CD43 is a highly glycosylated transmembrane protein that regulates T cell activation. CD43−/− T cells are hyperproliferative and the cytoplasmic tail of CD43 has been found to be sufficient to reconstitute wild-type proliferation levels, suggesting an intracellular mechanism. In this study, we report that upon TCR ligation CD43−/− T cells demonstrated no increase in tyrosine phosphorylation but a decreased calcium flux. Interestingly, CD43−/− T cells preferentially differentiated into Th2 cells in vitro, and CD43−/− T cells show increased GATA-3 translocation into the nucleus. In vivo, CD43−/− mice exhibited increased inflammation in two separate models of Th2-mediated allergic airway disease. In contrast, in Th1-mediated diabetes, nonobese diabetic CD43−/− mice did not significantly differ from wild-type mice in disease onset or progression. Th1-induced experimental autoimmune encephalomyelitis to MOG35–55 was also normal in the CD43−/− mice. Nonetheless, the CD43−/− mice produced more IL-5 when restimulated with MOG35–55 in vitro and demonstrated decreased delayed-type hypersensitivity responses. Together, these data demonstrate that although CD43−/− T cells preferentially differentiate into Th2 cells, this response is not sufficient to protect against Th1-mediated autoimmune responses.
The CD4+ Th cells differentiate into distinct subsets that play critical roles in the immune response to various antigenic challenges. Th1 cells secrete IFN-γ and IL-2 and are involved in cellular immunity to intracellular pathogens and autoimmunity. Th2 cells secrete IL-4, IL-5, IL-10, and IL-13 and are involved in immunity to extracellular pathogens, allergy, and atopic diseases. The differentiation of naive precursor Th cells into Th1 or Th2 cells is a complex process that is not completely understood, although cytokines are known to play a dominant role, with IL-12 and IL-4 directing Th1 and Th2 differentiation, respectively (reviewed in Ref. 1).
Other factors such as antigenic dose, TCR-MHC/Ag complex affinity, and duration of the TCR signal have also been shown to play a role in differentiation (reviewed in Ref. 2). Furthermore, different calcium (Ca2+) profiles can induce Th cell differentiation and remain characteristic of differentiated Th1 and Th2 cells (3). Th1 cells exhibit a high Ca2+ peak followed by rapid clearance, whereas Th2 cells have a relatively low Ca2+ peak but maintain an increased Ca2+ level. Finally, the nature of the costimulatory signal received by the T cell can influence differentiation. CD28 costimulation promotes Th2 responses, as it has been shown that Ag-stimulated T cells from CD28−/− mice fail to produce IL-4 (4). ICOS has also been suggested to play a role in Th2 differentiation, since ICOS−/− mice do not mount Th2 responses (5). Conversely, it has been shown that LFA-1 costimulation during priming inhibits IL-4 production and hence directs differentiation toward Th1 (6, 7). Thus, the convergence of many factors determines Th cell differentiation.
CD43 (sialophorin, leukosialin) is a large transmembrane glycoprotein extending ∼45 nm from the cell surface. It is highly expressed on many hematopoietic cells and is one of the most abundant molecules on T cells. Due to its large size and negative charge, it was thought that CD43 acted as a barrier to T cell-APC interactions. However, the negative regulatory effect of the CD43 molecule depends only on the intracellular tail, suggesting that CD43 functions to regulate intracellular events in T cell activation (8). Since CD43 localizes away from the T cell-APC interaction site via its interaction with the ezrin-radixin-moesin family of cytoskeletal adaptor proteins (9, 10, 11, 12), we have proposed that CD43 may function to remove intracellular proteins from the immunological synapse (9).
To determine the intracellular mechanisms by which CD43 may regulate the proliferation of T cells, we examined events downstream of T cell activation in CD43−/− T cells. TCR stimulation of protein tyrosine phosphorylation was not affected by CD43 deficiency. Interestingly, a defect in Ca2+ flux was seen in the CD43−/− T cells. Due to previous reports that lower Ca2+ fluxes can lead to preferential Th2 differentiation and that other costimulatory molecules can influence Th differentiation, we examined Th differentiation in the DO.11.10 CD43−/− T cells (DO.CD43−/−). The CD43−/− T cells showed a bias toward Th2 differentiation that could be reversed by reintroduction of CD43. We also show that CD43−/− T cells have an increase in GATA-3 translocation upon stimulation. In vivo, the CD43−/− mice demonstrated significantly increased Th2-mediated inflammatory responses, but were not protected from Th1-mediated autoimmune responses. These results demonstrate that CD43 regulates Th cell differentiation in vitro and in vivo.
