IL-23 plays a critical role establishing inflammatory immunity and enhancing IL-17 production in vivo. However, an understanding of how it performs those functions has been elusive. In this report, using an IL-17-capture technique, we demonstrate that IL-23 maintains the IL-17-secreting phenotype of purified IL-17+ cells without affecting cell expansion or survival. IL-23 maintains the Th17 phenotype over multiple rounds of in vitro stimulation most efficiently in conjunction with IL-1β. However, in contrast to Th1 and Th2 cells, the Th17 phenotype is not stable and when long-term IL-23-stimulated Th17 cultures are exposed to Th1- or Th2-inducing cytokines, the Th17 genetic program is repressed and cells that previously secreted IL-17 assume the cytokine secreting profile of other Th subsets. Thus, while IL-23 can maintain the Th17 phenotype, it does not promote commitment to an IL-17-secreting lineage.
The cytokine environment generated in an immune response to self or foreign Ags directs the development of discrete Th cell subsets that mediate specific types of inflammation as a mechanism for combating pathogens, with Th1 cells mediating immunity to intracellular bacteria and parasites, Th2 cells mediating immunity to extracellular parasites, and Th17 cells involved in resistance to extracellular bacteria and fungal infections (1, 2). Cytokines promote the development of effector subsets by activating STAT proteins that, in concert with lineage specific transcription factors, initiate a subset-specific effector program. Thus, IL-12-activated Stat4, and IFN-γ-stimulated induction of T-bet, promote Th1 differentiation, and IL-4 activated Stat6, which induces GATA-3 expression, promotes Th2 development (3). Although rapid progress has been made in understanding Th17 biology, and there are many parallels with Th1 and Th2 cells, many aspects of Th17 development and function are still unclear. TGF-β and IL-6, through the induction of IL-21, promote the development the IL-17-secreting phenotype (1, 2) by activating Stat3 and inducing Th17 lineage defining transcription factors including RORγt and RORα (4, 5). Other cytokines have been shown to contribute to the Th17 phenotype, though their precise function is not yet clear.
IL-23 is required for the in vivo function of Th17 cells. Although some Th17 cells can develop in the absence of IL-23, mice deficient in IL-23p19 have a greatly diminished ability to mediate inflammation (1, 2). The initial descriptions of Th17 cells were based on cultures stimulated with IL-23, although only a small percentage of cells within these cultures were IL-17-positive (6, 7). This resulted from the inability of IL-23 to prime naive cells to become Th17 cells, largely due to a lack of IL-23R on naive cells (8). In short-term cultures, IL-23 maintains a population of IL-17-secreting T cells, which has been interpreted as a function of promoting Th17 cell expansion or survival (9). However, this has not been extensively tested and in some early experiments IL-23 function was compared with control cultures incubated with IL-2, a cytokine now known to inhibit the Th17 phenotype (10). IL-23 induces the expression of IL-22 (11, 12), and most recently, IL-23 has also been shown to maintain a pathogenic Th17 phenotype compared with cells cultured in TGF-β and IL-6 due to an inability of IL-23 to induce IL-10 production (13). However, the long-term effects of IL-23 on Th17 cells, and whether IL-23 mediates commitment to the Th17 phenotype, as defined by the ability of these cells to maintain the Th17 phenotype in the presence of a Th1 or Th2 promoting environment, remain unclear. In this study, we define the effects of IL-23 on IL-17-secreting T cells through multiple rounds of Ag receptor stimulation. Using cytokine-capture assays, we demonstrate that IL-23 maintains the Th17 phenotype without affecting proliferation or survival. Results suggest that while IL-23 is a maintenance factor for the Th17 phenotype, it does not mediate commitment to this phenotype.
Materials and Methods
Indiana University Institutional Animal Care and Use Committee approved all animal studies in this report. C57BL/6 or BALB/c mice were from Harlan Sprague Dawley.
