taking into account that SLE is a systemic autoimmune disease, T and B cell subsets were quantified in the secondary lymphoid organs and aortic roots after 11 weeks of HCD. While CD4 mRNA levels were not affected either in the spleen or in the aortic arch of Apoe −/− Nba2.Yaa mice compared with Apoe −/− mice (Fig. 5a,b), gata3 and il4 mRNA expression was significantly increased both in the spleen and aortic arch (Fig. 5c,d). mRNA expression of Th2 marker gata3 was significantly upregulated while il10 increased in the aortic arch of Apoe −/− Nba2.Yaa mice compared with Apoe −/− Scientific Reports | (2020) 10:18324 | https://doi.org/10.1038/s41598-020-74579-8 www.nature.com/scientificreports/ mice in advanced atherosclerosis (Table 2). il10 mRNA expression in the spleen was, however, significantly downregulated in Apoe −/− Nba2.Yaa mice versus Apoe −/− mice on HCD, whereas it was similar in the lymph nodes of both mice strains (Table 3). In parallel the gene expression of Th1, associated gene tim3 prominently increased in the spleen of Apoe −/− Nba2.Yaa mice versus Apoe −/− mice on HCD. bcl6 mRNA was upregulated in the spleen and aortic arch of Apoe −/− Nba2.Yaa mice compared with Apoe −/− mice on HCD (Fig. 5e,f). Consistently, cd20 mRNA gene expression was upregulated both in the spleen and aortic arch of Apoe −/− Nba2.Yaa mice in comparison with Apoe −/− mice in advanced atherosclerosis (Fig. 5g,h). In conclusion, our results suggest that the Nba2.Yaa mutation in Apoe −/− mice leads to the induction of mRNA markers of Th2 and T FH cell populations, which in turn stimulate the induction of CD20 + B cells in Apoe −/− Nba2.Yaa mice on HCD.

Discussion
The pathogenesis of atherosclerosis disease is characterized by systemic inflammation. In this context SLE is associated with a significant risk of cardiovascular disease (CVD) 31 . The Framingham Offspring Study showed that the probability of myocardial infarction (MI) matched for age and gender is over 50 times higher in female SLE patients 32 . In this light, we generated an atherosclerosis-lupus-prone Apoe −/− Nba2.Yaa mouse model to investigate the mechanisms involved in the exacerbation of atherosclerosis in SLE. The present study reveals that atherosclerosis-lupus-prone mice develop vulnerable atherosclerotic plaques in correlation with the production of anti-ApoA-1 IgG and anti-dsDNA IgG antibody levels, both known to have critical pro-atherosclerotic activity 33 .
Macrophages have a key function in the pathogenesis of atherosclerosis by engulfing modified lipoproteins and transforming into lipid-laden foam cells with impaired efferocytosis, leading to necrotic core formation and subsequent plaque destabilization 34 . Pronounced macrophage activation was observed in large cohorts of patients with SLE 35 . Moreover, macrophages in SLE display a defective phagocytic function, enabling an aberrant accumulation of apoptotic debris, leading to the induction of inflammatory responses and breakdown of B cell tolerance 36 . In line with these findings, we found that atherosclerosis-lupus-prone Apoe −/− Nba2.Yaa mice exhibited increased macrophage accumulation in the aorta. One of the features of the Nba2.Yaa lupus-prone mouse is an increase in circulating monocyte counts which could explain the observed increase in plaque macrophages, without affecting plaque size 26 . Consistently, recent publications have shown that more than 50% of foam cells responsible for lipid deposition in the plaque arise from a phenotypical switch of smooth muscle cells 37 . However, the accumulation of macrophages can result in the build-up of uncleared apoptotic bodies, an important source of autoantigens and inflammation which in turn can promote plaque destabilization. www.nature.com/scientificreports/ Neutrophils have been shown to have a deleterious effect in atherosclerosis, promoting plaque instability 38,39 . Furthermore, neutrophil extracellular traps (NETs) formation promotes cytotoxic and prothrombotic effects implicated in arterial thrombosis 40,41 , endothelial cell death, lipoprotein modification, and inflammasome activation [42][43][44] . Moreover, neutrophils from SLE patients exhibit apoptosis and secondary necrosis 45 . Interestingly, we found that neutrophils accumulation in the aortic roots was reduced in Apoe −/− Nba2.Yaa mice. This could be an effect of neutrophil cell death known to be associated with SLE disease activity and elevated anti-dsDNA antibody levels 46 .