Materials and Methods
TCR Tg DO11.10xCD43−/− have been previously described (9). To produce CD43-deficient NOD mice, NOD.CD43−/− mice were backcrossed five times to the NOD/LtJ strain (The Jackson Laboratory). All NOD mice used were female. CD43−/− mice were backcrossed eight times to the C57BL/6NTac (B6) strain to produce the B6.CD43−/− strain and are bred at Taconic Farms specific pathogen-free barrier facilities. Only female mice were used in the experimental autoimmune encephalomyelitis (EAE)4 model. DBA/2NCr were purchased from the Animal Production Area (National Cancer Institute-Frederick Cancer Research and Development Center). All mice housed at the University of Chicago were bred and/or maintained in a specific pathogen-free condition in barrier facilities. The studies reported here conform to the principles outlined by the Animal Welfare Act and the National Institutes of Health guidelines for the care and use of animals in biomedical research.
Abs and reagents
Affinity-purified anti-CD28 (PV-1 (13)), anti-CD3 (145-2C11), and anti-phosphotyrosine (P-Tyr) supernatants (FB2 (14)) are all available from American Type Culture Collection and were prepared in our laboratory. Anti-phosphotyrosine (4G10; Upstate Biotechnology), anti-GATA-3, and anti-lamin A (Santa Cruz Biotechnology); goat anti-hamster IgG (Valeant Pharmaceuticals); anti-mouse HRP (Jackson ImmunoResearch Laboratories); anti-CD3-PE, anti-CD4-FITC, anti-CD44-PE, anti-CD45RB-PE, anti-CD62L-PE (BD Pharmingen); and anti-CCR3-FITC (R&D Systems) Abs were titrated before use. OVA323–339 peptide was produced by the University of Chicago Peptide Synthesis Facility. Myelin oligodendrocyte glycoprotein (MOG) 35–55 (MEVGWYRSPFSRVVHLYRNKG) was purchased from Peptides International. Purities (>97%) were confirmed by mass spectroscopy.
Primary DO.11.10 and DO.CD43−/− T cells were isolated and activated as described previously (15). For proliferation by CFSE dilution, cells were preincubated with 5 μM CFSE for 5 min as previously described. Conversely, cells were pulsed with 1 μCi/well [3H]thymidine for 8–12 h before harvest as described previously (16).
Nylon nonadherent lymph node T cells (10 × 106) were incubated for 15 min at 4°C with 5 μg/ml 2C11 and cross-linked at 37°C with goat anti-hamster Ab for various times. Cells were lysed (0.5% Triton X-100, 50 mM Tris (pH 7.6), 150 mM NaCl, 5 mM EDTA, 1 mM Na3VO4, 10 μg/ml leupeptin, 10 μg/ml aprotinin, and 1 mM PMSF), precipitated with FB2-coated protein A beads (Zymed) and samples were resolved on SDS-PAGE gels, transferred to Immobilon-P polyvinylidene difluoride membranes, and immunoblotted with 4G10. Signal was detected with SuperSignal (Pierce) on Kodak Biomax-MR film. Nuclear preparations were performed by lysing cells in 10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA (pH 8), 0.1 mM EGTA (pH 8), and 0.4% Nonidet P-40 on ice for 15 min. The nuclear pellet was extracted with 20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, and 1 mM EGTA, vortexed for 2 min at 4°C, and clarified by centrifugation to generate the nuclear fraction. Samples were separated by SDS-PAGE and proteins were detected using standard Western blotting protocols.
Ca2+ flux assay
Ex vivo lymph node cells (5 × 105/sample) were loaded with 10 μM Fura Red and 5 μM Fluo 3 (Molecular Probes). Directly before flow cytometry, cells were incubated with 5 μg/ml 2C11 for 10 min at room temperature. Cells were acquired for 1 min before addition of goat anti-hamster Ab. Cells were replaced on the LSR (BD Biosciences) and acquisition continued. Data were analyzed using FlowJo software (Tree Star).
Differentiation cell cultures
T cells (>90% CD3+) were stimulated for 7 days with irradiated DBA splenocytes and 1.0 or 0.1 μg/ml OVA peptide. For 48-h cytokine analyses, T cells were stimulated, harvested, centrifuged through Ficoll, and plated at 2.5 × 105/well in 48-well plates previously coated with 1 μg/ml 2C11 and incubated at 37°C for 48 h. Supernatants were collected and assayed by ELISA according to the manufacturer’s protocol (BD Pharmingen).
