Conditional deletion of Ccl2 in smooth muscle cells does not reduce early atherosclerosis in mice

Background and aims C–C motif chemokine ligand 2 (CCL2) is a pro-inflammatory chemokine important for monocyte recruitment to the arterial wall and atherosclerotic plaques. Global knockout of Ccl2 reduces plaque formation and macrophage content in mice, but the importance of different plaque cell types in mediating this effect has not been resolved. Smooth muscle cells (SMCs) can adopt a potentially pro-inflammatory function with expression of CCL2. The present study aimed to test the hypothesis that SMC-secreted CCL2 is involved in early atherogenesis in mice. Methods SMC-restricted Cre recombinase was activated at 6 weeks of age in mice with homozygous floxed or wildtype Ccl2 alleles. Separate experiments in mice lacking the Cre recombinase transgene were conducted to control for genetic background effects. Hypercholesterolemia and atherosclerosis were induced by a tail vein injection of recombinant adeno-associated virus (rAAV) encoding proprotein convertase subtilisin/kexin type 9 (PCSK9) and a high-fat diet for 12 weeks. Results Unexpectedly, mice with SMC-specific Ccl2 deletion developed higher levels of plasma cholesterol and larger atherosclerotic plaques with more macrophages compared with wild-type littermates. When total cholesterol levels were incorporated into the statistical analysis, none of the effects on plaque development between groups remained significant. Importantly, changes in plasma cholesterol and atherosclerosis remained in mice lacking Cre recombinase indicating that they were not caused by SMC-specific CCL2 deletion but by effects of the floxed allele or passenger genes. Conclusions SMC-specific deficiency of Ccl2 does not significantly affect early plaque development in hypercholesterolemic mice.


Introduction
Atherosclerosis is a low-grade, non-resolving inflammatory disease driven by retention of low-density lipoproteins (LDL) in the arterial wall [1].Inhibition of inflammatory pathways reduces plaque progression in experimental models and protects against the incidence of clinical atherosclerotic events in patients with prior myocardial infarction [2,3].Blocking inflammation broadly in the body, however, adversely affects defenses against infection.Research is needed to identify mechanisms that locally control inflammatory cells to devise potent anti-inflammatory therapies that are specific or selective for atherosclerosis [1,4].
C-C motif chemokine ligand 2 (CCL2) is a pro-inflammatory chemokine secreted by multiple cell types, which is central in the recruitment of monocytes to tissues and has a clear causal role in atherosclerosis [4].Global deficiency of Ccl2 in Apoe − /− mice, Ldlr − /− mice, and mice overexpressing apoB consistently inhibited atherogenesis and reduced plaque content of macrophages [5][6][7].Blocking Ccl2 activity in mice with advanced atherosclerosis reduced further plaque progression and changed plaque composition to contain fewer macrophages, more SMCs, and more collagen [8].These studies indicate a continued role of CCL2 that extends beyond disease initiation.CCL2 is also expressed in human atherosclerotic plaques and associated with features of plaque vulnerability [9], and CCL2 polymorphisms and elevated CCL2 plasma levels are associated with an increased risk of cardiovascular disease and mortality [9][10][11].
Expression in myeloid cells accounts for a part of the atherosclerosispromoting effect of CCL2 in mice [12].However, single-cell gene expression profiling of atherosclerotic arteries and other studies has revealed widespread expression of CCL2 in SMC lineage cells, suggesting that SMC-derived CCL2 may carry some of the proatherogenic effect [13][14][15].Two recent reports, published while we were finalizing the present study, investigated this in Apoe − /− mice but reached different conclusions.One report found a pro-atherogenic role of locally secreted CCL2 from plaque SMC-derived cells, which in some contexts was offset by an inhibitory effect of bone marrow SMC/pericyte-derived CCL2 on monocytosis [16].The other found exacerbated plaque inflammation and reduced cap SMCs after SMC-specific deletion of CCL2, and suggested that SMC-derived CCL2 is crucial for maintaining niches of homeostatic macrophages in the plaque that dampen chronic inflammation [14].
In the present study, we assessed the importance of SMC-secreted CCL2 for early atherosclerosis in murine atherosclerosis induced by PCSK9 overexpression.We find no effect of SMC-derived CCL2 on plaque development in this model when accounting for background effects of the engineered Ccl2 allele or its passenger gene variants.

Materials and methods
Please see an expanded methods section in the Supplemental Material.

