Statin-induced Kruppel-like factor 2 expression in human and mouse T cells reduces inflammatory and pathogenic responses

The transcription factor kruppel-like factor 2 (KLF2) is required for the quiescent and migratory properties of naive T cells. Statins, a class of HMG-CoA reductase inhibitors, display pleiotropic immunomodulatory effects that are independent of their lipid-lowering capacity and may be beneficial as therapeutic agents for T cell–mediated inflammatory diseases. Statins upregulate KLF2 expression in endothelial cells, and this activity is associated with an antiinflammatory phenotype. We therefore hypothesized that the immunomodulatory effects of statins are due, in part, to their direct effects on T cell KLF2 gene expression. Here we report that lipophilic statin treatment of mouse and human T cells increased expression of KLF2 through a HMG-CoA/prenylation–dependent pathway. Statins also diminished T cell proliferation and IFN-γ expression. shRNA blockade of KLF2 expression in human T cells increased IFN-γ expression and prevented statin-induced IFN-γ reduction. In a mouse model of myocarditis induced by heart antigen–specific CD8⁺ T cells, both statin treatment of the T cells and retrovirally mediated overexpression of KLF2 in the T cells had similar ameliorating effects on disease induction. We conclude that statins reduce inflammatory functions and pathogenic activity of T cells through KLF2-dependent mechanisms, and this pathway may be a potential therapeutic target for cardiovascular diseases.

Kruppel-like factor 2 (KLF2) is a member of a transcription factor family, with homology to the drosophila kruppel transcription factor. It is expressed in lung, endothelial cells, and lymphocytes and is essential for blood vessel integrity and lung development (1). Klf2-null mice exhibit abnormal blood vessel formation, resulting in embryonic hemorrhage and death (1). In addition to its role in lung and vascular development, KLF2 appears to play several roles in the regulation of T cell function. One of the first of these roles to be described is the inhibition of proliferation. Ectopic expression of the KLF2 gene in the constitutively proliferative human T cell leukemia line Jurkat decreases mitotic activity of these cells (2, 3). Furthermore, gene-targeted KLF2-deficient mouse T cells have a hyper proliferative phenotype (2, 3). Several lines of evidence indicate that KLF2 is required for the maintenance of T cell quiescence. KLF2 mRNA is expressed in naive and memory T cells and is rapidly downregulated upon TCR stimulation of these cells (4, 5). Although most of the functions ascribed to KLF2 indicate that KLF2 is required to maintain the nonactivated phenotype, some data suggest a more complicated set of functions. For example, KLF2 may also play a role in promoting the very early stages of T cell activation, at which time its expression is transiently increased in Jurkat cells, and it transactivates IL-2 promoter activity (6). Furthermore, the transition from effector to memory stages of T cell responses may involve KLF2 expression in effector cells before the memory phenotype is established, as described in mouse CD8⁺ T cells (5).

Due to the embryonic lethality of global KLF2 deficiency, the function of KLF2 in T cells has been studied in mice with selective deficiency of KLF2 only in hematopoietic cells (7) or only in lymphocytes (8–10). In all these cases, there is relatively normal T cell development in the thymus but a severe T cell deficiency in the periphery. This deficiency has been attributed to defective expression of sphingosine-1-phosphate (S1P) receptor 1 (S1PR1), which is required for S1P-mediated egress of T cells from the thymus and peripheral lymphoid organs. Other T cell homing defects in these mice have also been attributed to a lack of KLF2-dependent CD62L expression, which is required for naive T cell migration into lymph nodes. Other abnormalities in KLF2-deficient T cell expression that have been reported in individual studies, such as enhanced Fas ligand–mediated apoptosis (8) and expression of inflammatory chemokine receptors, leading to constitutive T cell migration into nonlymphoid tissues (9), have not been consistently seen in other studies (10). Overall, work performed with KLF2-deficient T cells in vivo indicates the importance of KLF2 expression for normal peripheral T cell recirculation but does not clarify how KLF2 modulates mature peripheral T cell function.

Statins, a class of HMG-CoA reductase inhibitors, display pleiotropic immunomodulatory effects, independent of their lipid-lowering effects. The antiinflammatory effects of statins may contribute to their atheroprotective actions, and clinical trials are in progress to test whether these drugs have benefit in various autoimmune diseases. Published studies suggest that statins may be beneficial for T cell–mediated diseases by suppressing inducible class II MHC expression and costimulators on APCs (11, 12), favoring Th2 versus Th1 differentiation of helper T cells (11, 13, 14), and augmenting circulating regulatory T cell numbers and their functional properties (15). However, the direct effects of statins on T cells remain poorly characterized. Statins are reported to bind to and block LFA-1 function, which is required for T cell interactions with APCs (16) and to block TCR signaling at Ras family GTPase-dependent steps, by interfering with prenylation of these signaling molecules (17, 18). Work with the human T cell leukemia line, Jurkat, suggests that statins may have antiproliferative effects on T cells independent of Ras, by uncoupling protein tyrosine kinases from TCR signal transduction pathways (19). An unproven hypothesis that has been put forward is that statins may impair T cell activation by altering membrane cholesterol levels and lipid raft formation, thereby perturbing the formation of the immune synapse between T cells and APCs (20). Statins upregulate KLF2 expression in endothelial cells (21, 22), and KLF2 expression is associated with a quiescent antiinflammatory phenotype of endothelial cells (23–25). In addition, there is a report that statins upregulate KLF2 expression in macrophages (26), and lentiviral-mediated KLF2 expression in macrophages mimics statin effects in reducing inflammatory chemokine expression (26). Taken together, these findings suggest that the beneficial effects of statins in treating cardiovascular disease include antiinflammatory actions independent of cholesterol lowering. We hypothesize that immunomodulatory effects of statins may also be related to their influence on T cell KLF2 expression. In the present study, we show that statins upregulate KLF2 expression in T cells and block activation-induced downregulation of T cell KLF2. Furthermore, statin treatment and exogenous Klf2 gene expression have very similar effects on T cell functions and on the pathogenic actions of T cells in a mouse model of myocarditis.