Construction of retroviral vectors
Full-length murine CD43 was fused with enhanced GFP (eGFP; CD43FLeGFP) at the 3′ end of CD43, cloned into pcDNA3 (BD Clontech) and transferred into the pBMN retroviral vector (a gift from Dr. E. Palmer, Basel Institute for Immunology, Basel, Switzerland). The construct was confirmed by sequence analysis at the University of Chicago Sequencing Facility.
Transduction of retroviral constructs into primary T cells
Cells were transduced as previously published with full-length murine CD43 that was fused with eGFP (CD43FLeGFP) (9). At 7–9 days, the transduced T cells were sorted and stimulated with OVA peptide for 48 h. Supernatants were collected and assayed for IFN-γ, IL-5, and IL-13 by ELISA (IFN-γ and IL-5; BD Pharmingen and IL-13; R&D Systems) according to the manufacturers’ instructions.
Ag sensitization and challenge
Mice were immunized i.p. with 25 μg of chicken egg OVA in 1 mg ppt alum per mouse. Mice were challenged with 6% nebulized OVA on days 7–10. Animals were sacrificed 4 days later and bronchoalveolar lavage (BAL) and histology were collected.
BAL, lung extracts, and histology
BAL, lung extracts, and histology were performed as previously described (19). Peribronchial and perivascular inflammation were scored on H&E sections, using a previously published scoring scheme (19), by a veterinary pathologist who was blinded to the experimental conditions. The score of peribronchiolar and perivascular inflammation was determined as follows: 0, normal; 1, few cells; 2, a ring of inflammatory cells one cell layer deep; 3, a ring of inflammatory cells two to four cells deep; and 4, a ring of inflammatory cells of more than four cells deep. Scoring was then performed by examining at least 20 consecutive fields. Mucus-containing cells were stained with periodic acid-Schiff and scored for the abundance of periodic acid-Schiff-positive goblet cells in each airway.
Ig serum levels
IgE and IgG2a (BD Pharmingen) serum levels were determined by ELISA according to the manufacturer’s protocol.
NOD.CD43+/− and NOD.CD43−/− littermates were screened for glycemia weekly beginning at 9 wk of age using OneTouch test strips for blood glucose levels (LifeScan). Mice were considered diabetic when glucose readings were >13.9 mmol/L (250 mg/dl) and were sacrificed when above this level for 3 consecutive wk.
EAE/delayed-type hypersensitivity (DTH) models
Each mouse received 100 μl of a CFA emulsion containing 200 μg of Mycobacterium tuberculosis H37Ra (Difco) and 200 μg of MOG35–55 distributed s.c. over three spots on the flank. Mice were evaluated for clinical symptoms on the following scale: 0, no abnormality; 1, limp tail or hind limb weakness; 2, limp tail and hind limb weakness; 3, partial hind limb paralysis; 4, complete hind limb paralysis; and 5, moribund.
Draining lymph nodes (axillary, inguinal, and popliteal) and spleen cells from mice primed with MOG35–55/CFA were removed on day 10. A total of 5 × 106 cells/well was plated onto 24-well flat-bottom plates in supplemented HL-1 and 0 μM or 25 μM MOG35–55. Supernatants were harvested at 24 or 48 h and analyzed for IFN-γ and IL-5 by capture ELISA. ELISA reagents were purchased from Endogen and were used according to the manufacturer’s protocol.
DTH responses to MOG35–55 were quantified using a 24-h ear-swelling assay. Prechallenge ear thickness was determined using a Mitutoyo model 7326 engineer’s micrometer (Schlessinger’s Tools). DTH responses were elicited by injecting 10 μg of peptide (in 10 μl of saline) into the dorsal surface of the ear using a 100-μl Hamilton syringe fitted with a 30-gauge needle. Twenty-four hours after ear challenge, the increase in ear thickness over prechallenge measurements was determined. Results are expressed in units of 10−4 inches ± SEM.
All statistics were done using the unpaired Student two-tailed t test. Error bars represent SEM.
