CD4+ T cell help is essential for primary CD8+ T cell responses to noninflammatory Ags. IL-2 is one of the principal cytokines made by naive CD4+ T cells, and we show in this study that it is an essential component of help. Adoptively transferred naive CD4+ TCR-transgenic OT-II cells supported endogenous primary CD8+ T cell responses, but IL-2-deficient OT-II cells were unable to provide help, although they responded to Ag in vivo and up-regulated CD40 ligand in vitro. Wild -type OT-II cells helped endogenous CD8+ T cell responses in IL-2-deficient mice, but not in IL-2Rα-deficient mice. Thus, CD4+ T cell-derived IL-2 is essential for CD8+ T cell responses to noninflammatory, cell-associated Ags. We suggest that it is also a critical component of help for CD8+ T cell responses to pathogens, because protective memory also requires CD8+ T cell stimulation by IL-2 during priming.
Help from CD4+ T cells can be essential for the generation of primary, memory, and protective secondary CD8+ T cell responses (1). Unhelped CD8+ T cells may fail to expand during the primary response (2, 3, 4), make primary responses but develop severely impaired memory (5, 6, 7), or develop into memory cells that fail to protect against pathogen challenge (8, 9).
It was first thought that IL-2 from Th cells was necessary to drive CD8+ T cell proliferation (10). This explanation fell from favor with the demonstration that help involved CD40-CD40 ligand (CD40L)3 interactions and could be replaced by ligation of CD40 on the APC (11, 12, 13). Furthermore, experiments with mice deficient for IL-2 or the IL-2Rα-chain suggested that IL-2 was essential for regulatory T cell function, but not for CD8+ T cell immunity (14, 15, 16). It was recently shown, however, that protective pathogen-specific CD8+ T cell memory failed to develop if the CD8+ T cells were IL-2Rα-deficient, even though primary and memory responses appeared normal (17, 18). This was reminiscent of memory primed in the absence of help (8, 9) and suggested that IL-2 might be an important component of help for the development of protective CD8+ T cell immunity to pathogens; however, the cellular source of IL-2 was not identified.
In this study, we developed an adoptive transfer model to ask whether CD4+ T cell-derived IL-2 was required for help-dependent CD8+ T cell responses. Mice primed with peptide-pulsed dendritic cells (DC) develop very long lasting CD8+ T cell memory; both primary and secondary responses are strongly help dependent (3, 19, 20). We found that IL-2 secreted by CD4+ T cells was essential for help-dependent primary CD8+ T cell responses and could restore these responses in IL-2-deficient mice. Thus, Th cell-derived IL-2 is a central component of help for CD8+ T cell responses.
Materials and Methods
C57BL/6 (B6), B6.PL, and OT-II mice (21) and mice deficient for IL-2 (22), IL-2Rα-chain (16), and CD40L (23) were purchased from Taconic Farms or The Jackson Laboratory. Male OT-II mice were crossed with gene-targeted female mice to generate male gene-targeted OT-II mice. Female IL-2−/− and IL-2Rα−/− littermates were used as recipients in Fig. 4. Experiments were approved by the University of Rochester Institutional Animal Care and Use Committee (Rochester, NY).
Naive CD62Lhigh CD4+ T cells were purified from OT-II splenocytes using magnetic beads (Miltenyi Biotec) and injected i.v. into the tail 24 h before immunization.
Splenic DC were cultured overnight in medium plus 1% autologous normal mouse serum, 1 ng/ml recombinant mouse GM-CSF, and peptides at 1 μM (OVA257–264), 2 μM (OVA323–339), or 10 μM (TRP2180–188) as described (3). Mice were immunized with 1–2 × 105 sex-matched DC. Initially mice were immunized i.p. as described (3); in later experiments mice were immunized s.c. and responses were measured in the draining lymph node on day 5 or in the spleen on days 6–8.
ELISPOT assays for IL-2 and IFN-γ were described previously (3). Peptides were used at 1 μM. Results are shown for individual mice. Spot counts are the mean ± SD of triplicate wells. Responses without peptide were very low (usually ≤11 spots (IFN-γ) or 16 spots (IL-2)/106 cells) and were not subtracted.
Flow cytometric analysis
Abs to CD4 (clone RM4-4), Thy1.2 (clone 30-H12), CD25 (clone 7D4), CD62L (clone Mel-14), CD69 (clone H1.2F3), anti-CD40L (clone MR1), and allophycocyanin-conjugated streptavidin were purchased from eBioscience or BD Pharmingen.
Results and Discussion
Help from adoptively transferred naive CD4+ T cells mimics help from endogenous CD4+ T cells
Mice immunized with peptide-pulsed DC make primary and secondary CD8+ T cell responses that are strongly help dependent (3, 19, 20). To test the importance of IL-2 secreted by Th cells, we established an adoptive transfer model using TCR-transgenic OT-II CD4+ T cells to provide help. The first experiments established that naive OT-II cells could provide help, which was most effective if the OT-II cells recognized Ag on the same APC as the endogenous CD8+ T cells and depended on OT-II expression of CD40L.
