Immune dysregulation accelerates atherosclerosis and modulates plaque composition in systemic lupus erythematosus

  1. Aleksandar K. Stanic*,
  2. Charles M. Stein†,‡,
  3. Adam C. Morgan*,
  4. Sergio Fazio*,§,
  5. MacRae F. Linton*,‡,
  6. Edward K. Wakeland,
  7. Nancy J. Olsen‖, and
  8. Amy S. Major*,**
  1. Divisions of *Cardiovascular Medicine and
  2. †Rheumatology and Immunology, Department of Medicine, and Departments of
  3. ‡Pharmacology and
  4. §Pathology, Vanderbilt University School of Medicine, Nashville, TN 37232-6300;
  5. Center for Immunology and Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX 75390; and
  6. ‖Division of Rheumatic Diseases, Department of Internal Medicine, University of Texas Southwestern Medical School, Dallas, TX 75390

  1. Communicated by Robert W. Mahley, The J. David Gladstone Institutes, San Francisco, CA, March 22, 2006 (received for review October 10, 2005)

 
Next Section

Abstract

Patients with systemic lupus erythematosus (SLE) have accelerated atherosclerosis. The underlying mechanisms are poorly understood, and investigations have been hampered by the absence of animal models that reflect the human condition of generalized atherosclerosis and lupus. We addressed this problem by transferring lupus susceptibility to low-density lipoprotein (LDL) receptor-deficient (LDLr−/−) mice, an established model of atherosclerosis, creating radiation chimeras with NZM2410-derived, lupus-susceptible, B6.Sle1.2.3 congenic or C57BL/6 control donors (LDLr.Sle and LDLr.B6, respectively). LDLr.Sle mice developed a lupus-like disease characterized by production of double-stranded DNA autoantibodies and renal disease. When fed a Western-type diet, LDLr.Sle chimeras had increased mortality and atherosclerotic lesions. The plaques of LDLr.Sle mice were highly inflammatory and contained more CD3+ T cells than controls. LDLr.Sle mice also had increased activation of CD4+ T and B cells and significantly higher antibody to oxidized LDL and cardiolipin. Collectively, these studies demonstrate that the lupus-susceptible immune system enhances atherogenesis and modulates plaque composition.

Three decades ago, Urowitz et al. (1) recognized that cardiovascular disease (CVD) and myocardial infarction were major causes of mortality in patients with systemic lupus erythematosus (SLE). More recently, Manzi et al. (2) reported that premenopausal women with SLE, a population usually protected from atherosclerosis, had a 50 times greater risk of a fatal vascular event compared with age- and gender-matched controls. In addition, we showed an increased prevalence of coronary atherosclerosis in SLE (3). Despite the fact that CVD is the most common cause of death in patients with SLE who survive the acute complications of the illness, little is known about the underlying mechanisms. It has been suggested that a combination of traditional risk factors, including hypertension, dyslipidemia, and lipid oxidation as well as nontraditional risk factors, such as autoantibodies and inflammation, may contribute to advanced vascular disease in SLE (4). Therefore, defining the autoimmune mechanisms that promote atherosclerosis is essential to optimize risk reduction and develop targeted therapeutics for prevention of CVD in SLE.

Atherosclerosis involves many cellular processes, and increasing evidence supports the role of inflammation and immunity in the pathogenesis of atherosclerosis (5). Macrophages and T cells make up a large percentage of the cells present in the atherosclerotic plaque (6). These cells contribute to the inflammatory process by producing cytokines that attract smooth muscle cells and lymphocytes that compromise plaque stability. B cell responses and autoantibodies to self-antigens such as oxidized LDL (oxLDL), heat-shock protein 60/65, and β-2-glycoprotein I have also been identified in humans with CVD and in animal models of atherosclerosis (7, 8). These antibodies can also be detected in humans and animals with autoimmune diseases such as SLE and the antiphospholipid antibody syndrome (9). However, whether autoantibody production is causally related to atherosclerosis is not known.

A factor that has limited understanding the relationship between inflammation and atherosclerosis in SLE is that animal models of lupus are genetically resistant to diet-induced atherosclerosis. The development of the NZM2410-derived congenic B6.Sle mouse strains made it feasible to examine lupus and atherosclerosis together on the susceptible C57BL/6 background. Morel et al. (10) described three major chromosome intervals in the NZM2410 mouse strain termed Sle1, Sle2, and Sle3 that are highly associated with lupus susceptibility. The investigators made a series of combined and single congenic mice on the C57BL/6 background. In general, Sle1 mediates loss of tolerance to nuclear antigens (11); Sle2 lowers the activation threshold of B cells leading to expansion of B-1 B cells and polyclonal IgM (12); and Sle3 is associated with decreases in the activation threshold of T cells, a concomitant increase in T cell-dependent polyclonal IgG production, and reduced activation-induced cell death (13). In bone marrow transfer studies to normal C57BL/6 animals, it was demonstrated that lupus susceptibility was carried and could be transferred by cells of hematopoietic origin (13, 14). Therefore, we exploited this ability to transfer lupus and made radiation chimeras of B6.Sle1.2.3 triple congenics with lethally irradiated, atherosclerosis-susceptible LDL receptor-deficient (LDLr−/−) mice and used this animal model to address the hypothesis that lupus-associated immune dysregulation promotes atherosclerosis.

