IFN-γ Regulates CD8+ Memory T Cell Differentiation and Survival in Response to Weak, but Not Strong, TCR Signals

Diana Stoycheva, Katrin Deiser, Lilian Stärck, Gopala Nishanth, Dirk Schlüter, Wolfgang Uckert and Thomas Schüler
J Immunol January 15, 2015, 194 (2) 553-559; DOI: https://doi.org/10.4049/jimmunol.1402058

Abstract

In response to primary Ag contact, naive mouse CD8+ T cells undergo clonal expansion and differentiate into effector T cells. After pathogen clearance, most effector T cells die, and only a small number of memory T cell precursors (TMPs) survive to form a pool of long-lived memory T cells (TMs). Although high- and low-affinity CD8+ T cell clones are recruited into the primary response, the TM pool consists mainly of high-affinity clones. It remains unclear whether the more efficient expansion of high-affinity clones and/or cell-intrinsic processes exclude low-affinity T cells from the TM pool. In this article, we show that the lack of IFN-γR signaling in CD8+ T cells promotes TM formation in response to weak, but not strong, TCR agonists. The IFN-γ–sensitive accumulation of TMs correlates with reduced mammalian target of rapamycin activation and the accumulation of long-lived CD62LhiBcl-2hiEomeshi TMPs. Reconstitution of mammalian target of rapamycin or IFN-γR signaling is sufficient to block this process. Hence, our data suggest that IFN-γR signaling actively blocks the formation of TMPs responding to weak TCR agonists, thereby promoting the accumulation of high-affinity T cells finally dominating the TM pool.

This article is featured in In This Issue, p.489

Introduction

Functional plasticity of individual CD8+ T cells is an indispensable prerequisite for the orchestration of adaptive immune responses (1). CD8+ T cells are equipped with a huge number of different cell surface molecules that function as receptors for soluble and cell-associated ligands provided by their environment. For example, the TCR-mediated interaction with self-peptide–MHC complexes and consumption of cytokines, such as IL-7, are essential to maintain survival and function of CD8+ naive T cells (TNs) (2). However, during the course of an infection, pathogen-related peptide–MHC complexes are produced at high amounts. CD8+ TNs equipped with appropriate TCRs undergo clonal expansion, convert into CD8+ effector T cells (TEs), and mediate pathogen clearance. Finally, most CD8+ TEs die during the contraction phase, and only a small fraction of memory precursor T cells (TMPs) give rise to long-lived memory T cells (TMs) (1). There is accumulating evidence that inflammatory cytokines, such as IL-12 and IFN-α/β, directly promote CD8+ T cell responses (3). This seems to result from the enhancement of proximal TCR signaling and the subsequent increase in Ag sensitivity of CD8+ T cells (4). The relative importance of IFN-γ for CD8+ T cell expansion is less clear. Although some studies demonstrated a direct, growth-promoting effect of IFN-γ for some CD8+ T cell clones (5), others expand independently of IFN-γ (6, 7). This suggests that IFN-γR signaling in CD8+ T cells affects their expansion in a context-dependent fashion, probably determined by clone-specific features or the nature of the pathogen.

This assumption is supported by the fact that IFN-γ controls immunodominance during the course of infection. In its absence, CD8+ TMs specific for subdominant epitopes accumulate to unusually high numbers (8). The relative abundance of certain CD8+ TM clones is also determined by early TCR signals. For example, initial TCR signal strength correlates positively with the likelihood that a certain CD8+ T cell clone enters the TM pool (9). However, it remains unclear whether a functional link exists among TCR signal strength, IFN-γR signaling, and CD8+ TM differentiation.

In this article, we demonstrate that IFN-γR signaling in CD8+ T cells regulates the differentiation of long-lived TMs. Surprisingly, this effect is only visible after stimulation with weak, but not strong, TCR agonists. In the absence of IFN-γR signaling, mammalian target of rapamycin (mTOR) signaling is impaired, correlating with elevated levels of the lymph node–homing molecule CD62L, the antiapoptotic molecule B cell lymphoma protein-2 (Bcl-2), and the TM-associated transcription factor eomesodermin (Eomes). As a result, the differentiation of long-lived CD62LhiBcl-2hiEomeshi CD8+ TMPs and subsequent TM formation in response to weak TCR agonists are more efficient for IFN-γR-deficient CD8+ T cells. This advantage is overcome by the reconstitution of mTOR signaling. However, the use of specific inhibitors revealed that the IFN-γ–dependent regulation of TMP-associated factors requires the coordinated action of JAK1, JAK2, and mTOR. In summary, we provide evidence for an unknown link between TCR and IFN-γR signaling in CD8+ T cells. According to our results, IFN-γR signaling, in combination with a weak TCR stimulus, suppresses the differentiation of long-lived CD8+ TMPs/TMs via mTOR-dependent and -independent signaling events. During the course of an immune response, this mechanism might facilitate the enrichment of high-affinity CD8+ TMs and the loss of low-affinity CD8+ T cells, thereby ensuring both efficient secondary immune responses and the prevention of autoimmunity by low-affinity, self-reactive CD8+ TMs.

Materials and Methods

Mice

C57BL/6J (B6), B6.SJL-Ptprca Pepcb/BoyJ (CD45.1-congenic), IFN-γ reporter (B6.129S4-Ifngtm3.1Lky/J), IFN-γ−/− (B6.129S7-Ifngtm1Ts), IFN-γR−/− (B6.129S7-Ifngrtm1Agt) (all CD90.2), B6.PL-Thy1a/Cy (CD90.1-congenic), Rag−/− (B6.129S7-Rag1tm1Mom/J), and OT-I mice expressing a transgenic (tg) TCR (Vα2Vβ5) specific for the chicken OVA-derived, H2-Kb–restricted peptide OVA257–264 (SIINFEKL) were bred in our animal facilities (Forschungseinrichtungen für Experimentelle Medizin [Charité Berlin] and Zentrales Tierlabor [Otto-von-Guericke University]). Mice were crossed to generate CD90.1/.2-disparate Rag−/−OT-I, Rag−/−IFN-γR−/−OT-I, and Rag−/−IFN-γ−/−OT-I mice. All mice were housed under specific pathogen–free conditions, and animal experiments were performed according to institutional guidelines (Landesamt für Gesundheit und Soziales Berlin and Landesverwaltungsamt Sachsen-Anhalt).

