Background
Both innate and adaptive immune responses contribute to the arterial inflammation that characterizes atherosclerosis [
1‐
3]. Mice lacking critical components of innate immunity, such as the Toll-like receptor (TLR) 2 and 4 and the TLR signaling protein MyD88, develop less atherosclerosis, indicating the involvement of pro-atherogenic endogenous TLR ligands [
4‐
6]. The role of adaptive immunity in atherosclerosis is more complex. While there is strong evidence that Th1 cells aggravate atherosclerosis [
7‐
9] the possible influence of Th2 cells is less clear [
10,
11]. Regulatory T cells (Tregs) [
12,
13] and B cells [
14] appear to have protective functions. A common feature of the studies that have revealed these associations is that atherosclerosis has been induced by hypercholesterolemia. Accordingly, it is likely that the immune responses that contributed to atherosclerosis development in these animals have been activated by ligands and antigens generated by hypercholesterolemia. The exact identity of these factors, as well as their mode of action, remains to be fully characterized. Attention has focused on the role of oxidized low-density lipoprotein (LDL) [
15]. LDL particles become oxidized by various enzymes and oxygen metabolites when entrapped in the extra cellular matrix of the artery wall [
16]. Oxidized LDL is targeted by both IgM and IgG autoantibodies [
17] and as much as 10% of the T cells present in atherosclerotic plaques are specific for antigens formed in oxidized LDL [
18]. T cells specific for oxidized LDL are also present in the circulation [
19] and transfer of CD4
+ T cells isolated from mice immunized with aldehyde-modified LDL results in a more aggressive development of atherosclerosis [
20] providing direct evidence for a pathogenic role of adaptive immunity against modified LDL in the disease process. Based on this knowledge attempts have been made to develop immunomodulatory therapy for prevention of cardiovascular disease and pilot vaccines containing apolipoprotein B (apo B) antigens have been shown to significantly reduce atherosclerosis in apolipoprotein E deficient (
Apoe
-/-
) mice [
21‐
23]. A limiting factor in the development of these therapies has been the poor understanding of the immune pathways activated by hypercholesterolemia [
24]. In the present studies we aimed to address this issue by characterizing the induction of adaptive immunity to hypercholesterolemia both systemically and in regional lymph nodes draining lesion-prone areas of the aorta. We used
Apoe
-/-
mice in which a primary immune response to hypercholesterolemia-associated antigens, such as oxidized LDL, develops spontaneously [
25]. To increase the antigen load we fed the mice a high-fat diet.
Methods
Animals
Female apolipoprotein E deficient mice on a C57BL/6 background were purchased from Taconic, USA. The animals were kept under controlled laboratory conditions in individually ventilated cages and food and water were provided ad libitum. All mice received chow diet until the age of 10 weeks. One group (n = 27) was then transferred to a high fat diet with 0.15% cholesterol and 21% fat (Lantmännen, Sweden) while the other group (n = 24) remained on chow diet. Mice were killed 4 weeks (high fat diet fed; n = 14, chow fed; n = 12) and 8 weeks (high fat diet fed; n = 13, chow fed; n = 12) after diet change, tissues were harvested and analyzed. The experiments were approved by the Animal Care and Use Committee of Lund University.
Analysis of plaque autoantibody, apolipoprotein B and oxidized LDL content
The heart and proximal part of the aortic arch was embedded in OCT (Tissue-Tek). Frozen sections of 10 μm were collected from the subvalvular region. For detection of IgG or IgM, slides were fixed in ice-cold acetone for 5 min and blocked with 10% mouse serum in PBS for 30 min. To detect IgM and IgG autoantibodies, slides were incubated with biotinylated anti-mouse IgM or IgG antibodies (Vector Laboratories) for 50 min at room temperature. For the detection of apo B and oxidized LDL, slides were fixed in ice-cold acetone for 5 min and blocked with 10% goat serum in PBS for 50 minutes and incubated overnight at 4°C with primary rabbit anti-mouse apo B (Abcam) or anti-human malondialdehyde (MDA)-apo B peptide (2D03, Bioinvent, Sweden) antibodies diluted in 10% goat serum in PBS. A biotinylated goat anti-rabbit IgG antibody (Vector Laboratories) diluted in PBS was used as secondary antibody. The reaction products were visualized with Vectastain ABC elite kit (Vector Laboratories) using DAB as substrate (Vector Laboratories). Slides were counter-stained with hematoxylin and omission of primary or secondary antibodies were used as controls. Immune-stained areas were quantified with BioPix iQ 2.0 software (Biopix, Sweden).
