Increased Inflammation in Atherosclerotic Lesions of Diabetic Akita-LDLr −/− Mice Compared to Nondiabetic LDLr −/− Mice

Background. Diabetes is associated with increased cardiovascular disease, but the underlying cellular and molecular mechanisms are poorly understood. One proposed mechanism is that diabetes aggravates atherosclerosis by enhancing plaque inflammation. The Akita mouse has recently been adopted as a relevant model for microvascular complications of diabetes. Here we investigate the development of atherosclerosis and inflammation in vessels of Akita mice on LDLr −/− background. Methods and Results. Akita-LDLr −/− and LDLr −/− mice were fed high-fat diet from 6 to 24 weeks of age. Blood glucose levels were higher in both male and female Akita-LDLr −/− mice (137% and 70%, resp.). Male Akita-LDLr −/− mice had markedly increased plasma cholesterol and triglyceride levels, a three-fold increase in atherosclerosis, and enhanced accumulation of macrophages and T-cells in plaques. In contrast, female Akita-LDLr −/− mice demonstrated a modest 29% increase in plasma cholesterol and no significant increase in triglycerides, atherosclerosis, or inflammatory cells in lesions. Male Akita-LDLr −/− mice had increased levels of plasma IL-1β compared to nondiabetic mice, whereas no such difference was seen between female diabetic and nondiabetic mice. Conclusion. Akita-LDLr −/− mice display considerable gender differences in the development of diabetic atherosclerosis. In addition, the increased atherosclerosis in male Akita-LDLr −/− mice is associated with an increase in inflammatory cells in lesions.


Introduction
Atherosclerosis is a chronic inflammatory disease characterized by formation of lesions in large-and medium-sized arteries.Diabetes is associated with increased atherosclerosis, and diabetic patients have a 2-4-fold increased risk of cardiovascular mortality [1,2].Moreover, stroke, coronary heart disease, and peripheral artery disease are more common and occur at an earlier age in diabetic patients than in nondiabetic persons [3].The proposed culprits responsible for the increased risk of atherosclerosis in diabetic patients include dyslipidemia, hypertension, endothelial dysfunction, oxidative stress, and increased generation of advanced glycation end-products (AGEs) [4].However, the underlying cellular and molecular mechanisms whereby diabetes accelerates atherosclerosis are still poorly understood, and one of the main reasons for this has been the lack of animal models of diabetes that replicate the disease as seen in humans.
One of the most widely used mouse models for cardiovascular disease in type 1 diabetes is the streptozotocininduced diabetes model in atherosclerotic apolipoprotein E-deficient (ApoE −/− ) or LDL-receptor-deficient (LDLr −/− ) mice.The streptozotocin induced diabetes model, however, has a drawback in the nonspecific toxic effects of streptozotocin to other organs than pancreas, since streptozotocin is known to be both hepatotoxic and nephrotoxic [5].In addition, streptozotocin can methylate DNA and has genotoxic effects by damaging DNA [6].Recently, streptozotocin was shown to have a direct toxic effect on lymphocytes in vitro, particularly on CD8-cells and B-cells [7].Streptozotocin treatment of mice also resulted in a relative increase in regulatory T-cells, and this effect was independent of hyperglycemia [7].CD8-cells [8,9], B-cells [10][11][12][13], and regulatory T-cells [14] have been shown to play important roles in atherosclerosis.Thus, these side effects of Experimental Diabetes Research streptozotocin could have an impact on the atherosclerotic disease, which is not related to the diabetic disease.
The ins2 akita/+ (Akita) mouse is a model of type 1 diabetes, characterized by a point-mutation causing proinsulin misfolding with subsequent endoplasmic reticulumstress leading to beta-cell apoptosis [15].The Akita mouse has previously been successfully used as a model of diabetic microvascular complications, including retinopathy, neuropathy, and nephropathy [16].Recently, increased atherosclerosis in Akita-Apo-E −/− mice and Akita-LDLr −/− mice compared to nondiabetic mice has been reported [17,18].Both of these studies primarily focused on the role of altered lipid metabolism.Although the cellular and molecular mechanisms behind why diabetes results in increased atherosclerosis are not known, one proposed mechanism is increased inflammation, due to increased oxidative stress in the atherosclerotic lesions.In atherosclerosis, subendothelial retention and oxidation of LDL induce expression of proinflammatory cytokines and recruitment of inflammatory cells to the vessel wall [19,20].In diabetes, increased glucose levels result in formation of AGE, which is believed to increase oxidative stress and inflammation [21,22].To see if the increased atherosclerosis reported in Akita-LDLr −/− mice is accompanied or driven by increased plaque inflammation, we measured the extent of inflammatory cells and cytokines in atherosclerotic vessels of Akita-LDLr −/− mice.We also asked whether Akita-LDLr −/− mice display an altered immunological T-cell profile compared to nondiabetic LDLr −/− mice.