Previous SectionNext Section

Results

Development of SLE in LDLr−/− Radiation Chimeras.

We made lupus-susceptible animals in the setting of atherosclerosis by transplanting lethally irradiated LDLr−/− mice with bone marrow from either B6 controls (LDLr.B6) or lupus-susceptible B6.Sle1.2.3 animals (LDLr.Sle). Sixteen weeks after transplantation, mice were placed on a high-fat Western diet. Animals were killed at 8 weeks after the initiation of diet (Fig. 1 A). At the time of killing, there was no difference in body weight (Fig. 1 B). However, the majority of LDLr.Sle mice had a urinary protein grade of 2+ or greater, significantly higher than the LDLr.B6 group (89% vs. 14%, respectively, P = 0.001) (Fig. 1 C). In addition, although many of the LDLr.Sle mice had serum creatinine and urea levels similar to those of controls, the mean concentrations were significantly increased in LDLr.Sle mice (Fig. 1 D and E), indicating a decline of renal function. The anti-dsDNA antibody titer was also substantially increased in the LDLr.Sle mice compared with controls (Fig. 1 F). These results indicate that the lupus disease state was successfully transferred to the LDLr−/− mice by bone marrow transplantation.

Fig. 1.
View larger version:
Fig. 1.

The lupus disease phenotype of B6.Sle1.2.3 mice can be transferred to LDLr−/− mice. (A) Lethally irradiated female LDLr−/− mice were reconstituted with either B6 or B6.Sle1.2.3 bone marrow. Sixteen weeks after transplantation, all animals were placed on a Western-type diet for 8 weeks. After this time (24 weeks after transplant), mice were killed and analyzed. (B) Body weight of LDLr.B6 (open bars) and LDLr.Sle (filled bars) mice. (C) Percentage of LDLr.B6 (open bars) and LDLr.Sle (filled bars) mice exhibiting protein in urine (1+, 30 mg/dl; 2+, 30–100 mg/dl; 3+, 100–300 mg/dl). (D) Levels of serum creatinine in LDLr.B6 (squares) and LDLr.Sle (circles) mice. (E) Levels of serum urea in LDLr.B6 (squares) and LDLr.Sle (circles) mice. (F) Levels of serum antibody specific for dsDNA as measured by specific ELISA in LDLr.B6 (squares) and LDLr.Sle (circles) mice. Bars represent the mean ± SEM of 12 LDLr.B6 and 9 LDLr.Sle mice. Shown is one of at least three experiments. In C–F, the P values were calculated by using a Mann–Whitney analysis. In D, the P value was calculated by using a χ2 analysis (see text).


Susceptibility to Lupus Exacerbates Atherosclerosis in LDLr.Sle Radiation Chimeras.

Next, we studied the size and composition of atherosclerotic lesions in the aortic sinus. After 8 weeks of a Western diet, the atherosclerotic lesion area was significantly increased in LDLr.Sle chimeras compared with control LDLr.B6 mice (Fig. 2 A).

Fig. 2.
View larger version:
Fig. 2.

Increased atherosclerosis in LDLr.Sle mice. (A) Oil-red-o analysis of the atherosclerotic lesion area in the aortic sinus of LDLr.B6 (squares) and LDLr.Sle (circles) mice; n = 17 for both groups. (B) Serum cholesterol (Left) and triglyceride (Right) levels before bone marrow transplantation (0), at the initiation of a Western-type diet (16 weeks), and at the time of killing (24 weeks) in LDLr.B6 (squares) and LDLr.Sle (circles) mice; n = 17 for both groups. (C) FPLC analysis of cholesterol lipoprotein distribution in serum of LDLr.B6 (squares) and LDLr.Sle (circles) mice. Shown are data from a pool of five to six mice in each group. Analysis of individual mice showed similar results. (D and E) Correlation of serum urea and creatinine levels and atherosclerotic lesion area in LDLr.Sle mice (circles). For comparison, LDLr.B6 mice (squares) are included on the graph but were not used in the calculations. (F) Measurement of systolic blood pressure by tail cuffing in LDLr.B6 (circles) and LDLr.Sle (triangles) mice. Data are the mean ± SEM of nine LDLr.B6 and seven LDLr.Sle mice. Depicted P values were calculated by Mann–Whitney analysis. Correlation was determined by Spearman’s analysis.


The levels of serum cholesterol and triglycerides at time of killing were slightly, but significantly, decreased (Fig. 2 B) in LDLr.Sle mice compared with controls. FPLC analysis of lipoprotein distribution showed that the difference was primarily because of a decrease in the non-high-density lipoprotein cholesterol fractions (Fig. 2 C). Serum creatinine and urea concentrations (see Fig. 1) did not correlate with atherosclerosis lesion size (Fig. 2 D and E). Also, the LDLr.Sle mice were not hypertensive compared with LDLr.B6 controls (Fig. 2 F). Therefore, mechanisms other than dyslipidemia, renal disease, and hypertension contribute to the advanced atherosclerosis in the lupus-susceptible animals.