Adoptive T cell transfer

CD8+ TNs were purified from spleen and lymph nodes of the respective donor mice using CD8α-specific MicroBeads and AutoMACS (Miltenyi Biotec). A total of 5 × 105–1 × 106 CD8+ T cells (purity > 97%) were injected i.v. into the tail vein of recipient mice. For some experiments, mice were immunized i.v. with 50 μg SIINFEKL (N4), SIIVFEKL (V4), SIITFEKL (T4), SIIQFEKL (Q4), and SAINFEKL (A2) (Biosynthan) 1 d after T cell transfer. Control animals were injected with Dulbecco’s PBS. Rapamycin was injected i.p. at 1.5 μg/mouse daily.

Cell culture

Spleen and lymph node cells of the respective OT-I mice were pooled and cultured at 5 × 105 OT-I T cells/ml in round-bottom 96-well plates in RPMI 1640 medium supplemented with 10% FCS, 1% penicillin/streptomycin, 2-ME (50 μM), nonessential amino acids, sodium pyruvate, and glutamate at 37°C with 5% CO2 for the indicated period. Where indicated, 1 μM of the respective altered peptide ligand (APL), 100 ng/ml rmIFN-γ (R&D Systems/BioLegend), 5 μg/ml purified anti–IFN-γ (from R4-6A2 hybridoma supernatant), 20 ng/ml rapamycin (Sigma-Aldrich), or 1 μM ruxolitinib (JAK1/2 inhibitor; Selleck Chemicals) was added. Supernatants of in vitro–cultured cells were collected and analyzed by an IFN-γ–specific ELISA (eBioscience), according to the manufacturer’s instructions.

OVA-expressing EG7 lymphoma cells were cultured in RPMI 1640 + 10% FCS medium with 0.4 mg/ml G418. A total of 1 × 106 cells was injected s.c. in the right flank of B6 mice. Tumor volume was measured every 2–3 d.

Retroviral transduction and bone marrow chimeras

Bone marrow chimeric mice were generated as described (10). Briefly, bone marrow stem cells were isolated from Rag−/−IFN-γR−/−OT-I (CD90.1) mice and purified with an EasySep Mouse SCA1 Positive Selection Kit (STEMCELL TECHNOLOGIES), according to the manufacturer’s instructions. Isolated stem cells were incubated for 3 d in StemPro-34 medium supplemented with 5% FCS, 100 U/ml penicillin/streptomycin, 2 mM l-glutamine, 10 ng/ml murine IL-3, 50 ng/ml murine IL-6, and 50 ng/ml murine stem cell factor (PeproTech). The myristoylated allele of protein kinase B/AKT fused to the estrogen receptor (myrAKTER) cassette (kindly provided by Prof. M.G. Kharas, Memorial Sloan Kettering Cancer Center, New York, NY) and the cDNA for GFP were linked by an internal ribosome entry site (IRES), and the resulting construct was cloned into an MP71 vector (11). Cultured bone marrow stem cells were transduced either with MP71-myrAKTER-IRES-GFP or MP71-IRES-GFP as control. Rag−/− (CD90.2) mice were sublethally irradiated with 5.5 Gy and reconstituted with 5 × 105–1 × 106 transduced stem cells/mouse.

Flow cytometry

The following Abs were used: anti-CD8α (53-6.8; BioLegend/eBioscience), anti-CD90.1 (OX-7; BioLegend/BD), anti-CD90.2 (53-2.1; BioLegend/BD), anti-CD62L (MEL-14; eBioscience/BD), anti-Ki67 (SolA15; eBioscience), anti–Bcl-2 (10C4; BioLegend), anti–T-bet (4B10; BioLegend/eBioscience), anti-Eomes (Dan11Mag; eBioscience), anti–IFN-γ (XMG1.2; BioLegend/eBioscience), anti-CD107a (1D4B; BD), and anti–p-mTORC1 (Ser2481; Cell Signaling). For intracellular cytokine staining, cells were restimulated for 3–4 h with 1 μM SIINFEKL in the presence of brefeldin A, monensin A, and anti-CD107a, fixed with the intracellular staining kit (BioLegend) according to the manufacturer’s instructions, and stained with anti–IFN-γ. For staining of nuclear Ags, cells were fixed using the Foxp3 Fixation Kit (eBioscience), according to the manufacturer’s instructions. Samples were measured on a FACSCalibur, FACSCanto, or LSR II flow cytometer and analyzed by FlowJo software.

Statistical analysis

Statistical analysis and graphical representations were done using Prism5 software (GraphPad Software). Statistical significance was determined using a nonparametric two-tailed Mann–Whitney or paired Student t test.

Results

The lack of IFN-γR signaling promotes the accumulation of memory CD8+ T cells in response to weak, but not strong, TCR signals

Self-peptide–MHC complexes are weak TCR agonists that usually do not activate mature peripheral CD8+ T cells in lymphocyte-competent mice. However, under lymphopenic conditions, self-peptide–MHC complexes, together with IL-7, promote CD8+ TN activation, lymphopenia-induced proliferation (LIP), and subsequent differentiation of TMs (12–15).

