Differential role of IL-2R signaling for CD8+ T cell responses in acute and chronic viral infections

Authors


Abstract

IL-2 is a cytokine with multiple and even divergent functions; it has been described as a key cytokine for in vitro T cell proliferation but is also essential for down-regulating T cell responses by inducing activation-induced cell death as well as regulatory T cells. The in vivo analysis of IL-2 function in regulating specific T cell responses has been hampered by the fact that mice deficient in IL-2 or its receptors develop lymphoproliferative diseases and/or autoimmunity. Here we generated chimeric mice harboring both IL-2R-competent and IL-2R-deficient T cells and assessed CD8+ T cell induction, function and maintenance after acute or persistent viral infections. Induction and maintenance of CD8+ T cells were relatively independent of IL-2R signaling during acute/resolved viral infection. In marked contrast, IL-2 was crucial for secondary expansion of memory CD8+ T cells and for the maintenance of virus-specific CD8+ T cells during persistent viral infections. Thus, depending on the chronicity of antigen exposure, IL-2R signaling is either essential or largely dispensable for induction and maintenance of virus-specific CD8+ T cell responses.

Abbreviations:
AICD:

activation-induced cell death

CpG:

oligonucleotide containing stimulatory CpG motif

LCMV:

lymphocytic choriomeningitis virus

MCMV:

mouse cytomegalovirus

VLP:

virus-like particle

Introduction

Interleukin-2 (IL-2) is an important cytokine exhibiting multiple functions during T cell responses. IL-2 was initially described as a T cell growth factor, since it strongly enhances T cell proliferation in vitro1, 2. Furthermore, there is in vivo evidence that expansion of T cells during primary immune responses against viruses is reduced in the absence of IL-2 35, and IFN-γ production by T cells has been reported to depend on the presence of IL-2 5; in contrast, acute responses against superantigens were reported to be normal 6. Furthermore, IL-2 has been described to enhance survival of T cells in later phases of the immune response during the establishment of T cell memory. IL-2-deficient Th cells failed to efficiently survive upon adoptive transfer 7, 8, and injection of recombinant IL-2 during the contraction phase of the T cell response enhanced T cell survival 9. Conversely, IL-2 has been implicated in activation-induced cell death (AICD), a process proposed to be important for the down-regulation of immune responses 6, 10. A further function of IL-2 is the reversal of T cell anergy; T cells that have been stimulated by antigen in the absence of CD28-mediated costimulation are refractory to further antigenic stimulation and become anergic or unresponsive. In vitro, addition of IL-2 overcomes this anergy and allows T cell proliferation to occur normally 11. Furthermore, the generation of mice deficient in IL-2 or its receptor has highlighted the importance of IL-2 in regulatory T cell (Treg) development. Both IL-2- and IL-2R-deficient mice are prone to develop autoimmunity and may develop lymphoproliferative diseases 1217. Thus, IL-2 is implicated in promoting as well as inhibiting T cell responses 18.

The role of IL-2 in regulating T cell responses has been studied in mouse strains lacking IL-2 or the IL-2Rα or β chain. Depending on the antigen used, a more or less important role for IL-2 was observed 35. However, all of these experiments should be interpreted with some caution, since i) the immune systems of these mice are dysregulated, and ii) competition between T cells for antigen and growth factors is an important component that is not present in mice completely lacking IL-2 or its receptors. The later phases of the immune response have been particularly difficult to study due to the various forms of autoimmune disease that occur in these mouse strains 19. Several recent studies investigating the role of IL-2 in healthy hosts do not support an essential role of IL-2 in promoting the priming and differentiation of effector T cells 2022 but indicate a role for IL-2R signaling in secondary expansion of memory CD8+ T cells 23.

In order to avoid the above mentioned problems, we generated mixed bone chimeras that received both normal and IL-2Rα chain-deficient bone marrow. In these mice, a mixed T cell compartment developed, consisting of both IL-2Rα+/+ and IL-2Rα–/– T cells. Importantly, these chimeric mice survived without signs of autoimmunity, indicating that sufficient Treg developed to maintain healthy T cell homeostasis. These chimeric mice allowed us to study the immune response mounted by IL-2Rα–/– CD8+ T cells in the presence of competing wild-type T cells in a non-perturbed Treg environment. We observed robust, albeit reduced, T cell responses mounted by IL-2Rα–/– T cells upon infection with low doses of lymphocytic choriomeningitis virus (LCMV), which leads to acute/resolved infection. Contraction of CD8+ T cell populations occurred similarly in both T cell subsets upon elimination of the virus. However, the ability of memory T cells to expand upon re-exposure to antigen was completely dependent upon IL-2 signaling, corroborating similar results obtained in a comparable experimental setting 23. In contrast to acute/resolved infection, where induction and maintenance of virus-specific CD8+ T cells was largely independent of IL-2R signaling, IL-2R signaling was crucial for the maintenance of virus-specific CD8+ T cells during persistent viral infection, as IL-2Rα–/– cells, despite being induced during the acute phase, were readily deleted during persistent infection.