Lupus Susceptibility Results in Changes in Lesion Composition.

We conducted histochemical and immunohistochemical analyses of the cellular content of the atherosclerotic plaques in the aortic sinus and observed no significant difference in macrophage content in LDLr.Sle and LDLr.B6 mice (Fig. 3 A and B). There was a 3-fold increase in the number of CD3+ T cells in lesions of the LDLr.Sle mice compared with controls (Fig. 3 A and B). Similar increases were observed in LDLr.Sle mice when lesions were stained for the CD4 T cell marker. Fluorescent staining for CD4 and CD3 revealed colocalization of these markers (Fig. 3 B). We determined the expression of the MHC class II molecule I-Ab in lesions. Consistent with the chronic inflammatory phenotype of lupus, we observed a dramatic increase in I-Ab expression in the LDLr.Sle mice compared with controls (Fig. 3 B). Collagen content did not differ between LDLr.Sle and LDLr.B6 mice (Fig. 3 A and B) and SMC (α-SMA) were present in the media but were not found in the lesion itself (data not shown). Collectively, these data demonstrate that the LDLr.Sle mice have an atherosclerotic plaque phenotype that is more cellular and contains more activated T cells.

Fig. 3.
View larger version:
Fig. 3.

Atherosclerotic plaque composition. (A) Quantitation of macrophages (CD3+ T and CD4+ T lymphocytes) in atherosclerotic plaques of LDLr.B6 (open bars) and LDLr.Sle (filled bars) mice. Macrophages are expressed as the percent macrophage staining area of total lesion area; n = 5 for both groups. CD3+ and CD4+ T lymphocytes are expressed as the percentage positive cells of total cells in the lesion; n = 3–5 for both groups. Collagen content was measured by Masson’s trichrome and expressed as a percentage of positive staining of total lesion area; n = 18 for the LDLr.B6 group and 19 for the LDLr.Sle group. Bars are the mean ± SEM from the average of four sections per mouse. (B) Immunohistochemical and immunofluorescent staining for macrophages (MOMA-2, magnification, ×50), Masson’s trichrome staining (magnification, ×50), MHC class II I-Ab (magnification, ×200), and colocalization of CD3 and CD4 (arrowheads) on T lymphocytes (magnification, ×400) in LDLr.B6 (Upper) and LDLr.Sle (Lower) mice. In the CD3/CD4 figures, the dotted line represents the internal elastic lamina, and the solid line designates the edge of the lesion. Shown are sections from one representative mouse of three to five in each group.


LDLr.Sle Mice Exhibit Hyperactive Immunity in the Periphery.

We determined whether LDLr.Sle mice showed increased peripheral immune activation. At the time of killing, the spleen weights in LDLr.Sle chimeras were significantly greater than those of controls (Fig. 4 A). Increased spleen weights corresponded to a slight, but significant, increase in cell numbers (Fig. 4 B). Flow cytometric analysis of LDLr.Sle splenocytes demonstrated an increase in the percentages (LDLr.B6 = 19.5 ± 1.9%; LDLr.Sle = 25.7 ± 0.3%, P = 0.03) and absolute numbers of CD4+ T cells compared with controls (Fig. 4 C and D). Numbers of CD8+ T cells, B cells (B220+,CD11c−), macrophages (Mac-3+), and NK1.1+ cells were not significantly affected. In addition, total CD11c+ dendritic cells (DC) were not different between the two groups. Although the plasmacytoid DC (B220+, CD11c+) cells were increased 2-fold in the LDLr.Sle mice (2.0 ± 0.6 × 106 cells) compared with the LDLr.B6 mice (1.0 ± 0.1 × 106 cells), the difference did not quite reach statistical significance (P = 0.06). Increases in these types of DC are consistent with the reported B6.Sle1.2.3 phenotype (15).

Fig. 4.
View larger version:
Fig. 4.

Analysis of splenic and thymic cells. (A and B) Comparison of wet spleen weights and cellularity in LDLr.B6 (open bars) and LDLr.Sle (filled bars) mice. (C) Contour plots of CD4+ T lymphocytes in the spleens of LDLr.B6 (Left) and LDLr.Sle (Right) mice. Cells were gated on total lymphocytes based on the forward vs. side scatter plot. Shown is one mouse from one experiment of two containing three mice per group. (D) Absolute numbers of designated spleen cells in LDLr.B6 (open bars) and LDLr.Sle (filled bars) mice. Numbers were calculated based on the total spleen cell count and the percent of each cell type. (E) Expression of CD69 on CD8α+CD4− single-positive and CD8α+CD4+ double-positive T lymphocytes isolated from thymi of LDLr.B6 (Left) and LDLr.Sle (Right) mice. (F) Representative contour plots of the T lymphocyte compartments of the thymus in LDLr.B6 (Left) and LDLr.Sle (Right) mice. Cells were gated on total lymphocytes based on the forward vs. side scatter plot. (G) Expression of the marker of activation CD69 on CD4+ T lymphocytes isolated from the spleen and thymus in LDLr.B6 (Left) and LDLr.Sle (Right) mice. Shown is one mouse from one experiment of two containing three mice per group. (H) Mean fluorescence intensity of Bcl-2 expression (Left) and percentage of Ki67+ cells (Right) in B cells isolated from LDLr.B6 (filled bars) and LDLr.Sle (open bars) mice. Shown is one experiment of two each containing three mice per group. Bars in graphs are the mean ± SEM. P values for flow cytometry experiments were calculated by using Student’s t test. P values for spleen weight and cellularity were calculated by using a Mann–Whitney test.