Previously, we studied the impact of IFN-γ on CD8+ TM generation and differentiation (7, 16). We showed that IFN-γR–deficient (IFN-γR−/−) and wild-type (WT) CD8+ TCR-tg OT-I T cells responding to their cognate peptide SIINFEKL did not differ with regard to CD62L expression, expansion, or TM formation (7). In contrast, the frequency of CD62Lhi IFN-γR−/− OT-I T cells was strongly increased when both OT-I populations underwent LIP in Rag-deficient (Rag−/−) mice (7). Because the self-peptides driving LIP of OT-I T cells are weak TCR agonists (13), we hypothesized that the lack of IFN-γR signaling in CD8+ T cells facilitates the differentiation of CD62Lhi TMs in response to weak, but not strong, TCR agonists. To test this assumption, we performed adoptive T cell–transfer experiments. Equal numbers of purified CD8+ T cells from Rag−/− CD90.1+ WT OT-I and Rag−/− CD90.1+/CD90.2+ IFN-γR−/− OT-I (IFN-γR−/− OT-I) mice were mixed and adoptively transferred into Rag−/− recipient mice. Different cohorts of mice were immunized with APLs of N4, which induce gradually increasing TCR signal intensities (V4 [SIIVFEKL] < T4 [SIITFEKL] < Q4 [SIIQFEKL] < A2 [SAINFEKL] < N4 [SIINFEKL]) in OT-I T cells (9). Control mice were injected with PBS, allowing LIP in response to self-peptide–MHC complexes. Twelve days later, most IFN-γR−/− CD8+ OT-I cells undergoing LIP in PBS-treated recipients expressed high levels of CD62L as opposed to WT CD8+ OT-I cells. This difference was still apparent in response to V4, but it finally disappeared after vaccination with stronger TCR agonists (Fig. 1A). Hence, the lack of IFN-γR signaling in CD8+ T cells favors the generation of CD62Lhi cells in response to weak, but not strong, TCR signals. Early CD62Lhi CD8+ T cells contain TMPs, which also express elevated levels of the antiapoptotic molecule Bcl-2 and the TM-associated transcription factor Eomes (17–19). Correlating with their high levels of CD62L, IFN-γR−/− CD8+ OT-I cells undergoing LIP also expressed elevated levels of Bcl-2 and Eomes compared with their WT counterparts. These differences remained stable after vaccination with the weakest APL V4, indicating that self-peptide–MHC complexes driving LIP and V4 induce comparable differentiation programs in CD8+ OT-I T cells. However, with increasing TCR signal strength, the expression of Bcl-2 and Eomes became indistinguishable in the OT-I populations, whereas T-bet levels were independent of IFN-γR signaling and TCR signal strength at this time point. High levels of CD62L, Bcl-2, and Eomes and low levels of T-bet indicate the long-term survival potential of CD8+ T cells (19, 20). Hence, our results suggested that IFN-γR signaling blocks the generation of long-lived CD8+ TMs in response to weak TCR agonists. To test this hypothesis, both T cell populations were transferred simultaneously into Rag−/− mice, which were either vaccinated with N4 or injected with PBS to allow LIP. As shown in Fig. 1B, WT and IFN-γR−/− CD8+ OT-I cells undergoing LIP in PBS-treated recipients were equally abundant until day 12. However, from days 30 to 60, the frequency of IFN-γR−/− CD8+ OT-I cells increased significantly, from ∼60 to 80%. This survival advantage of IFN-γR−/− CD8+ OT-I cells was not observed after vaccination with N4. Under these conditions, WT and IFN-γR−/− CD8+ OT-I cells had similar frequencies at any given time point (Fig. 1C). Importantly, 60 d after adoptive transfer into PBS-treated Rag−/− mice, IFN-γ+ CD107a+ cells had similar frequencies among WT and IFN-γR−/− CD8+ OT-I cells (Fig. 1D). Furthermore, the majority of WT and IFN-γR−/− CD8+ OT-I cells expressed high levels of CD62L (Fig. 1E), and the frequency of Ki67+ proliferating cells was indistinguishable in the OT-I populations at this time point of LIP (Fig. 1F). Hence, although different in number (Fig. 1B), WT and IFN-γR−/− CD8+ OT-I TMs differentiated equally well into day-60 TMs. This suggested that early, rather than late, IFN-γR signals determine the size of CD8+ TM clones primed by weak TCR agonists. Therefore, we hypothesized that the lack of early IFN-γR signaling facilitates the differentiation of long-lived CD62LhiBcl-2hiEomeshi TMPs after contact with weak, but not strong, TCR agonists (Fig. 1A), thereby determining the size of the resulting TM pool (Fig. 1B).

FIGURE 1.
FIGURE 1.

IFN-γR signaling regulates CD8+ T cell memory formation in response to weak, but not strong, TCR signals. (A–F) Equal numbers of WT and IFN-γR−/− OT-I T cells were transferred simultaneously into Rag−/− mice. One day later, mice were injected i.v. with PBS or the indicated APLs. (A) Twelve days after T cell transfer, splenocytes were analyzed by flow cytometry. Shown are relative fluorescence intensities for WT and IFN-γR−/− OT-I T cells. Data are representative of two independent experiments with four or five mice/group. T-bet levels were determined once. (B and C) After PBS or N4 injection, the relative abundance (± SEM) of WT and IFN-γR−/− cells within the OT-I T cell pool was determined by flow cytometry at the indicated time points. Data are representative of two independent experiments with three to five mice/group. (D–F) Sixty days after T cell transfer, the phenotype of splenic WT (black) and IFN-γR−/− (red) memory OT-I T cells was determined by flow cytometry. (D) Splenocytes were restimulated with N4 for 3–4 h prior to the measurement of CD107a and IFN-γ. (E and F) CD62L and Ki67 levels were determined directly after isolation. Shown are relative fluorescence intensities for the indicated molecules. Data are representative of two independent experiments with three or four mice/group. *p < 0.05.