Th cell differentiation
Naive CD4+CD62L+ cells isolated from spleens using a MACS isolation system (Miltenyi Biotec) were activated with soluble anti-CD3 (4 μg/ml 145–2C11) and soluble anti-CD28 (1 μg/ml; BD Pharmingen) in the presence of CD4+ depleted irradiated splenocytes (1:5). These conditions were found to be optimal over plate-bound anti-CD3 for Th17 generation, particularly in long-term cultures. Th1 and Th2 primed cultures were activated with plate bound anti-CD3 and CD4+6 cells/ml) and stimulated with plate-bound anti-CD3 (0.5 μg/ml) in the presence of CD4+ depleted irradiated splenocytes (1:5) and cultured with the original concentration of cytokines and Abs. For long-term Th17 cultures, cells were replated (0.5 × 106 cells/ml) and stimulated with soluble anti-CD3 (1 μg/ml) in the presence of CD4+
Th cell analysis
On day 5 of each round of stimulation, 1 × 10614). For intracellular cytokine staining, T cells were stimulated for 4 h with PMA (50 ng/ml, Sigma-Aldrich) and ionomycin (500 ng/ml, Sigma-Aldrich) in the presence of GolgiStop (BD Pharmingen) at the recommended concentration. Cells were fixed with 2% paraformaldehyde for 15 min and permeabilized in PBS containing 0.1% Saponin. Cells were stained with PE anti-IL-17 and allophycocyanin anti-IFN-γ (BD Pharmingen) before analysis. T cells were stained with 10 μM CFSE (Molecular Probes) for 10 min at 37°C before culture as indicated.
Th17 cells were activated with plate-bound anti-CD3 (4 μg/ml). After 4 h, Th17 cells were labeled with 75 μl of previously crosslinked (Controlled Protein-Protein Crosslinking Kit; Pierce) anti-CD45 (clone 30-F11; BD Pharmingen) and anti-IL17 (clone Tc11–18H10; BD Pharmingen) Abs (0.2 mg/ml) for 5 min on ice. Labeled cells were then diluted in prewarmed complete RPMI 1640 to a concentration of 105 cells/ml and rotated for 1 h at 37°C. After capture, the Th17 cells were stained with 100 μl of biotin-labeled anti-IL-17 (0.2 mg/ml; clone Tc11–8H4.1; BD Pharmingen) for 15 min before washing and incubating 10 min with Streptavidin-PE (BD Pharmingen). IL-17 captured cells were sorted using a FACS Aria cell sorter (BD Biosciences).
Phospho-STAT protein analysis
T cells were stimulated for 30 min with IL-23 (10 ng/ml) or IL-4 (10 ng/ml). Following stimulation cells were fixed and then permeabilized by methanol. Cells were then stained for phospho-Stat3 or phospho-Stat6 (BD Pharmingen) before analysis. For immunoblot analysis, cells differentiated for three rounds of stimulation were restimulated with IL-12 (5 ng/ml) for 1 h. Whole-cell protein lysates were immunoblotted with phospho-Stat4 (Zymed), total Stat4 (C-20; Santa Cruz Biotechnology) and actin (Calbiochem) as a control.
IL-23 maintains IL-17 secretion without affecting Th17 cell proliferation or expansion
Although the requirement for IL-23 in the function of Th17 cells in vivo is established, the precise role of this cytokine in affecting the Th17 phenotype is unclear. Among other functions, IL-23 was proposed to act as a Th17 cell proliferation or survival factor. To directly test these functions, we developed a cytokine capture assay for IL-17-secreting cells to compare IL-23 functions in enriched IL-17-high and low secretor populations. Naive CD4 T cells were cultured with TGF-β1+IL-6+IL-1β for 5 days before stimulation with anti-CD3 before selection of IL-17-high and low cells by cell sorting (Fig. 1⇓A). Following sorting there was a 10–12-fold enrichment for IL-17 secreting cells in the IL-17-high population (Fig. 1⇓A). The separation of cells into distinct populations was confirmed by demonstrating segregated expression of IL-17, IL-22, Il23r, and Rorc, while IFN-γ production was indistinguishable between the two populations (Fig. 1⇓, B and C). Intracellular staining for IFN-γ in these populations demonstrated less than 0.5% in any population (data not shown).
One of the proposed functions of IL-23 is promoting the proliferation or expansion of Th17 cells. To test this directly, naive CD4 T cells were cultured with TGF-β1+IL-6+IL-1β for 5 days, separated based on IL-17 production and labeled with CFSE (Fig. 2⇓A). CFSE-labeled IL-17-high and low populations were cultured in blocking Abs (anti-IFN-γ and anti-IL-4) alone or blocking Abs plus IL-23 for 24 or 48 h to assess proliferation. Although IL-17-high cells had an intrinsically higher rate of proliferation than IL-17-low cells, there was not a significant difference in proliferation between cultures incubated with or without IL-23 (Fig. 2⇓, B and C), suggesting that IL-23 does not promote a robust proliferative response.
