Batf3-Dependent Dendritic Cells Promote Optimal CD8 T Cell Responses against Respiratory Poxvirus Infection

During respiratory infection with vaccinia virus (VacV), a member of Poxviridae family, CD8⁺ T cells play important role in resolving the primary infection. Effector CD8⁺ T cells clear the virus by accumulating in the infected lungs in large numbers and secreting molecules such as IFN-γ that kill virally infected cells. However, precise cell types that regulate the generation of effector CD8⁺ T cells in the lungs are not well defined. Dendritic cells (DCs) are a heterogeneous population of immune cells that are recognized as key initiators and regulators of T-cell-mediated immunity. In this study, we reveal that a specific subset of DCs that are dependent on the transcription factor Batf3 for their development regulate the magnitude of CD8⁺ T cell effector responses in the lungs, thereby providing protection during pulmonary VacV infection.

ABSTRACT

Respiratory infection with vaccinia virus (VacV) elicits robust CD8⁺ T cell responses that play an important role in host resistance. In the lung, VacV encounters multiple tissue-resident antigen-presenting cell (APC) populations, but which cell plays a dominant role in priming of virus-specific CD8⁺ effector T cell responses remains poorly defined. We used Batf3^−/− mice to investigate the impact of CD103⁺ and CD8α⁺ dendritic cell (DC) deficiency on anti-VacV CD8⁺ T cell responses. We found that Batf3^−/− mice were more susceptible to VacV infection, exhibiting profound weight loss, which correlated with impaired accumulation of gamma interferon (IFN-γ)-producing CD8⁺ T cells in the lungs. This was largely due to defective priming since early in the response, antigen-specific CD8⁺ T cells in the draining lymph nodes of Batf3^−/− mice expressed significantly reduced levels of Ki67, CD25, and T-bet. These results underscore a specific role for Batf3-dependent DCs in regulating priming and expansion of effector CD8⁺ T cells necessary for host resistance against acute respiratory VacV infection.

IMPORTANCE During respiratory infection with vaccinia virus (VacV), a member of Poxviridae family, CD8⁺ T cells play important role in resolving the primary infection. Effector CD8⁺ T cells clear the virus by accumulating in the infected lungs in large numbers and secreting molecules such as IFN-γ that kill virally infected cells. However, precise cell types that regulate the generation of effector CD8⁺ T cells in the lungs are not well defined. Dendritic cells (DCs) are a heterogeneous population of immune cells that are recognized as key initiators and regulators of T-cell-mediated immunity. In this study, we reveal that a specific subset of DCs that are dependent on the transcription factor Batf3 for their development regulate the magnitude of CD8⁺ T cell effector responses in the lungs, thereby providing protection during pulmonary VacV infection.

INTRODUCTION

Vaccinia virus (VacV), a double-stranded DNA (dsDNA) virus belongs to the family Poxviridae, which also constitutes threatening human pathogens such as variola virus (causative agent of smallpox) and monkeypox virus (1, 2). Eradication of smallpox using VacV vaccination was regarded as a medicinal landmark, making VacV vaccination a “gold standard” in vaccine design (3). Since then, VacV has been extensively studied in terms of its biology and has revealed remarkable insights in understanding virus immune evasion mechanisms (4). Additionally, VacV has been widely utilized to infect mice via various routes to study disease pathology and immune response, as a surrogate model for human infections with smallpox and monkeypox (5). Particularly, the mouse-adapted Western Reserve strain of VacV (VacV-WR) has been pivotal in identifying determinants of disease severity in the respiratory tract (5–8).

