|Home | About | Journals | Submit | Contact Us | Français|
Statins are widely used as cholesterol-lowering agents that also decrease inflammation, and target enzymes essential for prenylation, an important process in the activation and intracellular transport of proteins vital for a wide variety of cellular functions. Here, we report that statins impair a critical component of the innate immune response, CD1d-mediated antigen presentation. The addition of specific intermediates in the isoprenylation pathway reversed this effect, whereas specific targeting of enzymes responsible for prenylation mimicked the inhibitory effects of statins on antigen presentation by CD1d as well as MHC class II molecules. This study demonstrates the importance of isoprenylation in the regulation of antigen presentation and suggests a mechanism by which statins reduce inflammatory responses.
The endogenous mevalonate pathway utilizes the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA), which is responsible for the biosynthesis of cholesterol and isoprenoids, and inhibiting HMG-CoA by statins potently suppresses this pathway (Fig. 1) (1). Statins have been used for the treatment of cardiovascular diseases over the years due to their lipid-lowering effects (2, 3). Their efficacy in the treatment of atherosclerosis has been attributed to both cholesterol-dependent and -independent activities (2, 4, 5). Given the cholesterol-independent effect of statins in inflammation, atherosclerosis and immunomodulation (2), it is not surprising that the immune response would also be affected independently of their lipid-lowering effects.
In addition to their ability to reduce the level of lipids, statins also inhibit the biosynthesis of isoprenoid intermediates such as geranylgeranyl pyrophosphate (GGPP) and farnesyl pyrophosphate (FPP) (6). Isoprenylation, the attachment of GGPP and FPP, is a post-translational modification of several proteins including the small GTP-binding proteins Ras, Rho and Rab (7). Isoprenylation plays some role in the activation and intracellular transport of proteins important for various cellular functions such as differentiation, proliferation, motility and the maintenance of cell shape (8). Moreover, prenylation enables GTPases to be targeted to the cell membrane and allows their subsequent interaction with downstream signal transduction effector molecules (9). Statins therefore impair the functioning of these small GTPases by preventing their correct membrane targeting (10).
Statins have been shown to affect both innate and acquired immune responses (11). Innate immune functions are inhibited via the targeting of TLR2 and TLR4 expression, preventing the activation of endothelial cells and macrophages (12). In contrast, acquired immune responses are affected by statin-dependent inhibition of antigen presentation by MHC class II molecules or T cell function and by the suppression of co-stimulatory molecules (e.g. CD40, CD80 and CD86) on macrophages, lymphocytes and endothelial cells (13, 14). In the adaptive immune response, the inhibition of prenylation plays an important role in the Th1 to Th2 switch and is protective against autoimmune diseases of the central nervous system, experimental arthritis, autoimmune myocarditis and systemic lupus erythematosus (15). Statins also inhibit pro-inflammatory gene expression in a number of leukocytes and vascular cells (15) and also diminish monocyte chemoattractant protein-1 in the neointima (16, 17). NK cell cytotoxicity in patients with cardiovascular disease has also been shown to be impaired by statins (18).
Natural killer T (NKT) cells are T lymphocytes with innate immune effector functions that are activated by lipid antigens presented by the MHC class I-like CD1d molecule (19-21). In the lipid-associated disorder atherosclerosis, T cells and NKT cells play crucial roles in the pathology of the disease (22-26). In advanced plaques, the numbers of DC, T cells, macrophages and HLA-DR expressing cells have been found in higher numbers than initial lesions or stable plaques (27). Further, NKT cells have been shown to contribute to the progression of atherosclerotic lesions (22, 28-30). Importantly, both CD1d expression and NKT cells have been detected in human atherosclerotic plaques (31-33). As statins have been found to reduce inflammation and stabilize atherosclerotic plaques by their activities independent of their lipid-lowering effects (34, 35), it is important to understand their ability to regulate antigen presentation. Interestingly, statins have been shown to alter MHC class II-mediated antigen presentation (36, 37), and although CD1d molecules traffic through some of the same endocytic compartments as MHC class II for antigen loading (19), their effects on antigen presentation by CD1d have not been studied. Thus, in the current study, the ability of statins to regulate lipid antigen presentation by CD1d molecules was investigated. It was found that statins significantly impair antigen presentation independent of their lipid-lowering properties, suggesting an important role for prenylation in the control of antigen presentation.
