Increased IL-10 expression following infection with LCMVClone13 but not LCMVARM
Increased levels of inhibitory cytokines such as Interleukin 10 (IL-10) have been demonstrated during persistent viral infections in humans [25, 26]. Infection of mice with LCMVClone13 induces anergy in CD4 and CD8 T cells, whereas infection with LCMVARM leads to robust immunity [10, 27]. We were interested to see if persistent LCMV infection led to increased levels of serum IL-10. We infected cohorts of mice with either LCMVARM or LCMVClone13, collected their sera at various time points post-infection (p.i.) and measured the IL-10 levels by ELISA (Figure 1a). Other than an early spike of IL-10 detectable on day 2 post-infection, LCMVARM-infected mice had low levels of IL-10 in their serum. IL-10 levels in LCMVARM-infected mice increased slightly above naïve mice as immune memory developed beyond day 20. In contrast, LCMVClone13-infected mice accumulated IL-10 in their serum. Beyond day 2 post-infection, LCMVClone13-infected mice had significantly higher levels of serum IL-10 compared to acutely infected mice, and the levels increased with time. The steady increase in serum IL-10 was concomitant with the functional inactivation, i.e. loss of cytolytic activity and IFN-γ production, in CD8 T cells ([28] and Figure 3). These data correlate the T cell anergy observed in persistently LCMV infected mice with an increase in serum IL-10.
CD8 T cells are critical for controlling chronic LCMV infection [27]. In order for IL-10 to directly effect CD8 T cells, they must express the IL-10 receptor. We confirmed IL-10R expression on CD8 T cells during the course of acute and chronic infections. In our hands, IL-10R is universally high at day 8 post-infection for both acute (LCMVARM) and chronic (LCMVClone13) infections. By day 30, IL-10R expression has decreased on a subset of memory CD8 T cells in acutely infected mice. Persistently infected mice, however, lose IL-10R expression on the majority of CD8 T cells, consistent with the deletion of the majority of anti-viral CD8 T cells. We next assessed IL-10R expression directly on LCMV-specific CD8 T cells. We looked at splenic GP33-specific CD8 T cells eight days post-infection, a time during which chronically infected mice contain virus-specific T cells with slightly impaired effector functions. GP33-specific CD8 T cells in both acutely and chronically infected mice had higher IL-10R expression when compared to naïve CD8 T cells. These data are consistent with the model that IL-10 directly affects CD8 T cells.
IL-10 blockade enhances IFN-γ production in persistently infected mice 8 days post-infection
We next wanted to determine if neutralizing IL-10 would have an effect on the developing T cell response during a chronic viral infection. We injected LCMVClone13-infected mice with anti-IL-10 antibodies or normal rat IgG on days 0, 2, and 4 after infection. Eight days following infection, CD8+ and CD4+ T cell responses were measured by peptide-induced IFN-γ production and cytolytic killing of peptide-pulsed target cells. LCMVClone13-infected animals exhibited depressed CD8 T cell responses: the frequency of cells responding to the immunodominant epitopes NP396 and GP33 was reduced by 60–80% when compared to LCMVARM-infected mice (Figure 2a). In contrast, mice that received IL-10 blockade exhibited enhanced T cell responses. The frequencies of NP396 and GP33 tetramer+ CD8 T cells were comparable to those observed in LCMVARM-infected mice (data not shown). The amount of IFN-γ production per cell, as measured by mean fluorescence intensity (MFI) was higher in anti-IL-10 treated mice than in LCMVClone13-infected, rat IgG treated controls for both LCMV peptides, GP33 (350 ± 12.5 vs. 248.5 ± 39.5 for anti-IL-10 treated vs. rat IgG treated) and NP396 (204 ± 30 vs. 116.7 ± 12.2).
LCMVClone13-infected mice receiving anti-IL-10 therapy exhibited a level of IFN-γ production that was indistinguishable from LCMVARM-infected mice (Figure 2a). Mice receiving anti-IL-10 therapy had larger spleens and significantly more (5–7 fold) virus-specific CD8+ T lymphocytes, relative to untreated LCMVClone13-infected control mice (Figure 2b).
CD4 T cells are essential for maintaining antiviral CD8 T cells [27, 29]. Therefore, we monitored the effect of anti-IL-10 treatment on rescuing virus-specific CD4 T cell function by testing their ability to produce IFN-γ upon stimulation with the LCMV MHC class II peptides GP61 and NP309. Anti-IL-10-treated mice exhibited higher levels of IFN-γ production by CD4+ T cells at 8 days p.i. (5.5% ± 1.2 vs. 1.5% ± 0.7 for anti-IL-10 treated vs. untreated LCMVClone13-infected mice, p = 0.033, Figure 2a). The absolute numbers of LCMV-specific CD4 T cells were also higher in the anti-IL-10 treated LCMVClone13-infected group as compared to LCMVClone13-infected mice that did not receive anti-IL-10 treatment (Figure 2b). Thus, neutralizing IL-10 during the time that T cell priming occurs drastically enhanced peak antigen-specific CD4 and CD8 T cell responses, both in absolute numbers and in function as measured by IFN-γ production.