Autoimmunity may be considered a disorder of decreased activation thresholds of specific immune populations and excessive activation during lymphocyte selection or in their mature state. Therefore, we investigated whether increased CD4+ T cell numbers in the spleen were the result of increased selection in the thymus. Analysis of surface levels of CD69, which is up-regulated on thymic CD4+CD8+ double-positive (DP) cells after successful positive selection, showed no differences on either DP cells or CD8+ cells between the groups (Fig. 4 E). Interestingly, total CD4+ T cell numbers were lower in thymi of LDLr.Sle mice (Fig. 4 F; LDLr.B6 = 18.5 ± 3.2%; LDLr.Sle = 8.3 ± 1.2%, P = 0.038) but were significantly more active (Fig. 4 G; CD69 staining LDLr.B6 = 30.0 ± 6.1%; LDLr.Sle = 59.5 ± 4.3%, P = 0.017). Once in the periphery, CD4+ T cells retained this overactive phenotype characterized by the expression of CD69 (Fig. 4 G; LDLr.B6 = 18.5 ± 1.8%; LDLr.Sle = 31.7 ± 3.6%, P = 0.03). We observed no difference in CD4+ T cell expression of the proliferation marker Ki67 (LDLr.B6 = 8.3 ± 1.0%; LDLr.Sle = 10.6 ± 0.6%, P = 0.13) but increased expression of the active apoptosis marker caspase 3 in LDLr.Sle mice (LDLr.B6 = 4.1 ± 0.6%; LDLr.Sle = 6.6 ± 0.6%, P = 0.03), perhaps indicating greater turnover of the CD4+ T cells.

In addition to increased CD4+ T cell activation, we also observed increased B cell activity. As reported in B6.Sle animals (12, 14), there were increases in the activation markers CD80 and CD86 on B cells isolated from spleens of LDLr.Sle mice compared with controls (data not shown). Increased B cell activation in LDLr.Sle mice was accompanied by increased proliferation, measured by Ki67 expression (Fig. 4 H Right) and an increase in Bcl-2 expression (Fig. 4 H Left) compared with controls. Collectively, these data indicate that LDLr.Sle mice have an increased B cell and T helper cell activation state characteristic of the B6.Sle1.2.3 congenic animals (16) and that exacerbation of atherosclerosis in LDLr.Sle mice occurs in the setting of dysregulated and overactivated B and CD4+ T cell populations.

Lupus-Susceptible Mice Have Higher Titers of Antiphospholipid Antibodies.

We measured serum antibody titers directed against oxLDL and cardiolipin and observed that LDLr.Sle mice exhibited significant increases in anti-oxLDL (Fig. 5 A) and anticardiolipin (Fig. 5 B) antibodies compared with controls. There was little to no detectable antibody against oxLDL or cardiolipin in either group at baseline (Fig. 5 A and B). Although it was not statistically significant, we did observe an increase in the anti-oxLDL IgM titer (Fig. 5 C Left). In addition, there was a significant increase in the anti-cardiolipin IgM titers (Fig. 5 D Left). There was no clear indication of either a Th-1 or Th-2 bias, because both IgG1 and IgG2a were increased in the LDLr.Sle mice compared with LDLr.B6 control animals (Fig. 5 C and D).

Fig. 5.
View larger version:
Fig. 5.

Anti-oxLDL and anticardiolipin antibodies. Total anti-oxLDL (A) and anticardiolipin (B) Ig in LDLr.B6 (squares) and LDLr.Sle (circles) mice at baseline, before bone marrow transplantation and at postdiet or time of killing. In both A and B, n = 10–14 mice per group. (C) Anti-oxLDL Ig isotypes in serum of LDLr.B6 (squares) and LDLr.Sle (circles) mice. (D) Anticardiolipin Ig isotypes in serum of LDLr.B6 (squares) and LDLr.Sle (circles) mice. Shown are the means ± SEM. Calculated P values were obtained by Mann–Whitney analysis.


