Regulation of CD8 T cell expansion and contraction is essential for successful immune defense against intracellular pathogens. IL-10 is a regulatory cytokine that can restrict T cell responses by inhibiting APC functions. IL-10, however, can also have direct effects on T cells. Although blockade or genetic deletion of IL-10 enhances T cell-mediated resistance to infections, the extent to which IL-10 limits in vivo APC function or T cell activation/proliferation remains unknown. Herein, we demonstrate that primary and memory CD8 T cell responses following Listeria monocytogenes infection are enhanced by the absence of IL-10. Surface expression of the IL-10R is transiently up-regulated on CD8 T cells following activation, suggesting that activated T cells can respond to IL-10 directly. Consistent with this notion, CD8 T cells lacking IL-10R2 underwent greater expansion than wild-type T cells upon L. monocytogenes infection. The absence of IL-10R2 on APCs, in contrast, did not enhance T cell responses following infection. Our studies demonstrate that IL-10 produced during bacterial infection directly limits expansion of pathogen-specific CD8 T cells and reveal an extrinsic regulatory mechanism that modulates the magnitude of memory T cell responses.
Interleukin 10 is a cytokine with complex activities that include down-regulation of T cell responses (1). Because IL-10 is produced by many different cell types, including T cells, B cells, macrophages, monocytes, and dendritic cells (DCs) (1), and because a similarly diverse array of cells express the heterodimeric IL-10R and thus respond to IL-10 (1, 2), it has been difficult to distinguish the truly important in vivo effects of this cytokine from the multitude of possible effects demonstrated in in vitro studies. With respect to IL-10-mediated effects on T cell responses, although some studies suggest that IL-10 acts directly on T lymphocytes (3, 4, 5, 6, 7, 8), other studies support the idea that IL-10 down-regulates APC function by decreasing cytokine production, MHC class II Ag presentation, and expression of costimulatory molecules (9, 10). More recently, a study demonstrated that IL-10 enhances in vitro CD8 T cell expansion during priming but diminishes T cell proliferation upon second exposure to Ag (11). Whether these findings hold true during in vivo CD8 T cell responses is unknown.
Although many cells can make IL-10, recent studies using an IL-10 reporter mouse demonstrated that T lymphocytes are the major in vivo source of this cytokine (12). T regulatory cells have been implicated as an important source of IL-10; however, infection of IL-10 reporter mice with Leishmania major or Toxoplasma gondii demonstrated that FoxP3-negative, IFN-γ-producing CD4 T cells were the major source of IL-10 (12, 13). Whether production of IL-10 by CD4 T cells responding to infection can directly influence pathogen-specific CD8 T cell responses is unknown.
L. monocytogenes is a Gram-positive, intracellular bacterial pathogen that invades the host cell cytoplasm by destroying the phagosomal membrane following bacterial phagocytosis. CD8 T cells play a prominent role in the clearance of L. monocytogenes from infected mice and provide long-term protective immunity (14). Primary infection with L. monocytogenes induces rapid CD8 T cell responses that contract into long-term memory populations. Rechallenge with L. monocytogenes induces very rapid and robust memory CD8 T cell responses (14, 15). In both settings, CD8 T cells are primed during the first 24–48 h of infection and then expand without further requirement for Ag (16, 17). Surrounding inflammatory responses, however, can influence the magnitude of CD8 T cell expansion (18, 19). Recent studies have demonstrated that IL-2-mediated signals transmitted during the early phase of T cell priming enhance the generation of L. monocytogenes-specific CD8 T cell memory (20).
Mice with genetic deletion of IL-10 are more resistant to L. monocytogenes infection (21, 22). Along similar lines, greater susceptibility of female mice to L. monocytogenes infection has been correlated with gender-specific increases in IL-10 production (23). Increased bacterial clearance in IL-10-deficient mice results from enhanced innate immune activation during the first 1–2 days of infection and thus precedes the development of pathogen-specific T cells. Apoptosis of nonspecific T cells, however, is induced by bacterial secretion of listeriolysin-O and serves as a major stimulus for IL-10 production in the infected spleen (24). CD4 and CD8 T cells from L. monocytogenes-infected IL-10−/− mice produce more IFN-γ than T cells from wild-type-infected mice, suggesting that pathogen-specific T cell responses are inhibited by the presence of IL-10 (22). A more recent study, however, demonstrated that IL-10-deficient mice had fewer L. monocytogenes-specific CD8 T cells during primary and memory responses, suggesting that IL-10 enhances pathogen-specific T cell responses (25).