Female C3H/HeJ and C57BL/6 mice were purchased from the Jackson Laboratory (Bar Harbor, ME). All procedures were approved by the Institutional Animal Care and Use Committee of the Indiana University School of Medicine.
Murine LMTK-CD1d1 and LMTK-CD1d1-DR4 cells were cultured in DMEM (BioWhittaker, Lonza, MD, USA) supplemented with 10% FBS (Hyclone, Logan, UT), 2 mM glutamine, and 500 μg/ml of G418. Simvastatin (sodium salt), FTI-277 and GGTI-298 were purchased from Calbiochem (San Diego, CA). Mevalonolactone, geranylgeranyl pyrophosphate (GGPP), farnesyl pyrophosphate (FPP) and squalene were obtained from Sigma-Aldrich (St. Louis, MO). The mouse CD1d-specific NKT cell hybridomas DN32.D3, N37-1A12, N38-2C12 and N38-3C3 were cultured in IMDM (BioWhittaker, Lonza, MD, USA) supplemented with 5% FBS and 2 mM L-glutamine. The HLA-DR4-restricted, human serum albumin (HSA)-specific murine T cell hybridoma, 17.9, was a gift from Dr. Janice Blum (Indiana Univ. School of Medicine, Indianapolis, IN). Mouse B cell lymphoma TA3 cells and the I-Ak-restricted, hen egg lysozyme (HEL)-specific T cell hybridoma 3A9 cells were kindly provided by Dr. Gail Bishop (Univ. of Iowa, Iowa City, IA). HEL was obtained from Sigma-Aldrich. Purified and biotinylated mAb specific for murine IL-2 were purchased from BD Biosciences (San Diego, CA). Recombinant mouse IL-2 used as a standard in the ELISA assays was obtained from PeproTech (Rocky Hill, NJ). Antibodies specific for Rho and Ras were purchased from Cell Signaling Technology, Inc. (Beverly, MA); antibodies against Rab5a (Santa Cruz Biotechnology, Santa Cruz, CA), Rab7 (Sigma-Aldrich) and Rab11 (BD Biosciences) were also used in the experiments described below.
Bone marrow cells from C3H/HeJ and C57BL/6 mice were cultured in RPMI 1640 supplemented with 2 mM L-glutamine, 50 μM 2-mercaptoethanol, 10% FBS and antibiotics, as well as 10 ng/ml each of murine GM-CSF and IL-4 as previously described (38). On day 7, the plates were gently flushed (3-4 times) to remove the loosely-adherent cells, which were subsequently used in analyses as BMDCs.
To generate cell lines that stably expressed siRNA targeted against GGTase I, GGTase II, or for the control (non-specific sequence), we used the pSuppressorNeo vector (Imgenex, San Diego, CA). Primers targeting the β subunit of GGTase I or GGTase II were designed using Imgenex’s siRNA retriever program. Primers for GGTase I were forward: 5′-TCGACAGGATAAAGAGGTGGTGCATGGAATTCGATGCACCACCTCTTTCTCCTGTTTTT-3′ and reverse: 5′- CTAGAAAAACAGGATAAAGAGG TGGTGCATCGAATTCCATGCACCACCTCTTTATCCTG-3′. Primers for GGTase II were forward: 5′-TCGACCGAGAAGAAATCCTGGTGTTGGAATTCGAACACCAGGATTTCTTCTCGGTTTTT-3′ and reverse: 5′-CTAGAAAAACCGAGAAGAAATCCTGGTGTTCGAATTCCAACACCAGGATTTCTTCTCGG-3′. Primers for each target were annealed, phosphorylated, and ligated into the pSuppressorNeo vector. A portion of the ligation reaction was used to transform chemically-competent DH5α E. coli. Drug-resistant clones were selected and propagated, and plasmids were isolated and screened using restriction endonuclease digestions for a unique site found in the siRNA primer sequence. Positive clones were sequenced to ensure the correct insert was present. LMTK cells containing CD1d in the pcDNA-zeo (Invitrogen) were transfected with either of the GGTase siRNA plasmids, or control sequence vector only. Cells containing the siRNA vector were selected for using 500 μg/ml G418, and drug-resistant cells were pooled and maintained as stable cell lines for experiments.