The ruptured plaques are often characterized by a thin and inflamed fibrous cap covering a necrotic core. The matrix metalloproteinases destabilize atherosclerotic plaques through the degradation of elastin, fibronectin, laminin and collagen in correlation with an increased rate of coronary events 47,48 . Apoe −/− Nba2.Yaa mice exhibited a pronounced increase in MMP-9 expression in the atherosclerotic roots and increased systemic levels of MMP-8 in parallel with a prominent reduction of collagen content in the atherosclerotic roots (Fig. 2). Collagen content is known to be critically important in preventing plaque rupture 49 . The matrix-degrading proteases MMP-8 and MMP-9 degrading components of the extracellular matrix are abundantly expressed in atherosclerotic plaques with a vulnerable histological appearance 47,48 . MMP-9 is a protein expressed and secreted in an inactive form named pro-MMP-9, which is then activated by proteolysis of the propeptide domain 50 . The level of pro-MMP-9 was reduced in the serum of Apoe −/− Nba2.Yaa mice, indicating an increase in the proteolysis of the pro-peptide MMP-9 in the serum of the Apoe −/− Nba2.Yaa mice in advanced atherosclerosis. CCL2, which is elevated in Apoe −/− Nba2.Yaa mice, attracts not only monocytes, but also T lymphocytes and NK cells 51 . Moreover, CCL2 is also known to contribute to plaque destabilization by exerting prothrombotic and inflammatory effects 52 . Taken together, the present findings indicate that plaque vulnerability in atherosclerosis-lupus-prone mouse model is increased, thereby providing a potential explanation for the higher prevalence of CV events in patients with SLE 1,2 .
Anti-ApoA-1 IgG antibodies initially found in SLE patients were associated with a higher prevalence and incidence of CAD with a worse prognosis, independently of traditional CV risk factors observed in autoimmune and non-autoimmune settings 6,12-18 . Anti-ApoA-1 IgG antibodies are described as active modulators of www.nature.com/scientificreports/ atherothrombosis and linked to a higher incidence of CAD [19][20][21][22]53 . Lupus-prone Nba2.Yaa mice produce high levels of nucleolar-related autoantibodies and exhibit splenomegaly in addition to a lethal form of lupus nephritis that causes 50% mortality by 14 months of age 24 . In agreement with this finding our atherosclerosis-lupus-prone mouse model showed spleen and lymph node hypertrophy, as well as elevated levels of anti-ApoA-1 antibodies which mediate pro-inflammatory, pro-arrhythmogenic and pro-thrombotic effects 33 . These results could be an  www.nature.com/scientificreports/ indication for a possible mechanism explaining the prominently increased of MI probability in SLE patients 32 . Furthermore, while one feature of atherosclerosis development is the dead cell accumulation leading to genetical material release, the amount of dsDNA correlates with the occurrence of CV events 28 . dsDNA is recognized by the Absent in melanoma 2 (Aim2) inflammasome found in human atherosclerotic lesions in proximity to the necrotic cores 27 , whose activation induces the release of the pro-inflammatory cytokines IL-1β and IL-18 54 . A recent study  www.nature.com/scientificreports/ demonstrated that the dsDNA Aim2 axis triggered a powerful cytokine response in lesional macrophages, thereby enhancing