|Home | About | Journals | Submit | Contact Us | Français|
Calcineurin (CN) is a phosphatase that activates nuclear factor of activated T cells (NFAT). While the CN inhibitors cyclosporine A (CsA) and tacrolimus (FK506) can prevent graft rejection, they also cause inflammatory diseases. We investigated the role of calcineurin using mice deficient in the CN catalytic subunit Aβ (CNAβ). Cnab−/− mice exhibit defective thymocyte maturation, splenomegaly and hepatomegaly. Further, as Cnab−/− mice age, they exhibit spontaneous T-cell activation and enhanced production of proinflammatory cytokines (IL-4, IL-6, and IFNγ). FOXP3+ Treg cells were significantly decreased in Cnab−/− mice likely contributing to increased T-cell activation. Interestingly, we found that CNAβ is critical for promotion of BCL-2 expression in FOXP3+ Treg and for permitting TGFβ signaling, as TGFβ induces FOXP3 in control but not in Cnab−/− T-cells. Together, these data suggest that CNAβ is important for the production and maintenance of Treg cells and to ensure mature T-cell quiescence.
CN, a serine/threonine protein phosphatase, is activated by Ca2+-calmodulin upon elevation of intracellular Ca2+ levels by T-cell receptor (TCR) signaling, and it activates NFAT molecules by dephosphorylation. CN is a well characterized drug target for prevention of graft rejection in transplant patients [1–3]. The immunosuppressive drugs FK506 and CsA inhibit NFATs which are involved in lymphocyte responses . It has been shown that FK506 and CsA inhibit positive selection but not negative selection of thymocytes suggesting that CN activation is critical for T-cell development in the thymus . Although CsA and FK506 are effective in treating transplant patients for graft acceptance, long-term side effects include frequent infections, kidney failure, inflammatory bowel disease (IBD) and lymphomas [6–8].
The administration of CsA to new-born BALB/c mice induces organ-specific autoimmunity, inhibits Foxp3 expression, and blocks Treg-cell generation, both in vitro and in vivo [9–11]. FOXP3 is induced by NFAT and AP-1 in human T cells, and CsA inhibits FOXP3 expression through inhibition of NFAT activation . Under optimal stimulatory conditions which activate AP-1, NFAT induces not only Il2, but also Il4 and Ifng resulting in a productive immune response in mice [12–14]. Together these observations suggest that low levels of TCR signaling leads to T-cell differentiation into Treg cells whereas higher levels of TCR signaling leads to effector T-cell generation. Consistently, CNAβ-deficient mice do not reject allogenic tumors, whereas CNAα-deficient mice succumb to inflammation and kidney failure [15;16]. A defect in FasL expression and impaired apoptosis of T cells leads to splenomegaly and lymphadenopathy in mice doubly deficient for Nfatc2 and c3 . These studies suggest that CN signaling is required for T-cell activation, activation-induced cell death (AICD) and T-cell tolerance [12;18].
Under tolerogenic conditions, NFAT forms a complex with FOXP3 and together induces Treg-cell generation by induction of Il2ra (CD25) and Ctla4 (CD152) . Recently, Tone Y et. al. have found that the Foxp3 enhancer contains SMAD3 and NFAT binding sequences, suggesting cooperative positive regulation of Foxp3 by TGFβ and TCR signals . CN signaling also seems to play an important role in T-cell regulation as the proportion of CD4+CD25+ cells are increased in Nfatc2 and Nfatc3 double knockout (DKO) mice. Although the suppressor activity of the DKO Treg cells (CD4+CD25++GITR++) is comparable to that of their wildtype (WT) CD4+CD25++GITR++ T-cell counterparts, their CD4+CD25− T cells exhibit activated phenotype suggesting that they are not inhibited by Treg cells in vivo. These data suggest that an intrinsic defect makes the DKO T cells resistant to tolerance induction . Since CsA inhibits not only T-cell activation but also Treg-cell generation in mice  and humans , whether the inhibitory effect of CsA is through inhibition of CN signaling or through its effect on other CN-independent signaling pathways is still unclear.