Total RNA was isolated from 1 × 106 cells using the High Pure RNA Isolation Kit (Roche; Indianapolis, IN). One μg of this RNA was used to generate cDNA using the Transcriptor First Strand cDNA Synthesis Kit (Roche), using random hexamers. The mixture of primers and template was denatured (10 min at 65°C) prior to the reverse transcription reaction. This cDNA served as the template in real-time PCR reactions. The Univeral Probe method (Roche) was used to determine expression levels. To the FastStart Universal Probe Master Mix (ROX), we added primers for the gene of interest or β-actin (purchased from IDT), cDNA, and one of the Universal Probes. Primers and probes were designed using ProbeFinder version 2.43 for mice. To analyze GGTase I expression, we used Universal Probe #11 with forward: 5′-TGCTTAGCAGGCTTGAGAGC-3′ and reverse: 5′-TTCA GGAACCGCACAGAAG-3′ primers. To analyze GGTase II expression, Universal Probe #67 was used with forward (5′-GCCTATGTTCAGAGCCTACA-3′) and reverse (5′-CACCGCACAAAATGAGAATC-3′) primers. For analysis of β-actin expression, Universal Probe #11 was used with forward (5′-ACTGCTCTGGCTCCTAGCAC-3′) and reverse (5′-CCTGCTTGCTGATCCACAT-3′) primers. Real-time PCR was performed using an Mx3000P (Stratagene) instrument with MxPro software (version 4.01). The following parameters were used: 1 cycle of 95°C for 10 min and 40 cycles of 95°C for 15 sec, then 60°C for 1 min during which time the analysis was done. Results were analyzed using the standard defaults except that the Mx4000 v1.00 to v3.00 algorithm was used for the adaptive baseline. The Ct of each transcript in the different cell lines was obtained. The ΔΔCt method was used with β-actin as the control to calculate changes in expression.
LMTK-CD1d1 cells were treated with various concentrations (0-50 μM) of simvastatin, GGTI-298 (0-10 μM) or FTI-277 (0-10 μM) for 24 h. Cells treated with vehicle only (DMSO) served as the negative control. In a separate set of experiments, cells were treated with simvastatin (50 μM) in the presence or absence of mevalonate (200 and 400 μM) for 24 h or α-GalCer (500 ng/ml for 1 h). The cells were then washed with PBS, fixed in 0.05% paraformaldehyde and co-cultured with the indicated NKT cell hybridomas as described previously (39). In parallel, BMDCs were treated with the indicated concentrations of simvastatin for 24 h, washed, fixed and used in NKT cell assays as above. LMTK-CD1d1 cells transfected with empty vector or GGTase-specific siRNA were also co-cultured with NKT cells as above. LMTK-CD1d1 cells were treated with various doses of β-MCD (0-20 mM) and nystatin (0-20 μg/ml) for 2 h. The cells were washed, fixed and co-cultured with the indicated NKT cell hybridomas as above.
LMTK-CD1d1 cells, treated with vehicle, simvastatin (25 and 50μM) or AY9944 (10 and 20μM) overnight, were scraped from 6 well plates. The cells (5 × 106) were washed twice with cold phosphate-buffered saline, and centrifuged at 2000 rpm. The pellet was resuspended in 0.5 ml of isopropyl alcohol, sonicated for 3 min and cleared by centrifugation at 10,000 rpm for 10 min. Supernatants were decanted and isopropyl alcohol was evaporated. A volume of 50 μl of isopropyl alcohol was added to each tube to resuspend the material and aliquots were used to measure the cholesterol level using the EnzychromTM assay kit (BioAssay Systems, Hayward, CA) with cholesterol standards used for calibration according to the manufacturer’s instructions.
