IL-1, generally considered an amplifier of adaptive immune responses, has been proposed for use as adjuvant during immunization with weak immunogens. However, its effects on memory T cell function remain largely undefined. Using the murine model of acute viral infection, in this paper, we show that in addition to augmenting the size of the Ag-specific pool, IL-1 signals act directly on CD8 T cells to promote the quality of effector and memory responses. Ablation of IL-1R1 or MyD88 signaling in T cells led to functional impairment; both the ability to produce multiple cytokines on a per cell basis (polyfunctionality) and the potential for recall proliferation in response to antigenic restimulation were compromised. IL-1 supplementation during priming augmented the expansion of Ag-specific CD8 T cells through the MyD88–IRAK1/4 axis, resulting in a larger memory pool capable of robust secondary expansion in response to rechallange. Together, these findings demonstrate a critical role of the IL-1–MyD88 axis in programming the quantity and quality of memory CD8 T cell responses and support the notion that IL-1 supplementation may be exploited to enhance adoptive T cell therapies against cancers and chronic infections.
Interleukin-1 is largely produced by innate immune cells during infection in response to cellular damage and distress sensing by Nod-like receptors (1–3). In addition to its well-established effects on innate immune cells, there is increasing evidence that IL-1 directly regulates diverse adaptive immune cells, particularly B cells and TH1, TH2, and TH17 cells (4). Signaling through the IL-1R1 by IL-1α and IL-1β ligands has been shown to be crucial for priming of CD8 T cells specific to influenza A virus (IAV) (5). Likewise, diminished CD8 T cell responses have been reported in the absence of IL-1R1 following infection with other intracellular pathogens such as Mycobacterium tuberculosis (6) and vaccinia virus (7). Thus, similar to its role as an amplifier of CD4 T cell responses, there is general consensus that IL-1 also serves to amplify CD8 T cell responses following infections, and IL-1 has been proposed for use as adjuvant during immunization with weak immunogens to augment CD8 T cell immunity (8).
It remains unknown whether IL-1 enhances CD8 T cell immunity by solely augmenting quantitative aspects of Ag-specific CD8 T cells or by also enhancing their functional capabilities. Because terminal effector and memory fates are instilled during early stages of CD8 T cell expansion when IL-1 production is at the highest, a specific role of IL-1 in programming of cell fate and function merits investigation. Moreover, direct effects of IL-1 signals on CD8 T cells and indirect effects through modulation of other immune cells known to impact CD8 T cell responses (such as CD4 T cells, dendritic cells [DCs], and NK cells) remain to be clearly understood.
In this study, we conducted a comprehensive characterization of Ag-specific CD8 T cell expansion, effector differentiation, and memory responses in the absence of IL-1R1 signaling using the murine model of infection with lymphocytic choriomeningitis virus (LCMV). We engaged complementary strategies of il1r1 and myd88 gene deletion, supplementation, and blockade of IL-1 signals to query the role of IL-1–MyD88 signaling axis in CD8 T cell immunity. Our studies show that ablation of IL-1R1 or MyD88, a key adaptor protein downstream of IL-1R1 (9) and TLRs, led to compromised expansion of Ag-specific cells in a CD8 T cell–intrinsic manner. In addition, effector cells generated in the absence of IL-1 signals exhibited reduced polyfunctionality and engendered a functionally impaired memory pool with reduced cytokine production on a per cell basis, compromised ability to coproduce IFN-γ, TNF-α, and IL-2, and impaired recall expansion upon antigenic rechallenge. Importantly, IL-1 signals regulated memory recall responses in a CD8 T cell–intrinsic manner, as exemplified by a head-to-head comparison of wild-type (WT) and IL-1R1– or MyD88-deficient CD8 T cells in the same infected milieu. Moreover, IL-1 supplementation during priming and expansion of Ag-specific murine CD8 T cells and CD19 chimeric Ag receptor (CAR)-modified human T cells augmented expansion. Together, these findings establish a key role of the IL-1–MyD88 signaling axis in promoting CTL quantity and function through cell-intrinsic effects and bears implications for the development of robust adoptive cell therapies against cancers and chronic infections.