IL-23 was also proposed to affect Th17 survival. However, in examining the overall cell growth in IL-17-high and low cells, we observed the increased proliferative capacity of IL-17-high cells that resulted in a 2–3-fold increase in cell number compared with IL-17-low cells, but only minor effects of IL-23 (Fig. 2⇑D). Similarly, IL-17-high and low cells cultured in the presence or absence of IL-23 had similar percentages of Annexin V+ cells after 2 or 4 days of culture with little effect of IL-23 culture (Fig. 2⇑E).
In contrast, we did note that in IL-17-high populations cultured with IL-23, a higher percentage of cells, with a higher intensity of IL-17 staining, was maintained compared with cells cultured in the absence of IL-23 (Fig. 2⇑B). The effects of IL-23 on maintaining the IL-17-secreting phenotype were even more dramatic when cells were activated with anti-CD3. After 2 days of activation, IL-17-high cells cultured in IL-23 still had >60% IL-17+ cells, while cultures incubated in the absence of IL-23 has <30% IL-17+ cells (Fig. 2⇑F). This effect was also observed in the IL-17-low cultures where IL-17+ cells comprised <10% of the population when cultured in the absence of IL-23 (Fig. 2⇑F). These data suggest that IL-23 maintains the IL-17-secreting phenotype without detectable effects on Th17 cell proliferation or expansion.
IL-23 maintains the Th17 phenotype in long-term cultures
To test the ability of IL-23 to maintain the Th17 phenotype over long term culture, we cultured naive CD4 T cells for 1 wk with TGF-β plus IL-6 plus IL-1β, the latter which we found amplifies IL-23 responsiveness in vitro (Ref. 9 and data not shown), and then for subsequent rounds of stimulation with either blocking Abs alone or IL-23 with blocking Abs. Despite a high level of IL-17 in the initial cultures, IL-23 was only partially effective in attenuating the decrease of IL-17 production from T cells following subsequent rounds of stimulation, compared with blocking Abs alone (Fig. 3⇓A). IL-23 was effective in limiting IFN-γ production from these cultures suggesting that the decrease in IL-17 production observed over multiple rounds of stimulation is not due to coincident increases in IFN-γ production or increases in the percentage of cells that are IFN-γ+.
Given the similarity in autoimmune disease phenotype between IL-1RI- and IL-23p19-deficient mice, and that IL-1 enhances IL-23 responsiveness (15, 16, 17), we next tested the ability of IL-1β to cooperate with IL-23 in long term cultures. Naive CD4 T cells were cultured as in Fig. 3⇑A for the first week and then cultured for subsequent rounds of stimulation with IL-23, IL-1β, or a combination of IL-23 and IL-1β. Although IL-1β was no more effective than IL-23 in attenuating the loss of IL-17 secretion over multiple rounds of stimulation, the combination of IL-23 and IL-1β was able to maintain a high level of IL-17 secretion over three rounds of stimulation (Fig. 3⇑B). There were not dramatic differences in the growth or survival among these cultures over several rounds of culture (data not shown), suggesting that the effects of these cytokines are not on survival or expansion, but rather on maintaining the phenotype of the cells. The combination of IL-23 and IL-1β was similarly capable of maintaining higher levels of IL-21 and IL-22 secretion than IL-23 alone, although there were decreases in these cytokines over rounds of stimulation (Fig. 3⇑B).