L-CD1d1-DR4 cells (40) were treated with 10 μM of HSA in the presence or absence of different concentrations of simvastatin for 24 h. The cells were then washed with PBS, fixed in paraformaldehyde and co-cultured with the 17.9 T cell hybridoma, with IL-2 production determined as above. Because these L cells also express CD1d1 molecules, an NKT cell co-culture was performed in parallel as a positive control. In other experiments, the murine B cell lymphoma TA3 cell line was treated with vehicle, simvastatin, FTI-277 or GGTI-298 in the presence or absence of 1 mg/ml of HEL and co-cultured with the HEL-specific, I-Ak-restricted 3A9 T cell hybridoma and IL-2 production was measured as above. BMDCs generated from C3H/HeJ mice were treated with LPS (1 μg/ml) overnight. The cells were washed and treated with the indicated concentrations of simvastatin, FTI-277 and GGTI-298 in the presence or absence of HEL. After two washes in PBS, the BMDCs were co-cultured with the 3A9 T cell hybridoma as above.
To determine the effect of prenylation inhibition on antigen presentation by MHC class I molecules, MC57G cells were treated with vehicle (DMSO), simvastatin (50 μM), FTI-277 (10 μM) or GGTI-298 (10 μM) overnight and then mock-infected or infected with vaccinia virus (VV; MOI=5) for 6 h in the presence or absence of inhibitors. The cells were washed, fixed and co-cultured for 24 h with splenocytes isolated from C57BL/6 mice on day 6 after a VV infection (1 × 106 pfu/mouse, i.p.). Supernatants were collected to measure the amount of IFN-γ in the co-cultures by ELISA.
LMTK-CD1d1 cells were treated with the indicated concentrations of simvastatin in the presence or absence of mevalonate. The cells were lysed in lysis buffer (25 mM HEPES buffer, pH 7.5, 150 mM NaCl, 1% NP-40, 10% glycerol, 25 mM sodium fluoride, 10 mM MgCl2, 1 mM EDTA, 10 mM sodium orthovanadate, 25 mM β-glycerophosphate), containing Complete® protease inhibitor tablets (Roche Diagnostics, Indianapolis, IN). The amount of protein in cell lysates was estimated using Bio-Rad protein assay reagents (Hercules, CA). Equal amounts of protein were loaded into each well and resolved on a 10% SDS-PAGE gel, and subsequently transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA). The blot was processed with antibodies specific for the indicated proteins and bands developed using chemiluminescence before exposure on film (41).
For assessing the relative distribution of GTPases in statin-treated cells, LMTK-CD1d1 cells were treated with simvastatin in the presence or absence of mevalonate. Cells were lysed in sucrose-containing HEPES buffer (200 mM sucrose, 20 mM HEPES, 1.5 mM MgCl2, 10 mM KCl, 1 mM EGTA, mM EDTA, 0.1 mM PMSF and Complete® protease inhibitor tablets) on ice for 15 min. The lysates were clarified by centrifugation at 1000 rpm for 5 min at 4°C. The resulting supernatant was centrifuged at 30,000 rpm at 4°C for 30 min. The resulting supernatant was then removed (cytosolic fraction) and the membrane pellet was washed in sucrose-containing HEPES buffer by centrifugation at 30,000 rpm for 30 min. Equal amounts of protein were loaded into each well and were resolved on a 10% SDS-PAGE gel, and transferred to a PVDF membrane. The blots were probed with antibodies specific for the individual GTPases, as well as markers for the cytosolic (GAPDH) and membrane (flotillin) fractions, and developed as above.
LMTK-CD1d1 cells were plated in sterile glass-bottom 35 mm dishes coated with collagen (MatTek, Ashland, MA) at a density of 1×106 cells per dish. Cells were treated with the indicated concentrations of GGTI-298, FTI-277, AY9944 or simvastatin with or without mevalonate for 24 h at 37°C. The cells were stained for CD1d and lysosome-associated membrane protein 1 (LAMP-1) followed by Texas Red- and FITC-conjugated secondary antibodies, and then analyzed by confocal microscopy as previously described (39, 40). The percent co-localization of CD1d and LAMP-1 was determined using MetaMorph software (Version 5; Molecular Devices, Sunnyvale, CA), in which six random fields were chosen from each picture.