KLF2 expression pattern during T cell activation. We examined the expression of Klf2 mRNA in TCR transgenic mouse CD8⁺ T (OT-1) cells before and after TCR stimulation using quantitative RT-PCR (qRT-PCR) (Figure 1A). Klf2 was highly expressed in naive T cells, but it was rapidly downregulated after 6 hours of αCD3 stimulation. Klf2 expression was very low to undetectable in mouse effector CD8⁺ T cells that were differentiated in vitro by a 5-day stimulation with αCD3, anti-CD28, and cytokines, but expression went up after 9 days of rest of the effector cells in IL-2–containing media. We observed a similar pattern of changes in Klf2 mRNA expression in wild-type mouse CD8⁺ T cells and wild-type or TCR transgenic CD4⁺ T cells (data not shown). Expression of Ccr7 and L selectin (Sell) mRNA was relatively high and Ccr5 expression was relatively low in naive CD8⁺ T cells, and this pattern was reversed in effector T cells (Figure 1, B–D). These data, which show a positive correlation of Klf2 expression with both Ccr7 and Sell expression and an inverse correlation between Klf2 and Ccr5, are consistent with published studies in various T cell populations (7, 9, 10), including KLF2-deficient T cells, which express more inflammatory chemokine receptors than wild-type T cells (9). Resting effector T cells expressed more Klf2, Sell, and Ccr5 and less Ccr7 than recently activated effector T cells (Figure 1, B–D).

Figure 1

Changes in Klf2 and chemokine receptor gene expression in T cells after stimulation. RNA was isolated from naive OT-1 CD8⁺ T cells, OT-1 cells after 6 hours of stimulation in vitro with αCD3, effector OT-1 cells generated by a 5-day culture, and effector T cells rested for 9 days after TCR stimulation in medium containing IL-2. (A) Klf2, (B) Ccr7, (C) Ccr5, and (D) Sell (Cd62l) RNA expression was determined by qRT-PCR analysis. Data represent the mean ± SEM of results from 3 experiments. *P < 0.05.

Effects of statins on T cell KLF2 expression. We predicted that statins would prevent KLF2 downregulation after T cell activation or would upregulate KLF2 expression in activated T cells. These predictions were confirmed by analyzing Klf2 mRNA in statin-treated OT-1 cells (Figure 2). The level of Klf2 expression in naive CD8⁺ T cells was significantly reduced 6 hours after initiation of αCD3 stimulation but not when fluvastatin, lovastatin, or simvastatin, all at 10 μM, were present in the culture (Figure 2A). Pitavastatin also blocked downregulation of Klf2 in naive OT-1 cells upon TCR stimulation, with a dose-dependent effect detectable from 0.5 to 10 μM (Figure 2B). We also examined the effect of statins on Klf2 expression in effector CD8⁺ T cells, which have very low levels in the absence of statins. Treatment of effector CD8⁺ T cells for 18 hours with 10 μM fluvastatin, lovastatin, or simvastatin (Figure 2C) or with 1, 5, or 10 μM pitavastatin (Supplemental Figure 1; supplemental material available online with this article; doi: 10.1172/JCI41384DS1) significantly increased Klf2 mRNA from low levels in vehicle-treated cells. Simvastatin treatment of effector OT-1 cells also increased S1pr1, Ccr7, and Cd62l expression (Figure 2D). Simvastatin treatment of naive and effector CD4⁺ T cells also resulted in elevated Klf2 mRNA levels (Supplemental Figure 2, A and B). In order to determine whether in vivo exposure to statins can influence T cell Klf2 expression, we performed 2 different types of experiments. First, we treated C57BL/6 mice with daily i.p. injections of 20 mg/kg lovastatin or vehicle (DMSO) for 3 consecutive days. Splenic CD8⁺ and CD4⁺ T cells were then purified and stimulated ex vivo with αCD3 in culture for 6 hours, before RNA isolation and qRT-PCR analysis. We found that resting CD8⁺ T cells from both control- and statin-treated mice had similar levels of Klf2 mRNA, but after ex vivo anti-CD3 stimulation, the CD8⁺ T cells from lovastatin-treated mice had 6.5-times higher Klf2 levels than the T cells from the control mice (Figure 2E). Klf2 expression was also significantly higher after ex vivo stimulation in CD4⁺ T cells from lovastatin-treated mice compared with control mice (Supplemental Figure 2C). We also tested whether statin treatment would affect Klf2 expression in T cells after in vivo antigen-induced activation. In order to obtain enough in vivo antigen-activated T cells, we elected to subcutaneously immunize OT-1/Rag2^–/– TCR transgenic mice with ovalbumin in complete Freund’s adjuvant and to harvest draining lymph node T cells for RNA isolation and analysis. One hundred percent of the lymph node T cells in these mice were specific for ovalbumin, and a significant proportion of the draining lymph node T cells were activated by immunization, allowing for detection of changes in Klf2 above the high Klf2 levels found in the fraction of naive T cells that did not respond to the immunization. Using this protocol, we did find a significant increase in Klf2 mRNA expression (Figure 2F).

