The Adaptor Molecule MyD88 Directly Promotes CD8 T Cell Responses to Vaccinia Virus¹

Vaccinia virus (VACV)³ is a large DNA virus and is a member of the genus Orthopoxvirus, which includes variola, monkeypox, buffalopox,, and cowpox. Variola, the etiological agent of smallpox, long considered to be the most deadly and persistent human pathogenic disease, was eradicated by 1977 through worldwide outbreak search and vaccination with live VACV (1). In humans and mice, immunization with VACV elicits a robust CD8 T cell response that plays an important role in the final clearance of primary infection and in subsequent protection from reinfection (2, 3, 4, 5). The potency and longevity of CD8 T cell responses induced in response to VACV has led to the development of recombinant VACV as a vaccine vehicle for a number of other infectious diseases including influenza, HIV, and malaria and for cancer immunotherapy. However, despite their great efficacy in inducing strong CD8 T cell responses, we know little about the immune mechanisms behind vaccines made with VACV as a vector. Defining those that govern the efficient generation of Ag-specific CD8 T cells is of importance for the development of safer and more effective vaccines.

It is thought that antiviral CD8 T cell responses are primarily regulated by innate signals induced by the invading virus. Professional APC, such as dendritic cells (DC), play a central role in activation of naive CD8 T cells. DC can sense viruses directly through various pattern recognition receptors such as members of the TLR superfamily (6, 7, 8, 9). The recognition of viruses by DC through TLRs leads primarily to their maturation, accompanied by changes in Ag presentation, costimulatory molecule expression, proinflammatory cytokine production, and migratory behavior. and secondarily to enhanced adaptive immune responses (10). MyD88 is an adaptor protein that is required for signal transduction by most TLRs and the IL-1R/IL-18R family (11, 12). The importance of this pathway to host response to pathogens has been demonstrated by enhanced susceptibility of MyD88-deficient (MyD88^−/−) mice to a variety of eukaryotic, bacterial, and viral pathogens (13, 14, 15, 16, 17). In the context of viral infections, MyD88 has been shown to be important in host defense to a number of viruses including lymphocytic choriomeningitis virus (LCMV) (14, 18), CMV (19), and HSV (15, 20). Although the protective role of MyD88 has largely been attributed to its importance in cells of the innate immune system, recent studies have shown that naive and Ag-primed CD8 T cells can express MyD88 and TLRs (21, 22, 23). This has added an additional layer of complexity to their potential role in antiviral immunity and led to the notion that MyD88-dependent pathways could be directly involved in promoting virus-specific T cell responses.

VACV has evolved elaborate strategies that counteract many of the innate and adaptive immune responses of the host. Of potential significance are the VACV proteins encoded by the A46R, A52R, and N1L genes. Early studies using in vitro systems and cell lines showed that the A46R protein can block IL-1R and TLR signaling by interacting with MyD88, TRIF (Toll-IL-1R domain-containing adaptor-inducing IFN-β), and TRIF-related adapter molecule and consequently interfere with downstream activation of MAPK and NF-κB (24, 25). The A52 protein has been implicated in blocking IL-1R-TLR-induced NF-κB activation and proinflammatory cytokine production by targeting TNFR-associated factor 6 and IL-1R-associated kinase 2 (24, 26, 27), while N1L protein suppresses signaling to NF-κB by TLR-IL-1R by associating with, and inhibiting, the IκB kinase complex (28). Importantly, VACV deletion mutants lacking either A46R (25), A52R (27), or N1L (29) genes are highly attenuated in vivo, providing direct evidence for an important role for TLR-IL-1R and MyD88-dependent pathways in anti-VACV immune responses. However, the precise in vivo role of MyD88 in innate and adaptive immunity to VACV is not clear.