To test this assumption, we established an in vitro priming system. For this purpose, naive WT and IFN-γR−/− OT-I cells were cocultured with V4 or N4 for 72 h. Flow cytometry revealed that IFN-γR−/− OT-I cells primed with V4 expressed significantly higher levels of Bcl-2 and Eomes compared with WT OT-I T cells (Fig. 2A, 2B). Importantly, expression levels of Bcl-2 and Eomes were indistinguishable between WT and IFN-γR−/− OT-I cells after stimulation with N4. Hence, the lack of IFN-γR signaling was associated with Bcl-2 and Eomes upregulation in response to weak, but not strong, TCR agonists.

FIGURE 2.
FIGURE 2.

In the absence of IFN-γR signaling, weak TCR signals promote the generation of CD62LhiBcl-2hiEomeshi TMPs. (A–D) Equal numbers of WT and IFN-γR−/− OT-I T cells were cocultured with V4 or N4 for 72 h and analyzed by flow cytometry. Representative graphs show relative fluorescence intensities for the indicated molecules (filled graphs: naive WT OT-I cells) (left panels). Bar graphs represent pooled data from 11 (A), 20 (B), 22 (C), and 9 (D) independent experiments performed in duplicates. (A–C) To calculate relative expression levels (+ SEM) of the indicated molecules, mean fluorescence intensities (MFIs) were determined for WT OT-I cells, values were set to 1, and MFIs measured for IFN-γR−/− OT-I T cells were calculated in relation. WT and IFN-γR−/− OT-I T cells were cocultured with the indicated peptides for 24 h (E) or 72 h (F) and analyzed by flow cytometry to determine the MFI for CD62L (E) or Eomes (F). Data in (E) are representative of four independent experiments performed in duplicates. Data in (F) are from one experiment performed in triplicates. (G and H) EG7 tumor cells were injected s.c. in the right flank of B6 mice. Three days later, mice received WT or IFN-γR−/− OT-I T cells stimulated with V4 for 72 h. Data in (G) are mean tumor size ± SEM for the indicated time points and are representative of two independent experiments with eight mice/group. (H) On the day of sacrifice, frequencies of WT and IFN-γR−/− OT-I cells were determined by flow cytometry in the spleens of tumor-bearing mice. Shown are the relative numbers of the CD8+CD90.1+ OT-I cell population within the total CD8+ T cell pool. *p < 0.05, ****p < 0.0001. ns, not significant.

After primary Ag contact, IFN-γ promotes the rapid upregulation of T-bet in CD8+ T cells (20). In accordance with this, T-bet levels were significantly lower in IFN-γR−/− OT-I cells, irrespective of whether they had been activated by V4 or N4 (Fig. 2C). However, T-bet is downregulated again in the course of memory CD8+ T cell differentiation (20, 21). This might explain why T-bet levels were identical in WT and IFN-γR−/− OT-I cells at later time points (Fig. 1A), whereas they differed significantly 72 h after priming (Fig. 2C).

Ki67+ proliferating cells had similar frequencies in the OT-I cell populations after stimulation with either peptide (Fig. 2D). In contrast, CD62L levels on WT and IFN-γR−/− OT-I cells correlated inversely with TCR signal strength. Importantly, however, CD62L was always more abundant on IFN-γR−/− OT-I cells stimulated with weak TCR agonists. This difference disappeared with increasing TCR signal strength and was not apparent after stimulation with N4 (Fig. 2E). Similar results were obtained for Eomes (Fig. 2F). In summary, our in vitro experiments suggest that early IFN-γR signals limit the differentiation of long-lived CD8+ CD62LhiBcl-2hiEomeshi TMPs in response to weak, but not strong, TCR agonists.

To test for the in vivo function of V4-primed WT and IFN-γR−/− OT-I cells, B6 mice were challenged s.c. with 106 EG7 tumor cells expressing N4. Three days later, mice were reconstituted with 106 WT or IFN-γR−/− OT-I cells activated in vitro with V4 for 72 h. Compared with WT OT-I cells, IFN-γR−/− OT-I cells efficiently delayed tumor growth (Fig. 2G). This was associated with their accumulation in the spleens of tumor-bearing mice (Fig. 2H). Thus, the improved survival capabilities of IFN-γR−/− OT-I CD62LhiBcl-2hiEomeshi TMPs generated in response to weak TCR ligands (Fig. 1B) are preserved under nonlymphopenic conditions (Fig. 2H) and correlate positively with tumor growth control (Fig. 2G).

Reduced mTOR signaling facilitates the formation of IFN-γR−/− TMPs

Signaling via mTOR limits CD8+ TMP/TM differentiation (18, 19). Furthermore, cell culture experiments with human cell lines and mouse embryonic fibroblasts demonstrated that IFN-γR signaling promotes mTOR function in a PI3K- and protein kinase B/AKT–dependent fashion (22–24). However, it remained unclear whether the modulation of mTOR by IFN-γ is also operative in CD8+ T cells and whether it is influenced by TCR signaling. To test this possibility, WT and IFN-γR−/− OT-I cells were stimulated with V4, and the degree of mTOR activity was quantified by flow cytometry. As shown in Fig. 3A, the levels of p-mTOR were significantly lower in IFN-γR−/− OT-I cells. No such difference was observed after coculture with N4 (Fig. 3B). These results suggested a causal link between impaired mTOR signaling and the more efficient generation of CD62Lhi IFN-γR−/− OT-I TMPs in response to weak TCR signals. To test this assumption, we created a system that allowed the selective upregulation of mTOR signaling in IFN-γR−/− OT-I cells.

FIGURE 3.
FIGURE 3.