CNAα-deficient mice have normal thymocyte maturation and T-cell responses to polyclonal stimulation, but they exhibit defective T-cell responses to antigenic stimulation [21;22]. CNAα KO mice die within 2 months after birth due to kidney failure . Although CNAα compensates to a certain extent for the loss of CNAβ in mature T cells, CNAβ plays a non-redundant role in T-cell homeostasis and T-cell regulation . The present study is a follow up to our recent study in which we have demonstrated that T cells in Cnab−/− mice are activated and produce increased levels of IL-6 and IFNγ . Since young CNAβ-deficient mice exhibit defective T-cell maturation  but they develop splenomegaly, and hepatomegaly we hypothesize that defective T-cell regulation increases inflammatory response resulting in lymphoid hyperplasia. In this study we present evidence that CNAβ deficiency affects Treg-cell-generation and leads to expansion of mature T cells with activated phenotype, suggesting that while CNAβ signaling is required for T-cell maturation and anti-tumor responses , it is also important for T-cell regulation and homeostasis.
Cnab−/− mice (C57BL/6) are obtained from Dr. Jeffery Molkentin (Cincinnati Children’s Hospital Medical Center)  were mated to DO11.10 TCR Tg Rag1−/− mice on a BALB/c background  . Mice were backcrossed 4 times to BALB/c mice. All experiments were reviewed and approved by the University of Arizona IACUC committee.
The genotype of newborn pups from double heterozygous matings was determined by PCR amplification of tail DNA and size fractionation on agarose gels .
Phenotype analysis of splenocytes was determined by four-color flow cytometry as described . FITC-, PE-, PerCP, PerCP-Cy5.5, Pacific Blue, or APC-conjugated antibodies to cell surface molecules were purchased from either BD Biosciences (San Diego, CA) or eBioscience (San Diego, CA). FOXP3 staining kit (clone FKJ-16s) was purchased from eBioscience (San Diego, CA). Thymocytes and splenocytes were surface stained with fluorochrome-conjugated antibodies after blocking with Fc blocking antibodies in FACS staining buffer at 4°C in the dark. Cells were washed once with FACS buffer, fixed in 2% paraformaldehyde and acquired using a BD-LSR or BD-LSR-II flow cytometer. Intracellular FOXP3 staining was performed according to manufacturer’s protocol as described in our earlier studies . Flow cytometry data were analyzed by Cell-Quest or FacsDIVA software.
For FOXP3 induction, splenocytes were isolated from Cnab−/− and control mice and 1×106 cells were cultured in RPMI-1640 medium supplemented with 10% FBS in the presence of anti-CD3/CD28 beads (Dynal Biotech, Invitrogen), IL-2 (100 IU/ml) and TGFβ1 (2–5 ng/ml) in 48 well plate for 3 days. Cells were harvested and stained for surface expression of CD4, CD25 and intracellular FOXP3 as described above. For cytokine measurements, culture supernatants were harvested and assayed for cytokines by ELISA kits from BD Biosciences as described in our recent study .
Splenocytes were isolated from Cnab−/− and control mice and 1×106 cells were cultured in s-MEM supplemented with 10% FBS in 96 well plates coated with anti-CD3/CD28 antibodies and Brefeldin A (10 µg/ml). Cells were harvested and stained for surface expression of CD4, washed, fixed in BD Cytofix/Cytoperm buffer (BD Biosciences, San Diego, CA) on ice, washed in Perm/wash buffer and resuspended in BD Cytoperm Plus buffer, washed with BD Perm/Wash buffer and then stained intracellularly with antibodies against FOXP3, IFNγ, and IL-17. Flow cytometry data were analyzed by FacsDIVA software.
The significance of the differences between wild-type and mutant mouse responses was calculated using Student’s t-test.
CNAβ deficiency causes a severe reduction in mature T-cell populations both in the thymus and periphery, and it also causes a defect in T-cell responsiveness . Cnab−/− mice also fail to reject allogenic tumors suggesting an impaired immune response . In order to understand the cause of the splenomegaly in older Cnab−/− mice we have analyzed T-cell populations at different ages. Consistent with earlier observations, flow cytometry analysis of thymocytes indicates decreased mature CD4+ and CD8+ single positive (SP) thymocytes in 2 wk-old Cnab−/− mice . Although the proportion of mature SP thymocytes is much smaller (30–50% of control) in Cnab−/− pre-weanlings (<2 wk) compared to their control mice (Figure 1A) and mature splenic T cells (Figure 1B, upper panels), the numbers steadily increase and reach comparable levels as the mice become older (Figure 1B, middle and lower panels). Thus, despite defects in thymocyte development and production, the total numbers of splenocytes (Figure 1C), CD4+ T cells (Figure 1D & E) and CD8+ T cells (data not shown) increase with age albeit to a lesser extent (compare upper left quadrants in Figure 1B) suggesting that with age the KO mice are overcoming the lymphopenia that is observed in younger KO mice.