CD43−/− T cells are hyperproliferative, yet have a decreased calcium flux
CD43−/− T cells are hyperproliferative in response to Con A, staphylococcal enterotoxin B, PMA/ionomycin, and anti-CD3 stimulation (20, 21). Using Ag-specific stimulus, we also found that DO.CD43−/− T cells exhibit significantly increased proliferation compared with wild-type (WT) T cells by both [3H]thymidine incorporation (Fig. 1,A) and CFSE intensity (Fig. 1,B). We obtained similar results in CD43−/− mice from the C57BL/6 background (data not shown). The increased proliferation of CD43−/− T cells led us to investigate whether CD43 deficiency affects early TCR-mediated events. Enriched ex vivo T cells were stimulated by cross-linking the TCR complex and immunoblotted with anti-P-Tyr. We observed no differences in gross P-Tyr levels between CD43+/− and CD43−/− T cells at 1, 5, or 10 min after cross-linking (Fig. 1,C). We have also observed no differences in the phosphorylation levels of specific downstream signaling molecules including Lck, CD3ζ, and LAT (data not shown). These data suggest that augmented phosphorylation at the level of TCR signal transduction proteins does not account for CD43−/− T cell hyperproliferation. To test whether downstream second messengers were affected, Ca2+ flux was tested. Unexpectedly, we found that CD43−/− T cells have a defect in Ca2+ flux in response to TCR cross-linking (Fig. 1 D). CD43−/− T cells have a delayed onset, and the amplitude is not as high as in WT T cells. However, the Ca2+ flux in CD43−/− and WT T cells plateau at similar levels. Thus, despite the hyperproliferative nature of CD43−/− T cells, they demonstrate a diminished Ca2+ flux upon TCR cross-linking.
CD43−/− T cells preferentially differentiate toward Th2 cells
TCR-induced Ca2+ flux levels have been tied to CD4+ T cell differentiation into Th1 and Th2 cells (3). Given the diminished Ca2+ flux of CD43−/− T cells and the putative role of CD43 as a costimulatory molecule, we hypothesized that the differentiation of CD43−/− T cells may be altered. To test this hypothesis, we compared the differentiation of DO.11.10 and DO.CD43−/− T cells in culture. Lymph node T cells were stimulated with Ag and APCs for 1 wk and restimulated with plate-bound anti-CD3 to induce cytokine production. CD43−/− T cells produced more IL-4, IL-5, and IL-13 than WT T cells (Fig. 2,A). However, IFN-γ production was not significantly altered, although in some experiments CD43−/− T cells tended to produce slightly less than WT T cells (Fig. 2 A and our unpublished observations). These results demonstrate that CD43−/− T cells produce greater amounts of Th2 cytokines upon restimulation, suggesting that they preferentially differentiate into Th2 cells.
Reintroduction of CD43 reverses preferential Th2 differentiation of CD43−/− T cells
To address the possibility that the preferential differentiation of CD43−/− T cells into Th2 cells was due to a potential role of CD43 in affecting T cell development, full-length CD43 fused to eGFP (CD43FLeGFP) was introduced into DO.CD43−/− T cells by retroviral transduction. CD43FLeGFP or control eGFP-transduced T cells were purified and restimulated with irradiated APCs and OVA peptide and assayed for cytokine production. CD43FLeGFP T cells produced equal or increased amounts of IFN-γ and decreased amounts of IL-4, IL-5, and IL-13 as compared with control-transduced T cells 48 h after Ag-specific stimulation (Fig. 2,B). Although it is common to see variation in the total cytokine production from experiment-to-experiment, these results are consistent with the relative differences in IFN-γ, IL-5, IL-4, and IL-13 production seen in WT and CD43−/− T cells (compare Fig. 2, A and B). Furthermore, this skewing of the CD43−/− T cell response also occurred when the cells were cultured in exactly the same manner as the transduction experiments (plate-bound anti-CD3 and anti-CD28 followed by Ag and APC restimulation at day 7; data not shown). We have also previously demonstrated that the hyperproliferation seen in CD43-deficient T cells can be reversed by re-expression of CD43 (15). Thus, Th2 differentiation of CD43−/− T cells is likely not a developmental effect and instead demonstrates that CD43 regulates differentiation of mature, peripheral T cells.
