Transient Role for CD1d-Restricted Natural Killer T Cells in the Formation of Atherosclerotic Lesions
Objective— CD1d-restricted natural killer T (NKT) cells are reported to play a proatherogenic role in the development of atherosclerosis. However, the contribution of NKT cells to mature lesion formation and the effector mechanisms through which they act are unknown.
Methods and Results— We measured lesion size in CD1d-null (CD1d−/−) mice on the low-density lipoprotein (LDL) receptor–deficient (LDLR−/−) genetic background after 4, 8, and 12 weeks of feeding on a Western diet. Lesions in CD1d−/−LDLR−/− mice were 47% smaller at 4 weeks than CD1d+/+LDLR−/− controls; however, there were no differences in lesion size between CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice at 8 or 12 weeks. We found that although NKT cells were present in the aortic arch of CD1d+/+LDLR−/− mice on the Western diet, no differences in mRNA abundance for Th1 or Th2 cytokines were observed between CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice.
Conclusions— CD1d-restricted NKT cells contribute to the formation of fatty streaks; however, their influence on lesion progression is transient, and they do not significantly affect the inflammatory cytokine milieu of mature lesions.
Accumulating evidence indicates that atherosclerosis is a progressive inflammatory condition in which innate and adaptive immune mechanisms play key roles in plaque formation. Recently, natural killer T (NKT) cells, a specialized subset of T lymphocytes that recognize glycolipid antigens presented by the nonpolymorphic CD1d protein,1,2 have been reported to accelerate atherosclerosis.
NKT cells are important in a diverse array of immune functions, including antimicrobial immunity, prevention of autoimmunity, tolerance induction, and tumor rejection. Upon stimulation, NKT cells rapidly secrete interferon (IFN)-γ or interleukin (IL)-4, resulting in the activation of conventional T3 and B4 lymphocytes, macrophages,5 and NK cells.6,7 Thus, NKT cells can influence the Th1/Th2 polarization of the immune response in the periphery.8–10 The majority of murine NKT cells that interact with CD1d express a semi-invariant T cell antigen receptor (TCR) consisting of a Vα14-Jα18 α-chain paired with a preferential β-chain repertoire and are referred to as invariant NKT (iNKT) cells.10,11 iNKT cells may also be defined by their reactivity to the potent antigen α-galactosylceramide (α-GalCer), an α-anomeric glycolipid purified from marine sponge.12
CD1 proteins that present lipid antigen to NKT cells have been identified in human atherosclerotic plaques.13 CD1 expression is localized to areas of T-cell infiltration, suggesting a functional association between CD1+ antigen presenting cells and T cells within the lesion.13 Further, foam cells generated in vivo with oxidized and acetylated low-density lipoprotein (LDL) can induce a proliferative response in CD1-restricted NKT cell lines.13 Consistent with these observations, CD1d-null (CD1d−/−) mice or wild-type irradiated mice reconstituted with CD1d−/− bone marrow have recently been reported to have reduced atherosclerosis after feeding on a Western diet.14 However, because these studies have focused on lesion formation at single early time points, it is unknown whether or not NKT cells contribute to the formation of mature plaques. Furthermore, the mechanism(s) through which NKT cells promote lesion development is unclear; in particular, it is not known if they alter the cytokine milieu of the plaque.
To address these questions, we crossed CD1d−/− mice with LDL receptor–deficient (LDLR−/−) mice and quantified the recruitment of NKT cells, lesion development, and cytokine mRNA in the aortic arch after feeding on a Western diet for 4, 8, and 12 weeks. Our data demonstrate that although lesion size at the earliest time point was reduced in CD1d−/−LDLR−/− mice, the effect was transient. Further, there was no significant difference in cytokine mRNA abundance in the aortas of CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice. These data indicate that the effects of NKT cells are limited to the formation of early fatty streak lesions.
CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice were generated by crossing CD1d−/− mice15 with LDLR−/− mice. CD1d−/− mice were provided by Dr Luc Van Kaer (Vanderbilt University, Nashville, Tenn), and LDLR−/− mice were purchased from Jackson Laboratories (Bar Harbor, Me). Mice were backcrossed at least 10× onto the C57Bl/6 genetic background before mating. Genotyping for LDLR was performed by polymerase chain reaction (PCR). Mice were weaned at 21 days of age and fed a chow diet containing 4.5% (w/w) fat (Ralston Purina) for 14 additional days. At 5 weeks of age, the mice were switched to a high-fat Western diet containing 1.25% (w/w) cholesterol (40% kcal from fat) without added cholate (D12108; Research Diets) and maintained for 4, 8, and 12 weeks. Mice were matched for gender in each control and experimental group.