IFN-γR signaling modulates CD8+ T cell responses to weak TCR signals via mTOR. WT (black) and IFN-γR−/− (red) OT-I T cells were cocultured in the presence of V4 (A) or N4 (B) and analyzed by flow cytometry 24 h later. Graphs show relative fluorescence intensities of p-mTOR for representative samples (filled graphs: naive WT OT-I cells). Data are pooled from 13 independent experiments performed in duplicates. (C and D) Sublethally irradiated Rag−/− mice were reconstituted with CD90.1+ IFN-γR−/− OT-I bone marrow transduced with MP71-myrAKTER-GFP vector. Splenocytes from chimeric mice were cultured for 24 h with V4 and the indicated combinations of tamoxifen (tam) and rapamycin (rapa). Tamoxifen was washed out 6 h after stimulation, and p-mTOR (C) and CD62L (D) levels were determined by flow cytometry. Graphs show relative fluorescence intensities for representative samples (filled graphs: naive IFN-γR−/−GFP+ OT-I cells). Relative p-mTOR and CD62L levels (+ SEM) were determined as described in Fig. 2. Data in (C) and (D) are pooled from two independent experiments performed in duplicates. *p < 0.05. ns, not significant.

For this purpose, bone marrow cells from Rag−/− CD90.1+ IFN-γR−/− OT-I mice were transduced with a retroviral vector encoding eGFP and myrAKTER (25). After binding of tamoxifen to myrAKTER, the protein becomes constitutively active and enhances mTOR signaling.

Sublethally irradiated Rag−/− mice were reconstituted with gene-modified bone marrow cells. After successful bone marrow engraftment, splenocytes from chimeric mice were stimulated with V4. The levels of p-mTOR and CD62L were determined in myrAKTER+eGFP+ IFN-γR−/− OT-I cells. The addition of tamoxifen led to increased mTOR activation, which could be blocked by rapamycin (Fig. 3C). Conversely, tamoxifen induced the downmodulation of CD62L (Fig. 3D). This effect was inhibited by rapamycin, thus excluding mTOR-independent effects of myrAKTER activation.

Together, the data presented so far suggest that the blockade of early IFN-γR signaling in CD8+ T cells responding to weak TCR signals limits mTOR-dependent signaling, subsequently favoring the generation of CD62LhiBcl-2hiEomeshi TMPs. This interpretation was further supported by experiments blocking the function of mTOR or autocrine IFN-γ action (Supplemental Figs. 1, 2). In contrast, strong TCR signals do not appear to be affected by IFN-γR signaling. This might be due to the fact that mTOR activity correlates positively with TCR signal strength (26). Hence, a strong TCR signal would be sufficient to drive mTOR-dependent effects, independent of IFN-γR signaling.

mTOR-related and -unrelated signaling events contribute to the IFN-γ–dependent regulation of Eomes and T-bet in CD8+ T cells primed with weak TCR ligands

The data presented in Fig. 3 and Supplemental Fig. 2 suggest that CD8+ T cell–derived IFN-γ modulates mTOR-dependent signaling in an autocrine fashion. However, it remained unclear whether rapamycin blocked TCR-driven IFN-γ production or IFN-γR signal transduction.

To study IFN-γ production, we first stimulated WT OT-I cells encoding an eYFP reporter transgene under control of the ifn-γ promoter with V4 and N4 in the presence or absence of rapamycin. As shown in Fig. 4A, ifn-γ promoter activity was strongly suppressed by rapamycin, correlating with a strong reduction in IFN-γ protein amounts in culture supernatants (Fig. 4B). Hence, rapamycin treatment might favor the differentiation of low-affinity TMPs by two different mechanisms: the inhibition of IFN-γ production and direct suppression of IFN-γR signal transduction.

FIGURE 4.
FIGURE 4.

mTOR-independent signals contribute to the IFN-γ–mediated repression of Eomes in response to weak TCR signals. (A and B) Splenocytes from OT-I mice encoding eYFP under control of the ifn-γ promoter were cultured for 24 h in the presence of V4 (filled bars) or N4 (open bars) without (black) or with (red) rapamycin. (A) Reporter gene activity was determined by flow cytometry after gating on CD8+ OT-I T cells. (B) IFN-γ in culture supernatants was quantified by ELISA. Data are representative (+ SD) from two independent experiments performed in duplicates. (C–J) IFN-γ−/− OT-I T cells were cultured for 72 h in the presence of the indicated peptides, rIFN-γ, rapamycin (rapa), and/or the JAK1/2-inhibitor ruxolitinib (JakI), as indicated. Eomes and T-bet levels were determined by flow cytometry. Data are representative graphs (filled graphs: naive IFN-γ−/− OT-I cells) and pooled results + SEM (bar graphs) from 10 (C), 13 (D), 8 (E, F, H, and I), or 4 (G and J) independent experiments performed in duplicates. **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant.

To further study the relative impact of both effects, we established an experimental system that enabled us to induce IFN-γR signaling at any given time point, independent of endogenous IFN-γ production. For this purpose, we made use of IFN-γ−/− OT-I cells, which cannot produce IFN-γ but are still sensitive to it. First, we verified that Eomes and T-bet regulation in IFN-γ−/− OT-I cells was comparable to that in WT and IFN-γR−/− OT-I cells. As shown in Fig. 4C, V4-stimulated, but not N4-stimulated, IFN-γ−/− OT-I cells reduced Eomes expression in response to rIFN-γ. However, T-bet was upregulated after rIFN-γ treatment, irrespective of TCR signal strength (Fig. 4D). Hence, IFN-γ controlled Eomes and T-bet levels in IFN-γ−/− OT-I cells, further confirming the data obtained with WT and IFN-γR−/− OT-I cells.