Preferential Th2 differentiation of CD43−/− T cells is not due to early IL-4 production by memory cells
One possible mechanism for the augmented Th2 differentiation of CD43−/− T cells is that the CD43−/− mice have greater numbers of memory cells that produce increased levels of IL-4 early in the T cell response. However, this mechanism is unlikely because no difference was found in the percentage of CD4+ memory T cells between WT and CD43−/− mice, as measured by CD45RBlowCD44highCD62L− cells (Fig. 3,A). Despite the equal numbers of memory T cells, it is possible that the CD43−/− memory T cells make more IL-4 than WT memory T cells. To investigate this possibility, CD62L-negative WT and CD43−/− T cells were stimulated with plate-bound anti-CD3 and the supernatants were analyzed at 48 h. The CD43−/− memory T cells did not produce greater levels of IL-4 and perhaps made slightly less than WT cells (Fig. 3,B). No significant difference in IFN-γ levels between the strains was found (our unpublished observations). Furthermore, total ex vivo T cells cultured with APCs and Ag were tested for early IL-4 production at 48 h and no difference was found (Fig. 3 C). Taken together, these data suggest that CD43−/− Th2 differentiation does not appear to be induced by early cytokine production.
CD43−/− T cells show increased levels of GATA-3
Another possible mechanism for the augmented Th2 differentiation of CD43−/− T cells is that IL-4 signaling may produce increased GATA-3 translocation into the nucleus in the absence of CD43. GATA-3 is a key regulator of both IL-4-dependent and IL-4-independent pathways to Th2 differentiation (22, 23), and changes in GATA-3 nuclear localization affects Th2 differentiation (24). DO.11.10 T cells were stimulated for 3 days with OVA peptide presented by irradiated splenocytes in the presence or absence of exogenous IL-4. We found that CD43−/− T cells show an increase in GATA-3 levels in the nucleus compared with WT T cells activated with peptide alone (Fig. 4, compare lanes 1 and 3). Upon addition of IL-4, GATA-3 levels increase dramatically in both WT and CD43−/− T cells (Fig. 4, lanes 1–4). Interestingly, CD43−/− T cells maintain higher levels of GATA-3 both in the absence and presence of IL-4 (Fig. 4, compare lanes 2 and 4). These results suggest that CD43 negatively regulates GATA-3 translocation perhaps by affecting IL-4R signaling directly.
CD43−/− mice exhibit increased baseline serum IgE
To determine whether a Th2 bias was present in vivo as well as in vitro, CD43−/− and CD43+/− littermates were tested for Th1 (IgG2a) and Th2 (IgE) serum Igs. BALB.CD43−/− mice housed in a conventional (nonspecific pathogen-free) facility with known endemic pathogens were found to have significantly higher baseline serum IgE levels than their heterozygous littermate controls. However, equal levels of serum IgG2a were found in both groups (Fig. 5). Thus, similar environmental challenges led to differential Ig isotype levels based on the CD43 genotype, suggesting that CD43−/− T cells may bias to a Th2 response in vivo.
CD43−/− mice have increased airway inflammation in two models of allergic airway disease
To further test whether CD43−/− mice have increased Th2 responses in vivo, we investigated whether CD43 deficiency had any effect on Th2-mediated allergic airway disease (AAD). BALB/c and BALB.CD43−/− mice were sensitized with an i.p. injection of OVA and challenged with 6% nebulized OVA for 3 consecutive days, starting 7 days after sensitization. Four days later, the animals were sacrificed and BAL was performed to determine the cellular content of the airways. In untreated animals, the primary immune cell type found in the lungs is macrophages with virtually no eosinophils, whereas treated animals develop airway inflammation characterized by eosinophilia (25). As seen in Fig. 6,A, CD43−/− mice develop significantly increased eosinophilia compared with WT mice. Furthermore, eosinophils made up a larger percent (62.78 ± 2.94 vs 49.57 ± 4.56, p < 0.02) of the total cells found in the BAL of CD43−/− mice. Eosinophilic airway inflammation has been shown to correspond to leukocyte inflammation around bronchioles and blood vessels in the lungs (25), and we observed that CD43−/− mice showed higher levels of both perivascular and peribronchiolar inflammation (Fig. 6,B). We also observed an increase in mucus metaplasia in CD43−/− mice (our unpublished observations). To determine whether the predisposition of the BALB/c background to a Th2-type response (19) was biasing our results, we used CD43−/− mice that were backbred to the Th1 predisposed strain B6 in a different Th2 model of AAD. The animals were sensitized with S. mansoni eggs and challenged intratracheally on day 7 with schistosome egg Ag (SEA). As found in the OVA model, B6.CD43−/− mice had significantly increased eosinophilia in their airways (Fig. 6,C). This finding is unlikely to be due to increased migration of Th2 CD43−/− cells, as we found no difference in the numbers of T cells in the BAL (Fig. 6,C). Examination of the local cytokine levels in the lung parenchyma of these mice demonstrated a significant increase in IL-5 but not IFN-γ in the CD43−/− mice (Fig. 6,D). Furthermore, the increased serum IgE levels are indicative of an increased systemic Th2 response (Fig. 6,E). CD43−/− mice had significantly increased levels of inflammation compared with that seen in WT mice, similar to the OVA model quantitated in Fig. 6,B (compare Fig. 6, F and G). These results demonstrate that CD43−/− mice have increased Th2-mediated airway inflammation compared with WT mice. This increase in AAD together with finding that CD43−/− mice have increased baseline IgE levels suggests an inherent bias toward a Th2 phenotype.