Figure 2

Statins increase mouse T cell KLF2 expression by a mevalonate/isoprenoid-dependent pathway. (A and B) Naive OT-1 T cells were treated for 18 hours with vehicle (–), fluvastatin (F) (10 μM), lovastatin (L) (10 μM), or simvastatin (S) (10 μM), or pitavastatin (Pitava) at indicated concentrations and then stimulated with αCD3 for 6 hours before RNA isolation and qRT-PCR Klf2 analysis. (C and D) OT-1 effector cells were treated with indicted statins for 18 hours, before RNA isolation and (C) qRT-PCR analysis of Klf2 or (D) S1pr1, Ccr7, and Sell. (E) C57BL/6 mice were injected i.p. with 20 mg/kg lovastatin or DMSO for 3 days, and then splenic CD8⁺ T cells were isolated and cultured with or without αCD3 for 6 hours, before RNA isolation and qRT-PCR analysis for Klf2. (F) OT-1/Rag-2^–/– mice were fed pitavastatin (30 mg/kg) or vehicle twice daily for 2 days, immunized subcutaneously with ovalbumin/CFA, and fed pitavastatin (30 mg/kg) or vehicle for 2 more days before sacrifice, RNA isolation from draining lymph node T cells, and qRT-PCR Klf2 analysis. (G) Effector OT-1 cells were cultured with vehicle, 10 μM simvastatin (Simva), 100 μM mevalonate, 10 μM farnesyl pyrophosphate (FPP), or 10 μM geranygeranyl pyrophosphate (GGPP), in indicated combinations for 8 hours, before RNA isolation and qRT-PCR analysis. (H) Effector OT-1 cells were cultured with vehicle, 20 μM FTI, or 20 μM GGTI for 8 hours, before RNA isolation and qRT-PCR analysis. (A and C–H) Data are mean ± SEM from 3 experiments. (B) Data are from 1 of 2 experiments performed, with similar results. *P < 0.05.

Previous studies on endothelial cells and T cells have indicated that statins exert at least some of their biological effects by blocking the formation of isoprenoids through the mevalonate pathway (21, 27). We therefore tested whether the effects of statins on murine T cell Klf2 expression were dependent on this pathway. We found that simvastatin-induced Klf2 upregulation in effector T cells in vitro was completely blocked by the addition of mevalonate. The statin effect was approximately 50% blocked by either farnesyl pyrophosphate (FPP) or geranygeranyl pyrophosphate (GGPP) (Figure 2G). Furthermore, treatment of effector mouse T cells with a geranyl-geranyl transferase inhibitor (GGTI), but not a farnesyl transferase inhibitor, reproduced the statin effect of increasing Klf2 expression (Figure 2H). These findings indicate that Klf2 expression in T cells is suppressed by a geranylation-dependent pathway, and statins release this suppression by blocking synthesis of mevalonate, which is required for the formation of GGPP.

In order to determine whether statins can influence KLF2 expression in human T cells, we isolated human blood CD3⁺ T cells from healthy donors, activated them in vitro with αCD3 or phytohemagglutinin (PHA) plus IL-2, and treated the cells with vehicle or statins, before preforming RNA isolation and qRT-PCR analyses. We found that TCR stimulation reduces KLF2 expression in the human T cells, but simvastatin or pitavastatin treatment blocks or reverses this reduction (Figure 3, A–C). We also examined the expression of KLF2 protein in control- and statin-treated human T cells using Western blot (Figure 3D). Both simvastatin and pitavastatin upregulated KLF2 protein expression in αCD3-stimulated T cells. We also determined, using qRT-PCR, that simvastatin increased human T cell expression of the complement regulatory molecule CD59 (data not shown), which is a protein previously shown to be upregulated by statins in endothelial cells (28). The addition of mevalonate also blocked the statin effect on human T cell KLF2 expression (Figure 3E), and GGTI reproduced the statin effect (Figure 3F). These findings are similar to our findings with mouse T cells (Figure 2), indicating that human T cell KLF2 expression is suppressed by an isoprenylation-dependent mechanism and that this mechanism is conserved between these 2 species.

Figure 3

Statins increase human T cell KLF2 expression by a mevalonate/isoprenoid-dependent pathway. Normal human blood T cells were stimulated with αCD3 (6 hours) or with IL-2 plus PHA for 4 days and/or either simvastatin or pitavastatin for 18 hours, before RNA or cytoplasmic lysate isolation, and (A–C) qRT-PCR analysis or (D) Western blot analysis for KLF2 expression was performed. Simvastatin was used at 10 μM and pitavastatin was used at 1 μM or indicated concentrations. (E) Normal human blood T cells were cultured with vehicle or 10 μM simvastatin alone or with 100 μM mevalonate for 8 hours, before RNA isolation and qRT-PCR analysis of KLF2. (F) Normal human blood T cells were cultured in the presence of vehicle, 20 μM FTI, or 20 μM GGTI for 8 hours, before RNA was isolated from the T cells for qRT-PCR analysis. (A) Data are the mean ± SD from 2 donors. (B) Data are from 1 of 2 donors, with similar results. (C) Data are the mean ± SEM from 3 different donors. *P < 0.05 (ANOVA). (D) Data are from 1 of 2 donors, with similar results. (E and F) Data are means ± SD of triplicate samples from 1 of 2 donors, with similar results.

Effects of statins on T cell functional responses. We next examined the influence of statins on functional responses of mouse and human T cells. We found that the proliferative responses of naive mouse CD8⁺ T cells to TCR simulation were markedly inhibited by simvastatin or pitavastatin (Figure 4, A and B). We also observed that effector CD8⁺ T cell proliferation (Figure 4C) and IFN-γ expression (Figure 4, D and E) were both markedly inhibited by statins. These data suggest that statins could impair both priming and effector phases of T cell responses. Simvastatin and pitavastatin also inhibited the modest proliferative responses of human blood T cells stimulated with αCD3 (Figure 4F) and the more robust proliferative responses of the human T cells to αCD3 plus anti-CD28 (Figure 4G). Human T cell IFNG expression, induced by PHA plus IL-2 stimulation, was also significantly reduced by simvastatin (Figure 4H).