Materials and Methods
Mice and infection
C57BL/6 mice (Thy1.2+, Thy1.1+, or Ly5.1+) were purchased from The Jackson Laboratory (Bar Harbor, ME). Thy1.1+ P14 mice bearing the H-2Db GP33 epitope-specific TCR (KAVYNFATM) were fully back-crossed to C57BL/6 mice and were maintained in our animal colony. CD4 TCR transgenic SMARTA mice, which have CD4 T cells specific for the gp67–77 (KGVYQFKSV) epitope of LCMV, were also maintained in our colony. To generate P14 or SMARTA chimeric mice, indicated numbers of Ag-specific CD8 (or CD4) T cells were adoptively transferred i.v. into naive mice ∼12 h prior to infection. IL-1R1−/− mice were purchased from The Jackson Laboratory along with B6.129SF2J control mice. MyD88flox/flox mice crossed with CD4-Cre mice (MyD88ΔT) with a T cell–specific deficiency in MyD88 were kindly provided by Dr. R. Medzhitov (10) (Yale University, New Haven, CT). All animals were used in accordance with University Institutional Animal Care and Use Committee guidelines. Armstrong strain of LCMV was propagated, titered, and used as previously described (11–13). For primary infections, mice were directly injected i.p. with 2 × 105 PFU LCMV.
Bone marrow chimeras
Mixed bone marrow chimeras were generated by reconstituting lethally irradiated (1000 rad) WT mice (Ly5.1) with a 1:1 mixture of WT (Thy1.1) and IL-1R1−/− (Thy1.2) bone marrow cells. Twelve weeks were allowed for immune reconstitution prior to experimentation.
All Abs were purchased from BioLegend (San Diego, CA) with the exception of granzyme B (Invitrogen). MHC class I (MHC-I) tetramers were made as described previously (13). Cells were stained for surface or intracellular proteins and cytokines as previously described (13). For analysis of intracellular cytokines, 2 × 106 lymphocytes were stimulated with 0.2 μg/ml GP33–41 peptide in the presence of brefeldin A for 6 h, followed by surface staining for CD8, Ly5.1, Thy1.1, or Thy1.2 and intracellular staining for IFN-γ, TNF-α, or IL-2. Flow cytometric analysis was performed on LSRII Fortessa (BD Biosciences, San Jose, CA). Single-cell suspensions of spleen cells, lymph nodes, lungs, livers, or PBMCs from mice were prepared, and direct ex vivo staining was carried out as described previously (12).
Murine T cell stimulation
Ag-specific CD8 T cells were purified through negative selection (MojoSort) and activated using GP33–41 peptide–loaded APCs for 4.5 d. Cells were expanded in 10% FBS RPMI media supplemented with IL-1β or IL-1β + IRAK1/4 inhibitor IN1449 as indicated. The cells were stimulated with 10 ng/ml of IL1-β for the first 3.5 d and 15 ng/ml of IL1-β for the last 24 h. IRAK1/4 inhibitor was used at 2 μg/ml. For in vivo follow-up experiments, 1 × 106 cultured Ag-specific cells were then transferred into naive C57/B6 mice.
Paired or unpaired Student t test (two-tailed) was used as indicated to evaluate differences between sample means. A Mann–Whitney U test was performed from pooled experiments where indicated. All statistical analyses were performed using Prism 5, and p values of statistical significance are depicted by asterisk per the Michelin guide scale. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and p > 0.05 was considered ns.
IL-1 signals drive robust CD8 T cell expansion during viral infection
We first confirmed the role of IL-1 signals in promoting CD8 T cell expansion using the murine model of LCMV infection. WT and IL-1R1−/− mice with LCMV were either sacrificed at day 8 postinfection to analyze total numbers and functional properties of effector CD8 T cells, or Ag-specific cells were followed longitudinally in blood using MHC-I tetramer staining to assess the kinetics of expansion, contraction, and memory differentiation. We examined DbGP33 as well as two other immunodominant epitope specificities, DbGP276 and DbNP396. Consistent with previous reports (14, 15), an analysis of the kinetics of Ag-specific CD8 T cell responses in blood showed evident impairment of CTL expansion in IL-1R1−/− mice at the peak of effector responses (day 8 postinfection) (Fig. 1A). In the absence of IL-1 signals, DbGP33-, DbNP396-, and DbGP276-specific effector CD8 T cells as well as the total Ag-specific CD8 T cell compartment was significantly smaller (Fig. 1B), and lower numbers were observed in both secondary lymphoid (spleen) and nonlymphoid (liver) organs (compared with WT CD8 T cells, IL-1R1−/− CD8 T cells showed ∼5–10-fold lower numbers in spleen and ∼2-fold lower numbers in liver) (Supplemental Fig. 1A, 1B). Decreased numbers of Ag-specific CD8 T cells were also confirmed by enumerating Ag-specific IFN-γ–producing CD8 T cells in response to in vitro stimulation with cognate peptide (Fig. 1C). We observed that reduced expansion of LCMV-specific CD8 T cells in the absence of IL-1 signals further extended to subdominant epitopes, and significantly lower (∼10-fold lower) numbers of Ag-specific cells for most epitope specificities were observed (Fig. 1C). These data demonstrate that IL-1R1 is crucial for CD8 T cell expansion across multiple dominant and subdominant epitope specificities during viral infection.