To define the mechanism for the ability of IL-1β to augment IL-23 function we first analyzed IL-23R expression in cells cultured with IL-23, IL-1β, or both cytokines for the second and third rounds of stimulation. Although culture with IL-23 maintained or enhanced Il23r mRNA expression, there was no increased expression in cells cultured with IL-1β alone or with both cytokines (Fig. 4⇓A). We then examined IL-23 signaling using flow cytometry to assess levels of phospho-Stat3 following acute stimulation of cultures incubated for three rounds in IL-23, IL-1β, or both cytokines. IL-23 stimulated Stat3 phosphorylation in cells from each of the conditions with insignificant differences in the phospho-Stat3 levels among the conditions, suggesting that the effect of IL-1β was not altering IL-23 signaling (Fig. 4⇓B). To test whether IL-1β had altered the Il17 gene to make it more responsive to IL-23, we took advantage of an assay we previously described for the acute stimulation of IL-17 production by a combination of IL-23 and IL-18 (18). Naive CD4 T cells primed with TGF-β plus IL-6 plus IL-1β for the first round of stimulation and cultured in IL-23 or IL-23 plus IL-1β for two rounds of stimulation were restimulated with IL-23 plus IL-18. Cells that were cultured with IL-23 plus IL-1β generated higher amounts of IL-17 than cells cultured in IL-23 alone in response to IL-23 plus IL-18 (Fig. 4⇓C). Because analysis of Il23r expression in these cultures (Fig. 4⇓A) showed only minor differences in expression, with slightly lower levels in the IL-23 plus IL-1β cultured cells, it suggests that culture in IL-23 plus IL-1β is directly affecting the responsiveness of the Il17 locus, although we did not find differences in the level of total histone acetylation between cells that were cultured with or without IL-1β (data not shown). IL-1β may enhance IL-23 function by activating cooperative transcription factors, or through indirect mechanisms including the ability to limit the inhibitory effects of IL-2 on Th17 development (19).
IL-23 does not mediate commitment to the Th17 lineage
Because IL-23 was capable of maintaining the IL-17-secreting phenotype, it allowed us to determine whether IL-23 mediated commitment to the Th17 lineage, with commitment being defined as the ability of cells to maintain the IL-17-secreting phenotype in the presence of cytokines promoting the development of other subsets. Th1 and Th2 cells are committed to their respective lineages after three rounds of stimulation (20). Naive CD4 T cells were primed with TGF-β plus IL-6 plus IL-1β for the first round, cultured for two rounds in IL-23 plus IL-1β, and then either maintained in IL-23 plus IL-1β or switched to IL-12 or IL-4 containing medium for the fourth round of stimulation before stimulation with anti-CD3 to assess cytokine production. Although cells cultured for the fourth round in IL-23 plus IL-1β maintained the ability to produce IL-17, cells switched to Th1 or Th2 promoting conditions showed diminished IL-17 production and the induction of IFN-γ and IL-4, respectively (Fig. 5⇓, A and B). Similar results were generated using cultures derived from C57BL/6 or BALB/c mice (data not shown). The levels of IL-4 induced following switching into Th2 conditions were lower than seen from cells cultured for 4 wk under Th2 conditions, though the levels of IFN-γ secreted by Th17 cultures switched to Th1 conditions were comparable to long term Th1 cultures (Fig. 5⇓, A and B). Expression of other Th17 genes including Il17f, Il23r, Il22, and Rorc were also diminished in cultures switched to Th1 or Th2 conditions (Fig. 5⇓, C and D). In contrast, naive CD4 T cells primed with TGF-β plus IL-6 plus IL-1β for the first round, cultured for two rounds in IL-23 plus IL-1β, and subsequently switched to culture conditions that promote Treg development were unable to develop into Foxp3-expressing cells (Fig. 5⇓E). Culture of cells with TGF-β plus IL-2 increased the percentage of IL-17+ cells. Thus, Th17 cells cultured with IL-23 can adopt some, but not all CD4 T cell lineages.
Although T cells cultured for three rounds under these Th17 conditions produce large amounts of IL-17, there is still some production of IFN-γ, and it remained possible that a contaminating population of cells was expanding and overtaking the Th17 cells upon switching cultures to Th1 or Th2 conditions. To eliminate this possibility, we used the cytokine capture protocol to isolate IL-17-high cells from Th17 cultures after three rounds of stimulation (Fig. 5⇑F) and then maintained the cultures with IL-23 plus IL-1β or switched to conditions promoting Th1 or Th2 development. As observed with unseparated Th17 cultures, IL-17-high cells maintained their phenotype with continued culture in IL-23 plus IL-1β, but showed decreased IL-17 and IL-22 production, and increased IFN-γ and IL-4 production in Th1 and Th2 conditions, respectively (Fig. 5⇑, G and H). Thus, IL-17-high cells are not stable secretors of IL-17 and upon exposure to conditions promoting the development of other Th subsets, they acquire new cytokine secreting characteristics.
