The action of type I IFN (IFN-I) on APCs is well studied, but their direct effect on CD4 T cells is unclear. To address this, we transferred IFN-I receptor-deficient (IFN-IR0) and -sufficient (wild-type, WT) TCR-transgenic CD4 T cells into WT mice and analyzed their response to immunization. In response to lymphocytic choriomeningitis virus immunization, WT CD4 T cells expanded ∼100-fold, whereas IFN-IR0 CD4 T cells expanded <10-fold. However, both WT and IFN-IR0 CD4 T cells expanded ∼10-fold after Listeria monocytogenes immunization. Poor expansion of IFN-IR0 CD4 T cells after lymphocytic choriomeningitis virus immunization was not due to a defect in proliferation or initial activation but to poor survival of the daughter cells. Thus, direct IFN-I signals can play either a critical or minimal role in CD4 T cell clonal expansion depending on the specific pathogen.
Efficient functioning of Ag-experienced CD4 T cells is dependent on two major processes: clonal expansion and effector differentiation. It is generally accepted that type I IFN (IFN-I)3, a set of innate antiviral cytokines produced in large quantities following infection (1), facilitate clonal expansion via their actions on APC (2, 3). The direct effects of IFN-I on T cells are less clear (4, 5, 6, 7, 8). We found recently that action of IFN-I on CD8 T cells is critical for clonal expansion in response to lymphocytic choriomeningitis virus (LCMV) infection in mice (9). However, the direct effect of IFN-I on CD4 T cell responses in vivo remains unclear. In this study, we address the consequences of direct IFN-I-mediated signals on CD4 T cells by comparing the response of adoptively transferred wild-type (WT) and IFN-IR-deficient (IFN-IR0) CD4 T cells in WT hosts using two well-characterized murine infection models: 1) acute infection with the intracellular bacterial pathogen Listeria monocytogenes (LM) and 2) acute infection with LCMV.
Materials and Methods
B6.J.129S2-Ifnar<tm1Agt>, IFN-IR0 mice on a 129/SvEv background described in (10) that were backcrossed to C57BL/6 mice as described (9) will be referred to hereafter as IFN-IR0. Dr. P. J. Fink (University of Washington, Seattle) provided B6J.Cg-Tg(TcraTcrb)425Cbn, the OVA323–339-specific OT-II CD4 TCR transgenic (Tg) mice on a C57BL/6 background. These mice will be referred to hereafter as OT-II. B6-Tg(TcrLMCV)1Aox, LCMV-GP61–80-specific SMARTA TCR Tg mice described in (11) were obtained on a C57BL/6 background from Dr. C. Surh (The Scripps Research Institute, La Jolla, CA) via Dr. M. J. Bevan (University of Washington, Seattle). These mice will be referred to hereafter as SMARTA. These mice were bred with C57BL/6 IFN-IR0 mice to generate IFN-IR0 OT-II and IFN-IR0 SMARTA mice. Congenically marked mice were generated by cross to B6.PL-Thy1a/CyJ (B6 Thy1.1) or B6.SJL-PtprcaPep3b/BoyJ (B6 Ly5.1) purchased from The Jackson Laboratory. Male OT-II or either male or female SMARTA mice at 4–12 wk of age were used for experiments. Sex-matched mice were used for adoptive transfers. All mice were maintained under specific pathogen-free conditions at the University of Washington animal care facility under the guidelines of the Institutional Animal Care and Use Committee.
Reagents, Abs, in vitro cultures, adoptive transfers, and staining
Universal IFN-I were from PBL Biomedical Laboratories. OVA was from Sigma-Aldrich. All Abs were purchased from either BD Biosciences or eBioscience. Splenocytes from WT and IFN-IR0 CD4 TCR Tg mice were cultured at a concentration of 0.1 × 106 cells per well in 96-well plates for 66 h, either with or without peptide (10 μg/ml), or peptide plus 1000 U/ml IFN-I. Intravenous adoptive cell transfers contained 1–2 × 105 Tg CD4 T cells unless otherwise indicated. Intracellular cytokine staining was done as described (9).
Virus, bacteria, and immunizations
A total of 2 × 105 PFU of LCMV (Armstrong) was injected i.p. 24–48 h after cell transfer. A total of 1–2 × 104 CFU of WT LM or rLM-Ova was injected i.p. OVA (1 mg) was injected by s.c., i.v., and i.p. routes on 1–3 subsequent days after infection.