IL-1 signals are necessary for optimal cytokine production by effector CTLs
Because effector and memory functional properties are imprinted during early stages of CD8 T cell priming and expansion when IL-1 production peaks (data not shown), we next assessed the functional competence of WT and IL-1R1−/− effector CD8 T cells to produce effector cytokines IFN-γ and TNF-α and also IL-2 (cytokine associated with central memory-fated CD8 T cells). As shown in Fig. 2A, polyfunctionality of Ag-specific effector cells at the peak of expansion was markedly reduced in the absence of IL-1 signals; all epitope specificities tested showed significantly reduced proportions of cells coproducing TNF-α and IFN-γ. Proportions of cells coproducing IL-2, TNF-α, and IFN-γ were also reduced, with highest differences observed in DbGP276-, DbGP92-, and DbNP205-specific IL-1R1−/− CD8 T cells (Fig. 2A). The level of expression of IFN-γ was also modestly reduced on a per cell basis for most epitope specificities in the absence of IL-1 signals (Supplemental Fig. 1C). Impaired polyfunctionality of Ag-specific CD8 T cells in the absence of IL-1R1 was associated with reduced proportions of CD127+ memory precursor effector cells in both spleen (Fig. 2B) and liver (Supplemental Fig. 1D) and by increased granzyme B expression (Supplemental Fig. 1E) and TCR stimulation, indicated by surrogate PD-1 expression (Supplemental Fig. 1F). These data demonstrate that IL-1 signals promote the expansion of memory-fated effector CTLs and also exert a critical role in programming the polyfunctionality of Ag-specific effector CD8 T cells.
IL-1 signals are needed for establishing an optimal memory CD8 T cell pool with robust polyfunctionality
To determine if absence of IL-1 signals also impacted memory CD8 T cell development, we next examined the quantity and quality of LCMV-specific CD8 T cells in the memory phase (day >30 postinfection). Consistent with diminished effector size and cytokine production, we found that Ag-specific memory CD8 T cell numbers were also diminished (by ∼2–6-fold in spleen) in the absence of IL-1 signals, as enumerated by both tetramer staining (Fig. 3A) and intracellular IFN-γ staining (Fig. 3B). All LCMV epitope specificities analyzed using IFN-γ production following peptide stimulation were lower in IL-1R1–deficient mice compared with WT LCMV-infected mice, with significantly lower GP33, NP396, and GP276 responses (18-fold for NP396, 5–7-fold for GP33 and GP276). Subdominant epitopes (such as GP118, GP92) also showed a modest decrease in IFN-γ+ CD8 T cells (Fig. 3B). Phenotypically, Ag-specific memory CD8 T cells in IL-1R1−/− mice exhibited impaired upregulation of memory marker CD127 and were found to be predominantly of the terminally differentiated CD127–KLRG-1+ phenotype (Fig. 3C) in both secondary lymphoid spleen location (Fig. 3C, Supplemental Fig. 1G) and in the peripheral liver site (Supplemental Fig. 1H). Importantly, Ag-specific memory CD8 T cells in IL-1–deficient mice presented an evident defect in polyfunctionality. For three major LCMV epitope specificities (GP33, NP396, and GP276), coproduction of IFN-γ, TNF-α, and IL-2 was evidently impaired; the frequencies of both double-producing (IFN-γ+ TNF-α+) and triple-producing (IFN-γ+ TNF-α+ IL-2+) CD8 T cells were markedly reduced (Fig. 3D). Furthermore, the amounts of IFN-γ and TNF-α cytokines produced in IL-1R1–deficient Ag-specific CD8 T cells were significantly lower on a per cell basis compared with their WT counterparts (Fig. 3E). Collectively, data presented thus far show that IL-1 signals are not only important for optimal effector responses but are also critical for the formation of a robust and functional memory CD8 T cell repertoire.