In this study, we demonstrate that T cell expression of MyD88 is necessary for the generation of large anti-VACV CD8 T cell populations. MyD88^−/− mice have a severe defect in mounting normal CD8 T cell responses against VACV. This is not due to a defect in the DC compartment as MyD88^−/− DC can mature and present viral Ag to CD8 T cells similar to wild-type (WT) DC. In contrast, adoptive T cell transfer studies show that CD8 T cells require intrinsic expression of MyD88 since WT CD8 cells can expand and differentiate into effector cells in a MyD88^−/− host, while MyD88^−/− CD8 T cells failed to expand in a MyD88-sufficient environment upon viral challenge. Our observations indicate a previously unappreciated requirement for MyD88 in the generation and accumulation of VACV-specific CD8 T cells.

The studies reported here conform to the animal Welfare Act and the National Institutes of Health guidelines for the care and use of animals in biomedical research. All experiments were done in compliance with the regulations of the La Jolla Institute Animal Care Committee in accordance with the guidelines by the Association for Assessment and Accreditation of Laboratory Animal Care. Eight- to 12-wk-old female C57BL/6, TLR2^−/−, TLR4^−/−, IL-1R^−/−, and RAG^−/− mice were all purchased from The Jackson Laboratory). OT-I TCR-transgenic mice were used as a source of Vβ5/Vα2 CD8⁺ T cells responsive to OVA-derived SIINFEKL peptide. TLR9CpG1^−/− mice and TrifLps2/Lps2^−/− mice generated on the C57BL/6 background by N-ethyl-N-nitrosourea mutagenesis were obtained from Dr. B. Beutler (The Scripps Research Institute, La Jolla, CA) (30). MyD88-deficient mice were obtained from Dr. S. Shresta (La Jolla Institute for Allergy and Immunology, La Jolla, CA) (constructed by Dr. S. Akira (Osaka University, Suita, Japan) and Dr. B. Beutler).

Vaccinia virus peptide epitopes used in this study were predicted and synthesized as described previously (31, 32). B8R (20-27; TSYKFESV), A3L (270-227; KSYNYMLL), A8R (189-196; ITYRFYLI), B2R (54-62; YSQVNKRYI), and A23R (297-305; IGMFNLTFI). MHC/peptide tetramers for the VACV Western Reserve (WR) epitope B8R (20-27; TSYKFESV)/H-2K^b, which were conjugated to allophycocyanin, were obtained from the National Institutes of Health Tetramer Core facility (Emory University, Atlanta, GA).

The VACV Western Reserve (VACV_WR) strain was purchased from the American Type Culture Collection, grown in HeLa cells, and titered on VeroE6 cells.

For most experiments, mice were infected i.p. with 2 × 10⁵ PFU of VACV. Effector responses were analyzed between days 4 and 7 postinfection, after restimulating in vitro with VACV peptides as described before (33).

DC were isolated essentially as described before (34, 35), with minor modifications. Briefly, popliteal lymph nodes or spleen fragments were digested for 20 min at room temperature with collagenase/DNase (1 mg/ml collagenase D and 1 μg/ml grade II bovine pancreatic DNase I; Boehringer Mannheim) and then treated for 5 min with EDTA to disrupt T cell-DC complexes. CD11c⁺ cells were enriched by positive selection using Miltenyi Biotec microbeads as per the manufacturer’s instructions and subsequently sorted into different subsets based on CD4 and CD8 expression by FACS.

For adoptive transfer experiments, 1 × 10⁵ or 1 × 10⁴ naive WT OT-I CD8 T cells were purified from spleens and lymph nodes of indicated naive donor mice with MACS technology (Miltenyi Biotec) and transferred into WT nontransgenic B6 or MyD88^−/− mice. One day later, mice were infected i.p. with rVACV expressing full-length OVA protein (VACV-OVA, 2 × 10⁶ PFU/mouse) or PBS as indicated. OT-I expansion and effector formation were detected by FACS staining of transgenic TCR α- and β-chains after gating on CD8 T cells and in some cases after restimulating in vitro with OVA (SINFEKL) peptide.

Lymph nodes (inguinal, brachial, axillary, superficial cervical, and mesenteric) were obtained from CD8⁺ OT-I TCR-transgenic mice and purified using positive selection with MACS beads (Miltenyi Biotec). Enriched cells contained 90–95% specific CD8⁺ TCR-transgenic T cells. These were labeled with CFSE (Molecular Probes) by incubating 10⁷ purified cells per ml with 5 μM CFSE for 10 min at 37°C. Cells were then washed three times in HBSS containing 2.5% FCS.