B6 mice were injected with OT-II cells and then immunized with DC pulsed with the Kb-restricted peptide OVA257–264 (pOVAI) plus the I-Ab-restricted peptide OVA323–339 (pOVAII), or with DC pulsed with pOVAI alone. Splenocytes were assayed 8 days later for IFN-γ responses to pOVAI and IL-2 responses to pOVAII (Fig. 1 A). Mice injected with OT-II cells plus pOVAI/pOVAII-pulsed DC made significant IFN-γ responses to pOVAI and IL-2 responses to pOVAII. Mice injected with OT-II cells plus pOVAI-pulsed DC did not respond to pOVAI or pOVAII, showing that OT-II cells helped only if they were stimulated by Ag in vivo. The pOVAII peptide is poorly immunogenic in B6 mice (26) and did not stimulate endogenous CD4+ T cell help in our experiments; pOVAI/pOVAII-pulsed DC failed to stimulate IFN-γ responses to pOVAI or IL-2 responses to pOVAII unless the mice were also injected with OT-II cells. Because the endogenous response to pOVAII did not provide help, we could evaluate help from gene-targeted OT-II cells without depleting host CD4+ T cells.
Similar results were observed (Fig. 1 B) using the Kb-restricted pTRP2I peptide, derived from the endogenous tumor Ag TRP2; mice injected with DC pulsed with pTRP2I/pOVAII made an IFN-γ response to pTRP2I only when OT-II cells were provided. Thus OT-II cells could help endogenous CD8+ T cell responses to weak self-Ags as well as strong xenogeneic Ags. This experiment also showed that help was most effective when OT-II cells and endogenous CD8+ T cells recognized Ag on the same DC (3, 10, 27). The IFN-γ response to pTRP2I was substantially diminished or absent when pOVAII and pTRP2I were presented by different DC populations, even though there was an excellent IL-2 response to pOVAII.
To confirm that help depended on CD4+ T cell expression of CD40L (11, 12, 13), wild-type (WT) or CD40L-deficient OT-II cells were injected into B6 mice, which were then immunized with pTRP2I/pOVAII-pulsed B6 DC (Fig. 1 C). Mice injected with WT OT-II cells made IL-2 responses to pOVAII and IFN-γ response to pTRP2I, but mice injected with CD40L-deficient OT-II cells, like mice that did not receive OT-II cells, made no significant IFN-γ response to pTRP2I. Similar results were obtained using the pOVAI epitope (data not shown).
IL-2-deficient OT-II cells are unable to help primary CD8+ T cell responses
Ligation of CD40 cannot always replace help, suggesting that help depends on additional CD4+ T cell effector functions (28, 29). Because IL-2 is the principal cytokine produced by naive T cells (30), we asked whether IL-2-deficient OT-II cells could provide help. CD4+ T cells from IL-2+/− OT-II and IL-2−/− OT-II littermates were transferred into B6 mice, which were then injected s.c. with pOVAI/pOVAII-pulsed DC (Fig. 2). Mice that received 1 × 106 IL-2+/− OT-II cells responded to both peptides; in contrast, mice that received IL-2−/− OT-II cells did not respond to either peptide. The importance of this cytokine was shown by the lack of any detectable IFN-γ response to pOVAI despite the high number of IL-2−/− OT-II cells transferred. Because IFN-γ is secreted by multiple effector subsets, the CD8+ T cell response was clearly very seriously impaired.
To determine whether pOVAI-specific CD8+ T cells failed to expand or survive in the absence of IL-2 or, alternatively, expanded normally but were functionally impaired, frozen spleen cells from the eight mice in Fig. 2 were stained with Kb/pOVAI tetramers. There was a highly significant difference (p = 0.03) in the percentage of tetramer-positive CD8+CD44high cells between the two groups (IL-2+/− OT-II cells, 15.60% ± 3.85%, n = 4; IL-2−/− OT-II cells, 2.35% ± 1.07%, n = 4). Thus, CD4+ T cell-derived IL-2, like help itself (3) was important for expansion and/or survival of Ag-specific CD8+ T cells in our experimental model.
IL-2-deficient OT-II cells respond to Ag in vivo and up-regulate CD40L in vitro
To determine whether IL-2-deficient OT-II cells could respond to Ag in vivo, IL-2+/− or IL-2−/− OT-II (Thy1.2+) cells were transferred into B6.PL (Thy1.1+) recipients, which were then immunized with DC pulsed with pOVAII plus keyhole limpet hemocyanin (KLH) to provide bystander help for the OT-II cell response (31). Draining lymph node cells were analyzed 5 days later for the presence of Thy1.2+CD4+ cells (Fig. 3 A). IL-2+/− OT-II cells expanded significantly in mice injected with pOVAII/KLH-pulsed DC, but not in mice injected with KLH-pulsed DC, showing that the expansion was pOVAII-specific. IL-2−/− OT-II cells expanded similarly as IL-2+/− OT-II cells; thus their inability to help CD8+ T cells could not be explained by developmentally impaired homing or Ag recognition.