We have investigated the CD8 T cell response to L. monocytogenes infection in mice lacking IL-10 or the IL-10R2 component of the IL-10R. We find that pathogen-specific CD8 T cells that are insensitive to IL-10 undergo greater in vivo expansion and give rise to larger memory T cell populations than CD8 T cells that are IL-10 responsive. CD8 T cell priming in mice with IL-10-unresponsive APCs, in contrast, is equivalent to priming in mice with APCs that respond to IL-10. Our studies indicate that IL-10 directly restricts the magnitude of primary CD8 T cell responses to L. monocytogenes and thereby limits the number of memory CD8 T cells.
Materials and Methods
Bacteria and infections
Recombinant L. monocytogenes strain-expressing OVA (LM-OVA), ActA-deficient LM-OVA (Act−/−LM-OVA), and L. monocytogenes strain 10403S L1942 was provided by Dr. H. Shen (University of Pennsylvania, Philadelphia, PA), Dr. J. Harty (University of Iowa, Iowa City, IA), and Dr. D. Portnoy (University of California, Berkeley, CA), respectively. Bacteria were grown in brain-heart infusion broth (BD Biosciences). Mice were immunized by i.v. injection of the indicated doses into the lateral tail vein. Spleens, blood, and liver were harvested at the indicated time points after immunization or cell transfer.
C57BL/6, B6. PL, C57BL/6-OT-1-transgenic (Tg),3 C57BL/6 IL-10−/−, C57BL/6 IL-10R−/−, and Rag1−/− mice were obtained from The Jackson Laboratory. The OT-1 lines were crossed onto the B6.PL and C57BL/6 IL-10R−/− background. Age-matched female mice were used in all of the experiments and maintained under specific pathogen-free conditions at the animal facilities of Memorial Sloan-Kettering Cancer Center New York, NY).
Adoptive transfer of TCR Tg CD8 T cells
Splenocyte suspensions containing the indicated number of Tg CD8 T cells from IL-10R−/− (Thy1.1/1.2) or wild-type (Thy1.1) OT-1 mice were adoptively transferred into congenic recipients by injection into the lateral tail vein. Before transfer, the naive surface phenotype of the T cells (CD44intCD62LhighCD69lowCD25low) was confirmed by flow cytometry.
Bone marrow-derived DCs were generated by culturing C57BL/6 or IL-10R−/− bone marrow cells in complete RPMI 1640 supplemented with 2% GM-CSF (provided by L. Denzin, Memorial Sloan-Kettering Cancer Center). Nonadherent cells were washed off after 2 days, and medium were replaced every 2 days. After 6 days, DCs were matured with 100 ng/ml LPS (Sigma-Aldrich) and washed multiple times. Cells were then pulsed with 10−6 M SIINFEKL peptide for 1 h at 37°C and washed. Before i.v. injection of 106 cells/mouse, the matured surface phenotype (CD80highCD86highMHC class IIhigh) of DCs was confirmed by flow cytometry.
In vitro activation of CD8 T cells
For polyclonal activation, CD8 T cells from C57BL/6 mice were MACS (Miltenyi Biotec) purified (negative selection) and activated with plate-bound anti- CD3- CD28 (30 μg/ml; BD Pharmingen) in 96-well plates.
Cells were washed three times with PBS and resuspended at 2 × 107 cells/ml in PBS containing 4 μM CFSE (Molecular Probes). After incubation at 37°C for 10 min in the dark, cells were immediately washed with cold RPMI 1640/10% FCS and counted before resuspension in PBS and i.v. injection into mice.