CD43−/− mice are not protected from Th1-mediated autoimmune responses
Although it is clear that the CD43−/− mice have increased Th2 responses, the effect of CD43 deficiency on Th1 responses was likewise investigated. Two possible outcomes were hypothesized. First, CD43−/− mice may have a generalized hyperresponsiveness leading to increased Th1 as well as Th2 responses. Second, the bias to Th2 in CD43−/− mice may inhibit Th1 responses such that CD43−/− mice would have delayed or decreased Th1-mediated responses. Indeed, it has been reported that Th1-mediated autoimmune diseases can be protected against by inducing Th2 differentiation (26). To address these hypotheses, two models of Th1-mediated autoimmune diseases were used. First, the role of CD43 deficiency in the onset and progression of insulin-dependent diabetes mellitus in NOD mice was tested. Blood glucose levels of female NOD.CD43+/− and NOD.CD43−/− littermates were tested weekly from 9 to 28 wk of age. Mice were considered diabetic when blood glucose levels were above 250 mg/dl. As seen in Fig. 7 A, there was no difference in disease onset or progression over time. Because these mice were only backcrossed five times to the NOD strain, only 60% of the mice became diabetic compared with the usual 80% penetrance found in normal NOD females. However, the animals comprising the NOD.CD43+/− and NOD.CD43−/− groups were littermates, and any genetic variability that might have affected onset of diabetes should be distributed randomly. These data demonstrate that there is no difference in the onset and progression of nonobese diabetes in NOD.CD43−/− and NOD.CD43+/− mice.
In an EAE model, B6 and B6.CD43−/− mice were injected s.c. with 200 μg of M. tuberculosis and 200 μg of MOG35–55 in CFA. The mice were evaluated for clinical symptoms of EAE and scored on a scale ranging from 0 to 5. There were no differences in either disease onset or severity in B6 and B6.CD43−/− mice (Fig. 7,B). At approximately day 10, both sets of animals developed EAE and disease progression was similar over the course of evaluation. This demonstrates that in an EAE model of a Th1-mediated immune response, CD43 deficiency neither exacerbates nor inhibits disease onset and severity. To examine the MOG35–55-specific response, lymph node and spleen cells were removed from mice primed with MOG35–55/CFA on day 10 and cultured with 25 μM MOG35–55 for 48 h, then analyzed for cytokine production. No cytokines were detected in unstimulated samples (our unpublished observations). CD43−/− T cells made similar, or slightly decreased, amounts of IFN-γ (Fig. 7,D and our unpublished observations); however, significantly more IL-5 was consistently produced (Fig. 7,E). To further determine the ability of CD43−/− mice to mount a Th1 response, DTH responses to MOG35–55 were determined by injecting 10 μg of peptide into the dorsal surface of the ear and measuring the increase in ear thickness 24 h after challenge. Although CD43−/− mice were able to mount a DTH response, it was significantly decreased compared with B6 mice (Fig. 7 C). Taken together, these data demonstrate that CD43−/− mice are capable of mounting Th1-mediated autoimmune responses that are similar to those of WT mice. However, closer examination suggests that even within the context of this strong Th1 response, CD43−/− mice produced more Th2 cytokines and had decreased DTH responses. Thus, although CD43 deficiency is able to direct increased Th2 differentiation, this response was not sufficient to protect against Th1-mediated autoimmunity in EAE or diabetes.