IFN-IR0 CD4 T cells are similar to WT CD4 T cells in phenotype and proliferation but are not inhibited by IFN-I during in vitro culture
Previous reports suggest that IFN-I exert an antiproliferative effect on anti-CD3-stimulated CD4 T cells during in vitro culture (6, 7). We asked whether IFN-I have similar effects when CD4 T cells were stimulated with cognate Ag in vitro. CD4 T cells derived from both WT and IFN-IR0 SMARTA mice were low for expression of CD44, CD25, and CD69, and were high for expression of CD62L (Fig. 1⇓A). Both up-regulated CD25 (Fig. 1⇓B) and underwent multiple cycles of proliferation (Fig 1⇓C, middle) when cultured in vitro with cognate antigenic peptide for 3 days. The addition of exogenous IFN-I during in vitro cultures strongly inhibited the proliferation of peptide-stimulated WT but not IFN-IR0 CD4 T cells (Fig. 1⇓C, bottom). Similar findings were made with OVA-specific OT-II CD4 T cells (data not shown). Together, the above observations raise the question of what would be the direct effect of IFN-I on CD4 T cells during an immune response in vivo.
Direct IFN-I action on CD4 T cells is important for clonal expansion in vivo following LCMV immunization
IFN-I have multiple effects on the cells of both the innate and adaptive immune systems. IFN-I also influence pathogen clearance and hence alter the antigenic load (10, 12, 13). Consequently, the comparison of CD4 T cell responses in infected WT and IFN-IR0 mice cannot specifically address the direct role of IFN-I on CD4 T cells during infection. Hence, we adoptively transferred a mixture of LCMV-specific WT and IFN-IR0-naive SMARTA CD4 T cells into WT hosts and compared their response to LCMV immunization (Fig. 2⇓A). As expected, in recipient mice that were left uninfected, WT and IFN-IR0 SMARTA donor cells were recovered at a low frequency in a ratio similar to the inoculum (Fig. 2⇓B, top). Donor SMARTA CD4 T cells expanded in response to LCMV (Fig. 2⇓B, bottom left), but ∼99% of the expanded donor cells were WT (Fig. 2⇓B, bottom right, and Fig. 2⇓C). IFN-IR0 CD4 donor T cells in the same host expand, but only marginally (Fig. 2⇓B, bottom, and Fig. 2⇓C). The difference in expansion between WT and IFN-IR0 SMARTA CD4 T cells was not restricted to the spleen, as it was seen in several tissues (Fig. 2⇓D). Thus, IFN-I produced in response to LCMV immunization act directly on virus-specific CD4 T cells and greatly contribute to their clonal expansion.
To determine whether the diminished expansion of IFN-IR0 CD4 T cells was due to a defect in initial activation, we compared the response of WT and IFN-IR0 SMARTA CD4 T cells early after LCMV immunization (Fig. 3⇓A). By 36 h after LCMV immunization, neither WT nor IFN-IR0 CD4 T cells expanded (Fig. 3⇓B, middle). The ratio of WT to IFN-IR0 CD4 T cells did not change substantially despite activation as indicated by CD25 up-regulation and CD62L down-regulation. However, by 7 days after LCMV-immunization WT SMARTA CD4 T cells expanded far better than IFN-IR0 SMARTA CD4 T cells (Fig. 3⇓B, bottom). Thus, direct signals mediated by IFN-I on LCMV-specific CD4 T cells have little influence on initial activation but greatly effect clonal expansion.
To assess whether diminished expansion of the IFN-IR0 CD4 T cells was due to a defect in proliferation, we compared the proliferation of CFSE-labeled WT and IFN-IR0 SMARTA CD4 T cells in WT mice after LCMV immunization. By 2.5 days postimmunization, both WT and IFN-IR0 CD4 T cells were recruited to proliferate (Fig. 3⇑C, middle). Despite their proliferation, the expansion of the IFN-IR0 CD4 T cells was greatly reduced at this time point (Fig. 3⇑D), indicating that most of the daughter cells generated from the donor IFN-IR0 CD4 T cells failed to survive.