Mice
Male Myh11-CreER T2 mice (B6⋅FVB-Tg(Myh11-cre/ERT2)1Soff/J, stock no.019079) expressing tamoxifen-inducible Cre recombinase (CRE-ERT2) under the SMC-specific myosin heavy chain 11 (Myh11) promotor [17], and Ccl2 flx/flx mice (B6N⋅129S1(FVB)-Ccl2tm1.2Tyos/J, stock no.023347) with loxP sites flanking exon 1-2 of the Ccl2 gene [18] were acquired from Jackson and intercrossed.Experimental mice of our study were male offspring of breeders heterozygous for the floxed Ccl2 allele.Only males were included because the Myh11-CreER T2 transgene is inserted on the Y chromosome.One of the male breeders was found without the Y-linked Myh11-CreER T2 , presumably due to the exchange in genetic material between pseudoautosomal regions of X and Y chromosomes [19,20], and was used to breed CreER T2 -negative male mice for control experiments.Animals were housed in groups with unlimited access to tap water and a standard laboratory diet in an enriched environment with standard bedding and nesting material under a 12/12 h day-night cycle in a temperature (20-25 • C) controlled facility.
At 6-8 weeks of age, male Myh11-CreER T2 x Ccl2 flx/flx and Myh11-CreER T2 x Ccl2 wt/wt littermates were injected i.p. with 1 mg tamoxifen (T5648, Sigma-Aldrich) dissolved in corn oil (P2144, Sigma-Aldrich) daily for 10 days to obtain mice with or without SMC-specific Ccl2 deletion (Ccl2 SMC− KO and Ccl2 SMC− WT ).Male Ccl2 flx/flx and Ccl2 wt/wt littermates were treated similarly.At 12 weeks of age, hypercholesterolemia was induced by a single tail vein injection of rAAV8-D377Y-mPCSK9 virus particles (rAAV-PCSK9, 1 × 10 11 vector genomes, produced at Vector Core at the University of North Carolina, Chapel Hill, North Carolina, USA) and shifted to a high-fat diet (D12079B, Research Diets Inc.) containing 21 % fat and 0.2 % cholesterol [21].After 12 weeks of feeding atherosclerosis-promoting diet, the mice were anesthetized with 5 mg pentobarbital and 4 mg lidocain i.p. and terminated by exsanguination.The arterial tree was perfused with Cardioplex for 30 s and 4 % phosphate-buffered formaldehyde (PFA) for 5 min at 100 mmHg through the left ventricle, followed by immersion-fixation in 4 % PFA for 24 h.Animal procedures were approved by The Danish Animal Experiments Inspectorate (license 2015-15-0201-00542) and conducted at Aarhus University following ARRIVE guidelines.
Aortae were cleaned of all periadventitial fat, cut open longitudinally, stained with Oil Red O for 10 min at 37 • C, washed in isopropanol and phosphate buffered saline (PBS), and mounted on a microscope slide with Aquatex for en face quantification of lesion coverage in the aortic arch down to the supreme intercostal artery.Analysis of the scanned slides was performed in Fiji.

Human atherosclerotic plaques
Human atherosclerotic plaques from the left anterior descending artery obtained from a previously published material [23] were used for immunohistochemical analysis of CCL2 localization using rabbit polyclonal anti-CCL2 (Lifespan Bioscience, cat.no.LS-B10540, 1:50) and Envision + System-HRP (DAB) (Dako, cat.no.K4010).The original material was collected after approval by the regional ethical committee [23].It was subsequently anonymized and did not require further ethical approval for the use in the present study.

Statistics
All statistical analysis was performed using GraphPad Prism 9 (GraphPad Software Inc.).Data was checked for normal and log-nomal distribution and equal variance and analyzed using the tests indicated in figure legends.Multiple regression analysis was used to analyze associations with plaque data using genotype and plasma cholesterol as independent variables and associations with plasma cholesterol levels (using genotype and plasma PCSK9 concentrations as independent variables).P values below 0.05 were defined as statistically significant.Data are shown as mean ± standard error of the mean (SEM) as indicated in each figure legend.The number of animals/samples analyzed is stated in figure legends.In some analyses, data points were not obtained for all mice because of technical failures.All procedures and analyses were performed blinded for the mouse genotype.