We then defined if the switch from Th17 to Th1 or Th2 was characterized by induction of the standard pathways and lineage determining factors. IL-4 signaling is qualitatively altered in Th1 cells through mechanisms that are still not clear but do not involve altered Stat6 activation (21), and IL-4 was able to activate Stat6 in cells cultured under Th1, Th2, or Th17 conditions (Fig. 6⇓A). During Th2 development, IL-12 signaling is extinguished, contributing to commitment in the Th2 lineage (22). However, while Il12rb2 expression was greatly decreased in cells cultured for three rounds in Th2 conditions, cells cultured in IL-23 plus IL-1β demonstrated expression of Il12rb2 similar to Th1 cells (Fig. 6⇓B). To determine whether IL-12 signaling was functional, we examined Stat4 expression and IL-12-induced phosphorylation of Stat4 in Th1, Th2, and Th17 cultures after three rounds of stimulation. Although Stat4 expression is reduced and IL-12-induced Stat4 phosphorylation is eliminated in Th2 cultures, normal expression of Stat4 was retained in Th17 cultures. IL-12-induced Stat4 activation was only modestly diminished in IL-23 plus IL-1β and IL-23 cultured cells compared with Th1 cultures (Fig. 6⇓C). Despite the reduction in IL-12-induced Stat4, IL-12 and IL-4 were still able to promote an increase in Tbx21 and Gata3 expression in Th17 cultures switched to Th1 or Th2 conditions, respectively (Fig. 6⇓D). Thus, Th17 cells, even after long-term culture, are competent to assume a Th1 or Th2 phenotype.
IL-23 is a critical cytokine in the development of inflammation. However, the cellular activities of IL-23 have not been well defined. In this report, we demonstrate that IL-23 maintains a Th17 phenotype in the absence of promoting proliferation or survival, but that IL-23-cultured cells are not committed and can assume other cytokine secreting phenotypes. There are several reasons that it may be advantageous to have a transient Th17 phenotype. In the absence of multiple levels of control, the potent proinflammatory activity of these cells, if left unrestrained, could result in excessive tissue damage in vivo. In that respect, requiring several cytokines to establish and maintain the phenotype, while the presence of IL-4 or IL-12 effectively diminishes the phenotype, potentially limits the effect of Th17 cells and allows Th1 or Th2 cells to mediate progressive inflammation. The sensitivity of Th17 cells to the cytokine environment is consistent with the earliest reports of Th17 cells that demonstrated the importance of neutralizing IFNs and IL-4 in the culture of IL-17-secreting T cells (6, 7). Our work takes this further by showing Th1- and Th2-promoting cytokines repress an established Th17 gene program. This sensitivity to other cytokines may also be reflected in vivo. In Experimental Autoimmune Encephalomyelitis models, where Th17 cells are clearly important for the initiation of disease, analysis of lymphocytes from the target organ show a heterogeneous population of IL-17- and IFN-γ-secreting cells (23). Similarly, while IL-17 is required at early time points in the establishment of allergic inflammation, Th2 cells establish the cytokine environment characteristic of the allergic response. IL-17 administered to mice with established allergic disease inhibits inflammation (24). Together, these data suggest that, as we have shown for in vitro cultured cells, established Th17 cells in vivo may be inhibited by the appearance of other Th subsets where a changing cytokine environment results in repression of the Th17 phenotype.
IL-23 production may promote IL-17 production in at least two ways, by maintaining the IL-17-secreting potential of Th17 cells following Ag receptor stimulation and by directly activating Il17 in conjunction with IL-18 (18). The mechanism of IL-23 function at the level of the gene remains unclear. IL-23 requires Stat3 for activity and several reports have noted that IL-23-induced Stat3 can bind to the Il17 promoter and this might play a role in chromatin remodeling and histone acetylation at the locus (18, 25). Conversely, enhanced Stat3 activity potentiates IL-23 activity (18, 25). However, as other Stat3 activating cytokines have distinct effects on Th cell cytokine production, IL-23 must provide a qualitatively unique signal. As the synergy between IL-12 and IL-18 depends on the engagement of MAPK and NF-κB signals to cooperate with Stat4 signaling (3), it is likely that IL-23-activated Stat3 collaborates with other pathways activated by IL-18 or IL-1β. It is also possible that IL-23 inhibits the expression or function of other factors that inhibit the Th17 phenotype. For example, T-bet is a potent inhibitor of IL-17 production and anti-CD3 induced T-bet expression is responsible for changing cells from a Th17 to a Th1 cytokine secreting pattern in IL-23-stimulated cultures (14). In that culture system Ag receptor stimulation was responsible for decreasing IL-17 production and cells cultured in IL-23 without Ag receptor stimulation maintained IL-17-secreting potential. Thus, in the absence of commitment, the balance of activating and inhibitory signals a Th17 cell receives may determine the level of cytokine it will subsequently produce. It is still possible that IL-23 has effects on Th17 cell expansion or survival in vivo, though our results suggest that those effects must be indirect.