Cell preparation and flow cytometry
Mice were fasted for at least 3 hours before anesthetization by intraperitoneal injection of Xylazine (Rompun, Bayer Health care), and Ketamine (Ketalar, Pfizer) followed by euthanization by heart puncture. The spleen, mediastinal, brachial, axial, renal, iliac, sacral lymph nodes and thymus were removed and meshed through a cell strainer (70 μm, BD Bioscience). The single cell suspension of the lymph nodes and thymus were washed in RPMI medium (Gibco) and resuspended in complete medium (RPMI 1640 supplemented with 10% FCS, 1% Sodium pyruvate, 1% Hepes, 1% Penicillin/Streptomycin, 1% L-Glutamine and 0.1% β mercaptoethanol [Gibco]). Splenocytes were pelleted and resuspended in red blood cell lysing buffer (Sigma) for 2 minutes at room temperature to remove erythrocytes. Cells were washed and distributed in a 96 well round bottom plate (Sarstedt) at a density of 2 × 106 cells/ml and incubated overnight in a humidified chamber at 37°C and 5% CO2. Day 2, the cells were incubated with 5 μg/ml brefeldin A (eBioscience), 50 ng/ml PMA and 1 μg/ml ionomycin (all from Sigma) for 4 hours. Cells were incubated with a Fc-receptor blocking antibody (FcR;CD16/32; clone 93, Biolegend) for 5 minutes followed by incubation with CD28-PE/Cy5 (clone 37.51), CD4-PE/Cy7 (GK1.5), CD25-APC (PC61) and CD8-APC/Cy7 (53-6.7 all Biolegend) at 4°C for 30 minutes. The cells were resuspended in Fix/Perm solution (eBioscience), washed with permeabilization buffer (eBioscience) and blocked with FcR block prior to incubation with FoxP3-PB (clone FJK-16 S, eBioscience), IFN-γ- FITC (clone XMG1.2, Biolegend), IL-10-PE (clone JES5-16E3, Biolegend) for 30 minutes at 4°C. Cells were washed with permeabilization buffer and resuspended in FC buffer (1% fetal calf serum [Gibco] and 0.5 mM EDTA in phosphate buffered saline). Measurements were performed using a Cyan ADP (Beckman Coulter), analyses were performed using Summit (Beckman Coulter, version 4.3) and gating was adjusted using a negative control staining.
Proliferation assay
Splenocytes (5 × 106) were centrifuged and resuspended in 1 ml PBS. The cells were incubated with 10 μL Carboxyfluorescerein Succinimidyl ester (CFSE; Invitrogen, diluted 1:10 in PBS) for 5 minutes. The reaction was stopped with addition of 1 ml FCS, cells were washed with PBS and resuspended in 5 ml complete medium (see above). Cells (1 × 106/ml) were transferred to a 96 well round bottom plate (Sarstedt) and incubated in presence or absence of 2.5 μg/ml Concanavalin A (ConA; Sigma) for 4 days at 37°C and 5% CO2. Cells were stained with FcR block (CD16/32; Biolegend) prior to staining with CD3-PE/Cy5 (clone 145-2C11), CD4-PE/Cy7 and CD8a-APC/Cy7 (all Biolegend). Stained cells were acquired in a gate comprising 50 000 cells and statistics were calculated on the number of CD4+ and CD8+ cells that had proliferated more than one time. Cells were acquired using a Cyan ADP (Beckman Coulter) and analyses were performed using Summit (Beckman Coulter, version 4.3). For cytokine analysis 5 × 105 splenocytes were cultured alone or with 2.5 μg/ml Con A for 72 hours and cytokine concentrations were measured in the cell culture supernatant using a Th1/Th2 9-plex (Meso Scale Discovery) according to the instructions of the manufacturer. The lower detection limit in this assay was 5 pg/ml.