The extent of atherosclerosis in the total aorta was quantitated by digital morphometry without knowledge of the genotype of the mice as described.16 For analysis of atherosclerotic lesions in the proximal aortic root, 8 CD1d−/−LDLR−/− and 8 CD1d+/+LDLR−/− mice were examined after 4 weeks on the Western diet. The extent of atherosclerosis was quantitated as described16 and is reported as the mean lesion area of all sections analyzed in each mouse.
Plasma samples were collected by cardiac puncture at the time of euthanization after a 4-hour fast. Total cholesterol was determined with a colorimetric assay (Spectrum Cholesterol Assay; Abbott Laboratories). Plasma samples from mice with similar total cholesterol levels were pooled and fractionated by fast-performance liquid chromatography on a Superose 6 column (Amersham Biosciences).
Quantitatative Real-Time PCR
Mice were perfused with RNAlater RNA stabilization solution (Ambion), and the proximal aortas from the brachiocephalic trunk to the left subclavian artery were placed in lysis buffer and homogenized. Total RNA was isolated with a QIAGEN fibrous tissue RNeasy Kit, and contaminating genomic DNA was removed with a DNA-free Kit (Ambion). The integrity and purity of the RNA samples was analyzed on an Agilent 2100 Bioanalyzer using the RNA 6000 laboratory chip.
Total RNA was reverse transcribed at 42°C with Superscript II (Invitrogen) and oligo d(T)12 to 18 primers for 120 minutes. RNase H (New England Biolabs) was then added to remove template RNA. Duplicate reactions in the absence of reverse transcriptase were also performed.
To quantitate the expression levels of specific genes, target cDNA was amplified in the presence of gene-specific primers (0.4 μmol/L) and fluorescent probe (100 nM) using a TaqMan universal PCR Master Mix (Applied Biosystems). For measurement of IFN-γ, 6 μL of cDNA was used for each control and experimental PCR reaction, whereas for all other genes 2 μL of cDNA was used. A pilot experiment confirmed linear amplification efficiency for the internal control gene within this range of input cDNA.
The IFN-γ, IL-12, IL-10, IL-4, and tumor necrosis factor (TNF)-α primers and probes were from Applied Biosystems. The primers and probes for cyclophilin A, CD68, CD69, and the Vα14-Jα18 α-chain were from Biosearch Technologies and were identical to published sequences.17,18 The sequences for CD69 were 5′-GCTCCCTGTGTGTCTGATCTTG-3′ (Fwd), 5′-GGCCTTTTTGTGAGGACAGTCT-3′ (Rev), and 5′-6-FAM-TAGGACCGCTTATAGCCCAAGGAACAGAGG-BHQ-1-3′ (Probe).
PCR reactions were performed on an ABI Prism 7900 Sequence Detection System as follows: 1 cycle of 95°C for 10 minutes and 45 cycles of 95°C for 15 seconds followed by 60°C for 1 minute. Reactions for each sample were performed in triplicate, and the expression levels of all genes were normalized to cyclophilin A. The lowest-expressing sample of the control group was taken as the calibrator, and all samples were expressed as a ratio of the calibrator sample calculated by the comparative cycle threshold method (2−ΔΔCt). All −reverse transcriptase control samples were assayed for the presence of genomic DNA with the cyclophilin A primer set and were found to be negative after 45 cycles of amplification.
Differences in the extent of atherosclerosis, plasma cholesterol levels, weights, and gene expression were analyzed with the Mann–Whitney nonparametric U test (InStat 2.01; GraphPad Software). P-values <0.05 were considered significant.
Atherosclerotic Lesion Formation Is Reduced in CD1d-Deficient Mice
To examine the role of CD1d-restricted NKT cells in the progression of atherosclerosis, CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice were fed a Western diet for 4, 8, and 12 weeks before euthanization and quantification of lesion size. After 4 weeks on the Western diet, lesions consisted of small fatty streaks covering an average of 1% of the total aorta in CD1d+/+LDLR−/− mice (Figure 1). In male mice, the lesions were 40.4% smaller in CD1d−/−LDLR−/− than in CD1d+/+LDLR−/− mice (P=0.016, Figure 1a). In female mice, there was a trend toward smaller lesions in the CD1d−/−LDLR−/− mice (41.5%), but the difference did not reach statistical significance. After 8 or 12 weeks on the Western diet, there were no significant differences in lesion area between the genotypes (Figure 1b). Oil red O–stained serial sections of the aortic root revealed that CD1d−/−LDLR−/− mice had 31.7% less (P=0.03) lipid staining than CD1d+/+LDLR−/− mice (Figure 2) at 4 weeks. Thus, both the surface area and thickness of the lesions were significantly smaller in CD1d−/−LDLR−/− than in CD1d+/+LDLR−/− mice at the earliest time point.