Figure 4

Statins reduce T cell proliferation and cytokine expression in vitro. (A and B) Naive or (C) effector mouse OT-1 cells were stimulated with αCD3 in the absence or presence of 10 μM simvastatin or the indicated concentrations of pitavastatin for 72 hours. Proliferation was measured by ³H-thymidine incorporation during the final 16 hours of culture. (D and E) Effector mouse OT-1 cells were simulated with αCD3 in the absence or presence of simvastatin for 48 hours, and (D) RNA was isolated and analyzed by qRT-PCR for Ifng expression, or (E) the cells were analyzed for intracellular IFN-γ by immunofluorescence staining and flow cytometry. (F) Human blood CD3⁺ T cells were stimulated with αCD3 or (G) mixture of αCD3, anti-CD28, and IL-2 for 48 hours in the absence or presence of 10 μM simvastatin or the indicated concentrations of pitavastatin, and proliferation was measured by ³H-thymidine incorporation during the final 16 hours of culture. (H) Human T cells were stimulated with PHA and IL-2 for 4 days, and then treated with 10 μM simvastatin for 18 hours, and IFNG expression was analyzed by qRT-PCR. Data are the mean ± SEM of triplicate experiments. P values are compared with vehicle control–treated groups. *P < 0.05. The FACS plot shown in E is representative, and the column chart shows mean ± SEM, of 12 separate analyses, with 2 different OT-1 preparations.

Statin effects on T cell function are dependent on upregulation of KLF2. In order to directly test whether statin effects on T cell function are mediated through modulation of KLF2 expression, we chose to study the influence of statins on human T cell functional responses when KLF2 expression was blocked. Reliable RNA knockdown in primary mouse T cells has proven very difficult to accomplish, and mice with T cell–specific deficiency of KLF2 have very few mature peripheral T cells (7, 9, 29). Therefore, in order to directly demonstrate that statin effects on mature effector T cell function are dependent on changes in KLF2 expression, we chose to block KLF2 translation in activated human T cells, using a lentiviral shRNA approach. We infected human blood T cells with lentiviruses expressing GFP and either KLF2-specific shRNA or random (control) shRNA. In order to accomplish efficient infection with shRNA-lentiviral vectors, it was necessary to first stimulate the cells with PHA plus IL-2. Since TCR stimulation markedly downregulated KLF2, we rested the T cells for 14 days before infection and statin treatment. As seen in Figure 5A, we found that statin induction of T cell KLF2 was diminished by KLF2 shRNA–expressing virus, compared with that of control virus. We were able to block KLF2 expression by between 80%–92% in statin-treated human T cells, in 4 separate experiments. Importantly, we found KLF2 shRNA–expressing virus completely blocked the ability of statins to reduce T cell IFNG expression (Figure 5B). Without shRNA knockdown, statin treatment reduced IFNG expression by 59%. The level of IFNG mRNA in KLF2 shRNA–infected cells was markedly higher than that in control-infected cells, and statin treatment did not reduce the elevated level. Therefore, the suppression of IFNG by the statin was completely dependent on KLF2 upregulation, establishing the necessity of KLF2 as a mediator of the T cell antiinflammatory effects of statins.

Figure 5

Blockade of KLF2 mRNA expression impairs statin suppression of IFN-γ. IL-2– and PHA-stimulated human blood T cells were rested for 14 days and then infected with control shRNA/GFP– or KLF2 shRNA/GFP–expressing lentivirus. GFP-positive cells, sorted 48 hours after infection, were further treated with or without 10 μM simvastatin for 18 hours and then analyzed using qRT-PCR for (A) KLF2 or (B) IFNG expression. mRNA was quantified as the percentage of β-actin, and then the data were normalized for basal gene expression in rested, control virus–infected cells without simvastatin treatment. Data are mean ± SEM from 4 separate experiments. *P < 0.05.

Statin treatment of CD8⁺ effector T cells reduces their pathogenicity in vivo. Because we found that statins reduce CD8⁺ functional responses, we predicted that CD8⁺ T cell–mediated pathology would be abrogated by statin treatment of T cells. In order to investigate this possibility, we took advantage of a model of mouse of CTL-mediated myocarditis, previously developed in our laboratory, in which effector OT-1 cells (ovalbumin specific) are transferred into cMy-mOva mice, which express ovalbumin in cardiomyocytes (30, 31). Accordingly, we treated effector OT-1 cells with 10 μM simvastatin or DMSO for 18 hours, before their adoptive transfer into cMy-mOva mice. Aliquots of T cells were removed before adoptive transfer, and analysis of these cells confirmed that the expression of Klf2 and S1pr1 was upregulated and that of Ifng was downregulated by simvastatin, as we observed before (Figure 6A). Likewise, the proliferative responses of the statin-treated OT-1 cells to αCD3 were reduced (Figure 6B). FACS analysis indicated that both DMSO- and simvastatin-treated cells were equally well labeled with CFSE, and they were all viable (data not shown). Five days after transfer of statin or vehicle control–treated OT-1 cells, cMy-mOva mice were sacrificed and analyzed. There was no significant difference in the recovery of CFSE-labeled cells from the spleens of mice receiving control- versus statin-treated T cells (Figure 6C), indicating equivalent viabilities over 5 days. The cardiac draining lymph node (CDLN), which we have previously described as a site reflecting cardiac inflammation (30, 31), was visibly larger in each of the mice that received control T cells compared with the nodes of the recipients of statin-treated T cells, and this was corroborated by total cell counts of the lymph nodes (7.7 × 10⁷ vs. 2.1 × 10⁷ cells/node in control vs. statin-treated groups, respectively). There were 2.3-times more vehicle control–treated OT-1 cells in the CDLNs than simvastatin-treated OT-1 cells (Figure 6D). There were also more B cells and CD11b⁺ cells in the nodes of the control group, compared with the mice that received simvastatin-treated OT-1 cells (data not shown), indicative of more cardiac inflammation in the control group. There were more inflammatory cells in the myocardium of the mice receiving the control T cells compared with mice receiving simvastatin-treated T cells, as indicated by histopathological scoring (Figure 6E). Immunohistochemical analyses of the presence of transferred OT-1 cells in the myocardium revealed significantly more control T cells compared with simvastatin-treated T cells in the hearts of recipient mice (Figure 6F). Serum troponin levels, which are directly related to the degree of myocyte damage, were over 6-times higher in mice that received control- versus simvastatin-treated T cells (Figure 6G). qRT-PCR analysis of RNA from heart tissue revealed significantly more Ifng, Vcam1, Cxcl10, and Ccl5 in the hearts from mice injected with control T cells compared with those injected simvastatin-treated T cells (Figure 6H).