Peptides, Abs, tetramers, and flow cytometry
Peptides were synthesized by the Memorial Sloan-Kettering Cancer Center microchemistry core facility. The following mAbs directed to mouse cell surface Ags were purchased from BD Pharmingen: anti-CD8-FITC, -PE, -PerCP, -allophycocyanin (clone 53-6.7), anti-CD62L-FITC, -PE (MEL-14), anti-CD44-PE (IM7), anti-CD69-PE (H1.2F3), anti-CD127-PE (SB)/199), anti-Thy1.1-PerCP (OX-7), anti-Thy1.2-allophycocyanin, -FITC (53-2.1), anti-CD80-FITC (16-10A1), anti-CD86-FITC (53-5.8), anti-CD210-PE (1B1.3a), anti-CD122-PE (TM-β1) anti-I-Ab-FITC (AF6-120.1), anti- CD11c-allophycocyanin (HL3), anti-V5 TCR-FITC (MR9-4), anti-IFN-γ-FITC, -PE (MG1.2), and anti-TNF-α-FITC (MP6-XT22). For flow cytometric analysis, 1–2 × 106 cells were aliquoted per staining well of a 96-well plate. After incubation at 4°C for 20 min with Fc block (BD Pharmingen) in FACS SB (PBS (pH 7.4), 0.5% BSA, and 0.02% sodium azide), cells were incubated for 30 min (1 h for stainings for tetramers) with saturating concentrations of the various mAbs and tetramers. Labeled cells were washed with SB and analyzed on a BD Biosciences LSR II flow cytometer and FlowJo software (Tree Star).
Murine IL-10 was quantified using an OptEIA kit from BD Pharmingen. To obtain lysates for cytokine assays, spleens were harvested at the indicated times postinfection, mascerated in ice-cold PBS containing 0.01% Triton X-100, and centrifuged at 10,000 × g. The supernatant was used for ELISA.
Statistical analysis was performed on Microsoft Excel software with the two-tailed Student t test. A p < 0.05 was considered significant.
IL-10 produced as a consequence of L. monocytogenes infection inhibits CD8 T cell responses
Although in vivo infection with L. monocytogenes is known to induce IL-10, the kinetics of IL-10 induction during early infection are incompletely defined (26). To determine how rapidly IL-10 is produced during infection, C57BL/6 mice were i.v. infected with LM-OVA and, at the indicated time points, spleen lysates were analyzed for IL-10 production by ELISA. Although IL-10 was undetectable in uninfected mice, IL-10 was detected as early as 12 h after infection and peaked 24–48 h after infection before declining 96 h following infection (last time point analyzed) (Fig. 1 A).
To investigate the effect of IL-10 signaling on L. monocytogenes-specific primary and secondary CD8 T cell responses, wild-type and IL-10−/− mice were infected with 5000 LM-OVA, and SIINFEKL-specific CD8 T cell responses were measured by H2-Kb SIINFEKL tetramer staining. Sixty days after primary infection, mice were rechallenged with LM-OVA and secondary expansion of CD8 T cells was analyzed 5 days later. As shown in Fig. 1B, in comparison to 7% SIINFEKL-specific CD8 T cell frequencies in spleens of wild-type control mice, we detected roughly 11% SIINFEKL-specific CD8 T cell frequencies in spleens of IL-10−/− mice 8 days after infection. The absolute number of CD62Llow SIINFEKL-specific splenic CD8 T cells was increased in IL-10−/− mice in comparison to wild-type mice (Fig. 1,C). Following rechallenge, IL-10-deficient mice had markedly increased secondary expansion, with ∼37% SIINFEKL-specific CD8 T cells, in comparison to reinfected wild-type mice, in which ∼25% of CD8 T cells were SIINFEKL specific. (Fig. 1, B and D). These results suggest that IL-10 produced early after L. monocytogenes infection restricts primary and secondary expansion of pathogen-specific CD8 T cells. This result might be explained by IL-10-mediated down-regulation of Ag presentation by APCs. Alternatively IL-10 might directly inhibit CD8 T cell proliferation by binding to IL-10R on T cells. A third possibility is that enhanced clearance of L. monocytogenes or increased expression of inflammatory cytokines in IL-10−/− mice might influence CD8 T cell priming.
IL-10R1 is transiently expressed on CD8 T cells following activation
As a first step to investigating whether IL-10 can directly impact CD8 T cell priming and/or expansion, we investigated surface expression of IL-10R1 on CD8 T cells. Purified CD8 T cells from C57BL/6 mice were activated with antiCD3-CD28 in vitro and, at the indicated time points, CD8 T cells were analyzed for surface expression of IL-10R1 by flow cytometry. Although naive CD8 T cells do not express surface IL-10R1, 24 h following in vitro activation CD8 T cells up-regulate surface IL-10R1. Up-regulation of IL-10R1 is transient, however, and 96 h after stimulation IL-10R1 expression becomes undetectable again (Fig. 2,A). This finding holds true in vivo, as demonstrated by adoptively transferring OT-1 (Thy1.1) CD8 T cells into LM-OVA-infected recipients. Similar to our in vitro finding, up-regulation of IL-10R1 was evident on activated OT-1 CD8 T cells by 24–48 h after infection followed by down-regulation at 96 h postinfection (Fig. 2 B). Our results confirm previous findings which demonstrated that IL-10R1 expression is regulated on activated T cells (27). These experimental results suggest that recently activated CD8 T cells might, for a transient period of time, be susceptible to IL-10-mediated suppression.