atherosclerotic lesion destabilization 55 . The level of anti-dsDNA IgG in Apoe −/− Nba2.Yaa mice could therefore be directly associated to unstable plaque vulnerability markers in Apoe −/− Nba2.Yaa mice on HCD. T cell response in SLE is very complex, and the Th2 cell polarization observed in Apoe −/− Nba2.Yaa mice in the present study might be associated with polyclonal B cell activation seen in SLE 56,57 . Moreover, it has been proposed that Th2 and the T FH cytokine IL-4, which is upregulated in the spleen and arch of the Apoe −/− Nba2.Yaa   www.nature.com/scientificreports/ mice, could rescue B cells from apoptosis and promote autoreactive B lymphocyte survival 58 . Consistently with our present findings, IL-4 treatment triggered the production of anti-dsDNA 58 antibodies and could thus increase the CD20 + B cell population in Apoe −/− Nba2.Yaa mice on HCD. In SLE murine models, IL-4 knockout mice produced less IgG1 and IgE serum Ig, suggesting a major role of this cytokine in the pathogenesis of the disease 59 . SLE is a systemic autoimmune disease that is known to be associated with polyclonal B cell hyperreactivity with a potential overactive germinal center (GC) and ectopic follicular activity enhancing memory B cells and plasmacytosis 60 . Maturation of the GC and the production of antibodies is dependent on Bcl-6 that is required for follicular T cell differentiation 61 , and Bcl-6 expression was significantly increased in the aortic arch and the spleen of Apoe −/− Nba2.Yaa mice (Fig. 5). In agreement with this finding, our results indicate that the Nba2.Yaa mutation in Apoe −/− mice resulted not only in Th2 cell polarization, but also T FH polarization. Thus, in turn the T FH cells stimulated the induction of CD20 + B cells and subsequently elevated anti-dsDNA and apoA-1 IgG antibody levels, leading all together to the exacerbation of atherosclerotic plaque vulnerability in Apoe −/− Nba2.Yaa mice on HCD. IL-10 is a cytokine known to be essential for regulating the immune response. However, IL-10 also improves B lymphocyte proliferation, Ig class switching and increases antibody secretion 62 . Interestingly, patients with SLE have high levels of IL-10 that are correlated both with the level of disease activity and anti-dsDNA antibody production 63 . In murine models, IL-10 blockade has shown to limit the renal damage and decrease the production of anti-dsDNA antibodies 63 . Although the augmentation of IL-10 was observed in the arch but not in the secondary lymphoid organ in Apoe −/− Nba2.Yaa mice on HCD, Srikakulapu and colleagues have, however, shown that autoantigen-dependent hypermutation, proliferation, affinity maturation, Ig class switching, memory cell generation, and differentiation into long-lived plasma cells may be carried out in the arterial wall 64 . They also speculate that immature B cells home to artery tertiary lymphoid organs (ATLOs) to undergo differentiation into mature B cells in the absence of the proper control mechanisms acting in the spleen, leading to the production of IgG from autoreactive atherosclerosis and SLE-specific B cells 64 . Altogether, these findings are consistent with the present study, showing a substantial increase of IL-10 expression in the arch of Apoe −/− Nba2.Yaa mice in parallel with increased levels of CD20 + B cells and anti-dsDNA antibodies.