Results

Generation and characterization of bone marrow chimeras

In order to generate chimeric mice harboring equal frequencies of IL-2Rα–/– and IL-2Rα+/+ T cells, Ly5.1+ C57BL/6 recipient mice were irradiated (950 radγ) and reconstituted with a 1:4 mixture of bone marrow derived from Ly5.1+ C57BL/6 mice and Ly5.2+ IL-2Rα–/– mice (test chimeras). This allowed the distinction between Ly5.1+ IL-2Rα+/+ T cells and Ly5.2+ IL-2Rα–/– T cells. Control chimeras were generated using a 1:1 mixture between bone marrow from Ly5.1+ and Ly5.2+ C57BL/6 mice (control chimeras). Chimerism was assessed in the blood 8–10 wks after reconstitution. This approach consistently yielded comparable levels of Ly5.1+ and Ly5.2+ T cells in the test and control chimeras (not shown). There was no sign of abnormal activation of IL-2Rα–/– T cells and mice appeared healthy. Moreover, cell counts were similar for test and control chimeras (not shown).

Anti-viral T cell responses in the absence of IL-2 signaling

Slightly reduced primary T cell responses have been reported for IL-2-deficient mice upon infection with LCMV 35. In order to assess anti-viral T cell responses in the absence of IL-2 signaling but in the presence of competing normal T cells and Treg, chimeric mice were infected with LCMV, and expansion of total CD8+ T cells was analyzed. While percentages of total IL-2Rα+/+ CD8+ T cells increased 7- to 8-fold in both test and control chimeras 8 days after infection, percentages of total IL-2Rα–/– CD8+ T cells increased only 2-fold (not shown). Expansion of LCMV gp33–41 (gp33)-specific CD8+ T cells was assessed in blood, spleen, liver and lung by tetramer staining. Frequencies of specific T cells were comparable within the IL-2Rα–/– CD8+ T cell compartment and the control CD8+ T cell compartment: roughly 13% of Ly5.2+ (IL-2Rα–/–) as well as of wild-type Ly5.1+ CD8+ T cells were specific for the LCMV-derived epitope gp33 (Fig. 1A). However, when the percentage of gp33-specific T cells within the total lymphocyte compartment was assessed in test chimeras, IL-2Rα–/– T cells were roughly 5-fold reduced compared to IL-2Rα+/+ T cells (Fig. 1A). Similarly, if absolute numbers of specific T cells per spleen were calculated, about 5-fold reduced numbers were found for IL-2Rα–/– T cells (Fig. 1A). Similar results were obtained for LCMV np396–404 (np396)-specific CD8+ T cells (not shown). Thus, IL-2Rα–/– LCMV-specific CD8+ T cells are efficiently primed but expand to only about 5-fold lower numbers than control T cells in a competing environment in chimeric mice.

Figure 1.

(A) LCMV-specific CD8+ T cell analysis 8 days after infection. Test (T) and control (C) chimeras were infected with 200 pfu LCMV WE, and gp33-specific CD8+ T cell frequencies were assessed by tetramer staining in blood, spleen, liver and lung. The upper row shows the percentage of tetramer (tet)+ cells among Ly5.1+CD8+ or Ly5.2+CD8+ T cells, and the middle row shows the percentage of tet+Ly5.1+CD8+ or tet+Ly5.2+CD8+ cells among total lymphocytes. The lower row shows total numbers of gp33-specific CD8+ T cells in spleen, liver and lung. Averages of three to six mice per group are shown from one of three independent experiments. Frequencies (among lymphocytes) and numbers of specific cells were corrected for chimerism. (B) Effector functions of LCMV-specific CD8+ T cells. Test (T) and control (C) chimeras were infected with 200 pfu LCMV, and 8 days later specific IFN-γ and TNF-α production was assessed by restimulation of spleen cells or liver lymphocytes with gp33 peptide, followed by intracellular cytokine staining. The percentage of IFN-γ- or TNF-α-producing cells among tet+ cells is shown. Averages of three mice per group are shown from one of two independent experiments (*p<0.05, **p<0.001).