Mice.
Akita-LDLr −/− (B6.Cg-Ins2 Akita LDLr tm1Her /J) and LDLr −/− mice were obtained from Jackson Laboratories and bred at our animal facilities.Akita mice were genotyped according to protocols provided by the Jackson Laboratories.Animals had free access to tap water and were fed a high-fat diet, containing 21% cocoa fat (weight%), 0.15% cholesterol (weight%), from six weeks of age, and were sacrificed at 18 or 24 weeks of age.Blood glucose was measured every three weeks using a One-Touch Glucometer (LifeScan Inc., CA, USA) in nonfasting mice.Plasma triglycerides and cholesterol were measured in nonfasting mice by colorimetric assays as described before [23].All animal experiments were approved by the Malmö-Lund Animal Care and Use Committee and the investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health.

Immunohistochemistry.
For assessment of atherosclerosis, plaque characteristics, myocardial fibrosis, and inflammation mice were sacrificed at 24 weeks of age.En face preparations of the aorta, Oil red O-staining, and quantification were performed as described before [23].Staining of monocytes/macrophages (MOMA-2, Biomedicals AG, Switzerland, detecting a glycoprotein located in the cytoplasm and on the cell surface on monocytes and macrophages), Tcells (anti-CD3) [24], collagen (Masson's trichrome), and AbcA1 (abcam) were performed as described before [23].Subvalvular plaque area was determined in haematoxylin stained sections of the aortic root.

Plasma Cytokines.
Cytokines in plasma from 24-weekold mice were measured using a Th1/Th2 9-plex assay (Meso Scale Discovery, USA) according to manufacturer's instructions.

Splenocyte Preparation and
Culture.Splenocytes were isolated as previously described [24].Briefly, splenocytes from 24-week-old mice were isolated, washed, and stimulated with Concanavalin A (2.5 μg/ml) or left unstimulated.After 72 hours, [methyl-3 H]-thymidine was added to wells.To quantify DNA synthesis, cells were harvested 16 hours after [methyl-3 H]-thymidine addition and measured using a liquid scintillation counter.

Statistical Analysis.
Values are presented as mean ± SD unless otherwise specified.Statistical analyses were performed with Graph-Pad 5 (Prism) or PASW Statistics 18 software.Statistical significance was determined using twoway ANOVA followed by Bonferroni post hoc tests unless otherwise specified.Correlation analyses were performed using Pearson (normally distributed variables) or Spearman (skewed variables).Linear regression analysis was performed with PASW Statistics 18.

Effect of Diabetes on Myocardial Fibrosis.
To determine if the increased atherosclerosis in Akita-LDLr −/− mice was associated with increased myocardial fibrosis, we measured collagen content in the muscular tissue of the heart, but we were only able to detect minor amounts of collagen in the tissue.Moreover, we did not find any differences between the groups (data not shown).In addition, we analyzed macrophage content in the heart muscle tissue, but did not find any signs of inflammation.4(a)).Moreover, both TNF-α (P = 0.089; Figure 4(b)) and the neutrophil activating chemokine KC (CXCL1; P = 0.059; Figure 4(c)) displayed trends towards being increased in male Akita-LDLr −/− mice compared to male LDLr −/− control mice.In a linear regression analysis model, IL-1β was associated with average glucose levels in diabetic male and female mice, but not with cholesterol levels.There were no significant changes in plasma levels of IFNγ, IL-2, IL-4, IL-5, or IL-10 between the groups (Table 2).Total IL-12, including both anti-inflammatory IL-12p40 and proinflammatory IL-12p70, was decreased in male Akita-LDLr −/− mice compared to male LDLr −/− control mice (Table 2).T-cell subsets have been shown to greatly influence the development of experimental atherosclerosis in mice [28].To assess whether the increased plaque development reflected a change in the balance between activated conventional (male)