Our studies demonstrate that CD43−/− T cells preferentially differentiate into Th2 cells (Fig. 2). We observed no abnormality in the percentage of memory T cells from CD43−/− mice and, when stimulated, these memory-type T cells did not produce greater levels of IL-4 than WT memory T cells (Fig. 3). Thus, naive CD43−/− T cells are not induced to differentiate into Th2 cells by memory T cells producing increased levels of IL-4 in their microenvironment. These data are consistent with an earlier report that showed no difference in thymic and peripheral T cell populations, as measured by cell surface markers, between CD43−/− and WT mice (20). Although the absence of CD43 has not been reported to affect T cell development (27), we addressed this directly by transducing CD43 into DO.CD43−/− T cells. The CD43-transduced cells were able to reconstitute the WT phenotype (Fig. 2 B), demonstrating that the preferential Th2 differentiation of CD43−/− T cells was not the result of a developmental change in CD43−/− mice. Additionally, our data address the possibility that other unknown genes from the 129 background are influencing the differentiation of the CD43−/− T cells, thereby isolating CD43 as the molecule involved.
Our data point to an intracellular mechanism by which CD43 deficiency leads to Th2 differentiation. The defect in Ca2+ flux in CD43−/− T cells (Fig. 1 D) could be such a mechanism. Lower Ca2+ fluxes have been reported to be characteristic of Th2 differentiation, whereas higher Ca2+ fluxes are characteristic of Th1 differentiation (3). Furthermore, manipulating T cell stimulation to mimic higher or lower levels of Ca2+ has been shown to induce naive CD4+ T cell differentiation toward Th1 or Th2 cells, respectively (28). Consistent with this Ca2+ effect are data suggesting that the translocation of different NFAT family members is involved in Th cell differentiation through their respective actions on the IL-4 promoter (29, 30, 31, 32, 33). Together, these reports support a model in which the decrease in Ca2+ flux in CD43−/− T cells leads to Th2 differentiation, although additional experiments are required to establish a direct link between the lower Ca2+ flux and Th2 differentiation in CD43−/− T cells. Furthermore, we cannot rule out the possibility that a developmental change has lead to the decreased Ca2+ flux in CD43−/− T cells since reconstitution experiments are not technically feasible.
The hyperproliferation of the CD43−/− T cells may be another factor in their preferential Th2 differentiation. Although from the first division after activation naive cells are competent to produce IFN-γ, the IL-4 locus requires at least four cell divisions before becoming accessible for transcription (34). Thus, the hyperproliferation of naive CD43−/− T cells could lead to more rapid accessibility of the IL-4 locus resulting in the presence of more IL-4 in the CD43−/− microenvironment. Although we did not find a difference in cytokine production at 48 h of culture, we cannot rule out this possibility because at this time point neither CD43−/− nor WT T cells have reached four divisions (our unpublished observations).
It has been reported that hyperresponsiveness in the absence of CD43 is an artifact of insufficient backcrossing (27). However, we and others consistently observe hyperproliferation of T cells from the same line of backcrossed B6.CD43−/− mice (21). Since reintroduction of CD43 into CD43−/− T cells leads to a reconstitution of WT proliferation levels (8, 15), genetic background cannot explain the hyperproliferative response. Moreover, here we demonstrate that CD43-transduced CD43−/− T cells regain normal differentiation patterns (Fig. 2 B). These data clearly demonstrate that the effect of CD43 deficiency on differentiation is due to CD43 itself and not due to differences in genetic backgrounds.
Consistent with the data from in vitro experiments, CD43 deficiency also leads to increased Th2 responses in vivo. Increased Th2-mediated allergic airway disease (Fig. 6) was found in the CD43−/− mice, whereas in two models of Th1-mediated autoimmune diseases, NOD and EAE, there were no differences in onset, progression, or severity of disease (Fig. 7, A and B). Although we cannot rule out the participation of non-T cells in increased Th2 responses in vivo, it is likely that Th2 bias in CD43-deficient T cells plays an important role. It has been suggested that CD43 plays a role in homing of T cells to sites of inflammation (35, 36, 37, 38, 39). Although one study found increased homing of CD43−/− T cells to secondary lymphoid organs (37), other studies have found a defect (38) or a delay (39) and yet others have shown no effect on homing (40). Additionally, Ab blocking of CD43 has been found to decrease homing to the islets in NOD mice (35, 36). Contrary to these reports, we have shown that NOD.CD43−/− mice have similar disease onset and severity (Fig. 7,A) and levels of lymphocytic infiltration in the islets (our unpublished observations) compared with WT. Moreover, despite studies by others that intracranial viral infection leads to delayed CD43−/− T cell migration (39), we failed to find any migration defect of lymphocytes to the CNS of MOG35–55-immunized CD43−/−mice (C. E. Smith and S. D. Miller, unpublished observations). It is possible that this difference is due to the immunization protocol or infectious agent. Although inflammation is significantly increased in the lungs of allergen-induced mice (Fig. 6, B, F, and G), our data that IgE levels are increased in environmentally challenged mice demonstrate a systemic Th2 bias. These data imply that the respective levels of inflammation seen in the in vivo models are due to changes in Th cell differentiation and not due to differences in homing.