In conclusion, our atherosclerosis-lupus-prone mouse model revealed a mechanism promoting atherosclerotic plaque vulnerability. The Nba2.Yaa mutation in Apoe −/− mice was associated with macrophage accumulation, plaque destabilization, the build-up of uncleared apoptotic bodies and subsequent induction of pro-atherogenic antibodies. In parallel, Apoe −/− Nba2.Yaa mice exhibited atherosclerotic plaque destabilization through a potential Th2 and T FH polarization known to stimulate the induction of CD20 + B cells and caused anti-dsDNA and apoA-1 IgG autoantibody production. Mice. B6.Nba2.Yaa mice were generated as described 65 . The Apoe −/− null mutation was introduced in B6.Nba2.Yaa mice by breeding. Eleven-week old Apoe −/− C57Bl/6 and Apoe −/− Nba2.Yaa mice were subjected to 11 weeks of high cholesterol diet (HCD) (20.1% fat, 1.25% cholesterol, Research Diets, Inc., New Brunswick, NJ), as a model of advanced atherosclerosis. The treatments and atherosclerosis protocols were well-tolerated by the mice, and no adverse events (such as weight loss and signs of systemic toxicity) were reported. At sacrifice, haematological parameters, serum triglycerides, total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), free fatty acids and glucose were routinely measured and expressed in mmol/l. Animals were euthanized by exsanguination after anesthesia with 4% isoflurane. All breeding and experimental protocols and procedures were reviewed and approved by the Institutional Animal Care and Use Committee of the Geneva University School of Medicine. Animal care and experimental procedures were carried out in accordance with the guidelines of the Institutional Animal Care and Use Committee of the Geneva University School of Medicine.

Measurements of serum inflammatory molecule levels.
Colorimetric enzyme-linked immunosorbent assay (ELISA) kits to measure serum CCL2, CXCL-1, pro-matrix metalloproteinase (MMP)-9, myeloperoxidase (MPO) and tissue inhibitor of metalloproteinase-1 (TIMP-1) levels (all from R&D Systems), and serum MMP-8 levels (Uscn Life Science Inc., Hubei, China) were used following manufacturer's instructions. Detection and quantification of the levels of urea nitrogen (BUN) in the serum were performed with Urea Nitrogen Colorimetric Detection Kit (Thermo Fisher Scientific) according to the manufacturer's instructions.
Determination of autoantibodies anti-apoA-1 by ELISA. Maxisorp plates (NuncTM, Denmark) were coated with purified, derived delipidated murine recombinant apolipoprotein A-1 (Biorbyt, United Kingdom) (20 mg/ml; 50 ml/well) for 1 h at 37 °C. After washing, all wells were blocked for 1 h with 2% bovine serum albumin (BSA) in phosphate buffer solution (PBS) at 37 °C. Then, samples were incubated for 1 h. Samples were also added to a non-coated well to assess individual non-specific binding. After washing 50 μl/well of signal antibody (alkaline phosphatase-conjugated anti-human IgG; Sigma-Aldrich, St Louis, MO) dilute 1:1000 in PBS/ BSA 2% solution was incubated 1 h at 37 °C. After washing, phosphatase substrate p-nitrophanyl phosphate disodium (Sigma-Aldrich) dissolved in diethanolamine buffer (pH 9.8) was added. Each sample was tested in duplicate and absorbance in optical densities (OD) was determined at 405 nm after 20 min of incubation at 37 °C (molecular DevicesTM Filtermax). Corresponding non-specific binding was subtracted from mean absorbance for each sample.
Determination of autoantibodies anti-dsDNA by ELISA. Salmon Sperm dsDNA was coated to ELISA plates precoated with poly L lysine (Sigma-Aldrich). Plates were then incubated with 1/100 diluted serum samples, and development performed with alkaline phosphatase-labelled goat anti-mouse IgM or IgG. Results are expressed in U/mL in reference to a standard curve.
Statistical analysis. Statistics were performed using GraphPad Prism 8, Statistica (version 13.0) and the SPSS statistical package (version 20.0). For comparison of two groups of continuous variables, two-tailed unpaired Mann-Whitney U-tests with a confidence level of 95% were conducted if data were non-normally distributed. Continuous variables were expressed as median (interquartile range [IQR]) and were compared between the two groups by nonparametric Mann-Whitney U test. Paired intergroup comparisons were performed by using Wilcoxon test. Spearman's rank correlation coefficients were used to assess correlations between variables. The number of mice used for each analysis is indicated in the figure legends. All data are presented as the mean ± SEM and the statistical significance threshold used is *p ≤ 0.05. **p ≤ 0.05; ***p ≤ 0.005.