IL-2 signaling has been reported to be important for homing of specific T cells to peripheral tissues 20, 21. We found, however, no difference between the ratios of IL-2Rα–/– and IL-2Rα+/+ LCMV-specific CD8+ T cells in spleen, liver and lung, indicating that tissue extravasation was not affected in the absence of IL-2 signaling during LCMV infection (Fig. 1A).

Next, effector functions of gp33- and np396-specific CD8+ T cells were assessed 8 days after infection by intracellular staining of spleen and liver lymphocytes for IFN-γ and TNF-α. No difference between IL-2Rα–/– and control T cells could be observed on a cellular level, indicating that IL-2 was not essential for differentiation of effector cells (Fig. 1B).

Contraction of specific T cells occurs normally in the absence of IL-2 signaling

LCMV-specific T cell responses peak around day 8–15 and subsequently decline. Since IL-2 has been implied in AICD 6, 10 as well as in enhanced T cell survival 7, 9, we assessed longitudinally the frequencies of gp33-specific CD8+ T cells after LCMV infection in test and control chimeras (Fig. 2A). No marked difference in the slope of decline could be observed between IL-2Rα–/– T cells and control T cells; gp33-specific CD8+ T cell frequencies declined by roughly 50% between day 15 and day 32, irrespective of whether they were IL-2Rα–/– or IL-2Rα+/+. These results indicate that the presence or absence of IL-2 is not critical for elimination of anti-viral T cells after riddance of the virus. We investigated the phenotype of LCMV-specific CD8+ T cells 32 days after infection, and almost 50% of the IL-2Rα–/– cells exhibited a central memory phenotype compared to only 30% of the wild-type cells (Fig. 2B), as defined by CD62L and CD127 expression 24, 25.

Figure 2.

(A) Expansion and contraction of LCMV gp33-specific CD8+ T cells. Test and control chimeras were infected with 200 pfu LCMV, and gp33-specific CD8+ T cell frequencies were longitudinally assessed by tetramer staining in blood. Frequencies of specific cells among lymphocytes were corrected for chimerism. (B) Phenotype of gp33-specific CD8+ T cells. IL-2Rα–/– and IL-2Rα+/+ CD8+ T cells were stained for CD62L and IL-7Rα expression in test (upper four dot plots) and control (lower dot plots) chimeras 32 days after LCMV infection.

Source of IL-2

Since IL-2Rα-competent CD8+ T cells expanded to 5-fold higher numbers than IL-2Rα-deficienT cells after infection with LCMV, we investigated whether CD4+ T cells were the relevant source of IL-2 in this setting. To this end, we depleted CD4+ T cells from test chimeras prior to LCMV infection and compared the longitudinal expansion and contraction of gp33-specific IL-2Rα+/+ and IL-2Rα–/– CD8+ T cells (Fig. 3A). Although expansion of IL-2Rα+/+ CD8+ T cells was reduced by about 25% in CD4-depleted test chimeras, it was still clearly higher than the expansion of IL-2Rα–/– CD8+ T cells, indicating that CD4+ T cells were not the only relevant source of IL-2 during LCMV infection. We confirmed that LCMV-specific T cells, and in particular CD4+ T cells, are able to produce IL-2 at early (day 7) and late (day 60) time points of LCMV infection (Fig. 3B).

Figure 3.

(A) Effect of CD4 depletion on primary expansion of LCMV-specific CD8+ T cells. Test chimeras were either depleted of CD4+ T cells (anti-CD4, lower graph) or left untreated (upper graph) before infection with 200 pfu LCMV WE, and the percentage of tetramer+Ly5.1+CD8+ or tetramer+Ly5.2+CD8+ cells among total lymphocytes is shown. Averages of six mice per group are shown. Frequencies of specific cells among lymphocytes were corrected for chimerism. (B) IL-2 and IFN-γ production by LCMV-specific CD4+ and CD8+ T cells. C57BL/6 mice were infected with 200 pfu LCMV WE, and 7 or 60 days later spleen cells were restimulated with the H-2Db-restricted gp33–41 (gp33) or the I-Ab-restricted gp61–80 (gp61) peptide, followed by intracellular assessment of IL-2 and IFN-γ production. Left graphs are gated on CD8+ T cells; right graphs are gated on CD4+ T cells.