IL-1 supplementation during priming augments expansion of virus-specific T cells
To confirm our conclusions from in vivo studies using IL-1R1–deficient mice that IL-1 signals are necessary for driving potent Ag-specific CD8 T cell expansion and for programming a robust memory CD8 T cell pool, we employed the minimalistic system of in vitro stimulation of LCMV GP33-specific TCR transgenic CD8 T cells in the presence or absence of IL-1 supplementation (Supplemental Fig. 2A). Compared with peptide stimulation alone, supplementation with IL-1 led to increased expansion of Ag-specific CD8 T cells (Fig. 4A), which correlated with modest enhancement of cell proliferation (as assessed by slightly greater extent of CFSE dilution) as well as modest increase in expression of prosurvival molecule Bcl-2 (Fig. 4B), whereas the expression of prosurvival molecule CD127 and activation marker CD25 remained largely similar (Supplemental Fig. 2B). To further confirm that IL-1R1 signal transduction is mediated through the downstream MyD88–IRAK-1/IRAK-4 molecular axis, we employed the IRAK-1/IRAK-4 inhibitor IN-1449. As predicted, stimulation with cognate peptide and IL-1 in the presence of IRAK-1/IRAK-4 inhibitor blocked the beneficial effects of IL-1 on CD8 T cells. The proliferation and expansion benefits provided by IL-1β signals were ablated upon addition of IRAK-1/IRAK-4 inhibitor, resulting in decreased numbers despite similar levels of Bcl-2 (Fig. 4A, 4B). Importantly, the programming effects of IL-1 supplementation during CD8 T cell priming also impacted CD8 T cell memory formation following adoptive transfer of equal numbers of Ag-specific CD8 T cells stimulated with cognate peptide in the presence or absence of IL-1 and IRAK-1/IRAK-4 inhibitor (Fig. 4C, Supplemental Fig. 2C). IL-1 supplementation led to ∼2–4-fold higher memory numbers in secondary lymphoid (spleen and inguinal lymph nodes) and nonlymphoid (lung) tissues compared with groups that did not receive IL-1 supplementation or were supplemented with IL-1 in the presence of IRAK-1/IRAK-4 inhibitor (Fig. 4C, Supplemental Fig. 2C). Nonetheless, the phenotypic properties of memory cells programmed in the presence or absence of IL-1 supplementation were largely similar as evidenced by similar upregulation of prosurvival memory marker CD127 and the lymph node homing marker CD62L during late memory stage (Supplemental Fig. 2D). Furthermore, when these memory cells were challenged with Listeria monocytogenes expressing LCMV peptide GP33, we found that cells that were originally treated with IL-1β underwent higher secondary expansion (Supplemental Fig. 2E). Based on the beneficial effects of IL-1 on virus-specific CD8 T cell expansion, we supplemented lentivirally transduced CAR T cells with IL-1 supplementation during CAR T cell generation to see if we could augment T cell recoveries. In line with our previous observations, we observed a modest increase in CAR T cell numbers with minimal changes to their phenotypic, survival, and homing markers (data not shown). Collectively, these observations demonstrate the beneficial programming effects of IL-1 supplementation on CD8 T cell expansion and memory and suggest that optimizing IL-1 signals could help boost therapeutic T cell products.
T cell–intrinsic MyD88 signals promote memory polyfunctionality and recall expansion
MyD88 is a key adaptor molecule downstream of IL-1R1 and most TLRs (9). Data presented in Fig. 4 support the notion that IL-1 signals largely transduce through the MyD88–IRAK-1/IRAK-4 molecular mediators. Previous studies have also reported reduced CD8 T cell expansion in MyD88-deficient mice, albeit its effects on cytokine production and recall expansion remain unclear. To determine if MyD88 expression in T cells is necessary for robust cytokine production by Ag-specific CD8 T cells, we used Myd88fl/fl cd4-cre mice in which Myd88 is ablated specifically in CD4 and CD8 T cells (10). Because CD4 T cells have been shown to regulate CD8 T cell responses in certain cases (16), we adoptively transferred WT TCR transgenic SMARTA CD4 T cells in Myd88fl/fl cd4-cre mice prior to infection to provide robust CD4 T cell help (Fig. 5A). Additionally, Myd88-deficient mice have been previously shown to mount reduced CD8 T cell responses to LCMV infection, thus resulting in impaired viral clearance. To ensure that our results related to CD8 T cell function were not confounded by viral persistence, we adoptively transferred WT TCR transgenic P14 CD8 T cells into WT and MyD88-deficient mice (Fig. 5A), which are expected to expand potently and effectively control the virus. This was confirmed by robust expansion (Supplemental Fig. 3A), lack of functional impairment (Supplemental Fig. 3B), and largely similar granzyme B (Supplemental Fig. 3C) and PD-1 (Supplemental Fig. 3D) expression on WT donor CD8 T cells isolated from both WT and Myd88fl/fl cd4 Cre recipients.