CCL2 is expressed in human atherosclerotic lesions
To evaluate the expression pattern of CCL2 during human atherogenesis, we performed CCL2 immunohistochemistry in human coronary lesions of different severity (Fig. 1A).CCL2 was found expressed in normal human coronary artery in both the medial and intimal layer.In early lesions, CCL2 was present in foam cells, but also in the underlying artery wall.In advanced atherosclerotic lesions, CCL2 was detected in areas adjacent to the necrotic core, while CCL2 expression in the medial layer was diminished.Staining of adjacent sections with ACTA2 showed considerable overlap with the pattern of CCL2 staining suggesting widespread SMC expression of CCL2 in normal media/intima as well as in developing plaques.

Human and murine SMCs in culture can be stimulated to express CCL2
Previous studies have reported CCL2 secretion from murine SMCs upon stimulation [24].To confirm this, we stimulated human VSMCs and a murine SMC line (MOVAS) with IL-1β and found that the secretion of CCL2 in the conditioned media of both increased dose-dependently (Fig. 1B and C).Also, assessed by flow cytometry, we found an increase in protein expression of CCL2 in response to IL-1β stimulation of MOVAS cells (Fig. 1D).

SMC-specific conditional knockout of Ccl2 in hypercholesterolemic mice
Modulation to a CCL2-secreting SMC phenotype by local IL-1β or other pro-inflammatory stimuli could facilitate ongoing inflammation in the developing atherosclerotic lesion.To test this hypothesis in vivo, we crossed mice with SMC-restricted CreER T2 expression and either homozygous floxed or wildtype Ccl2 alleles to obtain mice with SMCspecific deletion of Ccl2 (Ccl2 SMC− KO ) and littermate controls (Ccl2 SMC− WT ) (Fig. 2A).Mean allelic recombination efficiency was 75.9 ± 1.3 % (mean ± SEM) (Fig. 2B).Because recombination was determined in aortic samples after mechanical removal of the adventitia and endothelium, the analyzed cells may have contained a small amount of non-SMC cell types causing some underestimation of recombination efficiency.

Increased plasma cholesterol levels in Ccl2 SMC− KO mice
Atherosclerosis was induced in Ccl2 SMC− KO and Ccl2 SMC− WT mice by rAAV-PCSK9 injection followed by feeding a high-fat diet for 12 weeks (Fig. 2C).Plasma cholesterol was similar in the two genotypes after 2 weeks but significantly higher in Ccl2 SMC− KO mice than Ccl2 SMC− WT mice at both 6 and 11 weeks (Fig. 3A).Analysis of size-fractioned lipoproteins measured at 12 weeks showed that the increased cholesterol in Ccl2 SMC− KO mice was in apoB-containing lipoproteins (VLDL and LDL) (Fig. 3B).
Sterol analysis was performed to provide clues to the source of the increased plasma cholesterol in Ccl2 SMC− KO mice (Fig. 3C) [25].Markers for the rate of cholesterol synthesis (lathosterol and desmosterol) and intestinal cholesterol absorption (campesterol, stigmasterol, sitosterol, lanosterol) were either unchanged or reduced in plasma of Ccl2 SMC− KO compared with Ccl2 SMC− WT mice.These observations instead pointed towards a potential difference in LDL clearance, which could be caused by a difference in rAAV8 transduction efficiencies leading to differences in PCSK9 levels.We therefore measured plasma PCSK9 levels in samples taken at 2 and 11 weeks after rAAV8.Plasma PCSK9 levels increased from 2 to 11 weeks, but no statistically significant differences were found between Ccl2 SMC− KO and Ccl2 SMC− WT mice (Fig. 3D).Furthermore, introducing plasma PCSK9 levels in multiple regression analysis did not weaken the link between genotype and plasma cholesterol levels (Supplemental Table 1).
Although the floxed Ccl2 mice were already backcrossed into the B6 mouse background 10 times when we acquired them and further 3-4 times in our facility, a region around the Ccl2 locus will be derived from the 129/SVJ embryonic stem cells in which the Ccl2 gene targeting was performed [26].This passenger gene region could potentially harbor plasma cholesterol regulating loci.Furthermore, the genetic modification of the Ccl2 locus could affect neighboring gene expression in any cell type, which in theory could impact lipoprotein metabolism.To explore this, we also analyzed plasma lipids in CreER T2 -negative Ccl2 flx/flx and Ccl2 wt/wt mice that had been subjected to a similar regimen of rAAV8-PCSK9 treatment and high-fat diet.We found a numerical increase in plasma cholesterol levels in Ccl2 flx/flx compared with Ccl2 wt/wt at the last time point (Fig. 3E).Although this difference was not significant (p = 0.08), it is consistent with genetic background effects, rather than the SMC-specific Ccl2 deletion, being the underlying cause of the differences in plasma cholesterol.