CD8 T cells lacking IL-10R2 undergo enhanced primary expansion in L. monocytogenes-infected mice
To directly address the role of IL-10R signaling in CD8 T cells responding to L. monocytogenes infection, we crossed OT-1 TCR-transgenic mice, specific for H2-Kb/SIINFEKL, to C57BL/6 mice with a genetic deletion of IL-10R2 (28). Naive IL-10R2−/− OT-1 CD8 T cells were similar to naive wild-type OT-1 CD8 T cells in terms of CD69, CD44, CD62L, and CD25 expression (data not shown). To determine whether the lack of IL-10R2 expression affects the proliferative capacity of OT-1 CD8 T cells, wild-type and IL-10R2−/− OT-1 CD8 T cells were CFSE labeled and cocultured with SIINFEKL-coated, T cell-depleted irradiated splenocytes. Three days after stimulation, T cells were analyzed for CFSE dilution by flow cytometry. As shown in Fig. 3 A, wild-type and IL-10R2−/− CD8 T cells proliferated similarly, suggesting that the inability to respond to IL-10 does not alter the in vitro capacity of OT-1 CD8 T cell populations to expand.
To evaluate the direct effect of IL-10-signaling on CD8 T cell responses to L. monocytogenes infection, IL-10R2−/− and wild-type OT-1 CD8 T cells were cotransferred (1:1 ratio) into congenic C57BL/6 mice that had been infected 24 h earlier with LM-OVA (Fig. 3,B). The relative expansion of adoptively cotransferred wild- type and IL-10R2−/− CD8 T cell populations in recipient mice was evaluated by flow cytometry. Interestingly, IL-10R2−/− OT-1 CD8 T cell populations underwent 1.5- to 2-fold greater expansion in comparison to wild-type OT-1 CD8 T cells 6 and 8 days following infection, and the ratio of these two T cell populations changed from 1:1 to nearly 2:1 (Fig. 3, C and D). Similar differences in the ratio between wild-type and IL-10R2−/− T cell populations were also evident in other tissues such as blood and liver (Fig. 3 E). Thus, the difference in expansion between IL-10R2−/− and wild-type CD8 T cells was not restricted to spleen. Similarly, IL-10R2−/− CD8 T cells undergo 1.6- to 2-fold greater primary expansion in comparison to wild-type controls when cotransferred in congenic recipients 24 h before infection (data not shown).
Although IL-10R2−/− OT-1 T cells underwent greater expansion when compared with wild-type OT-1 T cells, the lack of IL-10 signaling did not influence their effector functions. As shown in Fig. 3 F, IL-10R2−/− OT-1 CD8 T cells obtained from L. monocytogenes-infected mice produced IFN-γ and TNF-α at levels and frequencies similar to those of wild-type OT-1 T cells. This observation holds true even when the cells were transferred 24 h before infection (data not shown). Thus, IL-10-mediated signaling in L. monocytogenes-specific CD8 T cells does not detectably impact IFN-γ or TNF-α production.
To determine whether naive CD8 T cells lacking IL-10R2 survive better in recipients than wild-type CD8 T cells, both types of OT-1 T cells were transferred into C57BL/6 mice in a 1:1 ratio in the absence of infection. At indicated time points after transfer, spleens of recipient mice were analyzed for relative frequency of adoptively transferred IL-10R2−/− and wild-type OT-1 T cells. A marginal reduction in the number of IL-10R2−/− CD8 T cells in comparison to wild-type control was evident at 144-h after transfer onward (data not shown). Thus, the increased frequency of IL-10R2−/− OT-1 T cells seen following infection cannot be attributed to nonspecific enhanced survival of IL-10-unresponsive T cells.