To test whether CD40L was expressed by IL-2-deficient OT-II cells, IL-2+/− and IL-2−/− OT-II cells were cultured in vitro with pOVAII-pulsed, IL-2−/− DC in the presence of anti-CD40L mAb. Cultures were harvested at 2, 4, and 6 h and stained with mAb to CD4, CD25, and CD69. CD40L was up-regulated by both populations and could be detected at all three time points (Fig. 3 B). Similar results were seen for IL-2+/− OT-II cells when IL-2 was added to the culture medium (data not shown). Thus the kinetics of CD40L expression were unaffected by the availability of IL-2.
OT-II cells help primary CD8+ T cell responses in IL-2-deficient but not IL-2Rα-deficient mice
These results were consistent with two distinct mechanisms. IL-2 secreted by OT-II cells might drive CD8+ T cell proliferation and differentiation (10). Alternatively, it might act mainly on the OT-II cells themselves; although IL-2 was not needed for CD4+ T cell expression of CD40L, it might drive other helper functions or stabilize CD40L expression (32). We reasoned that if OT-II-derived IL-2 acted mainly on CD8+ T cells, then normal OT-II cells would help CD8+ T cell responses in IL-2−/− mice, but not IL-2Rα−/− mice. Alternatively, if this IL-2 acted mainly on the OT-II cells themselves, then OT-II cells would fail to provide help in either strain (if the CD8+ T cells also required IL-2 stimulation) or would provide help in both strains (if the CD8+ T cells did not require IL-2).
OT-II cells were therefore injected into IL-2−/− and IL-2Rα−/− mice. The results (Fig. 4) show that OT-II cells were able to support CD8+ T cell responses in IL-2−/− mice, but not IL-2Rα−/− mice. Four of six IL-2−/− mice injected with OT-II cells plus pOVAI/pOVAII-pulsed DC made substantial IFN-γ responses to pOVAI, and significant IL-2 responses to pOVAII. The CD8+ T cell response to pOVAI in these mice was help dependent, because IL-2−/− mice that received OT-II cells but were injected with DC pulsed with pOVAI alone did not make any detectable pOVAI-specific IFN-γ response. In contrast, pOVAI/pOVAII-pulsed DC failed to prime any pOVAI-specific IFN-γ response in IL-2Rα−/− mice, even though exceptionally strong Ag-specific IL-2 responses to pOVAII were seen in all four mice (due perhaps to the absence of regulatory T cells or to elevated levels of endogenous IL-2). We therefore concluded that these help-dependent CD8+ T cell responses depended on the ability of host cells to respond to IL-2 and that this IL-2 could be provided by OT-II Th cells.
Our results show that in the absence of overt inflammation, help-dependent primary CD8+ T cell expansion and/or survival, and possibly effector function as well, depend not only on APC stimulation by CD40L (11, 12, 13) but also on Th cell-derived IL-2. The importance of CD4+ T cell-derived IL-2 for primary CD8+ T cell responses probably correlates inversely with the degree of pathogen-mediated inflammation. TLR stimulation together with CD40 ligation results in type I IFN secretion, driving CD8+ T cell expansion and differentiation in the absence of CD4+ T cell help (33), thus explaining why IL-2-deficient mice often have apparently normal primary responses (14). Studies using viral infection models have shown that CD8+ T cells must be stimulated by IL-2 during priming to develop into long-lived protective memory cells, even though the primary response is IL-2-independent (17, 18). IL-2 may program protective memory functions into Ag-specific CD8+ T cells that have proliferated in response to pathogen-mediated inflammatory signals; alternatively, IL-2 may be essential for the primary expansion and programming of a specialized subpopulation of protective memory cells. In either case, our results support the possibility that this cytokine is provided by CD4+ Th cells.
We thank Drs. Fernando Ontiveros and Ernest Wang for discussion, Drs. Nick Crispe, Tim Mosmann, and David Topham for critical review of the manuscript, and Dr. Joe Hollenbaugh for tetramer staining.
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 (NIH) Grant AI48721 (to A.M.L.) and NIH Training Grant AI007285 (to E.B.W.).
Abbreviations used in this paper: CD40L, CD40 ligand; B6, C57BL/6 mice; DC, dendritic cell; KLH, keyhole limpet hemocyanin; pOVAI, Kb-restricted OVA257–264 peptide; pOVAII, I-Ab-restricted OVA323–339 peptide; pTRP2I, Kb-restricted tyrosinase-related protein-2 (TRP2180–188) peptide; WT, wild type.