We used the EAE model to further explore the in vivo consequences of preferential Th2 differentiation in CD43−/− mice. Given their Th2 bias, it was possible that the CD43−/− mice would be protected against EAE. Consistent with our in vitro findings, CD43−/− T cells are capable of differentiating into Th1 cells (as seen by the equal IFN-γ production, Fig. 7,D), and CD43−/− mice develop EAE similar to WT mice (Fig. 7,B). The development of EAE in CD43−/− mice occurred despite their predisposition to Th2 differentiation, which is evident by their increased production of the Th2 cytokine IL-5 in response to MOG35–55 immunization (Fig. 7). In contrast to a previous report (41), we found that the Th2 bias was not sufficient to inhibit EAE disease course. However, the Th2 bias did play a role in the significant and consistent decrease in DTH responses in MOG-immunized CD43−/− mice.
It remains to be determined how the presence or absence of CD43 on T cells is regulating Th cell differentiation. Studies in human T cells have shown that CD43 costimulation in the presence and absence of TCR stimulation can induce IFN-γ production and promote Th1 differentiation (42, 43, 44, 45). However, these findings have not been replicated in murine systems, making it unclear whether this plays a role in our model. Interestingly, we showed that the absence of CD43 leads to higher levels of GATA-3 in the nuclear fractions of activated T cells (Fig. 4). How CD43 affects GATA-3 translocation remains unknown. However because CD43 plays a role in regulating GATA-3 even in the presence of exogenously added IL-4, the effect of CD43 on Th2 differentiation may be due to modulation of the responsiveness of IL-4R on T cells. Recent studies (9, 10, 11, 12) have demonstrated that CD43 is actively removed from the immunological synapse through interactions with the cytoskeletal linker proteins ezrin and moesin. These ezrin-radixin-moesin family members define a novel complex of proteins at the distal pole consisting of many “cargo” proteins including CD43, RhoGDI, and SAP-97 (Ref. 9 and our unpublished observations). This process has been postulated to be involved in assembling a polarized intracellular architecture conducive to efficient T cell responses. It is interesting to speculate that recruitment of CD43 to this complex may affect T cell activation and differentiation. Although this may be a direct effect of CD43, it is also possible that other proteins require association with CD43 for their localization away from the T cell-APC interaction site. Release of these proteins from CD43 may deregulate their function and affect T cell differentiation. Several proteins have been identified to interact with the cytoplasmic tail of CD43. For instance, Green and colleagues (46) have identified a serine/threonine kinase, HPK1, which binds to the cytoplasmic tail of CD43. Furthermore, cross-linking CD43 has also been shown to lead to association of Fyn with the cytoplasmic tail of CD43 in human T cells (47). Finally, we have observed that a fusion protein of GST linked to the cytoplasmic tail of CD43 can coprecipitate several other unknown proteins (J. L. Cannon and A. I. Sperling, unpublished observations). Thus, a model can be envisioned in which CD43-binding proteins would not be removed to the distal pole complex in CD43−/− T cells and, consequently, could remain at the immunological synapse where they might affect Ca2+ flux and other signaling pathways.
We thank Brian Gray, John Sedy, Donna Decker, Eric Allenspach, Aric Frantz, Allen Nelson, Chan Park, and Ryan Duggan for their contributions to these studies, Drs. David Eliot and Joel Weinstock for providing S. mansoni eggs and Ag, and Drs. Maria-Luisa Alegre, Yang-Xin Fu, and Janis K. Burkhardt for critical review and helpful discussion of these studies.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grants R01 AI44932 (to A.I.S.), and RO1 NS/AI-026543 (to S.D.M.). J.L.C. is supported by a fellowship from the Irvington Institute Fellowship Program of the Cancer Research Institute.
Abbreviations used in this paper: EAE, experimental autoimmune encephalomyelitis; P-Tyr, phosphotyrosine; MOG, myelin oligodendrocyte glycoprotein; eGFP, enhanced GFP; BAL, bronchoalveolar lavage; DTH, delayed-type hypersensitivity; WT, wild type; AAD, allergic airway disease; SEA, schistosome egg Ag.