Following respiratory VacV-WR infection, virus initially replicates in the lungs by infecting multiple cell types, resulting in an exponential rise in viral titers within a few days (2, 9, 10). This leads to gross pulmonary damage, including perivascular and peribronchial inflammation that severely compromises respiratory functions (6, 8, 10). Subsequently, viral progenies enter the bloodstream and disseminate to other organs, such as lymph nodes, spleen, liver, gastrointestinal tract, and ovaries, causing systemic infection and lethality at high viral doses (5). Recovery from a respiratory VacV-WR infection requires a tightly coordinated response by both NK and CD8⁺ T cells (10, 11). Early in infection, local gamma interferon (IFN-γ) release by NK cells limits VacV replication in the lung (11). Several days later, as the adaptive immune response develops, virus-specific CD8⁺ T cells play a vital role in restricting virus dissemination to visceral tissues and are necessary for complete clearance of virus and protection against death (10, 12). Like NK cells, CD8⁺ T cells contribute to host resistance via the release of IFN-γ, and there is evidence that IFN-γ is sufficient to protect mice in the absence of CD4 and B lymphocytes (10, 12).

VacV-specific CD8⁺ T cells can be detected in the lung starting at day 4 postinfection (p.i.) and peak around day 8 (10, 13), correlating with the time when the virus is completely cleared (10). At the peak of the effector phase, ∼30% of total CD8⁺ T cells are specific for VacV (10). These cells then contract and decline gradually until they stabilize by day 60 as a memory population and are maintained for >2 years postinfection (13). Recently, we showed that inflammatory monocytes (IMs) provide key signals for the persistence of circulating and tissue-resident memory (TRM) CD8⁺ T cells following respiratory infection with VacV (14). Intriguingly, although IM numbers were increased in the lungs between days 4 and 8 postinfection in wild-type (WT) mice, in mice that are defective in IM recruitment to the lungs, CD8⁺ T cell expansion, accumulation, and differentiation were normal during the acute phase of infection (14). This indicated that a separate antigen-presenting cell (APC) population is responsible for generating CD8⁺ T cell effector responses in the lung tissue following respiratory VacV infection.

Lung contains a heterogeneous population of dendritic cells (DCs) categorized as plasmacytoid DCs, monocyte-derived DCs, and conventional DCs (cDCs) (15). Conventional DCs can be further separated into two main subsets namely, cDC1 (CD103⁺ CD11b⁻) and cDC2 (CD103⁻ CD11b⁺) (16). In the context of respiratory viral infections, both DC subsets have been reported to acquire viral antigen and traffic to local draining lymph nodes (DLNs) (17). CD103⁺ DCs and their ontogenically related lymphoid tissue-resident CD8α⁺ DCs are known to participate in cross-presenting viral antigens to CD8⁺ T cells (18, 19). These two DC subsets uniquely express C-type lectin receptor, DNGR1 (dendritic cell NK lectin group receptor 1), which plays important role in phagocytosis of dying cells and facilitates cross-presentation (20, 21). Furthermore, both these DC subsets are dependent on the transcription factor Batf3 (basic leucine zipper transcription factor, ATF-like 3) for their development (22–24). Thus, mice deficient in Batf3 selectively lack only lymphoid-resident CD8α⁺ DCs (23) and related peripheral CD103⁺ DCs (22), without hampering the development of other DC subsets (23). Previously, CD103⁺ DCs isolated ex vivo from VacV-infected mice were demonstrated to have superior capacity to induce in vitro T cell proliferation (25). More recently, Batf3-dependent DCs (CD8α⁺ DCs and CD103⁺ DCs, henceforth referred as Batf3-DCs) were reported to provide unique signals for tissue-resident memory (TRM) generation in the skin and the lung but not circulating memory CD8⁺ T cells following VacV infection (26). However, whether Batf3-DCs also modulate the effector phase of antiviral CD8⁺ T cells during respiratory VacV infection has not been directly demonstrated.