A total of 5 × 10⁴ CD8⁺-purified CFSE-labeled TCR-transgenic cells was added to 1.25 × 10⁴ fluorescence- activated cell-sorted DC in 200 μl of RPMI 1640 containing 10% FCS, 50 μM 2-ME, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (complete medium) in 96-well V-bottom plates (Costar; Corning). Each culture was performed in duplicate. Cultures were analyzed for proliferation after 60 h. Cells were stained with anti-CD8α-PerCP (53-6.7; BD Pharmingen), anti-Vα2-PE (B20.1; BD Pharmingen), and anti-Vβ5-allophycocyanin cells from the entire well were analyzed for proliferation by flow cytometry.

Cytokine production in T cells was performed as previously described (36), with some modifications. Briefly, after lysing RBC, splenocytes from infected mice were resuspended in RPMI 1640 medium (Life Technologies) supplemented with 10% FCS (Omega Scientific), 1% l-glutamine (Invitrogen), 100 mg/ml streptomycin, 100 U/ml penicillin, and 50 mM 2-ME (Sigma-Aldrich). One to 2 × 10⁶ cells were plated in round-bottom 96-well microtiter plates in 200 μl with medium or the indicated VACV peptides at 1 μg/ml for 1 h at 37°C. GolgiPlug (BD Biosciences) was then added to the cultures according to the manufacturer’s instructions and the incubation was continued for 7 h. Cells were stained with anti-CD8 (PerCP; 53-6.7) and CD62L (PE; MEL-14), followed by fixation with Cytofix/Cytoperm (BD Biosciences) for 20 min at 4°C. Fixed cells were subjected to intracellular cytokine staining in BD Perm/Wash buffer for 30 min at 4°C. Anti-TNF (FITC; MP6-XT22) and IFN-γ (allophycocyanin; XMG1.2) were obtained from eBiosience and used at a 1/100 dilution. Samples were analyzed for their proportion of cytoplasmic cytokines after gating on CD8⁺CD62L^low T cells by a FACSCalibur flow cytometer using CellQuest (BD Biosciences) and FlowJo software (Tree Star).

Statistical significance was analyzed by Student’s t test. Unless otherwise indicated, data represent the mean ± SEM, with p < 0.05 considered statistically significant.

Over 7 days of primary i.p. infection with VACV-WR, we quantified the number of virus-specific CD8 T cells in spleens by using a tetramer of B8R, the immunodominant class I epitope of VACV. MyD88^−/− mice generated greatly reduced B8R-reactive CD8 T cell responses compared with WT mice, based on analyses of percentages and absolute numbers of tetramer-positive cells (Fig. 1,a). Reduced percentages of CD8 T cells capable of producing IFN-γ in response to B8R were also observed, although TNF production was not strongly affected. However, because of the reduced numbers of CD8 T cells generated, this translated into significant deficiencies in the accumulation of both IFN-γ- and TNF-producing effector T cells (Fig. 1,b). MyD88^−/− mice also showed a similar defect in generating CD8 T cell populations responsive to a range of subdominant VACV epitopes, which included A3L, A8R, A23R, and B2R (Fig. 1,c). To assess whether the differentiation state of VACV-specific CD8 T cells was impacted by MyD88, the expression of CD25, CD43, CD62L, and CD127 was analyzed on CD8⁺B8R⁻tetramer⁺ cells in the spleen 5 days postinfection with VACV-WR. Virus-specific CD8 T cells from MyD88^−/− mice expressed slightly higher levels of CD25, but other activation indicators including CD69, CD43, CD62L, and CD127 were comparable to WT cells, suggesting early activation of T cells upon virus stimulation is normal in MyD88^−/− mice (Fig. 1 d).