Figure 6

Statin treatment reduces pathogenicity of T cells. Effector OT-1Thy1.1 cells were treated with DMSO (cntrl) or 10 μM simvastatin for 18 hours, and then left untreated or restimulated with αCD3 for 24 hours before either (A) performing qRT-PCR analysis of Klf2, S1pr1, and Ifng expression or (B) analyzing proliferative responses, as described in previous figures. Aliquots of the same DMSO- or simvastatin-treated OT-1Thy1.1 cells were labeled with CFSE and 5 × 10⁴ cells were adoptively transferred into cMy-mOvaThy1.2 mice. At day 5, the mice were sacrificed, the number of OT-1 cells in (C) spleen and the (D) CDLN was determined by FACS, (E) myocardial inflammation was scored on histological sections of heart, and (F) the relative numbers of OT-1 cells in the myocardium were determined by Thy 1.1–specific immunohistochemistry. (G) Troponin I (TnI) levels were determined in serum collected from 5 mice of each group, at the time of sacrifice. (H) qRT-PCR analyses of Ifng, Vcam1, Cxcl10, and Ccl5 were performed on myocardial RNA. (A and B) Data are from 1 of 2 OT-1 preparations, with similar results. (B) Data are mean ± SD of quintuplicate determinations. (C, D, and H) Data are the mean ± SEM of samples from 5–9 mice in total, from 2 separate experiments with different OT-1 preparations. (E–G) Horizontal bars represent the mean of each group. Original magnification, ×100; ×400 (insets).

In order to determine whether in vivo statin treatment could suppress pathogenic effects of T cells, we transferred OT-1 T cells into pitavastatin- or vehicle-treated cMy-mOva mice and sacrificed them 5 days after transfer. We found that cardiac damage, as assessed by serum troponin levels, was markedly reduced in the statin-treated mice (Supplemental Figure 3). This finding is consistent with direct statin effects on T cells but may also reflect effects on APCs.

Ectopic KLF2 expression in CD8⁺ effector T cells mimics effects of statin treatment and reduces their pathogenicity in vivo. In light of the findings that statins increased the expression of Klf2 in OT-1 T cells and reduced their cardiac pathogenicity and that the statin effects are mediated, at least in part, through KLF2 expression, we predicted that ectopic expression of Klf2 would have similar effects as statin treatment. Accordingly, we prepared KLF2/GFP-expressing and control GFP–expressing retroviruses (KLF2-RV and Cntrl-RV, respectively), infected effector OT-1 cells with these viruses, FACS-sorted GFP⁺ cells, and transferred them into cMy-mOva mice. Klf2 mRNA expression in the KLF2-RV–infected T cells was about 6% of that of β-actin (Actb) mRNA levels, which is approximately 16-fold higher than in Cntrl-RV–infected T cells (Figure 7A). These levels, achieved by forced expression of ectopic Klf2, were about the same as those in naive OT-1 cells and statin-treated effector OT-1 cells (~4%–5% of Actb mRNA; see Figures 2, A, B, and Figure 6A). S1pr1 mRNA levels were 3-fold higher in KLF2-RV–infected T cells compared with Cntrl-RV–infected T cells (Figure 7A). KLF2-RV–infected T cells also proliferated less in response to αCD3 than Cntrl-RV–infected T cells (Figure 7B). At the time of sacrifice, there was the same number of CFSE-labeled cells in the spleens of mice receiving Cntrl-RV– versus KLF2-RV–infected T cells (Figure 7C), indicating equivalent viabilities over 5 days. Five days after adoptive transfer, there were 8-times more Cntrl-RV–infected OT-1 cells in the CDLNs than KLF2-RV–infected OT-1 cells (Figure 7D). There were also more B cells and CD11b⁺ cells in the nodes of the Cntrl-RV group compared with the KLF2-RV group (data not shown). Examination of the histological sections of cMy-mOva hearts revealed more inflammatory cell infiltration of the myocardia of the mice receiving the Cntrl-RV–infected T cells compared with mice receiving KLF2-RV–infected T cells (Figure 7E). Immunohistochemical analyses of the hearts from the cMy-mOva mice revealed significantly more OT-1 cells in mice receiving Cntrl-RV–infected T cells compared with mice receiving KLF2-RV–infected T cells (Figure 7F). Serum troponin levels were over 30-times higher in mice that received Cntrl-RV–infected versus KLF2-RV–infected T cells (Figure 7G). qRT-PCR analysis of RNA from heart tissue revealed significantly more Ifng, Vcam1, Cxcl10, and Ccl5 expression in the hearts from mice injected with Cntrl-RV–infected T cells compared with those injected with KLF2-RV–infected T cells (Figure 7H). Overall, the effects of KLF2 overexpression in OT-1 cells, in the cMy-mOva myocarditis model, were remarkably similar to the effects of ex vivo statin treatment of the T cells shown in Figure 6.