The plasticity of Th subsets is of considerable interest and is important in understanding how T cells regulate inflammation over time. In the long-term cultures we have described, the percentages of IL-17+ cells, and total IL-17 secretion, decrease over rounds of stimulation. Importantly, in the absence of cytokines that promote skewing to other phenotypes, cells that were secreting IL-17 do not appear to take on other phenotypes. We did not observe increases in IFN-γ-secreting or Foxp3+ cells (data not shown) and the population seems to acquire cells that do not secrete any of the cytokines examined. Whether this is conversion of Th17 cells to nonsecretor phenotypes or expansion of uncommitted cells is unclear. Following exposure to skewing conditions, Th17 cells appear to have differing abilities to acquire other phenotypes. For example, while IL-23-cultured IL-17+ Th17 cells may become Th1 or Th2 phenotype cells, we observe that culture of long-term Th17 cells with TGF-β plus IL-2 did not convert them into Foxp3 expressing cells, but rather increased IL-17 production (Fig. 5⇑), suggesting that IL-23-cultured cells are not capable of subsequently becoming adaptive regulatory T cells (Tregs).4 Conversely, natural Tregs have been shown to adapt a Th17 phenotype when cultured with IL-6 (26, 27), though there is still not a consensus on the ability of adaptive Tregs to undergo this change of phenotype (27, 28). Because of the phenotypic overlap, IL-23 may also be important for polarizing Th cells to distinguish the Th17 and Tfh phenotypes (29, 30, 31). Even with our culture conditions optimized for maintaining IL-17 secretion, IL-21 levels still decrease over time, and IL-22 levels are lower than in some previous reports (Fig. 3⇑) (11, 12). Although some of these effects might arise from the specific culture conditions, including the concentrations of cytokines added to the culture, it is possible that IL-23, as it maintains IL-17 production, limits IL-21 production, as it does for IL-10 production (13). Thus, even though IL-23 does not limit conversion to Th1 or Th2 lineages, it might restrict cells from adopting Treg or Tfh phenotypes.
The ability of IL-23 to promote inflammation in vivo likely involves several of the functions we have described, including the acute induction of IL-17, functioning in concert with IL-18 or IL-1, and the ability of IL-23 to preserve the IL-17-secreting potential of T cells. However, in the absence of IL-23 and the presence of other cytokine environments, IL-17 production can be diminished in favor of establishing new patterns of cytokine secretion. The transient Th17 phenotype may be a key to understanding how Th cell control may evolve during inflammation.
We thank J. Blum, C.-H. Chang, A. Mathur, and A. Dent for their suggestions on this manuscript and M. Mwanthi for technical assistance.
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.
↵1 This work was supported by U.S. Public Health Service Award AI45515 (to M.H.K.) from the National Institutes of Health. G.L.S. and N.Y. were supported by a Training Grant in Immunology and Infectious Disease (T32AI060519) and N.Y. was supported by the DeVault Fellowship from the Indiana University Cancer Biology Training Program.
↵2 G.L.S. and N.Y. contributed equally to this study.
↵3 Address correspondence and reprint requests to Dr. Mark H. Kaplan, Department of Pediatrics, and Microbiology and Immunology, Wells Center for Pediatric Research, Indiana University School of Medicine, 702 Barnhill Drive, RI 2600, Indianapolis, IN 46202. E-mail address:
↵4 Abbreviations used in this paper: Treg, regulatory T cell.
- Received June 17, 2008.
- Accepted August 28, 2008.
- Copyright © 2008 by The American Association of Immunologists