In this study, we found that Batf3^−/− mice were more vulnerable to disease after VacV-WR infection, as seen by significant reduction in their body weights compared to WT mice. This correlated with enhanced viral titers and dramatic defect in effector CD8⁺ T cell response in both the lungs and spleens of Batf3^−/− mice. Antigen-specific CD8⁺ T cells in the DLNs of Batf3^−/− mice showed strikingly reduced expression of Ki67 and CD25 as well as transcription factor T-bet. This suggests that in the absence of Batf3-DCs, virus-specific CD8⁺ T cells are suboptimally primed, which leads to their impaired expansion and accumulation in the lungs, thereby compromising host resistance against acute respiratory VacV infection.

RESULTS

Batf3^−/− mice lack CD8⁺ and CD103⁺ DCs following respiratory infection with VacV.

Mice deficient in transcription factor Batf3 lack the CD103⁺ DC subset in the peripheral tissues (22) and CD8α⁺ DC subset in the lymphoid tissues (23). However, during some infections, these DC subsets can be induced in Batf3^−/− mice by a compensatory mechanism driven by inflammatory cytokines (27, 28). Hence, we first decided to investigate whether intranasal VacV infection affects Batf3-DC subsets by inducing factors that allows their spontaneous generation, bypassing the need for Batf3. To this end, we infected WT and Batf3^−/− mice with VacV-WR intranasally and a few days later isolated the lungs and spleen to examine DC subsets. We found almost complete reduction in the frequencies and absolute numbers of CD103⁺ DCs in the lungs of Batf3^−/− mice compared to WT mice at day 6 postinfection (Fig. 1A). The absolute numbers of the other two lung DC subsets (double-negative [DN] DCs and CD103⁻ CD11b⁺ DCs), however, remained unchanged (Fig. 1A). Among the splenic DC subsets, a severe defect was observed in the frequencies and absolute numbers of the CD8α⁺ DC subset in Batf3^−/− mice (Fig. 1B). A concomitant increase in the frequencies and absolute numbers of DN DCs and CD4⁺ DCs was noted in Batf3^−/− mice compared with WT mice (Fig. 1B). Thus, we confirmed that CD103⁺ DCs in the lungs and CD8α⁺ DCs in the spleen remain depleted in Batf3^−/− mice following intranasal VacV-WR infection.

FIG 1.

Absence of CD103⁺ DCs in the lung tissue and CD8α⁺ DCs in the spleen tissue following intranasal VacV infection of Batf3^−/− mice. WT C57BL/6 and Batf3^−/− mice were infected with rVacV-WR-OVA intranasally (2 × 10⁴ PFU), and at day 6 postinfection, the lung and spleen tissues were harvested. (A) Lung cells were stained with LIVE/DEAD dye followed by CD3, B220, CD11b, CD11c, major histocompatibility complex class II (MHC-II), PDCA, and CD103. Prominent cells were excluded, and CD11c⁺ MHC-II⁺ cDCs were segregated using CD103 and CD11b. Absolute numbers of CD103⁺ CD11b⁻, CD103⁻ CD11b⁻, and CD103⁻ CD11b⁺ cells were quantified. (B) Similarly in the spleen tissue, after exclusion of prominent cell types, CD11c⁺ MHC-II⁺ cDCs were segregated using CD8α and CD4. Absolute numbers in each of the DC subsets were quantified. Data are representative of at least 3 independent experiments, and results are mean ± SEM (n = 3 mice/group). *, P < 0.05; **, P < 0.01. ns, not significant.

Batf3^−/− mice exhibit greater morbidity and impaired generation of IFN-γ-producing CD8⁺ T cells in the lungs following respiratory VacV infection.