IL-2 signaling is essential for secondary T cell responses

The ability of IL-2Rα+/+ or IL-2Rα–/– LCMV-specific T cells to perform in vivo recall responses was tested next. For primary immunization we used gp33-modified replication-incompetent virus-like particles (VLP). We have previously shown that gp33 coupled to VLP derived from the bacteriophage Qβ and loaded with oligonucleotides containing a stimulatory CpG motif (CpG) efficiently induces specific CD8+ T cell responses 26. Furthermore, induction of these CD8+ T cells was Th cell-independent 27, 28. Test and control chimeras were primed with gp33-VLP and were challenged 30 days later with recombinant vaccinia virus expressing the LCMV GP (VVG2) or with 104 pfu LCMV. Control CD8+ T cells vigorously expanded in gp33-VLP-primed mice, while no secondary expansion of IL-2Rα–/– T cells could be observed (Fig. 4). Since 1× gp33-VLP immunization only inefficiently primed IL-2Rα–/– CD8+ T cells, we changed the priming regimen and immunized mice every second day with gp33-VLP over a period of 8 days; this induced slightly increased primary expansion of gp33-specific IL-2Rα–/– CD8+ T cells, but secondary expansion of these cells was still severely impaired after challenge with 104 pfu LCMV.

Figure 4.

Primary and secondary expansion of gp33-specific CD8+ T cells. Test and control chimeras were immunized once s.c. with 150 μg gp33-VLP (1×) or every second day with 10, 50, 100, 100 and 50 μg gp33-VLP (5×). Mice were challenged 30 days later with 5 × 106 pfu recombinant vaccinia virus expressing the LCMV glycoprotein (VVG2) (indicated by arrows) or with 104 pfu LCMV WE (WE4). gp33-specific CD8+ T cells were measured 5 or 7 days after challenge. The percentage of tetramer+Ly5.1+CD8+ or tetramer+Ly5.2+CD8+ cells among total lymphocytes is shown. Averages of three to four mice per group are shown from one of two independent experiments. Frequencies of specific cells among lymphocytes were corrected for chimerism.

Since in general gp33-VLP priming only inefficiently primed IL-2Rα–/– CD8+ T cells, we also analyzed the recall proliferative potential of LCMV-primed IL-2Rα–/– CD8+ T cells. Because LCMV challenge in LCMV-primed mice leads to rapid control of the secondary infection and thus to only minimal secondary expansion of LCMV-specific CD8+ T cells, we chose to analyze secondary expansion after adoptive transfer of LCMV-primed cells into naive recipients. Chimeric mice were infected with LCMV, Ly5.1+ IL-2Rα+/+ and Ly5.2+ IL-2Rα–/– CD8+ T cells were FACS-sorted 45 days later, and 5000 gp33-specific CD8+ T cells were adoptively transferred into naive Ly5.1+ or Ly5.2+ C57BL/6 recipient mice. Recipient mice were challenged with LCMV, and expansion of the transferred memory T cells was analyzed 8 days later (Fig. 5A). While primed LCMV gp33-specific IL-2Rα+/+ CD8+ T cells dramatically expanded within 8 days after challenge, IL-2Rα–/– cells almost completely failed to do so (Fig. 5B); there was a 30- to 40-fold difference between the frequencies of IL-2Rα+/+ and IL-2Rα–/– LCMV-specific CD8+ T cells, a ratio that is significantly higher than the 5-fold difference observed after primary infection. Thus, IL-2 signaling plays a crucial role in secondary CD8+ T cell expansion upon LCMV infection, while IL-2 is partly dispensable for primary CD8+ T cell responses.

Figure 5.

Secondary expansion of gp33-specific CD8+ T cells upon adoptive transfer and LCMV challenge. (A) Test chimeras were infected with 200 pfu LCMV, and Ly5.1+ or Ly5.2+ CD8+ T cells were FACS-sorted 45 days after infection; 5000 gp33+ CD8+ T cells were adoptively transferred into naive Ly5.1+ C57BL/6 or Ly5.2+ C57BL/6 mice. One day after transfer, mice were challenged with 200 pfu LCMV, and expansion of adoptively transferred gp33-specific CD8+ T cells was assessed 8 days later in blood and spleen. (B) The percentages of gp33-specific CD8+ T cells among total lymphocytes are shown. Averages of three to four mice are shown from one of two independent experiments (*p<0.05).