Calcineurin signaling has been shown to be important for T-cell maturation and activation as evidenced by gene ablation studies and use of pharmacological agents that target calcineurin [16;27;28]. As we found increased numbers of T cells in older Cnab−/− mice, we analyzed T-cell activation by flow cytometry. Already by 2 months of age, we found that the frequency and total numbers of activated CD4+ T cells were significantly increased in Cnab−/− mice as assessed by expression of CD44 and CD62L (Figure 2A–C). As we have previously shown that Bcl-2 levels decrease in T cells activated by strong stimuli in vivo , and a recent report suggested that CNAβ was required for expression of BCL-2  we next measured Bcl-2 levels within T cells from Cnab−/− versus control mice by intracellular flow cytometry using a BCL-2-specific antibody (see supplemental figure 1). We found that both naïve and activated T cells in Cnab−/− mice have significantly increased levels of BCL-2 compared to WT controls (Figure 2 D,E), consistent with their increased accumulation. Further, most of these cells also expressed IL-7Rα at relatively high levels (data not shown), and a small percentage of them express CD69 at this time point . However, the increase in BCL-2 levels did not correlate with T-cell accrual as BCL-2 levels were increased to a greater extent in naïve T cells that did not accumulate compared to activated T cells which did accumulate (Figure 2 D, E). Thus, our data suggest that CNAβ is critical to maintain peripheral T-cell quiescence, but not to maintain normal expression of BCL-2 in conventional CD4+ T cells.
CN activates NFAT and NFAT induces Il2 and Foxp3 expression and cooperates with TGFβ signaling for Treg-cell generation and T-cell tolerance . To test whether T-cell activation in Cnab−/− mice is due to a defect in Treg-cell generation, we have analyzed mRNA for FOXP3 via RT-PCR and protein expression of FOXP3, CD25 and other T-cell markers via flow cytometry. Our RT-PCR data show that Foxp3 expression is down-regulated in the thymus of 2-week-old Cnab−/− mice (Figure 3A). We further analyzed Treg-cell generation in two-three-month-old adult mice. The proportion of FOXP3+CD4+ CD25+ Treg cells is decreased in KO mice, and the absolute numbers of FOXP3+ Treg cells are significantly reduced in Cnab−/− mice at 2–3 months of age (Figure 3B & C). Interestingly, we found that intracellular levels of BCL-2 within FOXP3+ Treg was decreased in Cnab−/− compared to WT controls (Figure 4A & B), consistent with the decreased accumulation of Treg in Cnab−/− mice. Since BCL-2 levels are decreased in activated T cells compared to naïve T cells, we have determined the activation status of FOXP3+ Treg cells. We have observed that the majority of FOXP3+ T cells are of the effector/memory phenotype, and that there is a significant increase of that population in Cnab−/− mice (Figure 4C). Together, these data suggest that decreased numbers of Treg cells is due to both production and survival. Further, these data suggest that the reduced numbers of Treg cells may contribute to the enhanced immune activation in Cnab−/− mice.
Recent studies have established that TGFβ1 induces FOXP3 expression and that the Foxp3 enhancer contains SMAD3 and NFAT binding sequences, suggesting cooperative regulation of Foxp3 expression by TGFβ and TCR signals . Mice with a T cell-specific deficiency in the CN regulatory subunit CNB have a block in thymic T-cell development at the double positive stage , and mice deficient in CN catalytic subunit alpha (CNAα) die early due to inflammation and kidney failure . Since mice deficient in CNAβ exhibit T-cell activation and splenomegaly later in the life, we have decided to determine whether CNAβ signaling is required for TGFβ1-induced FOXP3 induction in T cells. Our data revealed that CNAβ deficiency significantly affects the generation of FOXP3+ iTreg-cells (Figure 5A & B). The defect in FOXP3 expression by TGFβ1 in Cnab−/− cultures could not be due to a defect in IL-2 production since addition of IL-2 (100 IU/ml) to the cultures did not restore FOXP3 expression in Cnab−/− cultures. Decreases in FOXP3+ Treg cells in the Cnab−/− cultures are correlated with an increase in IFNγ (Th1), IL-4 (Th2) and IL-10 (Tfh, Tr1 and Th9) cytokine secretions (Figure 5C (7-wk-old); data not shown). These data suggest that CNAβ signaling is critical for prevention of spontaneous activation of T cells and maintenance of tolerance through inhibition of inflammatory response.