To assess CD8 T cell–intrinsic effects of MyD88, we focused on endogenous DbGP33-, DbNP396-, and DbGP276-specific CD8 T cell populations in WT and MyD88-deficient mice. As in the case of IL-1R1–deficient mice, these analyses uncovered significant defects in the expansion of MyD88-deficient CD8 T cells compared with their endogenous WT counterparts; endogenous (GP33-, NP396-, and GP276-specific) CD8 T cells lacking MyD88 expanded ∼5–10-fold lower than their WT counterparts (Fig. 5B), and their absolute numbers were reduced in most lymphoid and nonlymphoid organs analyzed (Fig. 5B) compared with corresponding Ag specificities in WT mice. Furthermore, the numbers of endogenous IFN-γ+ CD8 T cells were reduced 5–10-fold in MyD88-deficient mice compared with WT mice for all immunodominant and immunorecessive epitopes analyzed (Fig. 5C). Importantly, similar to IL-1R1–deficient CD8 T cells (Figs. 2, 3), Myd88-deficient effector (day 8) as well as memory (day >30) CD8 T cells were functionally impaired and produced lower levels of IFN-γ on a per cell basis (Fig. 5D). Polyfunctionality of Ag-specific effector (Supplemental Fig. 3E) and memory (Supplemental Fig. 3F) CD8 T cells, as marked by their ability to coproduce multiple cytokines IFN-γ, TNF-α, and IL-2, was also reduced upon CD8 T cell–specific ablation of MyD88. However, granzyme B expression was similar between WT and MyD88−/− DbGP33-, DbNP36-, and DbGP276-specific CD8 T cells (Supplemental Fig. 3G), suggesting a specific impact of MyD88 signals on the production of effector cytokines but not effector molecule granzyme B. In contrast to endogenous CD8 T cell responses, which were quantitatively and functionally impaired in MyD88-deficient mice compared with WT mice, WT GP33-specific donor CD8 T cells exhibited potent expansion (Supplemental Fig. 3A, 3B) and produced robust antiviral cytokines IFN-γ, TNF-α, and IL-2 (Supplemental Fig. 3B) and effector molecule granzyme B (Supplemental Fig. 3C) similarly whether they differentiated in MyD88-deficient or WT environments. These observations support the notion that CD8 T cell–specific MyD88 expression is critical for driving robust effector CTL expansion and cytokine production. Notably, lack of CD8 T cell–specific MyD88 expression also led to impaired expansion of memory cells on a per cell basis in response to secondary challenge (Fig. 5E). Collectively, these observations demonstrate that IL-1 and MyD88 signals are crucial for optimal effector and memory responses with respect to quantity and functional competence (cytokine production and recall expansion).