Role of IL-2 signaling during persistent LCMV infection

IL-2 signaling is largely dispensable for the primary expansion and contraction of LCMV-specific CD8+ T cells after acute/resolved infection. The situation might, however, be different in the continuous presence of antigen, such as during chronic LCMV infection. One could speculate that under such conditions, IL-2-augmented AICD might be relevant, and, consequently, IL-2Rα–/– CD8+ T cells might have a survival advantage 6, 10. Alternatively, IL-2 signaling might be crucial for the survival and turnover of virus-specific CD8+ T cells in a setting of constant antigen exposure 7, 9. To address this important question, we infected chimeras with 106 pfu LCMV strain Docile, which leads to chronic high-level LCMV infection (Fig. 7B). While there was no apparent difference between the expansion of IL-2Rα–/– and IL-2Rα+/+ gp33-specific CD8+ T cells 8 days after infection, IL-2Rα–/– gp33-specific CD8+ T cells declined dramatically, to almost below the detection limit, by day 15 and day 30 after infection (Fig. 6A and 7A). In sharp contrast, IL-2Rα+/+ gp33-specific CD8+ T cells were maintained at constant levels after day 15 of infection. Thus, during chronic high-level LCMV infection, IL-2 signaling is crucial for the physical maintenance of gp33-specific CD8+ T cells. However, despite their physical maintenance, the function of these cells was severely impaired with respect to IFN-γ production but not degranulation (Fig. 6B). Dysfunction of virus-specific CD8+ T cells with respect to cytokine production but to a lesser extent with respect to degranulation seems to be common during high-level chronic viral infection (2937 and Agnellini et al., unpublished observations).

Figure 7.

(A) LCMV gp33-specific CD8+ T cells after chronic LCMV Docile infection. Averages of longitudinal stainings of three to four mice per group are shown (from one of two independent experiments). Frequencies of specific cells among lymphocytes were corrected for chimerism. (B) LCMV titers in spleens of test (open circles) and control (filled circles) chimeras at the indicated time points after infection with 106 pfu LCMV Docile. (C) CD25 expression on LCMV-specific CD8+ T cells. Naive Ly5.1+ TCR-transgenic CD8+ T cells (104) specific for the LCMV gp33 epitope were adoptively transferred into naive Ly5.2+ recipients. One day after transfer, mice were infected with 200 pfu (low dose) or 106 pfu (high dose) LCMV Docile. Spleen cells were isolated 22 days later and incubated in vitro for 2 days in the presence (filled grey area) or absence (solid line) of gp33 peptide, followed by staining for CD8, Ly5.1 and CD25. Histograms are gated on CD8+ Ly5.1+ TCR-transgenic cells. Numbers indicate the MFI of the cell populations. One representative staining of three mice and two independent experiments is shown.

Figure 6.

(A) LCMV gp33-specific CD8+ T cells after chronic LCMV Docile infection. Test and control chimeras were infected with 106 pfu LCMV Docile, and gp33-specific CD8+ T cell frequencies were assessed by tetramer staining in blood. Dot plots of representative mice at day 8 and day 30 after infection are shown, with gating on CD8+Ly5.1+ or CD8+Ly5.2+ cells; numbers indicate the percentage of tetramer+ cells (upper right quadrant). (B) Functional capacity of gp33-specific CD8+ T cells at day 30 after LCMV Docile infection. Spleen cells of test chimeras were restimulated in vitro with gp33 peptide and assessed for degranulation (CD107) and IFN-γ production. Dot plots are gated on CD8+ T cells, and numbers indicate the percentage of cells in the right quadrants.

If physical maintenance of LCMV-specific CD8+ T cells during persistent infection is dependent on IL-2Rα signaling, it would imply that LCMV-specific CD8+ T cells are able to express IL-2Rα. To test this assumption, we assessed IL-2Rα expression upon stimulation of LCMV-specific CD8+ T cells from chronically infected mice (Fig. 7C). Indeed, LCMV-specific CD8+ T cells were able to up-regulate IL-2Rα expression upon gp33 stimulation, albeit to lower levels compared to memory CD8+ T cells originating from mice with resolved LCMV infection.

Role of IL-2 signaling during persistent MCMV infection

To address whether the requirement for IL-2 signaling is also operational during a low-level chronic infection, we assessed the dynamics of virus-specific CD8+ T cells upon infection with murine cytomegalovirus (MCMV). We focused our analysis on two different MCMV epitopes in the H-2b background: m45 and m38 38. Both are immunodominant epitopes, and whereas m45-specific CD8+ T cells expand and contract to constant memory levels, m38-specific CD8+ T cells show a continued accumulation over time 39. It is believed that this “CD8+ T cell inflation” is caused by recurrent antigen encounter 3942. To investigate the IL-2 dependence of m45- and m38-specific CD8+ T cell dynamics, test and control chimeras were infected with 106 pfu MCMV, and frequencies of specific cells were analyzed over a period of 28 days (Fig. 8). Interestingly, the IL-2 dependence of the m45- and m38-specific CD8+ T cells differed substantially; while m45-specific CD8+ T cells followed comparable dynamics irrespective of IL-2Rα deficiency, maintenance and accumulation of m38-specific CD8+ T cells was largely IL-2R-dependent after day 6 of infection. Thus, IL-2Rα-deficient m38-specific CD8+ T cells behaved comparably to LCMV gp33-specific CD8+ T cells during chronic infection, whereas IL-2Rα-deficient m45-specific CD8+ T cells behaved comparably to LCMV gp33-specific CD8+ T cells during resolved LCMV infection. These data further support the notion that IL-2R signaling is essential for the maintenance of T cell memory during chronic antigenic exposure.