The data were analyzed by a one way ANOVA and unpaired two-tailed Student’s t-test using GraphPad PRISM software (version 5.0 for Windows; GraphPad, San Diego, CA). A P value below 0.05 was considered significant. The error bars in the bar graphs show the standard deviation from the mean of triplicate samples.
In order to determine if prenylation is important for CD1d-mediated antigen presentation, murine LMTK-CD1d1 fibroblasts were treated with various concentrations of the statin simvastatin for 24 h and co-cultured with NKT cells. LMTK-CD1d1 cells treated with simvastatin showed a dose-dependent reduction in the stimulation of NKT cells, with an almost 50-60% decrease in antigen presentation at the highest concentration used, as compared to vehicle-treated cells (Fig. 2A). To ensure that the effect of prenylation inhibition on CD1d-mediated antigen presentation could be observed in primary antigen presenting cells, bone marrow-derived dendritic cells (BMDCs) were treated with different concentrations of simvastatin as above. As found with LMTK-CD1d1 cells, simvastatin treatment of BMDCs also caused a substantial reduction in antigen presentation by CD1d (Fig. 2B). Simvastatin did not alter CD1d cell surface expression on either LMTK-CD1d1 or BMDCs (Supplementary Figs. 1A, B). Nonetheless, inhibiting prenylation could have caused some qualitative changes in the functional expression of CD1d molecules. To test this possibility, simvastatin-pretreated LMTK-CD1d1 cells were incubated with the CD1d-specific ligand α-galactosylceramide (α-GalCer) to determine if there was an effect on exogenous antigen presentation. As expected, α-GalCer substantially enhanced the stimulation of Vα14+ (e.g., DN32.D3), but not Vα14− (e.g., N37-1A12) NKT cells. Interestingly, simvastatin-treated LMTK-CD1d1 cells could not stimulate NKT cells to the same extent as vehicle-treated cells in the presence of α-GalCer (Fig. 1C). Therefore, these results suggest that simvastatin treatment impairs both endogenous and exogenous lipid antigen presentation.
To determine whether simvastatin affected CD1d-mediated antigen presentation through its effects on cholesterol biosynthesis, LMTK-CD1d1 cells were treated with a highly specific inhibitor of cholesterol biosynthesis, AY9944. Notably, whereas AY9944 did not alter antigen presentation by CD1d molecules even at the highest concentration used (Fig. 2D), it did substantially inhibit cholesterol biosynthesis (Supplementary Fig. 2A). Moreover, treatment of LMTK-CD1d1 cells with cholesterol-depleting agents such as nystatin and β-methyl cyclodextrin (β-MCD) had only a modest effect on CD1d-mediated antigen presentation (Supplemental Fig. 2B, C). The latter results of β-MCD are in line with those recently reported (42). However, it is important to note that β-MCD concentrations above 10 mM are toxic (data not shown).
Mevalonate, the precursor for the biosynthesis of cholesterol and isoprenoids, is considered a therapeutic target for autoimmune diseases (15, 43). Simvastatin inhibits the biosynthesis of mevalonate by inhibiting the enzyme HMG-CoA reductase. Thus, to determine whether the simvastatin-induced inhibition of CD1d-mediated antigen presentation was mevalonate pathway-dependent, exogenous mevalonate was added to simvastatin-treated CD1d+ cells. Although exogenous mevalonate was able to substantially reverse the effect of simvastatin on CD1d-mediated antigen presentation (Fig. 3A), this approach could not distinguish between whether simvastatin impaired antigen presentation by inhibiting lipid biosynthesis or by isoprenylation (or both). To address this question, LMTK-CD1d1 cells were treated with simvastatin in the presence or absence of the geranylgeranylation precursor, geranylgeranyl pyrophosphate (GGPP), farnesylation precursor farnesyl pyrophosphate (FPP) or the cholesterol biosynthesis precursor squalene. The addition of GGPP, but not FPP, significantly reversed the inhibitory effects of simvastatin on CD1d-mediated antigen presentation (Fig. 3B, C). In contrast, squalene had no effect (Fig. 3D). These results strongly suggest that the simvastatin-induced inhibition of antigen presentation by CD1d is solely prenylation-dependent.