IL-1R1 functions in a CD8 T cell–intrinsic manner to drive the generation of a robust memory CD8 T cell pool
Our data presented diminished effector as well as compromised memory repertoire of CD8 T cells in IL-1R1–deficient mice. In IL-1R1–deficient mice, all cell types lack the expression of IL-1R1. Hence, defective effector and memory CD8 T cell responses could result from indirect effects of IL-1 signaling in non-CD8 T cells, such as DCs, CD4 T cells, etc. Additionally, viral clearance was slightly delayed in the IL-1R1−/− mice (data not shown), as also indicated by increased granzyme B and PD-1 expression on Ag-specific CD8 T cells (Supplemental Fig. 1E, 1F). To determine if IL-1 signals are required in a CD8 T cell–intrinsic manner during acute infection (as suggested by our data using MyD88-deficient CD8 T cells [Fig. 5]) without any confounding effects of prolonged stimulation associated with delayed pathogen clearance, we employed the strategy of mixed bone marrow chimerism, where WT and IL-1R1−/− CD8 T cells were analyzed in the same environment. We first generated mixed bone marrow chimeras by adoptively transferring equal numbers of WT and IL-1R1−/− bone marrow cells into lethally irradiated WT recipients (Fig. 6A). Following immune reconstitution, WT and IL-1R1–deficient CD8 T cells were found in similar proportions in the periphery (data not shown). Thereafter, chimeric mice were infected with LCMV, and Ag-specific CD8 T cell responses were evaluated after gating on WT or IL-1R1−/− donor cells. We observed significantly lower proportions of DbGP33-specific IL-1R1−/− effector (Fig. 6B) and memory (Fig. 6C) CD8 T cells compared with WT cells in the same mouse. Diminished memory size of IL-1R1−/− CD8 T cells was evident in all lymphoid and nonlymphoid organs analyzed (Fig. 6C), albeit CD127+ and KLRG-1+ proportions remained similar in effector and memory cells (Supplemental Fig. 4A, 4B), likely due to absence of any confounding factors of delayed viral clearance observed in Figs. 1 and 2. Reduced memory numbers of IL-1R1−/− CD8 T cells were confirmed by both tetramer staining (Fig. 6C) and peptide stimulation assays (Supplemental Fig. 4C) and were observed for multiple epitope specificities as evidenced by proportions of IFN-γ+ CD8 T cells following in vitro peptide stimulation (Supplemental Fig. 4C). It is noteworthy that the level of production of IFN-γ and the proportion of double cytokine-producing (IFN-γ and TNF-α) CD8 T cells were decreased in IL-1R1–deficient CD8 T cells (Fig. 6D) compared with their WT counterparts, indicating that both size and function of memory CD8 T cells are regulated by direct IL-1 signals in CD8 T cells. Data shown in Supplemental Fig. 1 and previous studies have all exhibited impaired viral clearance in IL-1R1−/− mice (14, 17). In this study, we uncoupled the effect of Ag persistence from the absence of IL-1 signaling. Because Ag clearance was not impaired, the compromised memory responses stem directly from the absence of IL-1R1 signaling in CD8 T cells.
In addition to polyfunctionality, robust recall expansion upon secondary challenge is another hallmark property of memory CD8 T cells. We found that IL-1R1–deficient memory CD8 T cells were further compromised in their recall proliferation potential; compared with WT CD8 T cells, IL-1R1−/− memory CD8 T cells did not mount a robust secondary expansion upon restimulation with cognate Ag (Fig. 6E) despite exhibiting similar phenotypic properties of CD127+ and KLRG-1− (S6B). CFSE dilution showed ∼2-fold decrease in proliferation of CD8 T cells when IL-1 signaling was ablated; BrdU incorporation, another measure of proliferation, also demonstrated lower DNA replication in the absence of IL-1R1 (Fig. 6E). These data are consistent with observations of enhanced secondary expansion of memory cells primed in the presence of IL-1 supplementation (Supplemental Fig. 2E). Although IL-1R1–deficient memory CD8 T cells showed impaired cytokine production and recall expansion following cognate Ag stimulation, they were unaffected in their ability to undergo proliferation in response to homeostatic cytokines such as IL-7 and IL-15 (Supplemental Fig. 4D). Collectively, these data demonstrate an intrinsic requirement for IL-1 signaling in CD8 T cells in driving robust primary expansion and imprinting hallmark memory properties of potent secondary expansion and polyfunctional cytokine production in response to secondary antigenic stimulation.
IL-1 signaling through the IL-1R1 is believed to act through DCs (e.g., IAV) (5) and T cells (e.g., protein immunization) (8, 18) to promote robust CD4 and CD8 T cell expansion. Our data provide evidence that the IL-1–MyD88–IRAK1/IRAK4 signaling axis in CD8 T cells promotes primary expansion and also serves to program robust polyfunctionality and secondary expansion of Ag-specific memory CD8 T cells during viral infection. Together with previous reports that IL-1 signals can act through infected and bystander DCs to prime CD8 T cell responses (5), our observations support the notion that induction of optimal magnitude and quality of CTL effector and memory responses requires a cooperative action of IL-1 on both innate cells and adaptive T cells.