Figure 8.

Expansion and contraction of MCMV m45- and m38-specific CD8+ T cells. Test and control chimeras were infected with 106 pfu MCMV, and m45- and m38-specific CD8+ T cell frequencies were longitudinally assessed in blood. Averages of three mice per group are shown.

Discussion

We analyzed the importance of IL-2 signaling during primary and secondary T cell responses as well as during chronic viral infections. As an experimental system we used chimeric mice harboring both IL-2R-deficient and control T cells; this enabled us to specifically study the role of IL-2 in a healthy host with a normal Treg compartment and in a natural competitive environment in which IL-2Rα-deficient T cells have to compete with normal T cells for proliferative niches. Acute/resolved LCMV infection of such chimeric mice allowed us to establish that IL-2Rα–/– T cells efficiently expand after infection, confirming earlier results that anti-viral T cell responses can occur in the absence of IL-2. However, our results demonstrated a roughly 5-fold reduced expansion of LCMV-specific T cells, indicating that IL-2 may be more important during primary T cell responses in a competitive environment than suggested by earlier studies in IL-2-deficient mice. Furthermore, this roughly 5-fold reduction was apparent in both lymphoid and non-lymphoid tissues, suggesting that recirculation, homing and extravasation to peripheral tissues was not affected by absent IL-2 signaling. Furthermore, IL-2 signaling was not essential for the differentiation of effector cell function, since IL-2Rα–/– T cells present in secondary lymphoid organs and in peripheral tissue 8 days after infection were normal with respect to cytokine production.

Antigen-specific CD4+ T cells are a potentially relevant source of IL-2. We confirmed that LCMV-specific CD4+ T cells are capable of producing IL-2 during acute (day 7) and during memory (day 60) phases after low-dose LCMV infection, and we addressed whether depletion of CD4+ T cells during the acute phase of LCMV infection would also lead to a 5-fold reduction of IL-2Rα+ T cells as observed for the IL-2Rα T cells in untreated test chimeras. However, in vivo depletion of CD4+ T cells in chimeric mice prior to LCMV infection did not lead to a 5-fold reduction of gp33-specific T cells as observed for IL-2Rα-deficient T cells, although it reduced gp33-specific CD8+ T cell frequencies by 25%. In addition, depletion of CD4+ T cells did not lead to reduced frequencies of m38- or m45-specific CD8+ T cells during MCMV infection, indicating that CD4+ T cells are not the only pivotal source of IL-2 (data not shown). Hence, other cell populations, for example activated DC 43, the CD8+ T cells themselves or other immune cells such as NKT cells may be additional relevant sources of IL-2.

It was shown previously that primary CD8+ T cell responses often occur normally in the absence of Th cells 44, 45. However, such unhelped CD8+ T cells fail to expand after re-exposure to the antigen 27, 28, 4649. Our present data are reminiscent of these earlier findings, since IL-2 was partly dispensable for primary CD8+ T cell responses but was essential for secondary responses. Hence, memory T cells primed in the absence of IL-2 signaling have a phenotype similar to “unhelped” CD8+ memory T cells. It is therefore conceivable that availability of IL-2 during the primary T cell response dictates the proliferative potential of memory T cells developing subsequently, as was indeed very recently demonstrated 23. Alternatively, IL-2 signaling might be crucial during secondary expansion of memory CD8+ T cells. It is interesting to note in this context that IL-2 is the key cytokine needed to overcome T cell anergy 50. T cells induced in the absence of CD28-mediated costimulation become anergic and fail to proliferate upon re-exposure to antigen; this phenotype may be overcome by addition of exogenous IL-2, which restores T cell proliferation 51. The data presented here suggest that priming of T cells in the absence of IL-2 may also result in an anergic phenotype. CD8+ T cells induced in the absence of CD28, IL-2 signaling or T help may therefore all have a similar phenotype and fail to proliferate during recall responses due to an inability to produce IL-2 or to perceive IL-2-mediated signals.