Atherosclerosis development
Atherosclerosis development was assessed in sections of the aortic root and by en face analysis of the aortic arch after 12 weeks of hypercholesterolemia (Fig. 4A and B).Aortic root plaque area was significantly increased in Ccl2 SMC− KO compared with Ccl2 SMC− WT mice (Fig. 4C).In the aortic arch, the lesion coverage showed an insignificant tendency towards an increase in Ccl2 SMC− KO compared with wildtype littermates (Fig. 4D).
Aortic root cross-sections were analyzed for plaque cellular composition using the macrophage marker LGALS3 and the SMC-marker ACTA2 (Fig. 5A).LGALS3 can be expressed on SMC lineage cells in addition to macrophages.Still, we have previously found that the level of misclassification by standard immune fluorescent staining in PCSK9induced atherosclerosis is minor as assessed by SMC lineage tracing [26].The number of LGALS3+ cells was increased in Ccl2 SMC− KO compared with wildtype littermates (Fig. 5B), while the number of ACTA2+ cells was unaffected (Fig. 5C).
Murine atherosclerotic lesions also contain a large population of SMC lineage cells that lose expression of ACTA2 and modulate to other forms [27].We therefore also quantified ACTA2/LGALS3 double-negative cells (ACTA2-/LGALS3-), which are enriched for modulated plaque SMCs, but found no differences between the two groups (Fig. 5D).
None of the differences in atherosclerosis measures between groups were statistically significant in a multiple regression model incorporating final plasma cholesterol (data not shown) or the area-under-thecurve of plasma cholesterol levels from 2 until 11 weeks (Supplemental Tables 2 and 3).This indicates that the observed changes were explained by the unexpected differences in plasma cholesterol between groups, which was unrelated to the deletion of Ccl2 in SMCs.This was confirmed by analysis of aortic root plaque area in CreER T2 -negative Ccl2 flx/flx and Ccl2 wt/wt mice (Fig. 4E).As observed for the Ccl2 SMC− KO mice, the plaque area was significantly increased in Ccl2 flx/flx mice compared to their wildtype counterparts, supporting the conclusion that secretion of CCL2 by modulated SMCs in the plaque is not an important mediator in plaque development.