Our study identifies a unique role of IL-1 in programming optimal functional properties of vigorous recall expansion of memory CD8 T cells in response to secondary challenge and polyfunctionality (coproduction of IFN-γ, TNF-α, and IL-2 cytokines). These findings in CD8 T cells parallel similar exertions of IL-1 in inducing IFN-γ, IL-23, and GM-SCF production in Th1 and Th17 CD4 T cells (19, 20) and uncover a functional dimension to IL-1 enhancement of protective CD8 T cell immunity, which was largely ascribed to increased Ag-specific CD8 T cell numbers previously. Previous studies have reported reduced expansion of Ag-specific CD8 T cells in IL-1R1−/− mice following a variety of intracellular infections such as IAV (5), vaccinia virus (7), M. tuberculosis (6), and even LCMV to varying degrees (14, 17). Our studies delineate a CD8 T cell–intrinsic role of IL-1R1 in driving optimal CD8 T cell quantity and function. In straight IL-1R1−/− mice, diminished CD8 T cell expansion has been attributed to suboptimal priming by IL-1R1−/− DCs, compromised CD4 T cell help, and even diminished type I IFN responses (21). Using the strategy of mixed bone marrow chimerism, our observations of impaired primary expansion, polyfunctionality, and recall expansion of IL-1R1–deficient CD8 T cells, compared with their WT counterparts in the same milieu, demonstrate that IL-1 signals act directly on CD8 T cells to promote various quantitative and qualitative aspects of CD8 T cell memory. Notably, our strategy of mixed bone marrow chimerism of IL-1R1–deficient and WT CD8 T cells or adoptive transfer of WT T cells into MyD88-deficient mice not only permitted directly querying the CD8 T cell–intrinsic role of IL-1R1 or MyD88 in regulating CD8 T cell responses but also allowed for doing so in the absence of any confounding factors related to impaired pathogen clearance previously reported in both IL-1R1 (14, 15) and MyD88 knockout (KO) mice (15, 17, 22). Effective pathogen control in the above systems dismisses any potential effects of repetitive antigenic stimulation in driving loss of memory function or mediating alterations in the expression of effector molecules (such as granzyme B).
In the minimalistic in vitro culture model, IL-1–induced CD8 T cell expansion was related to modestly increased proliferation and/or survival. Modest reduction in cell cycling and no differences in expression of Bcl-2 in the presence of inhibitor suggest that IL-1 may support CD8 T cell expansion through enhanced proliferation and cell survival through pathways other than Bcl-2. Loss of IL-1 enhancement of CD8 T cell expansion upon blockade of IRAK1/IRAK4, typically functioning downstream of MyD88, are consistent with our data in T cell–specific MyD88 deficiency model and are in line with the purported MyD88–IRAK1/IRAK4 signaling cascade downstream of IL-1R1 (15, 17, 22). Importantly, enhanced memory numbers following adoptive transfer of activated CD8 T cells supplemented with IL-1 clearly delineate a role of IL-1 signals during the programming phase and are consistent with previous reports of augmented T cell immunity to weak immunizations by IL-1 (8, 18).
MyD88 is a key adaptor protein required for intracellular signal transduction through IL-1R1 as well as most TLRs (4). Our observations of reduced expansion in the absence of CD8 T cell–specific MyD88 ablation is consistent with previous reports using MyD88−/− mice (15, 17, 22–24). Likewise, our observation of T cell–intrinsic requirement for MyD88 in promoting CD8 T cell expansion is also consistent with a previous study involving conditional MyD88 KO mice (24). Nonetheless, the role of MyD88 signals in promoting CD8 T cell polyfunctionality has thus far remained unexplored during the expansion and memory programming phase. Combined with a previous report that conditional ablation of MyD88 during the memory phase does not impact cytokine production (24), our observations of impaired polyfunctionality following T cell–specific ablation of IL-1R1 or MyD88 prior to T cell activation support the notion that IL-1 and downstream MyD88 signals specifically act during the expansion and programming phase to imprint optimal cytokine production potential on Ag-specific CD8 T cells. Polyfunctional CD8 T cells capable of robust recall expansion are an important goal of most vaccines against intracellular pathogens. Hence, our studies have direct relevance to immunization strategies as well as immunotherapies aimed at eliciting protective CD8 T cell immunity.
We thank Dr. Ruslan Medzhitov for providing the conditional MyD88 KO mice and for scientific discussions, and Laura Penny for excellent technical assistance.
This work was supported by research funding from the Seattle Children’s Research Institute to S.S. and V.K.
The online version of this article contains supplemental material.
Abbreviations used in this article:
chimeric Ag receptor
influenza A virus
lymphocytic choriomeningitis virus
MHC class I
The authors have no financial conflicts of interest.