IL-10 blockade restores CTL activity and lowers viral load
Cytolytic killing is a hallmark of activated antigen-specific CD8 T cells, and it is one of the first properties to be lost in persistently infected mice [10, 27]. We wanted to know if IL-10 blockade restored the cytolytic activity of CD8 T cells. We measured the cytolytic activity of CD8 T cells by 51Cr release assay using target cells pulsed with the immunodominant LCMV peptides NP396 and GP33 (Figure 2c). Killing of target cells was dramatically impaired in LCMVClone13-infected mice at 8 days p.i., but neutralizing IL-10 in vivo restored cytotoxic activity nearly to levels observed in LCMVARM-infected mice (Figure 2c). We then tested the mice to determine whether enhanced T cell responses induced by neutralizing IL-10 led to a reduction in viral load. Despite the improved CTL and cytokine responses in mice receiving IL-10 blockade (Figure 2a–c), they still had infectious virus (Figure 2d). However, the level of viremia was significantly reduced (> 1 log) in mice receiving IL-10 blockade (1.6 × 105 ± 3.9 × 104 pfu/ml, n = 5, for LCMVClone13 infections vs. 2.3 × 104 ± 6.7 × 103 pfu/ml, n = 5, for LCMVClone13 plus anti-IL-10, p = 0.0076). The lowered viral titers observed at day 8 p.i. could be due to enhanced cytolytic T cell activity or due to an enhanced antibody response to LCMV in anti-IL-10 treated mice, or a collaboration of the two effects. To distinguish between these two possibilities we compared the anti-LCMV antibody responses at day 8 post-infection in groups of LCMVClone13-infected mice with or without anti-IL-10 treatment. Anti-IL-10 blockade did not result in enhanced virus-specific humoral responses (Figure 2e). LCMV-specific IgG levels in the IL-10-blocked, persistently infected mice were not significantly different than untreated controls at day 8 post infection (p = 0.2). The data suggest that enhanced cytotoxic CD8 T cell responses and not humoral responses are responsible for the lowered viral titers observed in the anti-IL-10 treated mice early in the response. These data further support the concept that host-produced IL-10 directly impacts LCMV-specific CD8 T cells early during the immune response.
Despite the initial enhanced response, mice receiving a short course of IL-10 blockade contain a mixture of functional and anergized CD8 T cells, and exhibit a low-level persistent viremia
In LCMVClone13-infected mice receiving IL-10 blockade, IFN-γ production and cytolytic killing by CD8 T cells were intact, and CD4 T cell responses were also enhanced at day 8 p.i. (Figure 2a). It was essential to determine if this potent antiviral immune response could develop into immune memory and resolve the viral infection. We analyzed CD8 T cells from immune LCMVClone13-infected mice (day 30) with and without IL-10 blockade for LCMV-specific MHC/tetramer binding (Fig. 3a) and IFN-γ production (Fig. 3b). In untreated LCMVClone13-infected mice, we observed large numbers of GP33 tetramer+ CD8 T cells that failed to produce IFN-γ, a characteristic phenotype of anergized virus-specific T cells [10]. CD8 T cells specific for the NP396 epitope were mostly deleted and the rare NP396 tetramer binding cells that were present failed to produce IFN-γ upon peptide stimulation. In contrast, mice receiving IL-10 blockade had a small fraction of CD8 T cells that produced IFN-γ when stimulated with GP33 peptide (0.5% ± 0.2 vs. 0.1% ± 0.1%, Fig. 3b), indicating that, at least on some level, functional CD8 T cells were present in persistently infected mice that received IL-10 blockade. Persistently infected, anti-IL-10 treated mice deleted NP396-specific CD8 T cells, similar to LCMVClone13-infected, rat IgG-treated control mice. Taken together, the tetramer and IFN-γ data for two immunodominant epitopes suggest that the generation of virus-specific memory cells was impaired although some GP33-specific CD8 T cells that retain IFN-γ production persisted in mice that received IL-10 blockade (Figure 3b). We also monitored CD4 T cell responsiveness at 30 days p.i. A small population of GP61/NP309-specific CD4+ T cells remained in IL-10 blocked mice, although the MFI of IFN-γ production was low (Figure 3b).
Anti-IL-10 treated, LCMVClone13-infected mice remained persistently infected at 30 days p.i. (Figure 3c). However, viremia was approximately 1.5 logs lower in LCMVClone13-infected mice that received anti-IL-10 therapy (2.4 × 103 ± 6.6 × 102 pfu/ml, n = 5, for LCMVClone13 plus anti-IL-10 vs. 4.2 × 104 ± 1.1 × 104 pfu/ml, n = 5, for LCMVClone13).