Interestingly, the defect found in MyD88^−/− mice was not found in Trif^Lps2−/− mice. Normal B8R-reactive CD8 T cell responses were generated in Trif^Lps2−/− mice (Fig. 2,a). Production of IFN-γ and TNF by CD8 T cells in response to a range of dominant and subdominant VACV epitopes was also not dependent on Trif expression (Fig. 2, b and c), indicating some specificity in the importance of TLR-IL-1R type adaptors.

To test whether the impaired priming of CD8 T cells might have been related to a defective capacity of APCs to mature during the initial stages of infection, a recombinant virus expressing GFP (VACV-GFP) was used to infect different cell populations in vitro. VACV infected both CD8⁺ and CD8⁻ DC and macrophages, with low/negligible amounts found in T cells, B cells, granulocytes, and NK cells (Fig. 3,a). We focused on DC because they have previously been shown to be responsible for priming naive CD8 T cells to VACV (35). Four hours after in vitro infection, there was no difference observed in the capacity of any subset of MyD88^−/− DC to be infected by VACV-GFP (Fig. 3,b). Although some death of MyD88^−/− DC occurred after 20 h of viral infection (Fig. 3,b), infected DC up-regulated comparable levels of B7.2 (Fig. 3,c) as well as displaying normal levels of CD40 and class II MHC (data not shown). B7.1 expression was slightly lower in infected MyD88^−/− DC compared with WT controls (Fig. 3,c). Although we could not use the recombinant VACV-GFP in vivo to specifically focus on virus-infected cells, due to rapid loss of expression of GFP, a similar conclusion was evident from analyzing total CD11c⁺ DC from VACV-WR-infected mice. B7.1, B7.2, CD40, and class II were equivalently expressed in the absence of MyD88 over 20 h, although some evidence of delayed kinetics was apparent (Fig. 3 d and data not shown). These data strongly suggested that MyD88^−/− DC can be infected and mature comparable to WT DC after VACV infection.

To further examine where MyD88 was functional, we transferred high or low numbers of OT-I CD8 T cells into MyD88^−/− or WT mice and 1 day later infected them with VACV-OVA. Significantly, strong expansion of OVA-specific OT-I CD8 T cells was observed by tracking the transgenic TCR, and many OVA-reactive IFN-γ-producing CD8 T cells were generated, regardless of whether they were transferred into MyD88^−/− mice (Fig. 4, a and b). Notably, similar results were found when low number of OT-I cells (1 × 10⁴) were transferred (Fig. 4, a and b), even though it was reported that a high number of Ag-specific CD8 cells might influence DC maturation. These data suggested that the absence of MyD88 in the host environment does not limit the expansion and differentiation of naive CD8 T cells driven by VACV infection and that any impaired activity was not due to defective Ag presentation. In line with this, WT OT-I CD8 T cells cocultured with MyD88^−/− DC, purified from a VACV-OVA-infected host, proliferated to OVA at a rate comparable to those cultured with WT DC in vitro (Fig. 4 c).

To test whether the impaired CD8 T cell response in MyD88^−/− mice reflected an intrinsic role for MyD88, we transferred WT or MyD88^−/− CD8 T cells into WT RAG^−/− mice and infected the recipient mice with VACV-WR. RAG^−/− mice were used as recipients to prevent any endogenous T cell response. A large number of T cells were transferred because previous studies have shown that this limits any homeostatic proliferation, allowing us to effectively examine the VACV Ag-induced response. Seven days postinfection, there were greatly reduced numbers of B8R tetramer-reactive CD8 T cells generated from the MyD88^−/− donor cells (Fig. 5,a). The overall number of IFN-γ-producing VACV-specific effector CD8 T cells generated from MyD88^−/− donor populations was also strongly reduced, as shown by ex vivo restimulation with dominant and subdominant peptides (Fig. 5,b). These data show that MyD88 specifically regulates CD8 T cell expansion after encounter with viral peptides. However, it was also possible that MyD88 was required for survival/maintenance of CD8 T cells in the absence of Ag. To address this question, we transferred MyD88^−/− CD8 cells into RAG^−/− mice without virus infection. One week after transfer, there were ∼4-fold fewer donor MyD88^−/− CD8 T cells that survived compared with WT T cells (Fig. 5,c), showing an intrinsic survival defect in the absence of MyD88. However, when these mice were subsequently infected and generation of VACV-reactive CD8 T cells was examined, a similar strong reduction in all VACV-specific CD8 T cell populations was observed (Fig. 5 d). These data show that MyD88 is directly required within CD8 T cells for survival as well as expansion to VACV-derived Ag.