Discussion
In this study, we characterize the Akita-LDLr −/− mouse as a model of diabetic atherosclerosis.We also investigate if the atherosclerotic disease in Akita-LDLr −/− mice is accompanied by increased inflammation in atherosclerotic lesions.We show that the Akita-LDLr −/− mouse has considerable gender differences with regard to metabolic profile, atherosclerotic disease, and inflammatory cells in atherosclerotic lesions.While male Akita-LDLr −/− mice exhibit severe hyperglycemia, hypercholesterolemia, and hypertriglyceridemia compared to nondiabetic LDLr −/− mice, female Akita-LDLr −/− mice show marked hyperglycemia, but only a modest increase in cholesterol levels and no changes in plasma triglycerides.Male Akita-LDLr −/− mice have a five-fold increase in aortic lesion area, a threefold increase in subvalvular lesion area, and increased areas of macrophage and T-cell infiltration in the lesions compared to male LDLr −/− mice, but no such differences were seen in female mice.
Estrogen and its receptors are regulators of glucose and lipid metabolism and rodent studies link estrogen to anti diabetic effects.Estrogen is also known to exert antiinflammatory effects in both humans and rodents [29].It is possible that the gender differences reported in our study at least partly are due to the effects of estrogen, for example, plasma IL-1β levels are increased in male Akita-LDLr −/− mice compared to female Akita-LDLr −/− Experimental Diabetes Research (male) Akita-LDLr −/− (female) Akita-LDLr −/− (male) Akita-LDLr −/− (female)  males; Bonferroni post-hoc test.Two-way ANOVA revealed interactions between diabetes and gender for IL-12 (P < 0.05), and significant effect of gender for IL-5 (P < 0.05).Values represent mean ± SD. mice (Figure 4(a)).The reason why female Akita-LDLr −/− mice, despite induction of hyperglycemia, have no or only a minor increase in atherosclerosis, macrophage, and T-cell infiltration compared to female LDLr −/− mice could be due to the anti-inflammatory effects of estrogen.In this respect, it is important to note that hyperglycemia is less pronounced in females than in males (16.9 ± 3.8 mM versus 27.7 ± 4.6 mM).It is also important to note that nondiabetic female LDLr −/− mice have a 1.5-fold increase in cholesterol levels and 3-fold increase in subvalvular lesion area compared to nondiabetic male LDLr −/− mice.In fact, female nondiabetic LDLr −/− mice have the same subvalvular lesion area as male diabetic Akita-LDLr −/− mice, but the lesion area in female mice is not further increased in response to hyperglycemia.It is possible that the already high cholesterol levels in female nondiabetic LDLr −/− mice mask the effect of hyperglycemia on lesion formation in female diabetic Akita-LDLr −/− mice.Such observations have been made in other mouse models of diabetic atherosclerosis.For example, Reaven et al. studied male LDLr −/− mice on a high fat diet, which were made diabetic using streptozotocin [30].In that study, diabetic and nondiabetic mice had similar plasma cholesterol levels (25.1 versus 25.9 mM); however, diabetic mice had increased glucose levels, increased triglyceride levels, and increased formation of AGE epitopes in the artery wall.Despite this, there were no differences in atherosclerosis between diabetic and nondiabetic mice.The authors suggest that under conditions of marked hypercholesterolemia, there is no effect of hyperglycemia and/or of enhanced AGE formation on atherogenesis in LDLr −/− mice.Thus the observed gender differences in atherosclerotic lesions and inflammation in our study may have two possible explanations: (1) the already high cholesterol levels in female nondiabetic LDLr −/− mice could mask the effect of hyperglycemia on lesion formation in female diabetic Akita-LDLr −/− mice, and (2) the less pronounced in female Akita-LDLr −/− mice compared to male Akita-LDLr −/− mice may result in less atherosclerotic disease and less inflammation.
Our results partly confirm a recent study of LDLrdeficient Akita mice published by Zhou et al. [18].Zhou et al. reported hyperglycemia and hyperlipidemia accompanied by increased atherosclerotic disease in male and female Akita-LDLr −/− mice.In that study, however, they found no difference in blood glucose levels between male and female mice (23.5 ± 9 mM versus 21.8 ± 7.2 mM), whereas we found that male Akita-LDLr −/− mice had higher glucose levels than females (27.7 ± 4.6 mM versus 16.9 ± 3.8 mM, P < 0.001).High-fat diet induces insulin resistance and diabetes in C57/Bl6 mice [31] and there is gender difference [32,33].However, it has also been reported that there is a large difference in glucose levels in male versus female Akita mice, even on a low-fat diet (27.3 ± 5.3 mM in males versus 13.6 ± 3.8 mM in females) [26], and the authors speculate that estrogen and prolactin play a protective role in the females.The levels reported by Yoshioka et al. are similar to the levels we report (27.7 ± 4.6 mM in males versus 16.9 ± 3.8 mM in females).In addition, whereas we found no significant difference in lesion area in female mice, neither in en face preparations of the aorta nor in subvalvular lesions, Zhou et al. reported a significant increase in subvalvular lesions of female Akita-LDLr −/− mice compared to female LDLr −/− mice.However, in agreement with our study, the increase in lesion area in diabetic mice compared to nondiabetic mice were larger in males (224%) than in females (30%).The mice in our study were fed a high fat diet containing 0.15% cholesterol, whereas Zhou et al. fed the mice a low-fat diet with 0.02% cholesterol, thus there was a considerable difference in the size of the subvalvular lesion area in our study compared to Zhou et al. (630 000 μm 2 versus approx.350 000 μm 2 , resp., for female diabetic mice).The difference in size of subvalvular lesions may also explain why Zhou et al. report significant differences in diabetic female Akita-LDLr −/− mice compared to nondiabetic LDLr −/− mice.The subvalvular lesions in our study, which is almost the double size compared to the ones reported by Zhou et al., may be at a later atherosclerotic stage, at which the differences in female mice may be evened.Thus, since the subvalvular lesions are one of the earliest atherosclerotic lesions in mice, there may be differences in lesion size at other locations.This is supported by our finding that female Akita-LDLr −/− mice displayed a trend towards larger lesion area in the aortic arch compared to female LDLr −/− mice.