Previous SectionNext Section

Discussion

Patients with SLE have a marked increase in the prevalence of atherosclerosis and its complications (4). However, the cellular and molecular mechanisms underlying this accelerated atherosclerosis are poorly understood and cannot be attributed to traditional CVD risk factors (3, 17). Mechanistic studies of accelerated atherosclerosis in lupus are difficult due in part to the lack of an appropriate animal model. Vasculitis and early fatty streak lesions are increased in MRL/lpr mice (18, 19), and fatal myocardial infarction due to noncellular vessel occlusion occurs in male BXSB mice (20, 21). However, because of their genetic background, mice traditionally used to study the pathogenesis of lupus are largely resistant to developing the large, multicellular atherosclerotic lesions observed in humans. To address this problem, we used lupus-susceptible congenic mice, B6.Sle1.2.3, originally described by Morel et al. (10), that are on the atherosclerosis-susceptible C57BL/6 background. The phenotype of autoimmune dysregulation characteristic of this model of SLE has been shown to transfer with bone marrow transplantation; thus, we designed transplantation experiments in LDLr−/− mice to examine the impact of SLE on atherogenesis. We created radiation chimeras with either C57BL/6 or B6.Sle1.2.3 donors to LDLr−/− recipients. Studies from our laboratory have demonstrated that LDLr expression on hematopoietic cells in LDLr−/− mice contributes significantly to foam cell formation (22); however, because both control and experimental groups received LDLr-sufficient bone marrow, the interpretation of the results should not be affected. We found that LDLr−/− mice reconstituted with the B6.Sle1.2.3 bone marrow developed not only hallmarks of autoimmune dysregulation and decreased renal function consistent with SLE but also large atherosclerotic lesions in the aortic sinus characterized by accumulation of lipid-filled macrophages and increased numbers of CD3+ T lymphocytes.

Traditionally, dyslipidemia correlates directly with the extent and progression of atherosclerosis in many mouse models and in humans. Chronic renal failure, an independent complication of lupus, has also been associated with increased atherosclerosis (23, 24). These studies of atherosclerosis and renal failure in apoE−/− and LDLr−/− mice have shown that decreased renal function results in significant increases in very (V)LDL and LDL cholesterol compared with controls. Therefore, in the setting of renal impairment, it is difficult to ascertain whether the increase in atherosclerosis is a secondary effect of atherogenic lipids. In contrast, we found that LDLr.Sle mice have increased atherosclerosis despite lower levels of pathogenic non-high-density lipoprotein (HDL) and triglycerides (Fig. 2 C). Not only were LDL and triglyceride concentrations lower, but the extent of renal compromise did not correlate with the size of atherosclerotic lesions in LDLr.Sle mice (Fig. 2 D and E). These data support the hypothesis that immune mechanisms are primarily responsible for increased plaque formation in the LDLr.Sle mice. Interestingly, we have reported accelerated coronary artery atherosclerosis in lupus patients whose LDL and HDL cholesterol concentrations were not higher that those of control subjects (3). Similar observations regarding carotid artery atherosclerosis were made by Roman et al. (25). Therefore, it appears that traditional associations between LDL cholesterol and severity of atherosclerosis do not apply in the case of mice or patients with SLE.

A central question in lupus is whether abnormal immune activation results in progression of atherosclerotic lesions. We observed significant differences in atherosclerotic plaque composition in LDLr.Sle mice, consistent with increased immune activation. Specifically, there were increased areas of intense cellularity in the LDLr.Sle plaques that were not present in the LDLr.B6 controls (Fig. 3). These cells were not MOMA-2+ macrophages but stained for CD3, a T cell marker. In fact, CD3+ and CD4+ T cell numbers were 3-fold higher in the lupus-susceptible animals compared with controls. Double-staining experiments showed that the majority of CD4+ cells were CD3+ T cells. Antigen-presenting cells corresponding to areas of MOMA-2 staining appeared to be more activated in the LDLr.Sle mice, as exemplified by the marked overexpression of I-Ab, an MHC class II molecule that is up-regulated by IFN-γ production. Previous studies have implicated T cell-rich lesions as more unstable and prone to rupture, leading to vessel occlusion and myocardial infarction (26). Therefore, not only is atherosclerosis increased in LDLr.Sle animals, but, also, the plaques are typified by a cellular composition of a more vulnerable phenotype.

Consistent with increased immune activity in the atherosclerotic plaque, both the peripheral CD4+ T cell and B cell compartment of the LDLr.Sle mice exhibited immune volatility. We observed an increase in the absolute numbers and activation state of splenic CD4+ T cells in LDLr.Sle mice compared with controls (Fig. 4). We also observed the characteristic increase in B cell activation and autoantibody production associated with SLE (27). Increased antibody production in LDLr.Sle mice was not restricted to lupus-associated antigens such as dsDNA; there were also significant increases in anti-oxLDL and anticardiolipin antibodies (Fig. 5). The increase of anti-oxLDL and anticardiolipin antibody corresponded to a significant increase in the T cell-dependent IgG1 and IgG2a isotypes. Given the fact that Sle2 mediates expansion of B-1 B cells, a major source of IgM antibody, we were surprised that there were not even greater differences between LDLr.Sle and LDLr.B6 mice for anti-oxLDL IgM. Of note, some aspects of the induced antibody response may have been affected by the bone marrow transplantation protocol. It has been shown that bone marrow transplantation does not replace the B-1 B cell compartment in the peritoneum (28). Much of the literature suggests that much of the ox-LDL antibody is B-1 B cell-derived (29, 30); thus, it is possible that our current experimental protocol did not transfer the hyperactive phenotype to this cell type. Additional experiments designed to specifically look at B-1 B cells in this compound disease model will be necessary in the future. It is possible that increased ox-LDL-specific IgG1 and IgG2a, which can modulate inflammation via immune complex binding of Fcγ receptors (31), may act to exacerbate atherosclerotic disease in the LDLr.Sle mice.