There were no statistically significant differences in total cholesterol or weight between any of the groups of mice (data not shown). Analysis of pooled plasma revealed slightly lower levels of very low–density lipoprotein (VLDL) in CD1d−/−LDLR−/− than CD1d+/+LDLR−/− mice (Figure 3). The reduced VLDL in CD1d−/−LDLR−/− mice did not correlate with lesion size, however, as the difference in lipoproteins remained present at 8 and 12 weeks when the average lesion area between the 2 groups was indistinguishable (Figure 1).
Macrophage Recruitment Is Unaffected in CD1d-Deficient Mice
To determine whether the reduced atherosclerosis measured at 4 weeks reflected impaired recruitment of macrophages, we examined the expression of CD68 in the aortic arch. Despite the smaller average lesion size in whole aorta (Figure 1) and aortic root (Figure 2) in CD1d−/−LDLR−/− compared with CD1d+/+LDLR−/− mice, real-time PCR analysis of CD68 in the aortic arch revealed no difference between the 2 groups after 4 weeks of feeding on the Western diet (Figure 4). Similarly, immunohistochemical analysis of serial sections of the aortic root for expression of the cell-surface macrophage protein MOMA-2 revealed no difference between CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice (data not shown).
CD1d-Restricted NKT Cells Accumulate in the Aortic Arch
The presence of iNKT cells in the proximal aorta between the brachiocephalic trunk and left subclavian artery was determined by quantifying Vα14-Jα18 TCR mRNA by real-time PCR. CD1d+/+LDLR−/− mice fed a Western diet for 4 weeks had detectable levels of Vα14-Jα18 TCR mRNA, indicating the presence of iNKT cells in atherosclerosis-prone areas of the aortic arch (Figure 5). The amount of Vα14-Jα18 mRNA did not increase significantly between 4 and 8 weeks, and was similar to that in CD1d+/+LDLR−/− mice fed standard chow (Figure 5), indicating that iNKT cell recruitment under hypercholesterolemic conditions was transient.
Cytokine mRNA Is Unchanged in Aortas of CD1d-Deficient Mice
To determine whether alterations in the cytokine profile in CD1d−/−LDLR−/− mice might have influenced lesion progression, we measured mRNA for the T-cell activation marker CD69 and Th1 and Th2 cytokines in the aortic arch of mice fed the Western diet for 4 and 8 weeks. Although CD69 mRNA was more abundant in the aortas of mice after 8 weeks as compared with 4 weeks on a Western diet, no differences could be detected between CD1d−/−LDLR−/− mice and CD1d+/+LDLR−/− controls (Figure 6a). IFN-γ mRNA levels were not significantly different between CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice, nor did IFN-γ mRNA levels increase between 4 and 8 weeks (Figure 6b). Similarly, there were no significant differences in TNF-α, IL-12, or IL-10 mRNA between the 2 genotypes at 4 or 8 weeks. (Figure 6c through 6e). IL-4 mRNA was not detected in any sample analyzed (data not shown). Staining for collagen content did not reveal differences between the genotypes (data not shown). These data indicate that the absence of CD1d-reactive NKT cells did not result in a reduction in mRNAs for proinflammatory cytokines in the aortas of LDLR−/− mice.
This study demonstrates that the absence of CD1d-reactive NKT cells attenuates lesion formation in mice only during early fatty streak formation. There were neither significant differences in lesion size during later stages of atherosclerosis nor decreases in proinflammatory cytokine mRNA in the aortic arch of CD1d−/−LDLR−/− mice at any of the time points assayed. These data extend recent studies describing a proatherogenic role for NKT cells and demonstrate that their effect is transient and limited to fatty streak lesions.