Figure 7

Ectopic KLF2 expression reduces pathogenicity of T cells. Effector OT-1^Thy1.1 cells infected with Klf2/GFP– or control GFP–expressing retroviruses (KLF2-RV or Cntrl-RV, respectively) were prepared, as described in Methods, and (A) analyzed using qRT-PCR for Klf2 and S1pr1 and (B) for proliferative responses to αCD3, as described in previous figures. Aliquots of the same KLF2-RV– or Cntrl-RV–infected OT-1^Thy1.1 cells (5 × 10⁴) were adoptively transferred into cMy-mOva^Thy1.2 mice. At day 5, the mice were sacrificed and analyzed for OT-1 cells (C) in spleen and (D) in the CDLN, (E) the amount of myocardial inflammation, and (F) the relative numbers of OT-1 cells in the heart. (G) Serum troponin levels were determined, as described in Figure 6. (H) qRT-PCR analyses of Ifng, Vcam1, Cxcl10, and Ccl5 were performed on myocardial RNA. (A and B) Data are from 1 of 2 experiments, with similar results. (B) Data are mean ± SD of quintuplicate determinations. (C, D, and H) Data are the mean ± SEM of samples from 5 mice in total. (E–G) Horizontal bars represent the mean of each group. Original magnification, ×100.

In this study, we show that statins directly modulate T cell function, in vitro and in vivo. We also show that statins upregulate the expression of KLF2 in T cells and the functional effects of statins on T cells are, at least in part, dependent on KLF2 expression. These findings are consistent with 2 independent lines of published work, showing that statins have inhibitory effects on T cells and ameliorate T cell–mediated autoimmunity (reviewed in refs. 32, 33) and that dynamic changes in KLF2 expression in T cells is tightly linked to activation status and migratory function of T cells (2, 6, 7, 10, 34). Previous work has also established that statins upregulate KLF2 expression in endothelial cells and that KLF2 expression is linked to an antiinflammatory phenotype of endothelium (21, 23).

Our studies establish direct effects of statins on T cell function in both mouse and human T cells and in both CD4⁺ and CD8⁺ T cells. Although statins have a variety of effects on myeloid APCs that may have significant impact on T cell activation, our in vitro experiments were conducted with polyclonal stimuli in the absence of APCs in order to discern direct T cell effects. A variety of lipophilic statin drugs showed similar effects. Others have shown direct effects of statins on T cells, and various mechanisms for these effects have been proposed, such as inhibition of isoprenylation of Rho/Rac family GTPases (33). We show here that the statin effect on T cell KLF2 expression is completely abrogated by mevalonate, the metabolite generated by HMG-CoA reductase, and partially abrogated by isoprenoids, the downstream products of mevalonate. Our finding that statins induce expression of a gene that directly downregulates T cell activation has not been previously shown to our knowledge.

We examined effects of statins on T cell KLF2 expression and function in vivo in 4 different ways. First, we treated mice with statin injections over 3 days and then examined Klf2 mRNA in T cells isolated from spleens. We found significant elevations in Klf2 mRNA levels in the T cells ex vivo after αCD3 treatment. Second, we gavage treated ovalbumin-specific TCR transgenic OT-1 mice with statins or vehicle before and after subcutaneous ovalbumin immunization, and we determined that Klf2 mRNA was upregulated in the draining lymph node T cells. Third, we treated effector OT-1 cells with vehicle or statins in vitro and then transferred them into cMy-mOva mice expressing ovalbumin in the myocardium. Using this established model of CD8⁺ T cell–mediated cardiac inflammation (30, 31, 35), we found that statin treatment markedly reduced the T cell pathogenicity. Fourth, we gavage fed cMy-mOva mice vehicle or statins before and after transferring OT-1 cells and again found that cardiac damage was reduced in the statin-treated group. The basis for the reduced pathogenicity of statin-treated T cells could reflect reduced proliferative activity, altered homing functions, and reduced effector functions. We found no consistent changes in expression of chemokine receptors for inflammatory chemokines between statin-treated cells and control T cells (data not shown). Statin treatment did upregulate Ccr7 and Sell mRNA, which encodes proteins that direct T cell homing to lymph nodes, but paradoxically, we found fewer statin-treated OT-1 cells in the heart draining lymph node than control T cells. Statin-treated OT-1 cells also had upregulated S1PR1. Because the S1PR1 is required for T cell efflux from lymph nodes, it is plausible that statin-treated effector T cells were more likely to enter the CDLN, become reactivated there, and exit the node. Nonetheless, they may have proliferated less in the nodes. Statin treatment also reduced IFN-γ expression by the T cells, but this is not likely explain reduced pathogenicity, because we have found that T cell IFN-γ deficiency is not protective in our model of myocarditis (35).

Our data unequivocally demonstrate that statins can induce KLF2 gene expression in mouse and human T cells. We show that statins block the reduction in Klf2 expression that normally occurs when naive T cells are activated through the TCR, and they significantly increase the usually minimal expression of Klf2 in effector T cells. Previously reported studies have shown that KLF2 has multiple profound effects on T cell function (6, 7, 9, 10, 34, 36). Overall KLF2 is required for the expression of genes that maintain the phenotype of naive T cells and resting memory T cells. If statin effects on T cell phenotype and function are dependent on the induced changes in KLF2 expression, then these effects should be at least partly recapitulated by constitutive expression of an exogenously introduced Klf2 gene. In fact, when we infected OT-1 effector cells with a Klf2-expressing retrovirus and increased constitutive expression over 16 fold compared with control virus–infected cells (a similar level to simvastatin-treated OT-1 effector cells), the resulting changes in the OT-1 cells were remarkably similar to the set of changes in OT-1 phenotype and function induced by statin treatment. The shared effects on OT-1 cells of ectopic KLF2 expression and statin treatment include elevated S1pr1 expression, reduced proliferation, reduced numbers of cells in the CDLN after transfer into cMy-mOva mice, reduced inflammation and numbers of OT-1 cells in the hearts of these mice, and reduced inflammatory gene expression in the hearts.