To further investigate the role of prenylation in the simvastatin-induced inhibition of CD1d-mediated antigen presentation, BMDCs were treated with simvastatin, the farnesylation-specific inhibitor FTI-277, or geranylgeranylation-specific inhibitor GGTI-298 for 24 h. CD1d-mediated antigen presentation was significantly reduced in cells treated with simvastatin or GGTI-298, whereas FTI-277 treatment had only a modest effect (Fig. 4A). Neither of the inhibitors altered the cell surface expression of CD1d molecules (data not shown). Geranylgeranylation is controlled by two enzymes: GGTase I and GGTase II (44). To determine whether GGTase I and/or GGTase II can regulate CD1d-mediated antigen presentation, the activity of these enzymes was reduced using an RNAi system. Consistent with the GGTI-298 treatment above (Fig. 3A), LMTK-CD1d1 cells transfected with RNAi for either GGTase I, or GGTase II in particular, were impaired in their antigen presentation ability (Fig. 4B).
Like CD1d molecules, MHC class II molecules traffic through late endocytic compartments (19), and statins have been shown to also affect antigen presentation by the latter pathway (36, 37), although the role of prenylation was not studied. Thus, to determine whether the inhibition of prenylation alters MHC class II-mediated antigen presentation under the same conditions as observed with CD1d, HLA-DR4-transfected LMTK-CD1d1 (L-CD1d-DR4) cells (40) were treated with human serum albumin (HSA) in the presence or absence of simvastatin and co-cultured with the HLA-DR-restricted, HSA-specific T cell hybridoma 17.9. These same cells were co-cultured with NKT cell hybridomas as a control side-by-side. MHC class II-mediated antigen presentation was reduced in a concentration-dependent manner without a change in cell surface MHC II molecule expression (Fig. 5A and data not shown), similar to that observed with CD1d. Statins have been shown to reduce the cell surface expression of induced (e.g., by CD40 ligation or IFN-γ) MHC class II molecules, whereas the constitutive expression of MHC class II was found to be unaltered (36, 37, 45). For the current study, the effect of prenylation inhibitors on murine MHC class II (I-Ak)-mediated antigen presentation was also determined by treating the mouse B cell lymphoma TA3 or BMDCs from C3H/HeJ (H-2k) mice in the presence or absence of hen egg lysozyme with different concentrations of simvastatin, FTI-277 or GGTI-298. Simvastatin and GGTI-298 caused a significant reduction in MHC class II-mediated antigen presentation, whereas FTI-277 had only a modest effect in TA3 cells (Fig. 5B) and BMDCs (Fig. 5C). Therefore, these results suggest that under similar conditions that reduce antigen presentation by CD1d, statins also alter MHC class II-mediated antigen presentation by targeting the geranylgeranylation pathway.
CD1d molecules are structurally similar to MHC class I molecules (19). Therefore, it was also important to know whether prenylation inhibition alters antigen presentation by MHC class I. To address this question, murine MC57G fibroblasts were treated with vehicle, simvastatin, GGTI-298 or FTI-277 overnight. The cells were mock-infected or infected with vaccinia virus (VV) for 6 h in the presence or absence of prenylation inhibitors. The cells were then washed, fixed and co-cultured with splenocytes from C57BL/6 mice infected 6 days previously with VV as a source of VV-specific CTLs. The production of IFN-γ was used as a measure of VV-specific T cell recognition. Although inhibiting prenylation had no effect on antigen presentation by MHC class I molecules, as expected, the proteosomal inhibitor lactacystin caused a significant reduction (Fig. 5D).