Having established that Batf3-DCs remain depleted in VacV-infected mice; we sought to investigate the impact of Batf3 deficiency on disease outcome following intranasal VacV-WR infection. To this end, we infected WT and Batf3^−/− mice intranasally with a sublethal dose of VacV-WR and monitored weight loss as a measure of disease. Starting at day 4 postinfection, both WT and Batf3^−/− mice began to lose weight (Fig. 2A). However, from day 6 to day 8, Batf3^−/− mice exhibited significantly greater weight loss than their WT counterparts (Fig. 2A), indicating that lack of Batf3-DCs compromised protection against disease. Both groups of mice, however, recovered from the infection and reached their original body weight (not depicted). This is consistent with our previous data showing that as few as 10⁴ CD8⁺ T cells in the lungs are sufficient in preventing lethality and facilitating complete recovery from intranasal VacV-WR infection (10, 29). We have previously demonstrated that reduction in body weight following intranasal VacV-WR infection directly correlates with viral titers and systemic dissemination to peripheral sites such as the ovaries (10). Consistent with greater weight loss, we found that Batf3^−/− mice had significantly higher viral titers in the ovaries than WT mice (Fig. 2B). These data suggest that Batf3-DCs play important role in protection against disease and viral dissemination following respiratory VacV-WR infection.

FIG 2.

Increased weight loss in Batf3^−/− and impaired accumulation of VacV-specific CD8⁺ T cells in the lungs of Batf3^−/− mice after VacV-WR infection. (A) WT C57BL/6 and Batf3^−/− mice were infected with VacV-WR intranasally (1.5 × 10⁴ PFU), and body mass was monitored daily. Data represent a pool of two independent experiments with 3 to 4 mice per group. (B) Ovaries from Batf3^−/− and WT mice were harvested at day 8 postinfection, and viral titers were determined by plaque assay. (C to G) At day 8 postinfection, lungs were harvested and stained for CD8, CD44, and B8R (the immunodominant epitope of VacV) and absolute cell numbers were quantified. (F and G) Lung cells were incubated with the respective vaccinia virus peptides or pooled vaccinia virus peptides for the stipulated time, and intracellular expression of IFN-γ and absolute numbers of IFN-γ-positive cells were determined. Data are representative of at least three independent experiments with similar results. Results are mean ± SEM (n = 3 to 4 mice/group). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

To determine whether the greater loss in the body weight and higher viral titers in Batf3^−/− mice correlated with impaired antiviral CD8⁺ T cell immunity, we examined virus-specific CD8⁺ T cell response in the lungs at day 8 postinfection, when significant differences in body weights were observed. We found severe reduction in both the frequencies and absolute numbers of total CD8⁺ T cells in Batf3^−/− mice compared with WT mice (Fig. 2C). Activated CD8⁺ T cells (CD44⁺ CD62L⁻) were also dramatically reduced in Batf3^−/− mice (Fig. 2D). Staining with B8R tetramer, the immunodominant epitope of VacV (11), revealed no differences in the percentages of B8R⁺ cells between the two groups, but the absolute numbers of B8R⁺ cells were significantly reduced in Batf3^−/− mice compared with WT mice (Fig. 2E), suggesting that lack of Batf3-DCs results in reduced accumulation of total activated CD8⁺ T cells. To further evaluate the quality of accumulated CD8⁺ T cells in Batf3^−/− mice, we examined their ability to produce IFN-γ in response to immunodominant B8R peptide, subdominant A8R peptide, and pooled peptides (B8R, A3L, A8R, A23R, and B2R) (10, 30). Strikingly, under all three conditions, CD8⁺ T cells from Batf3^−/− mice were severely impaired in their ability to generate IFN-γ cells, as seen by their reduced percentages and absolute numbers compared with WT mice (Fig. 2F and G). This indicates that Batf3-DCs are important for accumulation of optimal numbers of functional CD8⁺ T cells in the lungs of VacV-infected mice.

Lack of Batf3-DCs severely affects the magnitude of expansion and accumulation of CD8⁺ T cells following respiratory VacV infection.