IFN-I promote IFN-γ production, at least in vitro, but this process is less efficient in murine CD4 T cells, compared with human CD4 T cells (4, 5). Therefore, we assessed whether IFN-IR0 SMARTA CD4 T cells that were responding to LCMV Ag in vivo could produce IFN-γ. Donor CD4 T cells were able to produce IFN-γ irrespective of whether they could receive IFN-I-mediated signals or not (Fig. 3⇑E, middle and bottom). The lack of IFN-I-mediated signals has no substantial effect on IFN-γ production (Fig. 3⇑F). In addition, both WT and IFN-IR0 SMARTA CD4 T cells populations contained TNF-α- and IL-2-producing cells (data not shown). Together, the above results show that the absence of IFN-I action on LCMV-specific CD4 T cells in mice did not have a major affect on their ability to become activated, their ability to proliferate, or their ability to differentiate into IFN-γ producing effectors, but greatly affected their ability to survive during Ag-driven proliferation and, as a consequence, drastically dampened clonal expansion.
IFN-I-mediated direct signals have a minimal contribution to the clonal expansion of CD4 T cells following LM immunization
We next asked whether Ag-specific CD4 T cells become similarly dependent on IFN-I under the conditions of immunization with an intracellular bacterial pathogen, LM. For this, we transferred CFSE-labeled OVA-specific WT and IFN-IR0 OT-II CD4 TCR Tg T cells into WT hosts and analyzed their response to immunization with recombinant LM expressing OVA (rLM-Ova). Both WT and IFN-IR0 OT-II CD4 T cells proliferated (Fig. 4⇓A, right) and expanded (Fig. 4⇓B) in response to rLM-Ova.
In a different set of experiments, we transferred WT and IFN-IR0 OT-II CD4 T cells together into WT hosts and gave exogenous OVA following either LM or LCMV immunization. Both WT and IFN-IR0 OT-II CD4 T cells expanded similarly to each other in response to exogenous OVA protein under the conditions of LM immunization (Fig. 4⇑C) and produced IFN-γ upon in vitro restimulation (as described in the Fig. 4⇑ legend). In contrast, in response to exogenous OVA under the conditions of LCMV immunization, WT OT-II CD4 T cells expanded markedly, whereas IFN-IR0 OT-II CD4 T cells did not (Fig. 4⇑D). The diminished expansion of IFN-IR0 OT-II CD4 T cells in response to exogenous OVA in LCMV-immunized mice was not due to a lack of proliferation, as determined by CFSE dilution (Fig. 4⇑F), but due to increased death during proliferation, as suggested by annexin staining (Fig. 4⇑G).
Thus, IFN-I produced in response to LCMV immunization provide direct survival signals to Ag-specific CD4 T cells, greatly contributing to their clonal expansion. In contrast, CD4 T cells primed under the conditions of LM are minimally dependent on direct IFN-I-mediated signals.
CD4 T cell expansion requires not only Ag-receptor-mediated signals but also costimulatory signals via APCs. It is well established that inflammatory cytokines induced by infection, such as IFN-I, contribute to T cell expansion via their ability to up-regulate costimulatory molecules on APC. Our study shows that the responding CD4 T cells also need direct survival signals mediated by IFN-I. In addition, we demonstrate that the pathogen determines the extent to which CD4 T cells are dependent on direct signals mediated by IFN-I. Thus, this study not only unravels a novel mechanism by which IFN-I contribute to sustaining CD4 T cell expansion in vivo, but also highlights how such fundamental immune mechanisms become redundant depending on the immunogen. These results have implications in dissecting the critical factors involved in host-pathogen interactions, microbial pathogenesis, and therapeutic and vaccination strategies.