Impact of IL-10-signaling on CD8 T cell contraction and memory maintenance
Wild-type and IL-10R2−/− OT-1 T cells underwent expansion for ∼8 days following transfer into infected recipient mice and then underwent contraction, giving rise to long-lasting memory cells (Fig. 4,A). The degree of contraction and the size of the resulting memory T cell population was similar for wild-type and IL-10R2−/− T cells and thus gave rise to IL-10R2−/− memory OT-1 T cell populations that were 1.5- to 2-fold larger in size than wild-type memory OT-1 T cell populations (Fig. 4 B). IL-10R2−/− and wild-type OT-1 memory T cells were stably maintained for 60 days following infection, suggesting that the effect of IL-10 on Ag-specific CD8 T cells is restricted to the priming phase of the T cell response. Consistent with this notion, IL-10R2−/− and wild-type OT-1 memory T cells express similar levels of CD122, CD127, and CD62L at days 14 and 60 after transfer (data not shown). Although our results suggest that IL-10-mediated effects on CD8 T cells responding to L. monocytogenes infection are restricted to the priming/expansion phase, it is unclear whether this reflects the narrow time period following infection that IL-10 is produced in vivo or the restricted expression of IL-10R2 on responding T cells.
IL-10R2−/− memory T cells undergo enhanced secondary expansion upon reinfection
To address the role of IL-10R2 signaling in responding CD8 T cells during secondary expansion, IL-10R2−/− and wild-type OT-1 T cells were cotransferred into congenic, LM-OVA- infected C57BL/6 mice and 60 days later mice were reinfected with 105 LM-OVA. Five days following rechallenge, spleens of recipient mice were analyzed for secondary expansion of adoptively transferred CD8 T cells by flow cytometry. As shown in Fig. 5, A and B, IL-10R2−/− OT-1 T cells underwent greater expansion than wild-type OT-1 T cells, achieving population sizes that exceeded wild-type populations by a factor of 3–4. Consistent with our findings during the primary T cell response, wild-type and IL-10R2−/− OT-1 T cells expressed IFN-γ or TNF-α to similar extents (Fig. 5 C).
Enhanced expansion of IL-10R2−/− T cells is attributable to IL-10
Because IL-10R2 can also signal in response to IL-22 (2), it was possible that diminished expansion of wild-type T cells might result from IL-22-mediated signals. In this case, greater expansion of IL-10R2−/− than wild-type T cells would be maintained in IL-10−/− recipients. In contrast, if IL-10-mediated signals account for the difference, then expansion of IL-10R2−/− and wild-type T cells should be similar in IL-10−/− recipient mice. We cotransferred IL-10R2−/− and wild-type OT-1 T cells at a 1:1 ratio into LM-OVA-infected wild-type or IL-10−/− recipient mice and determined the ratio of responding T cells at the indicated time points by flow cytometry. As previously demonstrated, in wild-type C57BL/6 recipients, IL-10R2−/− T cells underwent greater expansion during primary and memory responses. In contrast, in IL-10−/− recipients, IL-10R2−/− and wild-type OT-1 T cells underwent similar primary and memory expansion, thus maintaining the 1:1 ratio of the initial T cell inoculum (Fig. 6). This result indicates that IL-10 produced following L. monocytogenes infection is responsible for the diminished expansion of wild-type compared with IL-10R2−/− T cells.
Effect of IL-10R2 deficiency is not restricted to OT-1 T cells
To determine whether our finding that IL-10R2−/− OT-1 T cells undergo enhanced expansion can be generalized to polyclonal T cell populations, we adoptively transferred purified CD8 T cells from IL-10R−/− and B6.PL mice at a 1:1 ratio into Rag1−/− recipient mice 24 h following infection with LM-OVA. Primary expansion of adoptively transferred CD8 T cells was analyzed 8 days after infection using H2-Kb SIINFEKL tetramer staining. On day 28 following primary infection, recipient mice were rechallenged with LM-OVA and secondary expansion of SIINFEKL-specific T cells was measured 5 days later. As shown in Fig. 7,A, while 5–6% of wild-type CD8 T cells were SIINFEKL specific 8 days following primary infection, ∼10–11% of IL-10R2−/− CD8 T cells were SIINFEKL specific. The difference in the proportion of SIINFEKL-specific T cells was also reflected in the absolute number of SIINFEKL-specific T cells that accumulated in the spleen (Fig. 7,C). Following rechallenge, polyclonal CD8 T cells lacking IL-10R2 also underwent greater secondary expansion, achieving SIINFEKL-specific CD8 T cell frequencies of 37%, while wild-type CD8 T cells only reached frequencies of ∼21% (Fig. 7, B and D). These results indicate that IL-10-mediated inhibition of CD8 T cell populations is not restricted to monoclonal OT-1 T cells but also holds true for polyclonal Ag-specific T cell populations.