We found that statins reduce IFN-γ expression in mouse and human effector T cells. In an in vivo model of graft arterial disease, statins were shown to decrease IFN-γ expression by T cells within the grafts (37). The ability of statins to reduce IFNG expression in each of 4 human T cell preparations was suppressed when KLF2 upregulation was blocked, establishing a causal link between the statin effects on T cell function and on KLF2 upregulation. We also observed increased IFNG gene expression in the human T cells in which KLF2 expression was blocked. Increased IL-4 but not IFN-γ expression was observed in KLF2-deficient mouse single-positive thymocytes (10). We did not detect IL4 mRNA in either the control or the KLF2-blocked human T cells. Although we cannot explain the apparent difference in the effects of KLF2 deficiency on cytokines between the mouse gene deletion study and our human knockdown study, our results are consistent with a role for KLF2 in regulating expression of effector cytokines by mature peripheral T cells.

In summary, statins reduce proliferative and functional responses of T cells and also induce higher levels of T cell KLF2 expression. These effects result in reduced pathogenicity of T cells in an in vivo model of CD8⁺-mediated cardiac injury. The effects of statins on T cell function are dependent on KLF2 upregulation. Our results are important for the understanding of the therapeutic mechanisms of HMG-CoA reductase inhibitors and establish KLF2 as a potentially important target for the treatment of diseases in which proinflammatory T cells play a role, including atherosclerosis, myocarditis, and various autoimmune diseases.

View Supplemental data

This work was supported by National Institutes of Health grants AI059610 and HL087282 (to A.H. Lichtman) and HL-076686 and HL-89940 (to G. García-Cardeña). We wish to thank Hong Pang for excellent technical assistance and Gail Newton for assistance in human blood T cell preparations.

Address correspondence to: Andrew H. Lichtman, Brigham and Women’s Hospital, NRB Room 752N, 77 Avenue Louis Pasteur, Boston, Massachusetts 02115, USA. Phone: 617.525.4335; Fax: 617.525.4333; E-mail: alichtman@rics.bwh.harvard.edu.

Conflict of interest: The authors have declared that no conflict of interest exists.

Citation for this article: J Clin Invest. 2010;120(6):1961–1970. doi:10.1172/JCI41384.