Why do CD4 T cells become highly dependent on direct IFN-I signals under the conditions of LCMV but not LM infection? We considered the possibility that the high-level antigenic load provided by LCMV due to its noncytopathic nature and its ability to infect a variety of cells could play a role. If this accounted for the difference that we observed, IFN-IR0 OT-II CD4 T cells should have expanded similarly to WT cells in response to exogenous OVA administered under the conditions of LCMV infection. However, this was not the case. Data in Fig. 4⇑E show that LCMV was not preferentially killing IFN-IR0 OT-II CD4 donor T cells. Moreover, we observed that IFN-IR0 SMARTA donor cells expand better in LCMV-infected IFN-IR0 hosts than in WT hosts, despite having a much higher viral load (data not shown). The profile of cytokines induced and the dynamics of its regulation differ depending on the infectious agent, characteristics of its tropism, antigenic load, persistence, and host. LCMV induces high levels of IFN-I and almost no IL-12 (10, 12). This cytokine balance is somewhat skewed toward higher levels of IL-12 in LCMV-infected IFN-IR0 mice (10). In contrast, LM induces IL-12 in addition to lower levels of IFN-I (14). IL-12 has been shown recently to prolong the survival of Ag-activated CD8 T cells (15, 16). Based on this, we predict that IL-12 might be involved in supporting clonal expansion of CD4 T cells under the conditions of LM infection and suggest that the IFN-I-dominated cytokine milieu induced by LCMV infection instructs the responding CD4 T cells to become highly dependent on direct signals mediated by IFN-I. An alternative possibility might be that high levels of IFN-I produced in response to LCMV may cause very high clonal expansion by directly signaling to CD4 T cells. In the absence of the IFN-I-mediated signal, the response is reduced to a level comparable to that seen for LM that does not cause high IFN-I production. However, we cannot rule out the involvement of other mechanisms and factors. For example, recent studies indicate that IFN-I can cause T cell death during LM infection via sensitization to listeriolysin O-mediated toxicity (17). This raises the possibility that IFN-IR0 CD4 T cells responding under the conditions of LM infection, although suffering from the lack of IFN-I mediated survival signals could be simultaneously eluding listeriolysin O-mediated death. The combination of these positive and negative effects may balance each other, resulting in the expansion of IFN-IR0 CD4 T cells similar to WT CD4 T cells under the conditions of LM infection. These issues, together with the consequences of the cytokine profile induced at the beginning of each infection and the way these dynamic profiles change over time, especially during chronic infection, remain unclear. Considering the difference in the amount of IFN-I and the composition of other cytokines induced in response to diverse viral and bacterial infectious agents, it is important to further understand how critical mediators of T cell survival, death, and expansion differ in different infections.
IFN-I are known to exert an antiproliferative effect on CD4 T cells at least in vitro (6, 7). Our results confirmed this observation, and yet we were unable to replicate this effect in vivo. One possible explanation for this is that the complex set of factors induced during host-pathogen interaction in vivo help to overcome the anti-proliferative functions of IFN-I. Interestingly, IFN-γ, which was originally thought to inhibit CD4 T cell proliferation in vitro, has recently been shown to support the expansion of adoptively transferred CD4 T cells in LCMV-infected WT mice (18, 19).
The mechanisms by which IFN-I affect T cell survival, especially in vivo, require further investigation. This is especially important given the widespread use of IFN-I as therapy for viral infection, cancer, and autoimmunity. Previous reports have established that the antiviral properties of IFN-I are conducted by STAT1 and STAT2 (1, 20). Recent studies have shown that experimental removal of STAT-1 allows IFN-I to provide prosurvival and proproliferative signals to T cells (6). Based on these studies, we suspect that IFN-I may be contributing to T cell expansion by modulating the balance of STAT signaling. It will be important to determine whether the dynamic changes in the level of IFN-I and TCR stimulation during clonal expansion alters this balance.
We thank Drs. C. B. Wilson and M. J. Bevan for helpful discussion and for critical reading of the manuscript, Dr. P. J. Fink for OT-II mice, Dr. C. Surh for SMARTA mice, Dr. R. Ahmed and members of the Kaja laboratory for helpful discussions and K. Madhavi 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 in part by National Institutes of Health Grants 1R01AI053146 and R21AI051386 (to K. M.-K.) and by funds from the Washington National Primate Center and the University of Washington Department of Immunology. C.H.D. is supported by a Cancer Research Institute training grant.
↵2 Address correspondence and reprint requests to Dr. Kaja Murali-Krishna, Department of Immunology and Washington National Primate Center, Health Science Building Box 357650, 1959 Northeast Pacific Street, Seattle, WA 98195. E-mail address:
↵3 Abbreviations used in this paper: IFN-I, type I IFN interfere; IFN-IR0, IFN-IR-deficient; LCMV, lymphocytic choriomeningitis virus; LM, Listeria monocytogenes; Tg, transgenic; WT, wild type.
- Received November 14, 2005.
- Accepted January 16, 2006.
- Copyright © 2006 by The American Association of Immunologists