IL-2 has been shown to be involved in AICD and may therefore promote cell death during ongoing T cell responses 6, 10. On the other hand, injection of exogenous IL-2 has been shown to reduce cell death after viral clearance and enhance the number of surviving specific T cells 9. The decline of specific T cells after acute/resolved LCMV infection occurred with normal kinetics in the chimeric mice, and IL-2R expression did not alter the extent of T cell death. Therefore, IL-2 signaling is not strongly involved in the decline of T cell responses or the size of the T cell pool established after viral clearance. This observation may be explained by the fact that i) T cells are eliminated at a point in time when antigen availability has dropped to minimal levels, i.e. too late for antigen-dependent AICD, and ii) levels of IL-2 late in the immune response are probably too low to affect T cell survival. The phenotype of memory IL-2Rα–/– and IL-2Rα+/+ CD8+ T cells after acute/resolved LCMV infection differed to a certain extent; an increased percentage of IL-2Rα–/– CD8+ T cells exhibited an IL-7Rα+ CD62L central memory phenotype compared to IL-2Rα+/+ CD8+ T cells. This observation suggests that central memory CD8+ T cells might have a selective survival advantage in the absence of IL-2Rα signaling, which may be due their superior responsiveness to homeostatic cytokines such as IL-7 and IL-15 52, 53.

In contrast to the induction and maintenance of CD8+ T cells during acute/resolved infection, which were relatively independent of IL-2R signaling, the relevance of IL-2R signaling for CD8+ T cell maintenance was substantial in the setting of chronic viral infections with continued antigen exposure. Under these circumstances, IL-2Rα signaling became crucial for the maintenance of virus-specific CD8+ T cells and, surprisingly, did not accelerate in vivo cell death through AICD. The reason(s) for this strict dependence of CD8+ T cell maintenance during chronic infection on IL-2R signaling remains to be determined, but it could relate to the very high turnover of virus-specific CD8+ T cells during chronic infection due to continued antigen encounter (40, 54 and Agnellini et al., unpublished observations), which might be dependent on IL-2R signaling. The source of the relevant IL-2 remains to be determined; however, it is rather unlikely that CD4+ or CD8+ T cells are the major source, since they are generally unable to produce measurable amounts of IL-2 during chronic viral infection (30, 31, 55 and Agnellini et al., unpublished observations). It is conceivable, however, that some IL-2 production by LCMV-specific T cells might occur during the very early phases of the chronic infection, which might allow the differentiation of CD8+ T cells that can be maintained during chronic infection. It is also possible that there are other relevant sources of IL-2 during the chronic phase of LCMV infection that are responsible for the maintenance of LCMV-specific CD8+ T cells via IL-2Rα, a question that we will investigate in more detail in future experiments.

Taken together, IL-2R signaling is largely dispensable for primary antiviral T cell responses. IL-2R signaling is, however, crucial for secondary CD8+ T cell responses and for maintenance of CD8+ T cells during persistent infection.

Materials and methods

Mice, viruses and peptides

Transgenic mice expressing a TCR specific for LCMV peptide gp33–41 were described previously 56. Ly5.2+ C57BL/6 mice were purchased from Janvier, and Ly5.1+ C57BL/6 and IL-2Rα–/– mice were purchased from Jackson and were maintained in a specific pathogen-free (SPF) facility. Mixed bone marrow chimeras were generated by lethal irradiation of Ly5.1+ C57BL/6 recipient mice (950 radγ), followed by i.v. adoptive transfer of bone marrow cells from Ly5.1+ IL-2Rα+/+ and Ly5.2+ IL-2Rα–/– mice (test chimeras) or bone marrow cells from Ly5.1+ IL-2Rα+/+ and Ly5.2+ IL-2Rα+/+ mice (control chimeras). Bone marrow donors were maximally 6 wks old. Reconstitution was analyzed in blood at 8–10 wks. In order to be able to express numbers and frequencies of specific T cells independently of the degree of chimerism of the mice, values were corrected for chimerism at the time point prior to immunization as assessed in the blood. As an example, if chimerism was 60% Ly5.1+ cells and 40% Ly5.2+ cells, values for Ly5.2+ cells were corrected by a factor of 1.5. Correction values were in the range of 1–2.0 for all experiments. Animal experiments were performed according to the regulations of the Cantonal Veterinary Office.

The LCMV isolates WE and Docile were originally provided by Dr. R. M. Zinkernagel (University Hospital, Zurich, Switzerland) and were propagated at a low multiplicity of infection on L929 or MDCK cells, respectively. Mice were infected i.v. with 200 pfu LCMV WE or 106 pfu LCMV Docile. LCMV titers were determined in spleen using the focus forming assay 57.

The H-2Db-restricted LCMV glycoprotein (GP) peptide aa 33–41 (gp33 peptide, KAVYNFATM) 58, the I-Ab-restricted LCMV GP-derived peptide aa 61–80 (gp61 peptide, GLNGPDIYKGVYQFKSVEFD) 59, the H-2Kb-restricted MCMV m38 epitope (316–323, SSPPMFRV) and the H-2Db-restricted MCMV m45 epitope (985–993, HGIRNASFI) 38 were purchased from NeoMPS (Strasbourg, France).