Interestingly, we also observed an as-yet-unreported decrease in the number of CD4+CD8− T cells in the thymi of the LDLr.Sle mice compared with controls. However, a larger percentage of the thymic CD4+ T cells expressed the CD69 surface antigen, indicating that they were more activated compared with control animals. It is not known why CD4+ T cells are decreased in the LDLr.Sle animals. Because we observed an increase in the numbers of CD4+ T cells in the periphery or LDLr.Sle mice compared with LDLr.B6 animals, it is possible that the CD4+ T cells are being exported from the thymus at a faster rate in the lupus-susceptible animals. Another possibility is that the thymic CD4+ T cells are undergoing a higher rate of activation-induced cell death.

In conclusion, in a mouse model of atherosclerosis in SLE, we have shown that an overactive immune system participates in the progression of atherosclerotic lesions and leads to modification of plaque cellular composition. These studies shed light on the pathogenic role of the SLE-affected immune system in atherosclerosis. By using this model, it will be possible to examine the effects of lupus drugs on the progression of atherosclerosis. Also, future studies using the Sle1, Sle2, and Sle3 single congenic mice will allow us to determine the contribution of different immune components on the changes in atherosclerotic lesion size and composition observed in this study. Such studies will facilitate the identification and development of therapeutic interventions to treat SLE and minimize the risk of accelerated atherosclerosis.

Previous SectionNext Section

Materials and Methods

Mice.

C57BL/6 and LDLr−/− mice were originally obtained from The Jackson Laboratory and maintained in our colony. The LDLr−/− mice have been backcrossed to the C57BL/6 background at least 10 times. Lupus-susceptible B6.Sle1.2.3 triple congenic mice have been described and extensively characterized (10, 14, 32, 33). The mice are homozygous for the three lupus-susceptibility chromosome intervals on chromosomes 1, 4, and 7, respectively. All mice were maintained in microisolator cages and used according to the guidelines of the Vanderbilt University Institutional Animal Care and Use Committee. Unless otherwise stated, mice were fed a normal chow diet.

Production of Radiation Chimeras.

Transfer of the lupus-susceptibility phenotype to female LDLr−/− mice was accomplished by production of radiation chimeras as described in ref. 34.

Atherosclerosis Studies.

Sixteen weeks after bone marrow transplantation, animals were placed on a high-fat Western-type diet (21% milk fat and 0.15% cholesterol) for 8 weeks (Fig. 1 A). At the end of 8 weeks, animals were killed and analyzed for the extent of atherosclerosis and the presence of SLE as described in ref. 34.

Immunohistochemistry.

Staining of macrophages (MOMA-2) and MHC class II (I-Ab) was done as described by using 5 μm of acetone-fixed frozen sections (35). CD3+ T lymphocytes were stained by using rat anti-mouse monoclonal antibody to the CD3 antigen (clone C363.29B; Southern Biotechnology Associates, Birmingham, AL) and FITC-conjugated anti-rat. CD4+ T cells were visualized by using biotin-conjugated rat anti-CD4 (clone RM4–5; BD Biosciences) and avidin Texas red (Vector Laboratories). α-SMA was detected by using a rabbit anti-α-SMA polyclonal antibody (Lab Vision, Fremont, CA). Staining was quantitated by using kinetic histometrix 6 imaging and analysis software (Kinetic Imaging).

Flow Cytometry.

For flow cytometric analyses, spleens and thymi were removed and processed through a 0.70-μm mesh screen as described in ref. 35. Cells were counted, resuspended in 1.0% FBS/PBS (0.1% sodium azide), and incubated with the appropriate antibody for 30 min at 4°C. Cells were washed and analyzed by using a FacsCalibur flow cytometer (Becton Dickinson) and flowjo software.

ELISAs.

Serum antibody titers against oxLDL and cardiolipin were measured as described in ref. 34. Titers of dsDNA antibodies were measured by using the method of Morel et al. (33).

Serum Lipoprotein Analyses.

Total serum cholesterol and triglyceride levels were measured by using a colorimetric assay as described in ref. 34. Lipoprotein distribution was determined by using FPLC.

Determination of Renal Disease.

Renal function was assessed by measuring protein in urine at the time of killing by using a urine Multistix 10SG (Bayer). Serum levels of creatinine and urea were measured by specific quantitative colorimetric assay as per manufacturer’s specifications (BioAssay Systems, Hayward, CA).

Measurement of Systolic Blood Pressure.