Absence of IL-10R signaling in DCs does not increase the magnitude of L. monocytogenes-specific CD8 T cell responses
To determine whether IL-10R2 signaling in DCs affects CD8 T cell activation, expansion, and differentiation, LPS matured, bone marrow-derived DCs from IL-10R2−/− and wild-type mice were coated with SIINFEKL and adoptively transferred into C57BL/6 mice that had received IL-10R2−/− OT-1 T cells and been infected with wild-type L. monocytogenes (i.e., not expressing OVA). In this experimental scheme, wild-type L. monocytogenes will induce IL-10 without priming OT-1 T cells, while SIINFEKL-coated IL-10R2−/− or wild-type DCs will serve as the exclusive APC (Fig. 8,A). IL-10R2−/− and wild-type DC expressed similar levels of MHC class II, CD80, and CD86 before immunization (data not shown). In vitro priming of OT-1 T cells by wild-type and IL-10R2−/− DCs was equivalent, as revealed by CFSE proliferation (data not shown). As shown in Fig. 8,B, 4 days following adoptive transfer, CD8 T cells primed by either IL-10R2−/− or wild-type DCs expressed comparable levels of CD62L and CD44 and had undergone similar extents of division, as measured by CFSE dilution. Along similar lines, the absolute number of OT-1 T cells per spleen was similar in mice immunized with IL-10R2−/− or wild-type DCs (Fig. 8,C) as was the proportion expressing IFN-γ following peptide restimulation (Fig. 8 D). Overall, these results suggest that IL-10R2 signaling in in vitro-matured, peptide-coated DCs has no effect on proliferation, activation, and effector differentiation of adoptively transferred CD8 T cells.
Previous studies have demonstrated that IL-10 inhibits maturation and down-regulates costimulatory molecule expression on DCs (10). Thus, in vitro activation of DCs, as performed in the preceding experiment, might bypass in vivo inhibitory effects of IL-10 on APC function and T cell priming. To address this possibility, we transferred IL-10R2−/− OT-1 T cells into either IL-10R2−/− or wild-type mice that had been infected with 105 ActA-deficient LM-OVA. We chose the higher dose of ActA-deficient LM-OVA to eliminate the marked disparities in Ag load that occur between IL-10R2−/− and wild-type mice infected with a sublethal inoculum of wild-type L. monocytogenes. In addition, since in vivo levels of IL-10 might differ between wild-type and IL-10R2−/− recipient mice, we adoptively transferred OT-1 T cells deficient in IL-10R2. As shown in Fig. 8,E, 8 days following transfer there was no significant difference in the total number of adoptively transferred OT-1 T cells in the spleens of IL-10R2−/− and wild-type recipient mice. In addition, the percentage of transferred OT-1 T cells producing IFN-γ was similar in IL-10R2−/− and wild-type recipient mice (Fig. 8 F). These results indicate that IL-10R2 signaling in APCs does not diminish the primary expansion of CD8 T cells following L. monocytogenes infection.
Although IL-10 is recognized as an important immunoregulatory cytokine, it has remained unclear which cells are inhibited by this cytokine in the context of a microbial infection. Our initial experiments demonstrated more robust L. monocytogenes-specific CD8 T cell responses in IL-10−/− mice, suggesting either enhanced APC activation and Ag presentation in the absence of IL-10 or direct inhibition of CD8 T cell expansion by IL-10 in wild-type mice. To distinguish between these two models, we used chimeric mice in which either Ag-specific T cells or APCs lacked IL-10R2 and thus were selectively unresponsive to IL-10. Our experiments demonstrate that IL-10 inhibits CD8 T cell responses by restricting T cell expansion during primary and memory responses to L. monocytogenes infection. In contrast, the inability of APCs to respond to IL-10 did not significantly impact the magnitude or kinetics of CD8 T cell responses to infection. Thus, our study provides the first in vivo demonstration that IL-10 directly modulates CD8 T cell responses to bacterial infection.