It has been reported that TLR2 agonists can directly enhance survival and proliferation of activated murine T cells in vitro and in vivo (21, 22, 37). We therefore first investigated whether a direct effect of this pathway on CD8 T cells could account for the MyD88 requirement. CD8 T cells from TLR2^−/− mice were transferred into RAG^−/− mice, followed by infection with VACV. Seven days postinfection, CD8 T cells from TLR2^−/− mice mounted a normal response to VACV compared with WT cells. These cells expanded normally upon viral stimulation and also produced comparable levels of virus-specific effector cytokines (Fig. 6). Consistent with this, VACV infection induced robust CD8 T cell responses to both dominant and subdominant VACV epitopes in TLR2^−/− mice that was comparable to that seen in WT mice (Fig. 7, a and b). To further confirm the redundant role of TLR2 on T cells during VACV infection, we infected splenocytes with VACV-GFP in vitro and found that, unlike DC, CD8 T cells could not be infected in vitro (Fig. 7,c). Moreover, cell surface expression of TLR2 was significantly up-regulated on DC following VACV infection in vivo, but TLR2 was undetectable on CD8 T cells (Fig. 7 d). Consistent with previous reports, the TLR2 agonist Pam3Cys directly promoted CD8 T cell proliferation in vitro when combined with anti-CD3, but a highly purified preparation of VACV did not enhance anti-CD3-induced proliferation (data not shown). Taken together, our data suggested that in our experiment model, TLR2 is not the upstream receptor of MyD88-dependent signaling pathways.

Besides TLR2, previous studies have also shown the involvement of TLR9 in direct regulation of CD4 T cell survival (38). Moreover, TLR4 mediates a protective innate immune response against VACV infection (39). Thus, we also investigated whether a direct effect of pathways linked by these two receptors on CD8 T cells could account for the MyD88 requirement. We transferred TLR9^−/− or TLR4^−/− CD8 T cells into RAG^−/− mice and challenged the recipient mice with VACV 1 day later. Similar to TLR2^−/− cells, CD8 T cells from these mice responded to the virus normally and generated comparable VACV-reactive effector CD8 T cell populations (Fig. 8). Since MyD88 also plays a role in signaling downstream of the IL-1R family and the cytokines using these receptors, IL-1 and IL-18, have been shown to influence T cell survival and effector function (40), we furthermore transferred IL-1R^−/− CD8 T cells into RAG^−/− mice. Similarly, there was also no difference in responsiveness to VACV (Fig. 8). Consistent with this, VACV infection induced robust CD8 T cell responses to both dominant and subdominant VACV epitopes in TLR9^−/− and IL-1R^−/− mice that was comparable to that seen in WT mice (data not shown). Therefore, although MyD88 plays a critical role in T cells during VACV infection, the upstream receptors of MyD88-dependent signaling pathways remain unclear.

In this study, we show that T cell expression of MyD88 is critical for the magnitude of primary CD8 T cell responses to both dominant and subdominant VACV epitopes, with MyD88^−/− T cells exhibiting profoundly reduced expansion and antiviral cytokine production. These experiments help define the precise regulatory mechanisms that govern the efficient generation of adaptive immunity. The adaptor molecule MyD88 participates in host defense to a number of viruses including RNA viruses such as LCMV and DNA viruses including herpesvirus.