Next, we tested the relative importance of mTOR for the IFN-γR–dependent regulation of the two transcription factors. Notably, rapamycin treatment of V4-stimulated IFN-γ−/− OT-I cells promoted Eomes expression and repressed T-bet expression (Fig. 4E, 4H). Thus, weak TCR signals modulate the expression of both transcription factors in an mTOR-dependent, but IFN-γ-independent, fashion. However, the addition of rIFN-γ partially reversed these effects of rapamycin (Fig. 4F, 4I), further supporting that mTOR-independent signals contribute to the IFN-γ–mediated regulation of Eomes and T-bet. IFN-γR signaling leads to STAT1-dependent gene transcription and PI3K/AKT-dependent activation of mTOR and subsequent mRNA translation (22–24, 27). The activation of both signaling cascades is JAK1 and JAK2 dependent. Consequently, the inhibition of JAK1/2 activity fully prevented the rIFN-γ–mediated regulation of Eomes and T-bet in V4-stimulated IFN-γ−/− OT-I cells (Fig. 4G, 4J). Together, these results demonstrate that mTOR-dependent and -independent signaling events contribute to the IFN-γR–mediated regulation of Eomes and T-bet in conjunction with weak TCR signals. The relative contributions of different IFN-γR signaling modules to the TCR-dependent modulation of Eomes and T-bet, as well as the subsequent CD8+ TMP/TM differentiation, remain to be elucidated.

Discussion

Eomes and T-bet are critical regulators of CD8+ T cell function and differentiation (28, 29). High levels of T-bet favor the differentiation of short-lived CD8+ TEs (20). Conversely, low levels of T-bet and high levels of Eomes support the differentiation of CD8+ TMPs and long-term TM survival (19, 20). mTOR was identified as an important signaling module in T cells that senses environmental cues, integrates signals from multiple receptor systems (30, 31), determines Eomes and T-bet levels, and modulates CD8+ T cell differentiation (18, 19). Although there is accumulating evidence that multiple signaling pathways converge at the level of mTOR, the context- and interdependent contribution of different pathways to the differentiation of CD8+ T cells is largely unclear. The fact that CD62L, Bcl-2, and Eomes levels were elevated in IFN-γR−/− OT-I cells primed with weak TCR agonists (Figs. 1, 2) suggests that IFN-γR signaling affects the generation of long-lived CD62LhiBcl-2hiEomeshi TMPs in a TCR signal strength–dependent and mTOR-related fashion. This assumption was confirmed by the fact that p-mTOR levels were reduced in IFN-γR−/− OT-I cells stimulated with weak, but not strong, TCR agonists (Fig. 3A, 3B). The reconstitution of mTOR signaling in IFN-γR−/− OT-I cells normalized CD62L levels in response to weak TCR agonists (Fig. 3D). This indicates that TCR and IFN-γR signals indeed converge at the level of mTOR. In the event that TCR signals are below a certain threshold, IFN-γR signals seem to compensate for this lack of TCR signal strength, thereby promoting mTOR activity and the generation of short-lived CD8+ TEs. Our results suggest that this mechanism is operative at an early stage of CD8+ T cell activation and has a long-term impact on the composition of the CD8+ TM pool (Fig. 1B).

Our results may explain why CD8+ T cells specific for subdominant epitopes derived from Listeria monocytogenes accumulate more efficiently in IFN-γ−/− mice than in WT mice (8). In agreement with this, the lack of IFN-γR signaling in CD8+ OT-I cells correlated with more efficient CD8+ TM formation postinfection with L. monocytogenes expressing V4 (Supplemental Fig. 3). However, and although to a lesser extent, the lack of IFN-γ also promotes the expansion of CD8+ T cell clones specific for dominant L. monocytogenes epitopes (8). Hence, it was not surprising that the long-term survival of IFN-γR−/− OT-I T cells also was more pronounced postinfection with L. monocytogenes expressing N4 (Supplemental Fig. 3).

Inflammation is an important modulator of CD8+ T cell activation, function, and memory formation (4, 20, 32–34). However, the composition of the inflammatory cytokine milieu differs strongly between pathogens (32, 35, 36). Therefore, we do not want to exclude that the relative importance of TCR–IFN-γR interactions for CD8+ TM formation varies between experimental systems. It might be context dependent and affected by various environmental factors (e.g., cytokines that directly modulate proximal TCR signaling) (4).

During the course of an immune response, high- and low-affinity CD8+ T cells are activated. However, TCR signal strength determines the fate of recently activated CD8+ T cells. For example, weak TCR signals are associated with the premature arrest of CD8+ T cell expansion and the early onset of contraction (9). As a result, high-affinity CD8+ T cells accumulate and dominate the CD8+ TM pool (9). This competition-based model might be further supported by the intracellular signaling events described in this article. Given that low-affinity CD8+ T cells are primed in the course of an immune response, early IFN-γ signals would help to initiate a differentiation program dominated by high levels of T-bet and low levels of CD62L, Eomes, and Bcl-2. This would favor the generation of short-lived CD8+ TEs and might further facilitate the accumulation of high-affinity CD8+ TMs.

The IFN-γ–mediated elimination of low-affinity CD8+ TMs might facilitate secondary immune responses by high-affinity CD8+ T cells, as well as prevent the unwanted activation of low-affinity CD8+ T cells in response to self-peptide–MHC complexes and subsequent autoimmunity (8). This interpretation is supported by our findings that, compared with WT OT-I cells (Fig. 2G), the survival capabilities of IFN-γR−/− OT-I cells (Figs. 1B, 2H) and their cytotoxic functions are more pronounced after priming with low-affinity TCR ligands.

In summary, we show that early IFN-γR signals in CD8+ T cells modulate TE/TMP differentiation and the size of the resulting TM pool in a TCR signal strength–dependent fashion. Our findings provide new insights into the regulation of CD8+ T cell responses by IFN-γ and suggest that TCR affinity determines cytokine responsiveness of CD8+ T cells.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Özen Sercan for helpful discussions; D. Femerling, J. Winckler, and E. Denks for excellent technical assistance; and M.G. Kharas for the myrAKTER construct. L. monocytogenes expressing V4 or N4 were kindly provided by D. Zehn.

Footnotes

  • This work was supported by the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich TR36, SFB854, and SCHU 2326/1-1).