To exclude the intrinsic effect of Batf3 deficiency on T cells, we utilized the T cell receptor (TCR) transgenic T cell system, where we adoptively transferred equal numbers of fluorescence-activated cell sorter (FACS)-sorted naive (CD8⁺ CD44^lo) TCR Vα2Vβ5 transgenic OT-I cells specific for H-2K^b/OVA_257–264 into naive WT and Batf3^−/− mice and 1 day later infected them intranasally with recombinant VacV-WR expressing the full-length ovalbumin protein (rVacV-WR-OVA) (31–37). Following intranasal rVacV-WR-OVA infection, virus is cleared from the lungs as early as day 3 postinfection with no overt sign of disease or weight loss (33), allowing us to determine the impact of Batf3-DCs on CD8 T cell responses independent of differences in viral replication. The OT-I CD8⁺ T cell response was then analyzed in the lungs and spleens of WT and Batf3^−/− recipient mice at day 8 postinfection. We observed significant reduction in the percentage of OT-I cells in the lungs of Batf3^−/− recipients compared with WT recipients (Fig. 3A). An ∼70% defect in total OT-I cell numbers was observed in Batf3^−/− recipients compared to WT recipients. A similar defect was also seen in the percentages and absolute numbers of OT-I cells in the spleens of Batf3^−/− recipients compared with WT recipients (Fig. 3B). Comparison of the fold expansion from day 6 to day 8 between the two recipient groups showed that OT-I cells in WT recipients expanded ∼9-fold, whereas OT-I cells in Batf3^−/− recipients expanded only ∼4-fold (Fig. 3C). Similarly, splenic OT-I cells in Batf3^−/− mice also underwent dramatically reduced expansion (34-fold in WT and 18-fold in Batf3^−/− mice) compared with their WT counterparts (Fig. 3D). To further examine the functionality of accumulated OT-I cells, we assessed their capacity to produce the antiviral cytokines IFN-γ and tumor necrosis factor (TNF) in response to ex vivo stimulation with OVA-derived SIINFEKL peptide at day 8 postinfection. Consistent with the VacV-specific response (Fig. 2E), we observed significant reduction in both the frequencies and absolute numbers of IFN-γ (Fig. 3E and F)- and TNF (Fig. 3G and H)-producing OT-I cells in Batf3^−/− recipients compared with WT controls. Thus, at the peak of antiviral CD8⁺ T cell response, Batf3-DCs play a major role in regulating the magnitude of expansion and accumulation of functional CD8⁺ T cells in both the lungs and the spleen.

FIG 3.

Batf3-dependent DCs regulate the magnitude of antigen-specific CD8⁺ T cell accumulation in the lungs and the spleen. Equal numbers (5 × 10⁴) of WT naive (CD44^lo) and OT-I (Vα2⁺ Vβ5⁺) transgenic CD8 T cells were adoptively transferred into C57BL/6 and Batf3^−/− mice and infected with rVacV-WR-OVA (2 × 10⁴ PFU intranasally) the following day. Lungs (A) and spleen (B) were harvested at day 8 postinfection and stained for CD8, CD44, Vα2, and Vβ5 extracellularly, and the frequencies and cell numbers of OT-I CD8 T cells were determined. (C and D) Similar analysis was done at day 6 postinfection, and fold expansion of OT-I cells from day 6 to day 8 was enumerated. (E to H) At day 8 postinfection, lung and spleen cells were restimulated in vitro with the SIINFEKL peptide of ovalbumin, and OT-I cells were stained for IFN-γ and TNF. Similar results were obtained in three independent experiments. Results are mean ± SEM (n = 3 mice/group). **, P < 0.01.

Intact differentiation of CD8⁺ T cells accumulating in the lungs of Batf3^−/− mice during respiratory VacV infection.