1. Kuo CT, Veselits ML, Barton KP, Lu MM, Clendenin C, Leiden JM. The LKLF transcription factor is required for normal tunica media formation and blood vessel stabilization during murine embryogenesis. Genes Dev. 1997;11(22):2996–3006.
   View this article via: PubMed CrossRef Google Scholar
2. Buckley AF, Kuo CT, Leiden JM. Transcription factor LKLF is sufficient to program T cell quiescence via a c-Myc–dependent pathway. Nat Immunol. 2001;2(8):698–704.
   View this article via: PubMed CrossRef Google Scholar
3. Wu J, Lingrel JB. KLF2 inhibits Jurkat T leukemia cell growth via upregulation of cyclin–dependent kinase inhibitor p21WAF1/CIP1. Oncogene. 2004;23(49):8088–8096.
   View this article via: PubMed CrossRef Google Scholar
4. Schober SL, Kuo CT, Schluns KS, Lefrancois L, Leiden JM, Jameson SC. Expression of the transcription factor lung Kruppel–like factor is regulated by cytokines and correlates with survival of memory T cells in vitro and in vivo. J Immunol. 1999;163(7):3662–3667.
   View this article via: PubMed Google Scholar
5. Grayson JM, Murali-Krishna K, Altman JD, Ahmed R. Gene expression in antigen-specific CD8+ T cells during viral infection. J Immunol. 2001;166(2):795–799.
   View this article via: PubMed Google Scholar
6. Wu J, Lingrel JB. Kruppel–like factor 2, a novel immediate-early transcriptional factor, regulates IL-2 expression in T lymphocyte activation. J Immunol. 2005;175(5):3060–3066.
   View this article via: PubMed Google Scholar
7. Carlson CM, et al. Kruppel–like factor 2 regulates thymocyte and T-cell migration. Nature. 2006;442(7100):299–302.
   View this article via: PubMed Google Scholar
8. Kuo CT, Veselits ML, Leiden JM. LKLF: A transcriptional regulator of single-positive T cell quiescence and survival. Science. 1997;277(5334):1986–1990.
   View this article via: PubMed CrossRef Google Scholar
9. Sebzda E, Zou Z, Lee JS, Wang T, Kahn ML. Transcription factor KLF2 regulates the migration of naive T cells by restricting chemokine receptor expression patterns. Nat Immunol. 2008;9(3):292–300.
   View this article via: PubMed CrossRef Google Scholar
10. Weinreich MA, Takada K, Skon C, Reiner SL, Jameson SC, Hogquist KA. KLF2 transcription-factor deficiency in T cells results in unrestrained cytokine production and upregulation of bystander chemokine receptors. Immunity. 2009;31(1):122–130.
    View this article via: PubMed CrossRef Google Scholar
11. Youssef S, et al. The HMG-CoA reductase inhibitor, atorvastatin, promotes a Th2 bias and reverses paralysis in central nervous system autoimmune disease. Nature. 2002;420(6911):78–84.
    View this article via: PubMed CrossRef Google Scholar
12. Kwak B, Mulhaupt F, Myit S, Mach F. Statins as a newly recognized type of immunomodulator. Nat Med. 2000;6(12):1399–1402.
    View this article via: PubMed CrossRef Google Scholar
13. Arnaud C, Braunersreuther V, Mach F. Toward immunomodulatory and anti-inflammatory properties of statins. Trends Cardiovasc Med. 2005;15(6):202–206.
    View this article via: PubMed Google Scholar
14. Hakamada-Taguchi R, et al. Inhibition of hydroxymethylglutaryl-coenzyme a reductase reduces Th1 development and promotes Th2 development. Circ Res. 2003;93(10):948–956.
    View this article via: PubMed CrossRef Google Scholar
15. Mausner-Fainberg K, et al. The effect of HMG-CoA reductase inhibitors on naturally occurring CD4+CD25+ T cells. Atherosclerosis. 2008;197(2):829–839.
    View this article via: PubMed Google Scholar
16. Frenette PS. Locking a leukocyte integrin with statins. N Engl J Med. 2001;345(19):1419–1421.
    View this article via: PubMed CrossRef Google Scholar
17. Blank N, et al. Atorvastatin inhibits T cell activation through 3-hydroxy-3-methylglutaryl coenzyme A reductase without decreasing cholesterol synthesis. J Immunol. 2007;179(6):3613–3621.
    View this article via: PubMed Google Scholar
18. Ghittoni R, et al. Simvastatin inhibits T-cell activation by selectively impairing the function of Ras superfamily GTPases. FASEB J. 2005;19(6):605–607.
    View this article via: PubMed Google Scholar
19. Goldman F, Hohl R, Crabtree J, Lewis-Tibesar K, Koretzky G. Lovastatin inhibits T-cell antigen receptor signaling independent of its effects on ras. Blood. 1996;88(12):4611–4619.
    View this article via: PubMed Google Scholar
20. Dykstra M, Cherukuri A, Sohn HW, Tzeng S-J, Pierce SK. Location is everything: lipid rafts and immune cell signaling. Annu Rev Immunol. 2003;21:457–481.
    View this article via: PubMed Google Scholar
21. Parmar KM, Nambudiri V, Dai G, Larman HB, Gimbrone MA Jr, Garcia-Cardena G. Statins exert endothelial atheroprotective effects via the KLF2 transcription factor. J Biol Chem. 2005;280(29):26714–26719.
    View this article via: PubMed Google Scholar
22. Sen-Banerjee S, et al. Kruppel-like factor 2 as a novel mediator of statin effects in endothelial cells. Circulation. 2005;112(5):720–726.
    View this article via: PubMed CrossRef Google Scholar
23. Atkins GB, Jain MK. Role of Kruppel-like transcription factors in endothelial biology. Circ Res. 2007;100(12):1686–1695.
    View this article via: PubMed Google Scholar
24. Garcia-Cardena G, Gimbrone MA Jr. Biomechanical modulation of endothelial phenotype: implications for health and disease. Handb Exp Pharmacol. 2006;(176 pt 2):79–95.
    View this article via: PubMed Google Scholar
25. van Thienen JV, et al. Shear stress sustains atheroprotective endothelial KLF2 expression more potently than statins through mRNA stabilization. Cardiovasc Res. 2006;72(2):231–240.
    View this article via: PubMed CrossRef Google Scholar
26. Tuomisto TT, et al. Simvastatin has an anti-inflammatory effect on macrophages via upregulation of an atheroprotective transcription factor, Kruppel-like factor 2. Cardiovasc Res. 2008;78(1):175–184.
    View this article via: PubMed CrossRef Google Scholar
27. Dunn SE, et al. Isoprenoids determine Th1/Th2 fate in pathogenic T cells, providing a mechanism of modulation of autoimmunity by atorvastatin. J Exp Med. 2006;203(2):401–412.
    View this article via: PubMed CrossRef Google Scholar
28. Kinderlerer AR, et al. KLF2-dependent, shear stress-induced expression of CD59: a novel cytoprotective mechanism against complement-mediated injury in the vasculature. J Biol Chem. 2008;283(21):14636–14644.
    View this article via: PubMed Google Scholar
29. Yamada T, Park CS, Mamonkin M, Lacorazza HD. Transcription factor ELF4 controls the proliferation and homing of CD8+ T cells via the Kruppel-like factors KLF4 and KLF2. Nat Immunol. 2009;10(6):618–626.
    View this article via: PubMed Google Scholar
30. Grabie N, et al. IL-12 is required for differentiation of pathogenic CD8+ T cell effectors that cause myocarditis. J Clin Invest. 2003;111(5):671–680.
    View this article via: JCI PubMed Google Scholar
31. Taqueti VR, et al. T-bet controls pathogenicity of CTLs in the heart by separable effects on migration and effector activity. J Immunol. 2006;177(9):5890–5901.
    View this article via: PubMed Google Scholar
32. Greenwood J, Steinman L, Zamvil SS. Statin therapy and autoimmune disease: from protein prenylation to immunomodulation. Nat Rev Immunol. 2006;6(5):358–370.
    View this article via: PubMed Google Scholar
33. Ghittoni R, Lazzerini PE, Pasini FL, Baldari CT. T lymphocytes as targets of statins: molecular mechanisms and therapeutic perspectives. Inflamm Allergy Drug Targets. 2007;6(1):3–16.
    View this article via: PubMed Google Scholar
34. Bai A, Hu H, Yeung M, Chen J. Kruppel-like factor 2 controls T cell trafficking by activating L-selectin (CD62L) and sphingosine-1-phosphate receptor 1 transcription. J Immunol. 2007;178(12):7632–7639.
    View this article via: PubMed Google Scholar
35. Grabie N, et al. Endothelial programmed death-1 ligand 1 (PD-L1) regulates CD8+ T-cell mediated injury in the heart. Circulation. 2007;116(18):2062–2071.
    View this article via: PubMed CrossRef Google Scholar
36. Stein JV. How to be naive. Immunity. 2009;31(1):9–11.
    View this article via: PubMed CrossRef Google Scholar
37. Yi T, et al. Amelioration of human allograft arterial injury by atorvastatin or simvastatin correlates with reduction of interferon-gamma production by infiltrating T cells. Transplantation. 2008;86(5):719–727.
    View this article via: PubMed Google Scholar
38. Grabie N, et al. Neutrophils sustain pathogenic CD8+ T cell responses in the heart. Am J Pathol. 2003;163(6):2413–2420.
    View this article via: PubMed Google Scholar
39. Young A, et al. Flow activation of AMP-activated protein kinase in vascular endothelium leads to Kruppel-like factor 2 expression. Arterioscler Thromb Vasc Biol. 2009;29(11):1902–1908.
    View this article via: PubMed CrossRef Google Scholar