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    APL
    altered peptide ligand
    B6
    C57BL/6J
    Bcl-2
    B cell lymphoma protein-2
    Eomes
    eomesodermin
    IRES
    internal ribosome entry site
    LIP
    lymphopenia-induced proliferation
    mTOR
    mammalian target of rapamycin
    myrAKTER
    myristoylated allele of protein kinase B/AKT fused to the estrogen receptor
    TE
    effector T cell
    tg
    transgenic
    TM
    memory T cell
    TMP
    memory precursor T cell
    TN
    naive T cell
    WT
    wild-type.

  • Received August 11, 2014.
  • Accepted November 7, 2014.

References

  1. ↵
    1. Rutishauser R. L.,
    2. S. M. Kaech
    . 2010. Generating diversity: transcriptional regulation of effector and memory CD8 T-cell differentiation. Immunol. Rev. 235: 219–233.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Takada K.,
    2. S. C. Jameson
    . 2009. Naive T cell homeostasis: from awareness of space to a sense of place. Nat. Rev. Immunol. 9: 823–832.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Curtsinger J. M.,
    2. M. F. Mescher
    . 2010. Inflammatory cytokines as a third signal for T cell activation. Curr. Opin. Immunol. 22: 333–340.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Richer M. J.,
    2. J. C. Nolz,
    3. J. T. Harty
    . 2013. Pathogen-specific inflammatory milieux tune the antigen sensitivity of CD8(+) T cells by enhancing T cell receptor signaling. Immunity 38: 140–152.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Whitmire J. K.,
    2. J. T. Tan,
    3. J. L. Whitton
    . 2005. Interferon-gamma acts directly on CD8+ T cells to increase their abundance during virus infection. J. Exp. Med. 201: 1053–1059.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Tewari K.,
    2. Y. Nakayama,
    3. M. Suresh
    . 2007. Role of direct effects of IFN-gamma on T cells in the regulation of CD8 T cell homeostasis. J. Immunol. 179: 2115–2125.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Sercan O.,
    2. D. Stoycheva,
    3. G. J. Hämmerling,
    4. B. Arnold,
    5. T. Schüler
    . 2010. IFN-gamma receptor signaling regulates memory CD8+ T cell differentiation. J. Immunol. 184: 2855–2862.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Badovinac V. P.,
    2. A. R. Tvinnereim,
    3. J. T. Harty
    . 2000. Regulation of antigen-specific CD8+ T cell homeostasis by perforin and interferon-gamma. Science 290: 1354–1358.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Zehn D.,
    2. S. Y. Lee,
    3. M. J. Bevan
    . 2009. Complete but curtailed T-cell response to very low-affinity antigen. Nature 458: 211–214.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Stärck L.,
    2. K. Popp,
    3. H. Pircher,
    4. W. Uckert
    . 2014. Immunotherapy with TCR-redirected T cells: comparison of TCR-transduced and TCR-engineered hematopoietic stem cell-derived T cells. J. Immunol. 192: 206–213.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Engels B.,
    2. H. Cam,
    3. T. Schüler,
    4. S. Indraccolo,
    5. M. Gladow,
    6. C. Baum,
    7. T. Blankenstein,
    8. W. Uckert
    . 2003. Retroviral vectors for high-level transgene expression in T lymphocytes. Hum. Gene Ther. 14: 1155–1168.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Ernst B.,
    2. D. S. Lee,
    3. J. M. Chang,
    4. J. Sprent,
    5. C. D. Surh
    . 1999. The peptide ligands mediating positive selection in the thymus control T cell survival and homeostatic proliferation in the periphery. Immunity 11: 173–181.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Goldrath A. W.,
    2. M. J. Bevan
    . 1999. Low-affinity ligands for the TCR drive proliferation of mature CD8+ T cells in lymphopenic hosts. Immunity 11: 183–190.
    OpenUrlCrossRefPubMed
    1. Schluns K. S.,
    2. W. C. Kieper,
    3. S. C. Jameson,
    4. L. Lefrançois
    . 2000. Interleukin-7 mediates the homeostasis of naïve and memory CD8 T cells in vivo. Nat. Immunol. 1: 426–432.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Goldrath A. W.,
    2. L. Y. Bogatzki,
    3. M. J. Bevan
    . 2000. Naive T cells transiently acquire a memory-like phenotype during homeostasis-driven proliferation. J. Exp. Med. 192: 557–564.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Sercan O.,
    2. G. J. Hämmerling,
    3. B. Arnold,
    4. T. Schüler
    . 2006. Innate immune cells contribute to the IFN-gamma-dependent regulation of antigen-specific CD8+ T cell homeostasis. J. Immunol. 176: 735–739.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Kaech S. M.,
    2. J. T. Tan,
    3. E. J. Wherry,
    4. B. T. Konieczny,
    5. C. D. Surh,
    6. R. Ahmed
    . 2003. Selective expression of the interleukin 7 receptor identifies effector CD8 T cells that give rise to long-lived memory cells. Nat. Immunol. 4: 1191–1198.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Araki K.,
    2. A. P. Turner,
    3. V. O. Shaffer,
    4. S. Gangappa,
    5. S. A. Keller,
    6. M. F. Bachmann,
    7. C. P. Larsen,
    8. R. Ahmed
    . 2009. mTOR regulates memory CD8 T-cell differentiation. Nature 460: 108–112.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Rao R. R.,
    2. Q. Li,
    3. K. Odunsi,
    4. P. A. Shrikant
    . 2010. The mTOR kinase determines effector versus memory CD8+ T cell fate by regulating the expression of transcription factors T-bet and Eomesodermin. Immunity 32: 67–78.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Joshi N. S.,
    2. W. Cui,
    3. A. Chandele,
    4. H. K. Lee,