As the inhibition of prenylation caused a reduction in CD1d-mediated antigen presentation without altering its cell surface level, this effect might have been due to intracellular changes caused by preventing prenylation. Thus, LMTK-CD1d1 cells were treated with simvastatin, FTI-277, GGTI-298, or the cholesterol biosynthesis inhibitor AY9944, and co-localization of CD1d with the late endosome/lysosome marker LAMP-1 was analyzed by confocal microscopy. A significant decrease in the co-localization of CD1d and LAMP-1 was observed in cells treated with either simvastatin or GGTI-298 (Fig. 5), whereas exposure to FTI-277 or AY9944 had no effect. In agreement with our NKT cell assay results, when cells were treated with simvastatin in the presence of mevalonate, there was a recovery in CD1d/LAMP-1 co-localization (Fig. 5). These data suggest that alterations in CD1d intracellular trafficking following inhibition of prenylation substantially impair antigen presentation by CD1d molecules.
The widespread clinical use of statins has decreased the rate of morbidity and mortality of persons suffering from cardiovascular diseases. Besides their lipid-lowering effects, statins also improve or restore endothelial function by decreasing inflammation and enhancing the stability of atherosclerotic plaques (45). In addition, statins also regulate prenylation and are immunomodulatory (6). For example, statins promote Th2 polarization and inhibit Th1 development, resulting in reduced inflammation in various model systems (46). Thus, these drugs have an advantage over other cholesterol-lowering agents in the treatment of cardiovascular disease, where inflammation plays a critical role in its progression. In the current study, inhibiting cholesterol did not reduce antigen presentation by CD1d. Further, squalene, a cholesterol precursor, was not able to reverse the inhibition of CD1d-mediated antigen presentation by simvastatin. This suggests that the ability of simvastatin and other statins to regulate antigen presentation by CD1d is independent of their effects on cholesterol biosynthesis. Besides cholesterol, other pathways targeted by statins are those that regulate the supply of isoprenoids, particularly farnesyl pyrophosphate and geranylgeranyl pyrophosphate (7). In the context of antigen presentation by CD1d, we found that inhibition of geranylgeranylation, but not farnesylation, significantly impaired CD1d- (and MHC class II)-mediated antigen presentation. Statins alter the prenylation of specific GTPases (47) (Supplementary Fig. 3), important for the trafficking of molecules through the endocytic pathway. Prior studies have shown that Rho and Rab GTPases can regulate intracellular trafficking of a variety of proteins and the efficiency of antigen processing and presentation within various endocytic compartments in dendritic cells and B cells (49, 50). Thus, inhibiting the prenylation of these GTPases may have contributed to the altered trafficking of antigen presenting molecules (i.e., both CD1d and MHC class II) through endocytic compartments and consequent effects on antigen presentation (19). Our observation that either simvastatin or GGTI-298 can affect the intracellular localization of CD1d is consistent with idea.
Atherosclerosis is an immuno-inflammatory disease of arterial walls where both innate and adaptive immune responses contribute to disease development and progression (22). Oxidized low-density lipoprotein (LDL) is a major cause of injury in atherosclerosis and adaptive immune responses to oxidized LDL contribute to the development of the disease. Lipid and peptide antigens derived from oxidized LDL bind to CD1d and MHC class II molecules to activate pro-inflammatory T cells (48). A reduction in antigen presentation by either of these pathways results in diminished atherosclerosis. NKT cells have been implicated as proatherogenic in a mouse model of atherosclerosis (49). Furthermore, CD1d-expressing vascular dendritic cells have been found to be in association with NKT cells in atherosclerotic plaques (31). Some DCs interact with T cells within atherosclerotic lesions, whereas others appear to migrate to regional lymph nodes and activate mainstream T cells (50). Such T cell activation requires a transition of DCs from immature to mature forms. In healthy arterial walls, DCs are in their immature form, whereas in atherosclerotic lesions, DCs express high levels of HSP70 that contribute to their activation and maturation (51). NKT cells appear to play a greater role in the progression of atherosclerosis due to their activation by both immature and mature DCs (20). The results of the present study show that the inhibition of prenylation reduces both CD1d- and MHC class II-mediated antigen presentation. Thus, one mode of the anti-atherosclerotic action of statins, in addition to their cholesterol-lowering activity, may be through their inhibitory effect on proinflammatory antigen presentation pathways mediated by both CD1d and MHC class II molecules. In line with this, it is notable that the adoptive transfer of NKT cells or their activation by the CD1d-specific ligand α-GalCer, aggravates disease in mouse models of atherosclerosis (22). The present study shows that inhibition of prenylation modulates the effector activity of NKT cells by impairing lipid antigen presentation by CD1d.