To date, most studies of MyD88 in antiviral immunity have focused on cells of the innate immune system, such as DC and macrophages. Depending on the experimental system, MyD88^−/− DC show profound defects in maturation, costimulatory molecule expression, proinflammatory cytokine production, and/or Ag-presenting capacity. In response to LCMV infection, precursor DC and conventional DC from MyD88^−/− mice produce less IFN-α (18) and other inflammatory cytokines such as MCP-1, IL-1, and IL-6 (14), which can play an important role in regulating CD8 T cell expansion and survival (41, 42). Moreover, DC are dependent on the presence of MyD88 to efficiently cross-present viral Ags from infected cells (43). In the context of VACV infection, a recent study demonstrated that, similar to LCMV, production of IL-1 and IL-6 by DC was mediated through TLR-2 in a MyD88-dependent but TRIF-independent manner (44). Notably, 4- to 6-fold higher VACV titers were detected in TLR-2^−/− or MyD88^−/− mice than in WT mice as early as 3 days after infection, implying that production of innate cytokines by DC through a TLR-2/MyD88 pathway may in part contribute to innate immune control of VACV (44). In the present study, we investigated the effects of VACV infection on DC function by focusing on their capacity to mature and activate CD8 T cells. We show that VACV can infect multiple DC subsets and induce their maturation as measured by cell surface up-regulation of CD80, CD86, CD40, and MHC class II. Splenic CD8α⁺ DC were the major subset infected by VACV and this correlated with their superior capacity to activate naive CD8 T cells. Using several complementary approaches, we found that the defective CD8 T cell responses observed in MyD88^−/− mice could not be explained by a defect in DC maturation and Ag presentation. First, MyD88^−/− DC infected with VACV were able to up-regulate several costimulatory molecules comparable to WT DC. Second, strong expansion and cytokine production of WT CD8 T cells responding to Ag in the context of VACV infection was observed after transfer into MyD88^−/− mice. Third, both VACV-infected WT and MyD88^−/− DC efficiently induced Ag-specific CD8 T cell responses ex vivo. Our studies then extend previous reports by showing that in contrast to production of IL-1 and IL-6, MyD88 is not required for DC to generate normal CD8 T cell responses directed against VACV epitopes. The capacity of MyD88^−/− DC to up-regulate costimulatory molecules and efficiently stimulate VACV-specific CD8 T cells might be related to production of inflammatory cytokines that can mimic and hence replace the need for the MyD88 signal. Potentially, consistent with this idea are recent reports showing that two cytokines implicated in DC maturation and CD8 T cell priming, namely, IFN-β (44) and IL-12 (45), can be induced in DC in response to VACV in a MyD88-independent manner. Together these results demonstrate that DC can respond to VACV infection via distinct mechanisms that involve MyD88-dependent and -independent pathways.

An important observation in our study is that VACV-specific CD8 T cells need to express MyD88 for optimal expansion and cytokine production. Although much evidence implicates MyD88 as a key signaling component of innate responses, this molecule is also expressed in CD8 T cells (21, 22). We show that WT CD8 cells can expand and differentiate into effector cells in a MyD88^−/− host, while MyD88^−/− CD8 T cells failed to expand in a MyD88-sufficient environment upon VACV challenge, fully supporting the conclusion that T cells are the direct recipient of MyD88 signals. Interestingly, a recent study demonstrated that although MyD88 was dispensable for early T cell division and effector differentiation, it played an important role in supporting the survival and accumulation of LCMV-specific CD8 T cells during clonal expansion (46). Thus, MyD88-dependence by antiviral CD8 T cells may be a general phenomenon that applies to a variety of viruses.

Recently, several groups have shown that both human and murine CD4 and CD8 T cells can express functional TLR. Engagement of TLR-1/2, TLR5, TLR7/8, and TLR9 directly on CD4 T cells augments their proliferation in vitro, in part, by increasing IL-2 production and maintaining IL-2Rα chain expression (38). Moreover, TLR3 and TLR9 signaling in CD4 T cells has been shown to promote survival by increasing the expression of antiapoptotic molecules such as Bcl-x_L through activation of PI3K and NF-κB pathways (37, 47). The effects of TLR engagement on CD8 T cells are less well characterized. Naive CD8 T cells can express TLR1, TLR2, TLR6, TLR7, and TLR9 mRNA, whereas TLR3, TLR4, TLR5, and TLR8 mRNA expression is very low or undetectable (21, 22). To date, only TLR1/2 engagement with lipopeptide Pam3CysSk4, a synthetic analog of bacterial and mycoplasmal lipoprotein, on CD8 T cells has been shown to enhance proliferation, survival, and cytokine production after TCR engagement (21, 22).