Recombinant vaccinia virus expressing LCMV glycoprotein (VVG2) was originally obtained from Dr. D. H. L. Bishop (Oxford University, Oxford, UK) and grown on BSC40 cells at a low multiplicity of infection. Quantification was performed as described 60.

Bacterial artificial chromosome (BAC)-derived MCMV MW97.01 was provided by Prof. U. H. Koszinowski (Munich, Germany). MCMV MW97.01 was previously shown to be biologically equivalent to MCMV Smith strain ATCC VR-194 (recently re-accessioned as VR-1399) and is here referred to as MCMV 61. MCMV was grown on mouse embryonic fibroblasts (MEF) according to established protocols 62. Mice were infected i.v. with 106 pfu.

GP33-VLP based on the peptide gp33 coupled to VLP derived from the bacteriophage Qβ have been described previously 26. Packaging of CpG oligonucleotides (5′-GGGGTCAACGTTGAGGGGGG-3′, thioester stabilized) into the gp33-VLP was performed as described previously 26.

Viral challenge

Viral challenges were performed either by i.p. infection with 5 × 106 pfu Vacc-GP or i.v. with 200 or 104 pfu LCMV WE.

Antibodies and peptide/MHC class I tetramers

Allophycocyanin (APC)- or PE-conjugated peptide/MHC class I tetrameric complexes were generated as previously described 63. The following anti-mouse mAb were purchased from Becton Dickinson, BD Pharmingen (Allschwil, Switzerland): anti-CD45.1 (PE or biotin), anti-CD127 (FITC), anti-CD107a (FITC), anti-CD62L (allophycocyanin), anti-IFN-γ (FITC, PE or allophycocyanin), anti-IL-2 (allophycocyanin), anti-TNF-α (FITC), anti-CD8 (PerCP or allophycocyanin), anti-CD25 (FITC) and anti-CD4 (FITC or PerCP).

Cell stimulation, immunofluorescent staining and analysis

Lymphocytes were isolated from lung and liver after perfusion as described 24. For direct staining, whole blood, single-cell suspensions from spleens or lymphocytes from liver and lung were used. Cells were incubated for 20 min at 4°C with peptide/MHC tetramers, anti-CD45.1 and anti-CD8 antibodies. For extracellular CD107 and for intracellular IFN-γ, IL-2 or TNF-α staining, lymphocytes were stimulated with 1 μg/mL peptide for 6 h in the presence of anti-CD107a antibodies and Brefeldin A, washed, surface stained at 4°C and fixed/permeabilized using 500 μL FIX/perm solution (FIX/perm solution: FACSLyse (BD) diluted to 2× concentration with H20 and 0.05% Tween 20 (Sigma, Buchs, Switzerland)). Cells were washed once and incubated at room temperature with directly conjugated antibodies specific for intracellular proteins. Cells were washed and resuspended in PBS containing 1% paraformaldehyde (Sigma). Four-color flow cytometric analysis was performed using a FACSCalibur flow cytometer (BD) with CellQuest software (BD). List mode data were analyzed using WinList software (Verity software house, Inc., Topsham, ME, USA).

Isolation of primed IL-2Rα–/– or IL-2Rα+/+ T cells and adoptive transfer

Spleen cells were isolated from test chimeras that had been infected with LCMV 45 days previously. CD8+ T cells were purified first by magnetic cell sorting (Miltenyi Biotech, Bergisch Gladbach, Germany) according to the manufacturer's instructions. For further purification of Ly5.1+ IL-2Rα+/+ and Ly5.2+ IL-2Rα–/– cells, they were FACS-sorted using a FACS Aria (BD) to a purity of >97%.

In vivo CD4 depletion

For in vivo depletion of CD4+ T cells, 0.2 mg purified YTS 191.1 mAb 64 was injected i.p. 3 days and 1 day prior to infection as well as 3 and 6 days post-infection. Depletion was analyzed by flow cytometry and was >99%.

Statistics

Independent sample student t-test analyses were performed using SPSS 14.0.

Acknowledgements

The authors would like to thank Eva Niederer for excellent assistance in cell sorting. This work was supported by the Roche Research Fund for Biology (A. O.), the ETH Zurich (A. O.), the Swiss National Science Foundation (A. O.) and the Vontobel Foundation (A. O.).

Note added in proof

The article cited in the text as “Agnellini et al., unpublished observations” has now been accepted for publication. The reference [ ]is:

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

  8. 8

    WILEY-VCH

Ancillary