Further, Zhou et al. determined liver specific expression of genes involved in lipid metabolism and inflammation by quantitative PCR [18].In a recent report Jun et al. showed increased atherosclerosis in male apoE-deficient Akita mice, but did not analyze female mice [17].In the latter study, the authors analyzed expression of lipoprotein receptors in the liver, lipid secretion from the liver and, plasma lipid profile.Thus, both these recently published studies have focused on lipid metabolism and thus mainly studied differences in the liver.To assess whether the increased atherosclerosis seen in diabetic mice is accompanied by increased inflammation in atherosclerotic vessels, we characterized and quantified inflammatory cells in subvalvular lesions by immunohistochemistry.In addition, we used quantitative PCR to analyze the expression of inflammatory genes in the brachiocephalic artery, which is one of the most plaque prominent locations in the arterial tree in mice.
Studies of plaques derived from both type 1 and 2 diabetes patients have shown increased accumulation of Tcells and macrophages [34].Similar to humans, subvalvular lesions of Akita-LDLr −/− male mice were enriched in both T-cells and macrophages compared to LDLr −/− controls.Surprisingly, LDLr −/− mice have a higher percentage of macrophages in the lesions than diabetic Akita-LDLr −/− mice, which probably reflects that the lesions are at an earlier stage, whereas more advanced lesions display increased infiltration of smooth muscle cells and more fibrous tissue.This is supported by increased collagen content in Akita-LDLr −/− male mice compared to controls.
It was previously shown that Akita-LDLR −/− mice displayed decreased mRNA levels in the liver of several genes involved in lipid metabolism (Srebp1a, Srebp1c, AbcA1, Lxrα, and Cyp7b1) [18].Some of these genes have been shown to modulate atherosclerotic disease.For example, since LXRs promote cholesterol efflux via upregulation of the ABC family [35], one would expect that LXRs would have antiatherogenic properties due to increased reverse cholesterol transport in the aorta.Indeed, LXRαβdeficient macrophages displayed increased accumulation of cholesterol [36], and apoE-and LXRα-deficient mice show increased atherosclerotic disease [37,38].In agreement with this, treatment with LXR agonist reduces atherosclerosis in mice [39][40][41][42].However, LXRs do not only affect atherosclerosis due to increased reverse cholesterol transport.It has been shown that LXRs are negative regulators of macrophage inflammatory gene expression.In vitro studies show that LXR ligands inhibit the expression of inflammatory genes such as IL-6, IL-β, MCP-1, MMP-9, and osteopontin [43][44][45].Moreover, in vivo treatment of apoE −/− mice with Experimental Diabetes Research LXR agonists resulted in substantially reduced MMP-9 gene expression in the aortas [44].In our study, quantitative gene expression analysis of either IL-6, IL-1β, MCP-1, or osteopontin in the brachiocephalic artery did not reveal any difference between Akita-LDLr −/− or control mice.There was a tendency to increased MMP-9 gene expression in Akita-LDLr −/− mice compared to LDLr −/− mice including both genders, though this increase was more evident in female mice than in male mice.Also ABCA1 has been reported to have anti-inflammatory properties shown by ABCA1-deficient macrophages, which displayed increased TNFα expression upon LPS stimulation [27].Moreover, bone marrow transplantations demonstrated an antiatherogenic function of ABCA1 in macrophages independently of changes in plasma lipids [46].In our study, total ABCA1 levels, analyzed by immunohistochemistry, were increased in subvalvular lesions of male Akita-LDLr −/− mice.Although the percentage of plaque area stained with anti-ABCA1 was decreased in male Akita-LDLr −/− mice, the same difference was present in female mice.In conclusion, in our study we find no evidence suggesting that downregulation of AbcA1 or Lxrα gene in lesions could explain the gender difference in inflammatory cells observed in subvalvular lesions of Akita-LDLr −/− mice.On the other hand, Zhou et al. reported increased TNFα, MCP-1, and IL-1β staining in lesions of Akita-LDLr −/− mice compared to LDLr −/− mice.Differences in our study compared to the study by Zhou et al., for example, the diet, location of the lesions, or protein versus mRNA analysis, may explain the reported differences in cytokines in atherosclerotic lesions.
Diabetes is a strong independent risk factor for heart failure [47,48].In a previous paper Basu et al. found that normolipidemic Akita C57Bl6 mice were characterized by diastolic dysfunction at three and six months of age, but preserved systolic function [49].Moreover, there was no evidence of myocardial hypertrophy or fibrosis in the diabetic mice.The latter is in agreement with our data, showing no or only minor collagen content in cardiac tissue and no differences between diabetic and nondiabetic mice.
Several papers have shown that Akita male mice (on normolipidemic background C57bl6 background) develop elevated systolic blood pressure [50][51][52], which is not present in female mice [50].Basu et al. measured heart rate in Akita C57BL/6J mice, but did not find any differences between diabetic and nondiabetic mice [49].In humans, increased blood pressure is a risk factor for the development of atherosclerotic disease; however, in mice the relationship between blood pressure and atherosclerosis is less clear.While several reports have demonstrated reduced atherosclerosis in mice with decreased blood pressure, other studies have shown that changes in blood pressure do not affect atherosclerotic disease (reviewed in [53]).Thus, it is difficult to predict whether differences in blood pressure could underlie the gender differences in atherosclerosis observed in Akita-LDLr −/− mice.
It is widely recognized that T-cells influence atherosclerosis in mice [28].For example, regulatory T-cells have in several studies been shown to protect against atherosclerosis [54].Since we did not find any differences in percentages in either the protective regulatory (CD4+CD25+FoxP3+) Tcells or the putatively harmful activated (CD4+CD69+) Tcells, we suggest that diabetic atherosclerosis in these mice is not induced by an overall immune activation.