In the present study, we investigated whether a direct effect of TLR2 pathway on CD8 T cells could account for the MyD88 requirement. Using an adoptive transfer system where we transferred purified naive polyclonal CD8 T cells from WT and TLR2^−/− mice into RAG^−/− mice that were then infected with VACV, we show that both percentages and absolute numbers of CD8 T cells reactive with the immunodominant (B8R) and four subdominant epitopes (A3L, A8R, A23R, B2R) of VACV were comparable between the two groups. The responses we chose account for up to 70–80% of the total VACV CD8 response (32). In all cases, we did not observe any requirement for TLR2 in either expansion or cytokine production by VACV-specific CD8 T cells. Similarly, infection of TLR2^−/− mice with VACV resulted in the generation of robust CD8 T cell responses directed against both dominant and subdominant VACV epitopes. Consistent with a lack of a role for TLR2 on CD8 T cells, we were also unable to detect any TLR2 expression on the surface of CD8 T cells at various times postinfection with VACV either in vivo or in vitro. Moreover, using recombinant VACV-expressing GFP, we show that while VACV readily infects CD11⁺ DC, it cannot infect CD8 T cells in culture. Moreover, we ruled out several other upstream receptors such as IL-1R, TLR4, and TLR9. Together, our results suggest that it is unlikely that MyD88 is downstream of TLRs sensing a VACV pathogen-associated molecular pattern directly on CD8 T cells. Similar to our results with VACV, mice deficient in TLR2, TLR4, TLR7, TLR8, TLR9, IL-1R, IL-18R, and caspase 1-deficient mice infected with LCMV do not reproduce the phenotype of MyD88^−/− mice since they mount normal LCMV-specific CD8 T cell responses (18, 46). Thus, although MyD88 plays a critical role in CD8 T cells during viral infections, the identity of its upstream receptor/s remain unclear. As has been proposed before (46) and also supported by our studies, one possibility is that MyD88-dependent viral recognition may occur through novel, as yet unidentified TLRs or other cell surface receptors. Another nonmutually exclusive possibility is that viruses such as LCMV and VACV can activate MyD88-dependent pathways in CD8 T cells indirectly, e.g., via cytokines (48).

While this manuscript was in the review process, Quigley et al. (49) also reported that VACV-specific CD8 T cells need to express MyD88 for optimal expansion/survival. Interestingly, however, although in our study we did not find any role for TLR2 in either expansion or cytokine production by VACV-specific CD8 T cells, the published article by Quigley et al. (49) concluded that TLR2 is the receptor that links to MyD88 and that TLR2 is required by these CD8 T cells. The discrepancies between our data with regard to TLR2 usage and Quigley’s finding (49) are not clear at present, but they are likely due to the different experimental systems. For example, C57BL/6 mice were used throughout in our study, whereas B10.D2 mice were used in Quigley’s study (49). Consistent with our data, Rahman et al. (46) have recently shown that mice deficient in TLR2 on the C57LB/6 background infected with LCMV also do not reproduce the phenotype of MyD88^−/− mice. Therefore, the backgrounds of the mice might simply explain the differences. Further studies are therefore required to examine the precise mechanisms by which the genetic background of mice can influence the differential usage of TLR2 by virus-specific CD8 T cells.

In conclusion, we demonstrate that although T cell expression of MyD88 is necessary for the generation of large anti-VACV CD8 T cell populations, MyD88 is not required for DC maturation and Ag presentation. Our finding that VACV-specific CD8 T cells directly require MyD88 for their expansion and survival together with other studies showing that VACV can modulate this pathway raises an intriguing idea that this virus may have evolved strategies to subvert antiviral immune responses by directly targeting CD8 T cells.

The authors have no financial conflict of interest.