VacV-specific effector CD8⁺ T cells undergo secondary differentiation in the inflamed lung tissue (13). This differentiation phenotype is characterized by upregulation of KLRG1 on the cell surface of effector CD8⁺ T cells in response to the lung inflammatory milieu (13, 38). This enables segregation of effector CD8⁺ T cells into short-lived effector cells (SLECs [KLRG1⁺ CD127⁻]) and memory precursor effector cells (MPECs [KLRG1⁻ CD127⁺]) (33, 39, 40). SLECs normally die during the contraction phase, whereas MPECs survive and mature to form long-lived memory CD8⁺ T cells (41). To investigate whether lack of Batf3-DCs affects the differentiation status of effector cells, we compared the expression of KLRG1 versus CD127 on OT-I cells recovered from Batf3^−/− and WT recipients. We found no differences in the percentages of SLECs and MPECs between lung OT-I cells of the two recipients (Fig. 4A). Similar results were noted in the spleen (not shown). We recently showed that lung-infiltrating effector CD8 T cells can be further divided into CXCR3^hi and CXCR3^lo subpopulations, where the CXCR3^lo population represents terminally differentiated cells (13). Both WT and Batf3^−/− recipients exhibited similar percentages of cells segregated by CXCR3 versus KLRG1, suggesting that once effector cells reach the lung their secondary differentiation is unimpaired in the absence of Batf3-DCs (Fig. 4B). Another scheme used to segregate terminally differentiated effector cells from central memory-like cells involves comparing expression of CD27 versus CD43 (13, 38, 42). CD27^hi cells give rise to long-lived central-like memory cells with superior recall proliferative capacity compared with CD27^lo cells (13, 38, 42). Again, the distribution of CD27^hi and CD27^lo OT-I cells appeared normal in Batf3^−/− recipients compared with WT recipients (Fig. 4C). Notably, no differences were observed in the expression of T-bet and Eomes between OT-I cells recovered from the lungs of WT and Batf3^−/− recipient mice (Fig. 4D and E). Moreover, OT-I cells isolated from Batf3^−/− recipients were proliferating to the same extent as those from WT recipients (Fig. 4F). Similar results were found in the splenic OT-I cells recovered from Batf3^−/− recipients and their WT counterparts (not shown). Overall, these data suggest that although accumulation of effector CD8⁺ T cells is severely compromised in Batf3^−/− mice (Fig. 3A), the differentiation status of the effector CD8⁺ T cells that accumulate in the lungs remains intact in the absence of Batf3-DCs.

FIG 4.

Intact differentiation of effector CD8⁺ T cells in the lungs. (A to F) Similar to Fig. 3, at day 8 postinfection, lungs from WT and Batf3^−/− recipients were harvested and stained with CD8, CD44, Vα2, or Vβ5. OT-I cells were also stained with KLRG1, CD127, CXCR3, CD27, and CD43 extracellularly (A to C) and Ki67, T-bet, and eomes intranuclearly (D to F). Data are representative of three independent experiments (n = 3 mice per group).

Batf3-DCs are required for optimal priming of antigen-specific CD8⁺ T cells in the lung draining lymph nodes.

Prior to CD8⁺ T cell accumulation in the lungs, antigen-bearing DCs, particularly CD103⁺ DCs, migrate to the lung draining lymph nodes (DLNs) and prime antigen-specific naive CD8⁺ T cells (17, 43). To investigate whether impaired CD8⁺ T cell accumulation at the peak of the effector phase in the lung was due to defective priming in DLNs, we analyzed OT-I CD8⁺ T cell responses at day 6 postinfection in the DLNs. We found significant reduction in the frequency and absolute numbers of OT-I cells in Batf3^−/− recipients compared with WT recipients (Fig. 5A). At this time point, very few OT-I cells accumulated in the spleen and no differences were observed between the two recipients (Fig. 5B). To analyze whether this early reduction in OT-I CD8⁺ T cell response in the DLNs was due to impaired expansion, we assessed the expression of proliferation marker Ki67 between the OT-I cells of the two recipient groups. Interestingly, OT-I cells in Batf3^−/− recipients had a reduced percentage of Ki67⁺ cells as well as lower mean fluorescence intensity (MFI) compared to WT recipients, suggesting that impaired proliferation possibly led to reduced OT-I cell accumulation in the DLNs (Fig. 5C). Since interleukin-2 (IL-2) signaling is known to play a major role in early expansion of CD8⁺ T cells (44, 45), we compared expression of CD25 (IL-2 receptor α [IL-2Rα]) and found profound reduction in the expression of CD25 on OT-I cells in Batf3^−/− recipients compared with their WT counterparts (Fig. 5D). To further assess the impact of Batf3 deficiency on transcription factors necessary for optimal effector CD8⁺ T cell response, we compared the expression of two canonical T-box transcription factors, T-bet and Eomes (39). Consistent with reduced Ki67 and CD25 expression, Batf3^−/− recipients generated significantly reduced proportions of T-bet⁺ Eomes⁺ double-positive (DP) cells compared to their WT counterparts (Fig. 5E). The MFI levels for T-bet but not Eomes were also significantly lower in OT-I cells recovered from Batf3^−/− recipients (Fig. 5E). This indicates that in the absence of Batf3-DCs, CD8⁺ T cells are suboptimally primed, resulting in their impaired expansion and full activation in the DLNs. Consequently, fewer CD8⁺ T cells accumulate in the lungs, thereby compromising protection against respiratory VacV infection.