As there was an increased frequency of activated T cells and a decreased frequency of Treg cells in Cnab−/− mice we next assessed the inflammatory profile of their T cells. To do this, we performed intracellular cytokine staining for IFN-g and IL-17 in CD4+ and CD4+ FOXP3+ T cells in Cnab−/− and control mice. After brief stimulation with anti-CD3/CD28, we found that the frequencies of CD4+ T cells that were IFNγ+ or IL-17+ were significantly increased in Cnab−/− mice (Figure 6A–C). Interestingly, we also found that a substantial fraction of CD4+FOXP3+ T cells were secreting IL-17, but not IFN-γ (Figure 6D–F), suggesting that in the absence of CNAβ signaling, a subpopulation of Treg cells may be pro-inflammatory rather than suppressive.
Since we have observed activation of T cells and defective Treg-cell generation in Cnab−/− mice, we hypothesized that the splenomegaly in these mice is due to expansion of activated T-cell populations. To test this, we mated Cnab−/− mice to DO11.10 Rag1−/− TCR transgenic mice which have no other lymphocytes except transgenic TCR-bearing CD4+ T cells that will recognize ovalbumin peptides in the context of I-Ad. We have observed that DO11.10 Rag1−/− Cnab−/− mice did not develop splenomegaly and lymphomas. However, T cells in these mice responded well to stimulation compared to Cnab+/+ DO11.10 Rag1−/− control T cells as assessed by CD25 expression, but exhibited a decrease in FOXP3 expression (26% in control vs. 13% in KO) upon addition of TGFβ1 (Figure 7). Presence of Cnab−/− B cells (Cnab−/− DO11.10 Rag1+/− mice) resulted in a severe reduction in FOXP3 expression in T cells (Figure 7, lower panels). These data suggest that FOXP3 induction by TGFβ1 is partly dependent on CNAβ signaling.
Our results presented here suggest a regulatory role for CNAβ in T cells in addition to its known function for T-cell maturation and development. Our data has shown that CNAβ deficiency results in divergent effects on Treg versus conventional T cells (Tconv). In Tconv, CNAβ is critical to promote T-cell quiescence and inhibit proinflammatory cytokine production. In Treg cells, CNAβ is critical for expression of BCL-2 and permitting responsiveness to TGFβ1. Combined, these defects translate to altered overall control of T-cell homeostasis and quiescence.
Our data suggest that CNAβ has multiple anti-inflammatory effects in the immune system. We found that CNAβ (i) limits spontaneous pro-inflammatory Th1 and Th17-cell generation; (ii) controls Treg-cell generation from the thymus; (iii) is critical for generation of inducible Treg cells; and reciprocally affects expression of BCL-2 in Tregs vs. conventional T cells. Additionally, Cnab−/− T cells produce increased amounts of IL-6 which is known to suppress lymphokine-activated killer (LAK)-cell response [23;31].
T-cell homeostasis is achieved by production and survival of distinct T-cell populations (naïve, effector and memory). Mice deficient in either BIM, BCL-2, FAS or FASL all develop defects in T-cell homeostasis [32–36]. Since CN mediates TCR activation signals and since T-cell activation down-modulates BCL-2 expression, we reasoned that CN also may regulate T-cell homeostasis by modulating expression of BCL-2. Indeed, a recent report showed that BCL-2 protein expression is downregulated in Cnab−/− mouse lymphocytes . Surprisingly, our data revealed that there is no such down-regulation of BCL-2 expression in naïve T cells in Cnab−/− mice compared to control mice. Since BCL-2 deficiency causes lymphopenia  and early lethality , and CNAβ-deficient mice develop splenomegaly and lymphomas, the previously observed loss of BCL-2 levels may have reflected a loss of a population of CD4+ T cells that express lower levels of BCL-2. Indeed, as we observed a significant decrease in BCL-2 levels in effector/effector-memory CD4+ T cells compared to naïve T cells (Figure 2D–E) in both control and CNAβ KO mice, it is possible that Western-blot analysis of total T cells will yield lesser signal intensity (BCL-2) in the CNAβ KO T cells since there is a significant increase in activated T cells in CNAβ KO mice. Nonetheless, our data clearly show that BCL-2 is expressed in conventional Cnab−/− T cells. Dysregulation of these normal survival/apoptotic processes could foster expansion of activated T cells, inflammation and inflammation-induced cancers such as lymphoma and IBD–associated colon cancer.