IL-10 blockade in mice with established persistent infections did not result in enhanced T cell responsiveness or lowered viral titers
Blocking IL-10 early in the infection during T cell priming (days 0–4) led to enhanced LCMV-specific T cell responses and lower viral titers. We wanted to know if blocking IL-10 was a useful strategy in lowering viral titers in mice with established persistent infections. We chose mice that were persistently infected with LCMVClone13 (day 30 p.i.) and injected them with normal rat IgG or anti-IL-10 antibodies on days 30, 32 and 34 post-infection. Two weeks following anti-IL-10 treatment we assayed these mice for CD4 and CD8 T cell function and measured serum viral titers. IL-10 blockade after establishment of viral persistence had no effect on LCMV-induced T cell tolerance. T cell anergy could not reversed (data not shown), and there was no change in viremia (Figure 3d). T cells from mice treated with anti-IL-10 late in the infection were non-functional and indistinguishable from untreated LCMVClone13-infected mice (data not shown). These data suggest that there is an early critical window during which IL-10 works to induce anergy in T cells.
Mice receiving early anti-IL-10 therapy exhibit CTL activity following in vitro re-stimulation
Next, we wanted to know if the anergy observed in anti-IL-10 treated, LCMVClone13-infected mice was complete and irreversible, or if the effector/memory T cells that remained could expand and function. Splenocytes from LCMVClone13-infected mice (day 30 p.i.) were stimulated in vitro with GP33 peptide, in the presence of IL-2, for 5 days and their ability to kill target cells was determined by 51Cr release assay (Figure 4a). Re-stimulation in vitro restored cytolytic activity of LCMVClone13-infected anti-IL-10 treated mice; their CTL activity was comparable to LCMVARM-immune mice. In contrast, GP33-specific CD8 T cells from LCMVClone13-infected, untreated mice were refractory to peptide re-stimulation and failed to kill target cells. These data clearly demonstrate that IL-10 blockade early in the viral infection led to incomplete tolerization of virus-specific CD8 T cells despite the fact that these mice had low-level viremia.
Mice receiving anti IL-10 therapy have an enhanced antibody response against LCMV 30 days post-infection
We assessed the antiviral antibody titers 30 days p.i. in LCMVClone13-infected mice with and without anti IL-10 treatment (Figure 4b). Interestingly, mice receiving IL-10 blockade had significantly greater levels of LCMV-specific antibodies than mice that did not receive anti-IL-10 therapy (O.D. 0.4201 ± 0.045, n = 8 vs. 0.1875 +/- 0.035, n = 6, p = 0.002). The level of antiviral IgG in LCMVARM immune mice was much greater than the LCMVClone13-infected mice that received anti-IL-10 blockade (1.151 ± 0.042, n = 10 versus 0.4201 ± 0.045, n = 8, p < 0.0001, not shown). These data, combined with the low antiviral antibody levels seen 8 days p.i., show that the LCMV-specific antibody response matured with time in the IL-10 blocked mice.
IL-10 knockout mice recapitulate the phenotype observed in anti-IL-10 antibody treated mice: early T cell enhancement followed by induction of anergy
The data presented so far clearly show that IL-10 blockade led to enhanced early T cell responses but not the eradication of virus. One possibility is that the T cell anergy observed at 30 days p.i. could be due to the inefficiency of the antibody treatment. Continued blockade of IL-10 throughout the course of persistent LCMVClone13 infection may result in complete viral clearance and the development of T cell immunity. To decisively determine the causal role of IL-10 in generating T cell anergy during persistent viral infections, and to avoid the vagaries of multiple injections of heterospecific neutralizing antibodies, we attempted to persistently infect IL-10 deficient mice. We first tested the ability of IL-10-/- mice to clear acute LCMVARM infections. Cohorts of IL-10-/- mice were infected with 2 × 105 pfu of LCMVARM and their ability to mount T cell responses and clear virus was assayed at 8 days post infection. IL-10-/- mice mounted robust CD8 T cell responses (Figure 5a) and cleared LCMVARM (Figure 5b) as efficiently as wild type mice. We then infected IL-10-/- mice with LCMVClone13 and examined their T cell responses and ability to clear virus at days 8 and 30 p.i. IL-10-/- mice exhibited exaggerated T cell responses at 8 days p.i., while wild type mice exhibited virus-induced immune suppression. Interestingly, heterozygous IL-10+/- littermate controls exhibited T cell responses intermediate between those of control IL-10+/+ and IL-10-/- mice (Figure 5c). However with time, IL-10-/- mice, as well as IL-10 heterozygotes, lost their virus-specific CD8+ T cells and became persistently infected by day 30 p.i. (Figures 5c, d). We also tracked the IFN-γ responses to three other LCMV epitopes, including subdominant epitopes, and the patterns were identical (data not shown); responses to all epitopes tested were absent 30 days p.i. in IL-10-/- mice. These data confirm and extend the observations made using anti-IL-10 antibody treatment and clearly demonstrate that IL-10 plays a direct and early role in generating T cell tolerance; however, additional mechanisms are operating to generate and/or maintain virus-induced anergy.