Systolic blood pressure was monitored by tail cuffing on conscious, preconditioned mice by using a BP-2000 (Visitech Systems, Apex, NC) and accompanying software, available at the Mouse Metabolic Phenotype Core facility at Vanderbilt University. A total of three measurements were done on each animal, and the average reading was used.

Statistical Analyses.

Statistical analyses were conducted by using prism 3.03 software. Unless otherwise stated, differences between groups were determined by using a Mann–Whitney nonparametric t test. Statistically significant differences in urine protein grade were determined by χ2 analysis. Correlation analyses were conducted by using a Spearman analysis. A P value of <0.05 was considered significant.

Previous SectionNext Section

Acknowledgments

We thank Drs. Luc Van Kaer and Mirsanda Stanic for thoughtful reading of the manuscript and Tiffany N. Crouch, Jennifer L. McCaleb, and Youmin Zhang for technical assistance. This work was supported by National Institutes of Health (NIH) Building Interdisciplinary Research Careers in Women’s Health (BIRCWH) Grant 5 K12 HD043483-04, American Heart Association Scientist Development Grant 0330412N, a small project grant from the Nashville Chapter of the Lupus Foundation of America, a grant from the Lupus Research Institute (to A.S.M.), and NIH Grants HL57986 and HL65709 (to S.F.) and HL65405 and HL53989 (to M.F.L). We also acknowledge the Vanderbilt Mouse Metabolic Phenotyping Centers (NIH Grant DK59637-01).

Previous SectionNext Section

Footnotes

  • **To whom correspondence should be addressed at:
    Division of Cardiovascular Medicine, Department of Medicine, Vanderbilt University School of Medicine, Room 383, PRB, 2220 Pierce Avenue, Nashville, TN 37232-6300.
    E-mail: amy.major{at}vanderbilt.edu

  • Author contributions: A.K.S., C.M.S., N.J.O., and A.S.M. designed research; A.K.S., A.C.M., and A.S.M. performed research; S.F., M.F.L., and E.K.W. contributed new reagents/analytic tools; A.K.S., A.C.M., N.J.O., and A.S.M. analyzed data; and A.K.S., C.M.S., and A.S.M. wrote the paper.

  • Conflict of interest statement: No conflicts declared.

  • Abbreviations:

    Abbreviations:

    CVD,
    cardiovascular disease;
    LDL,
    low-density lipoprotein;
    LDLr,
    LDL receptor;
    LDLr.B6,
    LDLr and C57BL/6 chimeras;
    LDLr.Sle,
    LDLr and B6.Sle1.2.3 chimeras;
    oxLDL,
    oxidized LDL;
    SLE,
    systemic lupus erythematosus.
Previous Section
 