Discussion
SMCs have traditionally been considered to have a protective role in atherosclerosis by producing fibrous tissue that stabilizes the plaque and protects it from rupture and thrombotic complications [28].Over the last decade, however, multiple studies have revisited the role of SMCs in atherosclerosis using mouse models and lineage-tracing techniques [27].Such experiments have shown that plaque SMCs derive from a small group of arterial SMCs that undergo clonal expansion and modulate to cells with fibroblast-like (aka.fibromyocytes) and chondrocyte-like (aka.chondromyocytes) phenotypes [29][30][31][32][33].Some of these show evidence of inflammatory signaling [34], but the causal role of this for plaque development is still sparsely understood.
In the present study, we tested whether the ability of SMCs to secrete the proinflammatory cytokine CCL2 is a mediator linking SMC modulation to arterial inflammation and early plaque development.CCL2 is a central cytokine involved in monocyte recruitment to tissues, and global knockout of Ccl2 reduces plaque formation and progression in mice [5][6][7]12].We confirmed the expression of CCL2 in human plaques in SMC-rich areas including the arterial media in the pre-diseased arterial wall.Furthermore, we confirmed the ability of both human and mouse SMCs to respond to an inflammatory stimulus with CCL2 expression.To Unexpectedly, we found increased hypercholesterolemia and atherosclerosis in mice with SMC-specific deficiency of Ccl2 compared with wildtypes littermates.Importantly, however, a critical control experiment in mice lacking the Cre recombinase showed that this paradoxical effect was not caused by the knockout of Ccl2 in SMCs.Rather it was explained by genetic differences between the floxed and wildtype Ccl2 allele; either the floxing of the gene itself or passenger gene variants in linkage disequilibrium with the Ccl2 locus originating from the 129/ SVJ embryonic stem cells used to create the mouse line [18,35].
The genetic differences appeared to accelerate atherosclerosis at least partly by increasing hypercholesterolemia. Plasma cholesterol was significantly increased in Ccl2 SMC− KO compared with Ccl2 SMC− WT mice, which was unexpected since previous studies of mice with global CCL2 knockout or overexpression did not show any effects on plasma cholesterol levels [5][6][7]12].Although the differences in plasma cholesterol levels in Ccl2 flx/flx and Ccl2 wt/wt mice (without the Cre transgene) were not statistically significant, they resembled those of the main study.The differences were not caused by differences in the efficiency of rAAV-PCSK9 transduction since the effect was not present after 2 weeks but developed over time and plasma PCSK9 levels were similar between groups in the two genotypes.
Recently, Owsiany et al. reported increased atherosclerosis and macrophage recruitment in Apoe − /− mice with heterozygous, but not homozygous, knockout of Ccl2 in SMCs [16].The increased atherosclerosis in heterozygous animals appeared to result from increased monocytosis in combination with a preserved ability to recruit monocytes into the arterial intima, which was lost in homozygous animals [16].Recombination rates were not quantified in the study of Owsiany et al. and hence it is difficult to compare our mice (with an app.75 % recombination rate of the floxed Ccl2 locus) with the two genotypes reported in that study.Notably, even though our findings may on the surface seem consistent with the increased atherosclerosis reported in the heterozygous mice in Owsiany et al., the proatherogenic effect in our experiments, as discussed above, was not caused by Ccl2 deletion.It is unlikely that the specific genetic background effects we saw in our study influenced the results of Owsiany et al., because they used a floxed Ccl2-mCherry mouse line, which was created using C57BL/6J embryonic stem cells [16,36].In this line, mCherry is transcribed with Ccl2 through a 2A aphtovirus cleavage site, and both transcripts are disrupted upon recombination.Pekayvaz et al. used the same Ccl2-mCherry line but found increased atherosclerotic plaque burden without monocytosis in mice with homozygous SMC-specific Ccl2 deletion compared with littermate controls [14].Reconciling the different findings from now three studies of SMC-specific Ccl2 knockout is complicated.We think it reflects not only the complexity of the biology of SMC/pericyte-derived CCL2, which has different functions in the arterial wall, bone marrow, and potentially other organs, but also the complexity of the Cre-LoxP mouse technology and its pitfalls.To our knowledge, it has not been tested whether the Ccl2-mCherry allele (without recombination) affects atherosclerosis, as we found for the floxed Ccl2 allele in the present study.This cannot be excluded since we have previously found that some SMC-expressed fluorescent proteins by themselves inhibit atherosclerosis [30].
Overall, we conclude that the knockout of Ccl2 in SMCs in our experimental context did not change atherogenesis.This may indicate that SMC-secreted CCL2 is not an important regulator of plaque inflammation.Alternatively, it may be the result of multiple opposite effects of SMC-derived CCL2 on atherosclerosis that outbalance each other.It should also be noted that upregulation of both Ccl7 (Mcp3) and Ccl12 (Mcp5) located downstream of the Ccl2 gene was originally demonstrated for the mouse line used in our studies upon Cre-mediated deletion of the floxed Ccl2-region [18].Given that both encoded proteins are also ligands for the C-C chemokine receptor type 2 (CCR2), potential compensating effects from these genes cannot be excluded.

Strengths and limitations
Our study has limitations.Recombination of CCL2 alleles in SMCs was not complete as is generally the case in conditional knockout studies, and it is possible more robust inhibition of CCL2 secretion from SMCs would have affected plaque development.Despite multiple attempts with different antibodies, we did not succeed in performing immunostaining for CCL2 to evaluate whether the loss of SMC-derived CCL2 was consequential for total CCL2 content in plaques.Finally, our studies were not designed to investigate the impact of SMC-specific knockout on monocyte trafficking, which would have been relevant on the background of the recent opposing findings by Owsiany et al. [16] and Pekayvaz et al. [14].

Conclusion
Conditional deletion of Ccl2 in SMCs does not inhibit the development of atherosclerosis in mice with PCSK9-induced hypercholesterolemia.