FIG 5.

Defective priming of antigen-specific CD8⁺ T cells in Batf3^−/− mice. At day 6 postinfection, DLNs (A) and spleen (B) of WT and Batf3^−/− recipient mice were harvested and stained with CD8, CD44, Vα2, and Vβ5. OT-I cell numbers were also enumerated. (C to E) Lymph node cells pregated on OT-I were stained with either Ki67 intranuclearly (C), CD25 extracellularly (D), or T-bet and Eomes intranuclearly (E). Data are representative of two independent experiments. Results are mean ± SEM (n = 3 mice/group). *, P < 0.05; **, P < 0.01. ns, not significant.

DISCUSSION

Following a respiratory VacV-WR infection, NK cells and innate IFN-γ play a crucial role in early host defense (11), followed by CD8⁺ T cells, and their ability to produce IFN-γ is required for viral clearance and recovery (10, 12). Recently we demonstrated that IMs regulate the persistence of memory CD8⁺ T cells but were dispensable for effector CD8⁺ T cell responses (14). Extending these studies, we now show that Batf3-DCs play a prominent role in generating protective CD8⁺ T cell effector responses against respiratory VacV-WR infection. We found that Batf3^−/− mice lost significantly more body weight than WT mice, which correlated with higher viral titers and significantly reduced accumulation of functional CD8⁺ T cells in the lungs of Batf3^−/− mice. Interestingly, CD8 T cell responses in the Batf3^−/− host were impaired during both virulent VacV-WR and attenuated rVacV-WR-OVA infections, suggesting that Batf3-DCs regulate antiviral CD8 T cell responses independent of virulence or disease severity. Early during the infection, CD8⁺ T cells in the DLNs of Batf3^−/− mice exhibited reduced expression of Ki67, CD25, and T-bet, suggesting that CD8⁺ T cell priming is impaired in the absence of Batf3-DCs. Interestingly, the small number of effector CD8⁺ T cells that accumulated in the lungs of Batf3^−/− mice showed intact differentiation. These findings reveal that Batf3-DCs specifically regulate the magnitude of expansion and accumulation of CD8⁺ T cells during the effector phase, thereby contributing to host resistance against acute respiratory poxvirus infection.