References

  1. ↵
    1. Urowitz M. B.,
    2. Bookman A. A.,
    3. Koehler B. E.,
    4. Gordon D. A.,
    5. Smythe H. A.,
    6. Ogryzlo M. A.
    (1976) Am. J. Med 60:221–225.
    CrossRefMedlineISI
  2. ↵
    1. Manzi S.,
    2. Selzer F.,
    3. Sutton-Tyrrell K.,
    4. Fitzgerald S. G.,
    5. Rairie J. E.,
    6. Tracy R. P.,
    7. Kuller L. H.
    (1999) Arthritis Rheum 42:51–60.
    CrossRefMedlineISI
  3. ↵
    1. Asanuma Y.,
    2. Oeser A.,
    3. Shintani A. K.,
    4. Turner E.,
    5. Olsen N.,
    6. Fazio S.,
    7. Linton M. F.,
    8. Raggi P.,
    9. Stein C. M.
    (2003) N. Engl. J. Med 349:2407–2415.
    Abstract/FREE Full Text
  4. ↵
    1. Frostegard J.
    (2005) J. Intern. Med 257:485–495.
    CrossRefMedlineISI
  5. ↵
    1. Wick G.,
    2. Knoflach M.,
    3. Xu Q.
    (2004) Annu. Rev. Immunol 22:361–403.
    CrossRefMedlineISI
  6. ↵
    1. Ross R.
    (1993) Nature 362:801–809.
    CrossRefMedline
  7. ↵
    1. Sherer Y.,
    2. Tenenbaum A.,
    3. Praprotnik S.,
    4. Shemesh J.,
    5. Blank M.,
    6. Fisman E. Z.,
    7. Motro M.,
    8. Shoenfeld Y.
    (2002) Cardiology 97:2–5.
    MedlineISI
  8. ↵
    1. George J.,
    2. Harats D.,
    3. Shoenfeld Y.
    (2000) Clin. Rev. Allergy Immunol 18:73–86.
    CrossRefMedlineISI
  9. ↵
    1. Bacon P. A.,
    2. Stevens R. J.,
    3. Carruthers D. M.,
    4. Young S. P.,
    5. Kitas G. D.
    (2002) Autoimmun. Rev 1:338–347.
    CrossRefMedline
  10. ↵
    1. Morel L.,
    2. Yu Y.,
    3. Blenman K. R.,
    4. Caldwell R. A.,
    5. Wakeland E. K.
    (1996) Mamm. Genome 7:335–339.
    CrossRefMedlineISI
  11. ↵
    1. Mohan C.,
    2. Alas E.,
    3. Morel L.,
    4. Yang P.,
    5. Wakeland E. K.
    (1998) J. Clin. Invest 101:1362–1372.
    MedlineISI
  12. ↵
    1. Mohan C.,
    2. Morel L.,
    3. Yang P.,
    4. Wakeland E. K.
    (1997) J. Immunol 159:454–465.
    Abstract
  13. ↵
    1. Mohan C.,
    2. Yu Y.,
    3. Morel L.,
    4. Yang P.,
    5. Wakeland E. K.
    (1999) J. Immunol 162:6492–6502.
    Abstract/FREE Full Text
  14. ↵
    1. Sobel E. S.,
    2. Mohan C.,
    3. Morel L.,
    4. Schiffenbauer J.,
    5. Wakeland E. K.
    (1999) J. Immunol 162:2415–2421.
    Abstract/FREE Full Text
  15. ↵
    1. Zhu J.,
    2. Liu X.,
    3. Xie C.,
    4. Yan M.,
    5. Yu Y.,
    6. Sobel E. S.,
    7. Wakeland E. K.,
    8. Mohan C.
    (2005) J. Clin. Invest 115:1869–1878.
    CrossRefMedlineISI
  16. ↵
    1. Wakeland E. K.,
    2. Wandstrat A. E.,
    3. Liu K.,
    4. Morel L.
    (1999) Curr. Opin. Immunol 11:701–707.
    CrossRefMedlineISI
  17. ↵
    1. Roman M. J.,
    2. Salmon J. E.,
    3. Sobel R.,
    4. Lockshin M. D.,
    5. Sammaritano L.,
    6. Schwartz J. E.,
    7. Devereux R. B.
    (2001) Am. J. Cardiol 87:663–666, A11.
    CrossRefMedlineISI
  18. ↵
    1. Gu L.,
    2. Weinreb A.,
    3. Wang X. P.,
    4. Zack D. J.,
    5. Qiao J. H.,
    6. Weisbart R.,
    7. Lusis A. J.
    (1998) J. Immunol 161:6999–7006.
    Abstract/FREE Full Text
  19. ↵
    1. Gu L.,
    2. Johnson M. W.,
    3. Lusis A. J.
    (1999) Arterioscler. Thromb. Vasc. Biol 19:442–453.
    Abstract/FREE Full Text
  20. ↵
    1. Mizutani H.,
    2. Engelman R. W.,
    3. Kinjoh K.,
    4. Kurata Y.,
    5. Ikehara S.,
    6. Good R. A.
    (1993) Blood 82:3091–3097.
    Abstract/FREE Full Text
  21. ↵
    1. Kirzner R. P.,
    2. Engelman R. W.,
    3. Mizutani H.,
    4. Specter S.,
    5. Good R. A.
    (2000) Biol. Blood Marrow Transplant 6:513–522.
    CrossRefMedlineISI
  22. ↵
    1. Linton M. F.,
    2. Babaev V. R.,
    3. Gleaves L. A.,
    4. Fazio S.
    (1999) J. Biol. Chem 274:19204–19210.
    Abstract/FREE Full Text
  23. ↵
    1. Shapiro R. J.
    (1993) Metabolism 42:162–169.
    CrossRefMedlineISI
  24. ↵
    1. Bro S.,
    2. Bentzon J. F.,
    3. Falk E.,
    4. Andersen C. B.,
    5. Olgaard K.,
    6. Nielsen L. B.
    (2003) J. Am. Soc. Nephrol 14:2466–2474.
    Abstract/FREE Full Text
  25. ↵
    1. Roman M. J.,
    2. Shanker B. A.,
    3. Davis A.,
    4. Lockshin M. D.,
    5. Sammaritano L.,
    6. Simantov R.,
    7. Crow M. K.,
    8. Schwartz J. E.,
    9. Paget S. A.,
    10. Devereux R. B.,
    11. Salmon J. E.
    (2003) N. Engl. J. Med 349:2399–2406.
    Abstract/FREE Full Text
  26. ↵
    1. van der Wal A. C.,
    2. Becker A. E.
    (1999) Cardiovasc. Res 41:334–344.
    FREE Full Text
  27. ↵
    1. Wakeland E. K.,
    2. Morel L.,
    3. Mohan C.,
    4. Yui M.
    (1997) J. Clin. Immunol 17:272–281.
    CrossRefMedlineISI
  28. ↵
    1. Hayakawa K.,
    2. Hardy R. R.,
    3. Herzenberg L. A.
    (1985) J. Exp. Med 161:1554–1568.
    Abstract/FREE Full Text
  29. ↵
    1. Palinski W.,
    2. Horkko S.,
    3. Miller E.,
    4. Steinbrecher U. P.,
    5. Powell H. C.,
    6. Curtiss L. K.,
    7. Witztum J. L.
    (1996) J. Clin. Invest 98:800–814.
    MedlineISI
  30. ↵
    1. Shaw P. X.,