Conclusion
In conclusion, both male and female Akita-LDLr −/− mice are hyperglycemic compared to control LDLr −/− mice.However, whereas male Akita-LDLr −/− mice have a 2-fold increase in plasma cholesterol and 3-fold increase in triglyceride levels, female Akita-LDLr −/− mice have only a modest diabetesinduced increase in cholesterol and no increase in triglyceride levels.This is accompanied by a dramatic increase in atherosclerosis as well as increased plaque inflammation in male Akita-LDLr −/− mice, but no significant changes in plaque size or inflammatory cells in lesions in female Akita-LDLr −/− mice compared to nondiabetic LDLr −/− mice.We propose that the Akita-LDLr −/− mouse is a promising tool for studying development of cardiovascular disease both in a setting of severe as well as a more moderate increase in cholesterol levels.

Figure 2 :
Figure 2: Male Akita-LDLr −/− mice have increased atherosclerosis compared to male LDLr −/− mice.Mice were sacrificed at 24 weeks of age and atherosclerosis were quantified both in en face preparations of the aorta (a) as well as in sections from the aortic root (b).Percentage plaque area of total vessel area in the aorta and representative en face preparations of the aortas stained with Oil red O (dark red-colored) are presented in (a) and subvalvular lesion areas are presented in (b).Values are presented as individual mice and as mean ± SEM.Two-way ANOVA revealed interactions between diabetes and gender ((a) * * , (b) * * * ), and significant effect of diabetes ((a) * * * , (b) * * * ) and of gender (b) * * * .Bonferroni post hoc test yielded * * P < 0.01, * * * P < 0.001.Scale bar = 2 mm.
* P < 0.05 versus LDLr −/− females; Mann-Whitney.Levels of mRNA were measured by real-time RT-PCR and were normalized to the expression of the housekeeping control gene cyclophilin B. values represent mean ± SD.