Plasma oxidized-LDL specific antibodies and Apolipoprotein B immune complex analysis
Plasma oxidized-LDL-specific antibodies were analyzed in Cu
2+ oxidized (10 μg/mL in PBS) or MDA-p210 peptide (KTTKQ SFDLS VKAQY KKNKH; 10 μg/mL in PBS) coated microtiter plates [
26]. Plasma antibodies were detected using antibodies recognizing mouse IgG, IgM (Jackson ImmunoResearch), IgG1 (557272, BD Pharmingen) and IgG2a (553389, BD Pharmingen; cross reacts with IgG2c). In the apolipoprotein B immune complex assay, microtiter plates were coated with rabbit anti-apolipoprotein B antibody (ABcam), and the plasma apo B containing immune complexes were detected with the anti-mouse IgG1 and IgG2a detection antibodies described above.
Analysis of cholesterol and triglyceride content
Total plasma cholesterol and plasma triglycerides were quantified with colorimetric assays, Infinity™ Cholesterol and Triglyceride (Thermo Electron).
Statistical analysis
Analysis of data was performed using unpaired t test or Mann Whitney test for skewed data. Data are presented as mean ± standard deviation. Analysis was performed using GraphPad Prism 5.01 (Graphpad software) and a level of P < 0.05 was considered significant.
Discussion
It is well established that adaptive immune responses induced by hypercholesterolemia play an important role in the development of atherosclerosis [
2,
3], but the pathways involved remain to be fully characterized. In the present study we assessed the response of CD4
+ and CD8
+ T cells, as well as humoral immunity, to a high-fat diet. At 4 weeks the immune response to an aggravated hypercholesterolemia was found to involve expression of IFN-γ by CD8
+CD28
+ T cells in plaque-draining lymph nodes and in the spleen. At the latter location this pro-inflammatory response was counterbalanced by a concomitant release of IL-10 from both CD4
+ and CD8
+ T cells, whereas no induction of IL-10 occurred in plaque-draining lymph nodes. However, following an extended 8 weeks of high-fat diet activation of CD4
+ T cells could be observed also in draining lymph nodes. As opposed to the CD4
+ T cells in the spleen that remained in a phenotype characterized by expression of IL-10, the mediastinal lymph node CD4
+ T cells by then had differentiated into IFN-γ secreting Th1 cells. Our findings imply that when antigens generated in a hypercholesterolemic state are acquired systemically and presented to T cells in the spleen this results in differentiation of CD4
+ T cells into IL-10 producing Th2 cells. This notion is in accordance with previous studies by Zhou and coworkers demonstrating that diet-induced hypercholesterolemia is associated with a switch in oxidized LDL-specific IgG towards a Th2 isotype [
27]. Our observation of an increased capacity of splenocytes from mice given high-fat diet to secrete IL-4 is also in accordance with a hypercholesterolemia-dependent shift towards Th2. In contrast, when antigens generated by hypercholesterolemia are acquired in atherosclerotic lesions and presented to T cells in draining lymph nodes this results in activation of Th1 cells. The reason why antigens generated by hypercholesterolemia activates a Th2 response systemically but a Th1 immune response when acquired by antigen presenting cells (APC) in atherosclerotic lesions remains to be elucidated. One possibility could be that the pro-inflammatory environment of the atherosclerotic plaque activates APC to express co-stimulatory molecules and cytokines in a manner that subsequently primes naïve T cells in draining lymph nodes to Th1 differentiation. Supporting this notion, dendritic cells pulsed with oxLDL aggravate atherosclerosis or are atheroprotective depending on the priming environment. Habets
et al. used infused DCs pulsed with copper-oxidized LDL into
LDLr
-/-
hosts which reduced carotid lesion size and increased plaque stability [
28]. On the contrary, MDA-LDL pulsed DCs aggravated atherosclerosis in
Apoe
-/-
mice [
29]. Interestingly, immunization with MDA-LDL in complete Freunds adjuvant is atheroprotective [
30], indicating that
ex vivo primed DCs generate a pro-inflammatory response whereas in vivo primed DCs are anti-inflammatory. Gautier
et al. reported that DCs in
Apoe
-/-
and
LDLr
-/-
mice are Th1 and Th17 disposed, while removal of DCs increase hypercholesterolemia, further supporting the dual role of DCs in atherosclerosis [
31]. Since the capability of DCs to prime CD4
+ T cells is equivalent in normo- and hypercholesterolemic mice [
32], DCs may initiate the Th1 priming associated with aggravated atherosclerosis as well as the Th2 response present in splenocytes of mice given high fat diet. It is also possible that Th1 priming of CD4
+ T cells in the mediastinal lymph node could be aided by the IFN-γ secreting CD8
+ CD28
+ T cells that develop in draining lymph nodes prior to the occurrence of CD4
+ Th1 cells.