Even though Treg cells are reduced in CNAβ-deficient mice, they are not absent. As we observed that T cells undergo activation in the presence of Treg cells, our data suggest that either natural Treg cells are not functioning normally, or that the observed defect in inducible Treg (iTreg)-cell generation may contribute to the immune dysregulation. A cardinal feature of the immune dysregulation in CNAβ-deficient mice we have observed is the increased production of IL-6 and IFNγ . Intracellular cytokine data presented here confirm that conventional CD4+ T cells are differentiated towards Th1 and Th17 in the absence of CNAβ (Figure 6), identifying a novel anti-inflammatory role for CNAβ. Our data would suggest that the loss of Treg cells in Cnab−/− mice leads to spontaneous activation of T cells and B cells that promotes an inflammatory environment that favors lymphoma development. In support of this, we find that (i) T cells from Cnab−/− mice have significantly increased production of IL-6 when stimulated ex vivo; (ii) that the majority of lymphomas that develop are B-cell lymphomas; and (iii) that mice overexpressing IL-6 generate B-cell lymphomas . Although more work is required to determine the role of IL-6 on lymphoma development, our data here are supportive of such a model and describe for the first time, an anti-inflammatory role for CNAβ in both conventional and Treg cells. It is possible that CNAβ signaling is required for normal selection of T cells in the thymus during development and that a defect in negative selection causes peripheral dysregulation.
Our observation that the splenomegaly was prevented in Rag1−/− Cnab−/− DO11.10 mice suggests that endogenous antigens are driving the defects in T-cell homeostasis in Cnab−/− mice. However, it is still unclear whether lymphomas develop in the absence of either of the lymphocyte subsets in Cnab−/− mice. The data also suggest that Cnab−/− mice that have no B cells do not produce abnormally high levels of proinflammatory cytokines, with the exception of IL-17. This could be an indirect effect of defective FOXP3 expression which is known to inhibit IL-17 induction through its interaction with RORγt . These data suggest that CNAβ signaling is involved in limiting the inflammatory response of T cells through induction of FOXP3.
Since CsA inhibits the expression of Il2 and Foxp3 , and since NFAT binds to promoter/enhancer elements of Tgfb1, Foxp3 and Il2 and induces their expression (IL-2 is required for Foxp3 induction by TGFβ1 in CD4+CD25− T cells), it is possible that the FOXP3+ Treg cells in Cnab−/− mice could also be defective in TGFβ1 production resulting in a functional deficiency of these cells. Understanding the specific roles of CN in Treg-cell generation and function and its role in T-cell homeostasis is important for finding the relationship between T-cell activation and B-cell lymphoma in CNAβ-deficient mice. Since Treg-cell deficiency is known to lead to T-cell activation and increased anti-tumor responses, it is generally believed that inhibition of Treg cells is a potential therapy for most cancers. However, Cnab−/− mice do not reject allogenic tumors . IL-6 and Th2 cells have been found to induce activation and transformation of B cells since chronic activation of B cells is observed in lymphomas and all the tumors observed in Cnab−/− mice are of B-cell origin [41–43]. These observations shed light on the CN role in T-cell regulation which is important for controlling inflammatory responses. Further studies are required for understanding the signaling mechanisms involved in lymphomagenesis under inflammatory conditions.
Since inflammation plays a significant role in tumorigenesis and inflammatory cytokines are elevated in transplant patients who are treated with CsA [31;42], developing therapeutic approaches to block the inflammation in these patients would definitely reduce their chance of developing lymphomas. Thus, the CNAβ-deficient mouse is a unique model to study the immune dysregulation and lymphomagenesis observed in transplant patients.
This study was supported by NIH AI067903 and U01CA084291 to TD, and ACS-IRG7400128, NCI-CCSG (CA 023074) and NCI-Lymphoma SPORE (P50CA3080501A1) Career Development Award from AZCC to RB. The authors thank Dr. Lisa M. Rimsza for reading and her thoughtful comments and suggestions.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.