The effect of IL-10 on defense against L. monocytogenes infection is not limited to restraining CD8 T cells responses. The first study to characterize L. monocytogenes infection in IL-10-deficient mice demonstrated that innate immune defense is more effective in the absence of this cytokine, with roughly 10- to 100-fold enhanced bacterial clearance during the first 24–48 h of infection. This same study also demonstrated enhanced T cell responses to L. monocytogenes infection in the absence of IL-10, suggesting that IL-10 inhibited T cell priming (22). Direct comparisons between T cell responses in wild-type and IL-10−/− mice are difficult, however, because disparities in bacterial survival were already occurring during the T cell-priming phase. A recent study by Seder and colleagues (25) attempted to circumvent this issue by infecting wild-type and IL-10−/− mice with high doses of L. monocytogenes and then treating them with antibiotics. However, because antibiotic administration was delayed for 24 h, concern about enhanced bacterial clearance in IL-10−/− mice before antibiotic administration persists. To directly address the effect of IL-10 on CD8 T cell responses, therefore, we transferred IL-10-responsive and IL-10-unresponsive CD8 T cells of defined specificity into the same mouse and compared their expansion, contraction, memory formation, and re-expansion upon L. monocytogenes infection. In contrast to previous studies, therefore, concerns about disparities in the innate immune response or the rate of bacterial clearance could be set aside. Our studies clearly demonstrated that IL-10 mediates a direct effect on CD8 T cells responding to L. monocytogenes infection. Confirming recently published findings from Seder and colleagues (25), we also see expression of IL-10R1 on recently activated CD8 T cells, albeit in our experiments expression was temporally restricted and limited to 24–48 h after TCR stimulation.
Although our studies demonstrate that endogenous IL-10 production during L. monocytogenes infection limits the expansion of pathogen-specific CD8 T cells, many questions remain. It is unclear, for example whether CD8 T cells undergo more rapid proliferation in the absence of IL-10, proliferate for a longer period of time (e.g., one extra round of division could explain the difference we see), or perhaps are less prone to apoptosis during the course of proliferation. Although in vitro studies suggest that IL-10 decreases expression of annexin V on activated CD8 T cells, it is difficult to extrapolate from in vitro studies to in vivo responses to microbial infection (11). Exposure of CD8 T cells to IL-10 during priming limits their response to Ag upon re-exposure, an effect that was attributed to diminished IL-2 production and proliferation (6). It is possible, therefore, that IL-10 produced during early stages of L. monocytogenes infection limits proliferation of CD8 T cells by a similar mechanism.
The results of in vitro cytokine studies are often difficult to reconcile with in vivo experimental findings. For example, although type I IFN and IFN-γ can inhibit CD4 and CD8 T cell proliferation in vitro, recent studies demonstrate IFN-mediated enhancement of CD4 and CD8 T cell expansion in LCMV-infected mice (29, 30, 31, 32). In these circumstances, a number of scenarios, including cytokine doses, exposure times, and the contribution of third-party cells or factors, might explain disparities between in vitro and in vivo findings. We believe, therefore, that these considerations may also provide an explanation for our findings that IL-10 inhibits CD8 T cell responses in vivo, while others have recently demonstrated that IL-10 enhances CD8 T cell proliferation in vitro (11). Other factors, such as the dosage of L. monocytogenes used to immunize mice, might also explain disparities between our results showing enhanced T cell memory in the absence of IL-10, and recent results demonstrating enhanced memory in the presence of IL-10 (25).
Recent studies have demonstrated that IL-10 inhibits Ag-specific CD8 T cell responses in mice with chronic LCMV infection (33, 34). In this setting, CD8α-ve DCs were found to support IL-10 production during LCMV infection and in vivo blockade of IL-10 with a neutralizing Ab resulted in rapid enhancement of IFN-γ production by antiviral CD8 T cells and resolution of the persistent infection (33, 34). Similar observation have also been made in chronic hepatitis C infection (35). These previous studies did not determine whether the effect of IL-10 blockade was on CD8 T cells or on APCs. Our experiments, however, suggest that IL-10 may mediate its effect directly on virus-specific CD8 T cells. Although further studies will be required to determine how IL-10 restricts CD8 T cell responses, we believe that the effects of IL-10 can be exploited to either restrict CD8 T cell responses in the setting of autoimmunity or enhance CD8 T cell responses in defense against microbial pathogens.
We thank Pamer laboratory members for helpful discussions.
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 AI 39031 and AI 42135 (to E.G.P.).
Abbreviations used in this paper: Tg, transgenic; DC, dendritic cell; LCMV, lymphocytic choriomeningitis virus.