Activated Th1 lymphocytes are a key source of effector cytokines that potentiate immunopathological processes and contribute to the pathogenesis of atherosclerosis. Previous studies using RAG−/−LDLR−/− knockout mice have documented that the complete absence of lymphocytes provided protection early in disease progression but not when lesions were larger and more complex.19 The observation that CD1d−/−LDLR−/− mice had significantly less atherosclerosis than CD1d+/+LDLR−/− controls at the earliest time point (4 weeks) is consistent with data that suggest NKT cells mediate early immune responses at inflammatory sites before the classical T-cell response. NKT cells can be stimulated in the periphery on primary TCR engagement of antigen, resulting in the rapid secretion of cytokines.9,10 Further, NKT cells are efficiently activated by immature and mature dendritic cells which constitutively express CD1 proteins.20 Thus, NKT cells can be stimulated before major histocompatibility complex II expression and activation of peptide-specific T cells. Taken together, this suggests that the absence of NKT cells provides protection from atherosclerosis under mildly hypercholesterolemic conditions or during early atherogenesis.19,21
At the 4-week time point, the VLDL fractions of lipoproteins were slightly lower in CD1d−/−LDLR−/− than CD1d+/+LDLR−/− mice, as reported in RAG−/−LDLR−/− mice fed a similar diet.19 These findings indicate that lipoprotein metabolism may be subtly influenced by the absence of lymphocytes in the liver. Because ≈30% of hepatic T lymphocytes in mice are NKT cells,22 they may contribute to the regulation of lipoprotein metabolism. However, for 2 reasons it is unlikely that the reduction in VLDL accounted for the reduced lesion size in the CD1d−/−LDLR−/− mice at 4 weeks. First, the VLDL profile of CD1d−/−LDLR−/− mice was identical to that of RAG-deficient mice in a similar study19 in which no effect on lesion size was reported. Second, the VLDL fraction is also reduced in CD1d−/−LDLR−/− mice at 8 and 12 weeks when lesion sizes were identical to those in CD1+/+LDLR−/− mice. The lack of correlation between lesion size and small changes in VLDL content may reflect the lower atherogenicity of VLDL than LDL in mice.23
To explore the mechanism of decreased lesion formation in CD1d−/−LDLR−/− mice, we measured macrophage recruitment in the proximal aorta by immunostaining with MOMA-2 and in the aortic arch by real-time PCR for CD68. We noted that, although macrophage recruitment was unaffected in CD1d−/−LDLR−/− mice after 4 weeks of feeding on the Western diet, plaque macrophages had accumulated 38% less lipid in CD1d−/−LDLR−/− mice than CD1d+/+LDLR−/− controls as determined by oil red O staining. Thus, the deficiency of NKT cells reduced lesion formation in a manner distinct from that of chemokines such as MCP-1 which directly recruit macrophages.24,25 It is possible that NKT cells act directly within the lesion to promote macrophage activation, resulting in the enhanced uptake of lipid.
We detected Vα14-Jα28 mRNA in the aortic arch of CD1d+/+LDLR−/− mice fed a chow or Western diet but not in the aortas of CD1d−/−LDLR−/− or wild-type CD1d+/+LDLR+/+ mice. This suggests that the mild elevation of serum cholesterol levels in CD1d+/+LDLR−/− compared with CD1d+/+LDLR+/+ mice (175 to 225 mg/dL versus 95 to 125 mg/dL) was sufficient to generate chemoattractants and recruit NKT cells. Interestingly, Vα14-Jα28 mRNA did not accumulate after the 4-week time point, consistent with our data showing that NKT cells affect atherosclerosis only during early fatty streak development. CD3 mRNA expression was similar in CD1d−/−LDLR−/− and CD1d+/+LDLR+/+ mice (data not shown).
The observation that NKT cells are present in the aortic arch during feeding on the Western diet prompted us to compare T-cell activation in CD1d−/−LDLR−/− and CD1d+/+LDLR−/− mice. Th1 cytokine release and exacerbated lesion size occur after injection of α-GalCer into apolipoprotein E−/− mice,26 indicating that activated NKT cells can potentially contribute to atherosclerosis. However, α-GalCer is not found in mammals, and its physiological relevance to atherosclerosis is unclear. Although activated NKT cells have been isolated from splenocytes of mice injected with α-GalCer,14 we sought to determine whether endogenous antigens activated NKT cells in the vessel wall under hypercholesterolemic conditions. We found no difference in CD69 mRNA between the genotypes, indicating that T cells within the lesion were not activated by the presence of NKT cells at these time points. This result further suggests that NKT cells are either activated very early in lesion progression or exhibit a subtle effect on T-cell activation below the level of detection for this assay.