How might this translate in vivo? In addition to its well known LDL-lowering properties, simvastatin is not very bioavailable upon oral administration due to its poor absorption (52), despite showing beneficial effects through immune-mediated mechanisms (53, 54). Further, altering antigen presentation by this class of cholesterol-lowering drugs may lead to somewhat increased susceptibility of statin-treated patients to various pathogens, particularly virus infections (3). Therefore, under some conditions, the long term use of statins may contribute to the impairment of both innate and adaptive immune responses, with consequent effects; such treated patients would simply need to be monitored more closely.
Simvastatin does not alter the cell surface expression of CD1d. LMTK-CD1d1 (A) and BMDCs (B) were treated with various concentrations of simvastatin for 24 h. The cells were then washed and stained with a PE-conjugated isotype control or anti-CD1d antibody (1B1) for 30 min at ice. The cells were washed and fixed in 1% paraformaldehyde for 15 min on ice before FACS analysis. Open histogram: isotype control. Filled histogram: anti-CD1d.
Cholesterol-depletion does not alter CD1d-mediated antigen presentation.A, LMTK-CD1d1 cells were treated with vehicle (DMSO), simvastatin (25 and 50 μM) or a specific inhibitor of cholesterol biosynthesis, AY9944 (10 and 20 μM) for 24 h. The cells were lysed and the total level of cholesterol was quantitated. LMTK-CD1d1 cells were treated or untreated with various concentrations of (B) β-methyl cyclodextrin (β-MCD) or (C) nystatin for 4 h. The cells were then washed and co-cultured with the indicated NKT cell hybridomas.
Effects of simvastatin on prenylation in LMTK-CD1d1 cells and its reversal by mevalonate. A, LMTK-CD1d1 cells were treated with simvastatin (50 μM) in the presence or absence of mevalonate (200 and 400 μM) and the prenylation of the indicated Rab GTPases and Ras proteins was analyzed by Western blot. Lane 1 (vehicle); Lane 2 (Mevalonate-200 μM); Lane 3 (Mevalonate-400 μM); Lane 4 (Simvastatin-50 μM); Lane 5 (Simvastatin + Mevalonate-200 μM); Lane 6 (Simvastatin + Mevalonate-400 μM). Lane 4 in (a) shows slower migration of the respective bands, indicative of prenylation inhibition. B and C: LMTK-CD1d1 cells were treated with simvastatin (25 and 50 μM) in the presence or absence of mevalonate (400 μM). The cytosolic (B) and membrane (C) fractions were analyzed for the presence of Rho, Rab5 and Ras by Western blot. GAPDH served as a control cytosolic protein, whereas flotillin was the control membrane protein used. Lanes: 1. Vehicle, 2. Mevalonate-400 μM, 3. Simvastatin-25 μM, 4. Simvastatin-25 + Mev-400, 5. Simvastatin-50 μM, 6. Simvastatin-50 + Mev-400.
We would like to thank Gail Bishop for the 3A9 and TA3 cells, Janice Blum for the 17.9 T cell hybridoma, Wen Tao and Janardhan Sampath for their advice and help with the RT-PCR, and Keith March for helpful discussions.
1This work was supported by National Institutes of Health grants RO1 AI46455 and PO1 AI056097 (to R.R.B.), and NSF CHE-0194682 from the National Science Foundation (to J.G.H.). G.J.R. and R.M.G. were supported by NIH training grants T32 DK007519 and T32HLO7910, respectively.
There are no financial conflicts of interest by any of the authors.