An unexpected finding in the present study was that the initial pro-inflammatory response to hypercholesterolemia was started by CD8
+ T cells and not by CD4
+ T cells. The role of CD8
+ T cells in atherosclerosis has not been extensively studied. CD8
+ T cells are present in atherosclerotic lesions of
Apoe
-/-
mice, but are less frequent than CD4
+ T cells [
33]. However, in advanced human lesions the CD8
+ T cells may account for up to 50% of the entire leukocyte population [
34]. Targeting of a CD8
+ T cell immune response against arterial smooth muscle cells has been shown to markedly aggravate atherosclerosis in
Apoe
-/-
mice [
35]. CD8
+ T cells have also been implicated in the accelerated atherosclerosis that occurs in patients with autoimmune diseases such as systemic lupus erythematosus [
36]. In a study by Elhage and coworkers [
37] CD8 deficiency did not influence atherosclerosis development in
Apoe
-/-
mice. However, in the same study it was unexpectedly found that deficiency of CD4 increased plaque formation. The authors concluded that this could be due to the expansion of CD8
+ T cell population that characterizes CD4
-/- mice.
Our findings do not permit any clear conclusion regarding the antigen(s) responsible for hypercholesterolemia-induced immune activation. It is generally assumed that oxidized LDL is a dominant autoantigen generated by hypercholesterolemia [
38]. In agreement with previous studies oxidized LDL IgG and IgM autoantibodies were present in chow-fed
Apoe
-/-
mice [
25] and the levels did not change significantly in response to high-fat diet. In contrast, mice fed a high-fat diet were found to have higher levels of apo B/IgM immune complexes in the circulation which could reflect an increase in the amount of circulating oxidized LDL as well as an enhanced generation of anti-oxidized LDL IgM. There was however no change in the anti-oxLDL IgG1/IgG2c ratio. This observation is in contrast to studies by Zhou
et al. [
27] demonstrating a shift towards Th2 isotype anti-oxidized LDL IgG in severe hypercholesterolemia. The reason for this discrepancy remains to be clarified but may involve the difference in antigens used in the ELISAs. However, the observation that high-fat diet increased the capacity of splenocytes to secrete IL-4 suggests that hypercholesterolemia stimulated a shift towards Th2 also in the present study. In order to further develop the assay, it would be interesting to determine if the origin of IL-4 production is CD4
+ or CD8
+ T cell specific. An important limitation that should be kept in mind when interpreting the present findings is that mouse models of atherosclerosis may not be entirely representative of the pathophysiology of human atherosclerosisis [
39]. Moreover, it is also unlikely that the sudden increase in plasma cholesterol levels observed in response to high fat diet in the present study would occur in humans. Nevertheless, in spite of these limitations, our findings provide support for a role of CD8
+ T cells in immune responses activated by hypercholesterolemia.
The accumulation of oxidized LDL in atherosclerotic lesions was correlated to lesion size. In spite of this there were no signs of activation of CD4
+ T cells in draining lymph nodes 4 weeks after diet change, suggesting a lack of reactivity against extracellular antigens such as oxidized LDL. The observation of increased IFN-γ expression in CD8
+CD28
+ T cells instead point to an initial immune response against cell antigens. Although immune responses against apoptotic cells generally are tolerogenic and inhibit atherosclerosis [
40] it can not be excluded that this represents an immune response against damaged vascular cells.
Acknowledgements
We thank Ingrid Söderberg, Irena Ljungcrantz and Ragnar Alm for expert technical assistance. This study was supported by grants from the Swedish Medical Research Council, the Swedish Heart-Lung foundation, the Crafoord foundation, the Knut and Alice Wallenberg foundation, the Söderberg foundation, VINNOVA, the Albert Påhlsson foundation, the Malmö University Hospital foundation, European Community's Sixth Framework Programme contract ("IMMUNATH") LSHM-CT-2006-037400, Swedish Foundation of Strategic Research and the Lundström foundation.
Authors' contributions
DK carried out the cell work, flow cytometry and data analysis. OHR participated in the cell work and flow cytometry. KEB participated in titration of assays and study design. JP carried out the immunohistochemistry. MW participated in the proliferation assays. HB, GNF and JN participated in the design and coordination of the study and JN wrote the manuscript. All authors read and approved the final manuscript.