Fig. 1 .
Fig. 1.CCL2 expression in SMC-rich regions of human plaques and CCL2 secretion in SMC cultures.(A) CCL2 (brown color) and ACTA2 (red color) detected by immunohistochemistry in adjacent sections of normal human artery wall and atherosclerotic lesions of different severity.CCL2 is expressed in media and intima of the normal arterial wall.CCL2 is also expressed in foam cells and underlying medial SMCs, and in active inflammatory areas adjacent to the necrotic core in the fibroatheroma.F, foam cells.I, Intima.M, Media, NC, necrotic core.Scale bars = 100 μm.CCL2 is secreted from cultured SMCs stimulated with IL-1β.CCL2 levels measured in conditioned media (CM) from (B) MOVAS cells (n = 6 replicates per group) and (C) hVSMCs (n = 5-6 replicates per group) treated with IL-1β for 24 h.(D) Flow cytometry measurement of intracellular CCL2 accumulation in MOVAS cells after blocking the secretory pathway by Brefeldin-A.IL-1β stimulation (1 ng/ ml) increases CCL2 accumulation measured as median fluorescent intensity (n = 3 replicates per group).Graphs show mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by Dunnett's T3 multiple comparisons test following Brown-Forsythe ANOVA (B,C; individual IL-1β dosage groups compared with control) and unpaired Student's t-test (D).(For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2 .
Fig. 2. Experimental design.(A) Illustration of the genetic alterations in Myh11-CreER T2 x Ccl2 flx/flx mice that lead to SMC-specific knockout of Ccl2 after tamoxifen injection.Black arrows flanking the left loxP site indicate primers used for detecting recombination by qPCR.(B) Recombination efficiency was measured as the reduction in (non-recombined) floxed CCL2 alleles in genomic DNA of the aortic media in Myh11-CreER T2 x Ccl2 flx/flx (n = 21, grey bar) compared with Ccl2 flx/flx (n = 15, black bar) mice.The graph shows mean ± SEM. ****p < 0.0001 by unpaired Welch t-test.(C) Schematic illustration of the mouse experiment showing the timing of tamoxifen injections (10 daily doses between 6 and 8 weeks of age), induction of LDL receptor-deficiency by a single tail-vein injection of rAAV8-D377Y-mPCSK9 at 12 weeks of age, and final analysis after feeding Western-type diet for 12 weeks.Blood samples were collected 2, 6, and 11 weeks after rAAV-PCSK9 injection.

Fig. 3 .
Fig. 3. Plasma cholesterol phenotype.(A) Plasma cholesterol levels were initially similar but significantly increased in Ccl2 SMC− KO (n = 22, orange circles) compared with Ccl2 SMC− WT (n = 24, blue squares) mice at 6 and 11 weeks after rAAV-PCSK9 injection.(B) Analysis of size-fractionated lipoproteins by FPLC at 12 weeks after rAAV-PCSK9 injection revealed increases in cholesterol in VLDL-and LDL-sized, but not HDL-sized, lipoproteins.(C) Sterol analysis for markers of cholesterol biosynthesis (lathosterol and desmosterol) or cholesterol absorption (campesterol, stigmasterol, sitosterol, lanosterol) showed no significant changes or changes opposite the direction that could explain the increased plasma cholesterol levels in Ccl2 SMC− KO (n = 23) compared with Ccl2 SMC− WT mice (n = 21).Plasma PCSK9 levels were not different between Ccl2 SMC− WT (n = 22) and Ccl2 SMC− KO mice (n = 24) (D).(E) Plasma cholesterol levels were numerical increased, though not statistically significant, in Cre-negative Ccl2 flx/flx (n = 15, black squares) compared with Ccl2 wt/wt (n = 17, grey triangles) mice 11 weeks after rAAV-PCSK9 injection.Graphs show mean ± SEM. *p < 0.05; **p < 0.01; NS, not statistically significant in a mixed-effects model for repeated measurements with Šidák's multiple comparison test (A,D,E) and Welch t-test (C).(For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4 .Fig. 5 .
Fig. 4. Quantification of atherosclerosis after 12 weeks of atherogenic diet feeding.Representative images of (A) hematoxylin-eosin-stained cross sections from the aortic root, and (B) Oil Red O-stained aortic arches opened for en face quantification.Scale bars = 400 μm (A) and 2 mm (B).(C) Plaque size was statistically significantly increased in Ccl2 SMC− KO (n = 14, orange circles) mice compared with Ccl2 SMC− WT (n = 14, blue squares) mice in the aortic root.(D) In the aortic arch, a similar tendency was seen but not statistically significant (Ccl2 SMC− KO , n = 24; Ccl2 SMC− WT , n = 21).Aortic root plaque size was also statistically significantly increased in Cre-negative Ccl2 flx/flx (black squares, n = 12) mice compared with Ccl2 wt/wt (grey triangles, n = 10) (E).Graphs show mean ± SEM. *p < 0.05, NS, not statistically significant by Welch t-test on log-transformed data.(For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)