Many studies have documented the pro- and antiatherogenic role of Th1 and Th2 cytokines, respectively, in atherosclerosis. Thus, we sought to determine whether NKT cells influenced the level of inflammatory and regulatory cytokines in the lesion as opposed to the systemic secretion of Th1 cytokines observed in α-GalCer–injected mice.14,26 There were no differences in IFN-γ, TNF-α, IL-12, and IL-10 mRNA levels in the aortas of CD1d−/−LDLR−/− as compared with CD1d+/+LDLR−/− mice, indicating that Th1 cytokines were not reduced with the absence of NKT cells. IL-12 and TNF-α mRNA expression trended lower in CD1d−/−LDLR−/− mice than in controls at 4 weeks, but the differences were not statistically significant. Interestingly, IFN-γ mRNA expression was not increased after 8 weeks on the Western diet. This suggests that the concomitant expression of IL-12 and IL-10 maintains the homeostasis of IFN-γ in the lesion, and NKT cells do not significantly alter this Th1/Th2 balance.
These data indicate that although CD1d-reactive NKT cells are detected in the vessel wall, they do not influence the cytokine milieu or the degree of inflammation in the plaque during advanced atherosclerosis. Taken together with recently published studies, these results suggest that the role of NKT cells is largely limited to fatty streak formation.
This work was supported in part by grants HL63894 and HL52773 (to I.F.C.) and HL4826 (to H.A.C) from the National Institutes of Health. We would like to acknowledge Dr Luc Van Kaer for providing the CD1d−/− mice and thank Dr Karl Weisgraber and Dr Matthias Schneider for reviewing the manuscript, John C.W. Carroll and John Hull for preparation of the figures, Naima Contos for manuscript preparation, and Stephen Ordway and Dr Gary Howard for editorial assistance.
- Received July 7, 2004.
- Accepted November 3, 2004.
Bendelac A, Lantz O, Quimby ME, Yewdell JW, Bennink JR, Brutkiewicz RR. CD1 recognition by mouse NK1+ T lymphocytes. Science. 1995; 268: 863–865.
Singh N, Hong S, Scherer DC, Serizawa I, Burdin N, Kronenberg M, Koezuka Y, Van Kaer L. Cutting edge: activation of NK T cells by CD1d and α-galactosylceramide directs conventional T cells to the acquisition of a Th2 phenotype. J Immunol. 1999; 163: 2373–2377.
Carnaud C, Lee D, Donnars O, Park SH, Beavis A, Koezuka Y, Bendelac A. Cutting edge: cross-talk between cells of the innate immune system: NKT cells rapidly activate NK cells. J Immunol. 1999; 163: 4647–4650.
Lantz O, Bendelac A. An invariant T cell receptor α chain is used by a unique subset of major histocompatibility complex class I–specific CD4+ and CD4−8− T cells in mice and humans. J Exp Med. 1994; 180: 1097–1106.
Kawano T, Cui J, Koezuka Y, Toura I, Kaneko Y, Motoki K, Ueno H, Nakagawa R, Sato H, Kondo E, Koseki H, Taniguchi M. CD1d-restricted and TCR-mediated activation of Vα14 NKT cells by glycosylceramides. Science. 1997; 278: 1626–1629.
Nakai Y, Iwabuchi K, Fujii S, Ishimori N, Dashtsoodol N, Watano K, Mishima T, Iwabuchi C, Tanaka S, Bezbradica JS, Nakayama T, Taniguchi M, Miyake S, Yamamura T, Kitabatake A, Joyce S, Van Kaer L, Onoe L. Natural killer T cells accelerate atherogenesis in mice. Blood. 2004; 1; 104: 2051–2059.
Trogan E, Choudhury RP, Dansky HM, Rong JX, Breslow JL, Fisher EA. Laser capture microdissection analysis of gene expression in macrophages from atherosclerotic lesions of apolipoprotein E-deficient mice. Proc Natl Acad Sci U S A. 2002; 99: 2234–2239.
Cao X, Sugita M, van der Wel N, Lai J, Rogers RA, Peters PJ, Brenner MB. CD1 molecules efficiently present antigen in immature dendritic cells and traffic independently of MHC class II during dendritic cell maturation. J Immunol. 2002; 169: 4770–4777.
Eberl G, Lees R, Smiley ST, Taniguchi M, Grusby MJ, MacDonald HR. Tissue-specific segregation of CD1d-dependent and CD1d-independent NK T cells. J Immunol. 1999; 162: 6410–6419.
Tupin E, Nicoletti A, Elhage R, Rudling M, Ljunggren HG, Hansson GK, Berne GP. CD1d-dependent activation of NKT cells aggravates atherosclerosis. J Exp Med. 2004; 199: 417–422.