    5. D. R. Urso,
    6. J. Hagman,
    7. L. Gapin,
    8. S. M. Kaech
    . 2007. Inflammation directs memory precursor and short-lived effector CD8(+) T cell fates via the graded expression of T-bet transcription factor. Immunity 27: 281–295.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Takemoto N.,
    2. A. M. Intlekofer,
    3. J. T. Northrup,
    4. E. J. Wherry,
    5. S. L. Reiner
    . 2006. Cutting Edge: IL-12 inversely regulates T-bet and eomesodermin expression during pathogen-induced CD8+ T cell differentiation. J. Immunol. 177: 7515–7519.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Lekmine F.,
    2. A. Sassano,
    3. S. Uddin,
    4. J. Smith,
    5. B. Majchrzak,
    6. S. M. Brachmann,
    7. N. Hay,
    8. E. N. Fish,
    9. L. C. Platanias
    . 2004. Interferon-gamma engages the p70 S6 kinase to regulate phosphorylation of the 40S S6 ribosomal protein. Exp. Cell Res. 295: 173–182.
    OpenUrlCrossRefPubMed
    1. Kaur S.,
    2. L. Lal,
    3. A. Sassano,
    4. B. Majchrzak-Kita,
    5. M. Srikanth,
    6. D. P. Baker,
    7. E. Petroulakis,
    8. N. Hay,
    9. N. Sonenberg,
    10. E. N. Fish,
    11. L. C. Platanias
    . 2007. Regulatory effects of mammalian target of rapamycin-activated pathways in type I and II interferon signaling. J. Biol. Chem. 282: 1757–1768.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Kaur S.,
    2. A. Sassano,
    3. B. Dolniak,
    4. S. Joshi,
    5. B. Majchrzak-Kita,
    6. D. P. Baker,
    7. N. Hay,
    8. E. N. Fish,
    9. L. C. Platanias
    . 2008. Role of the Akt pathway in mRNA translation of interferon-stimulated genes. Proc. Natl. Acad. Sci. USA 105: 4808–4813.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Kharas M. G.,
    2. R. Okabe,
    3. J. J. Ganis,
    4. M. Gozo,
    5. T. Khandan,
    6. M. Paktinat,
    7. D. G. Gilliland,
    8. K. Gritsman
    . 2010. Constitutively active AKT depletes hematopoietic stem cells and induces leukemia in mice. Blood 115: 1406–1415.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Yao S.,
    2. B. F. Buzo,
    3. D. Pham,
    4. L. Jiang,
    5. E. J. Taparowsky,
    6. M. H. Kaplan,
    7. J. Sun
    . 2013. Interferon regulatory factor 4 sustains CD8(+) T cell expansion and effector differentiation. Immunity 39: 833–845.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Nguyen H.,
    2. C. V. Ramana,
    3. J. Bayes,
    4. G. R. Stark
    . 2001. Roles of phosphatidylinositol 3-kinase in interferon-gamma-dependent phosphorylation of STAT1 on serine 727 and activation of gene expression. J. Biol. Chem. 276: 33361–33368.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Pearce E. L.,
    2. A. C. Mullen,
    3. G. A. Martins,
    4. C. M. Krawczyk,
    5. A. S. Hutchins,
    6. V. P. Zediak,
    7. M. Banica,
    8. C. B. DiCioccio,
    9. D. A. Gross,
    10. C. A. Mao,
    11. et al
    . 2003. Control of effector CD8+ T cell function by the transcription factor Eomesodermin. Science 302: 1041–1043.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Intlekofer A. M.,
    2. N. Takemoto,
    3. E. J. Wherry,
    4. S. A. Longworth,
    5. J. T. Northrup,
    6. V. R. Palanivel,
    7. A. C. Mullen,
    8. C. R. Gasink,
    9. S. M. Kaech,
    10. J. D. Miller,
    11. et al
    . 2005. Effector and memory CD8+ T cell fate coupled by T-bet and eomesodermin. Nat. Immunol. 6: 1236–1244.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Powell J. D.,
    2. G. M. Delgoffe
    . 2010. The mammalian target of rapamycin: linking T cell differentiation, function, and metabolism. Immunity 33: 301–311.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Waickman A. T.,
    2. J. D. Powell
    . 2012. Mammalian target of rapamycin integrates diverse inputs to guide the outcome of antigen recognition in T cells. J. Immunol. 188: 4721–4729.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Thompson L. J.,
    2. G. A. Kolumam,
    3. S. Thomas,
    4. K. Murali-Krishna
    . 2006. Innate inflammatory signals induced by various pathogens differentially dictate the IFN-I dependence of CD8 T cells for clonal expansion and memory formation. J. Immunol. 177: 1746–1754.
    OpenUrlAbstract/FREE Full Text
    1. Pham N. L.,
    2. V. P. Badovinac,
    3. J. T. Harty
    . 2009. A default pathway of memory CD8 T cell differentiation after dendritic cell immunization is deflected by encounter with inflammatory cytokines during antigen-driven proliferation. J. Immunol. 183: 2337–2348.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Zehn D.,
    2. S. Roepke,
    3. K. Weakly,
    4. M. J. Bevan,
    5. M. Prlic
    . 2014. Inflammation and TCR signal strength determine the breadth of the T cell response in a bim-dependent manner. J. Immunol. 192: 200–205.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Dalod M.,
    2. T. P. Salazar-Mather,
    3. L. Malmgaard,
    4. C. Lewis,
    5. C. Asselin-Paturel,
    6. F. Brière,
    7. G. Trinchieri,
    8. C. A. Biron
    . 2002. Interferon alpha/beta and interleukin 12 responses to viral infections: pathways regulating dendritic cell cytokine expression in vivo. J. Exp. Med. 195: 517–528.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Obar J. J.,
    2. E. R. Jellison,
    3. B. S. Sheridan,
    4. D. A. Blair,
    5. Q. M. Pham,
    6. J. M. Zickovich,
    7. L. Lefrançois
    . 2011. Pathogen-induced inflammatory environment controls effector and memory CD8+ T cell differentiation. J. Immunol. 187: 4967–4978.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section