The broad consensus that has emerged from previous studies with other viral infections is in line with our findings. Systemic infections with West Nile virus (WNV) (23) and murine cytomegalovirus (MCMV) (46), as well as mucosal infections such as influenza virus (47) and rotavirus (48), led to impaired CD8⁺ T cell effector responses in Batf3^−/− mice. This suggests that Batf3-DCs are paramount in generating antiviral CD8⁺ T cell effector responses regardless of the nature of the virus, the route of infection, and tissue tropism. However, the mechanistic rationale underlying the defective CD8⁺ T cell effector responses in Batf3^−/− mice is unclear. One key observation we reported was that the fold change in the accumulation of CD8⁺ T cells in the lungs and the spleen from day 6 to day 8 postinfection in Batf3^−/− mice was drastically less than that in WT mice. This suggested that Batf3-DCs possibly interact and regulate CD8⁺ T cells during or before that time frame. Batf3-DCs could either impart priming signals that regulate naive CD8⁺ T cell activation and proliferation, as well as their trafficking to the lungs, or impart survival signals once activated CD8⁺ T cells reach the lungs. We observed reduced expression of Ki67 in the DLNs of Batf3-DCs, suggesting suboptimal early proliferation might be responsible for reduced accumulation of activated CD8⁺ T cells in the lungs. Since, it is well described that TCR ligation and costimulation (CD28-B7) primed antigen-reactive CD8⁺ T cells to undergo proliferation in an IL-2/IL-2R-dependent manner (44, 45), we examined the expression of CD25 and found it to be markedly reduced on CD8⁺ T cells in mice lacking Batf3-DCs. Moreover, canonical CD8⁺ T cell transcription factors T-bet and Eomes were also expressed at suboptimal levels on these cells, further corroborating that Batf3-DCs are required for optimal priming of antigen-reactive CD8⁺ T cells for their full activation and expansion. These observations are consistent with a previous report that showed suboptimal expression of CD25 and T-bet on CD8⁺ T cells in Batf3^−/− mice when infected intranasally with influenza virus (49). Similarly, during rotavirus infection of Batf3^−/− mice (48), the percentages of Ki67⁺ CD8⁺ T cells in the lamina propria were also significantly reduced compared to those in WT mice, again suggesting that in mice lacking Batf3-DCs, CD8⁺ T cells are primed suboptimally, resulting in their reduced expansion and accumulation in target tissue. In addition to defective expansion, another possibility is that CD8 T cells that do not receive signals from Batf3-DCs can have survival defects resulting in reduced numbers at day 8 postinfection. Accordingly, we have also done experiments staining cells with annexin V; however, detection of apoptotic cells in vivo is sometimes challenging due to their rapid clearance by phagocytic cells. In several experiments, we were unable to detect an increase in the percentage of annexin V⁺ cells in the OT-I cells recovered from Batf3^−/− mice early in the response to VACV-WR, but this does not prove or disprove a hypothesis regarding survival.

Upon activation, virus-specific CD8⁺ T cells upregulate the expression of chemokine receptor CXCR3 (13, 38, 50). As CXCR3-expressing CD8⁺ T cells enter the inflamed lungs of VacV-infected mice, they receive additional signals that result in their secondary differentiation, whereby some cells downregulate CXCR3 and localize near the vasculature, compared to CXCR3^hi cells that are positioned at the airway epithelium (13). This gave rise to the concept of “layered immunity,” whereby CXCR3^hi cells positioned near the virus entry site provide the first layer of defense against the incoming virus particles, whereas CXCR3^lo cells localized near vasculature provide the second layer of defense against escaped viral particles, thereby inhibiting their systemic spread and preventing mortality. Intriguingly, the small number of virus-specific CD8⁺ T cells that were still present in the lungs of Batf3^−/− mice underwent proper secondary differentiation, as noted by the intact proportions of CXCR3 versus KLRG1 or KLRG1 versus IL-7Rα subsets in Batf3^−/− recipients compared to WT mice. Also, expression of transcription factors T-bet and Eomes, as well as proliferation marker Ki67, was normal in these cells, further corroborating that the CD8⁺ T cells that reach the lungs of Batf3^−/− mice undergo intact secondary differentiation. This suggests that a distinct APC subset other than Batf3-DCs, present in the lungs, might regulate secondary differentiation of accumulating virus-specific CD8⁺ T cells. Previously, it was shown that during influenza virus infection, CD8⁺ T cells trafficking into the lungs interact with lung DCs and receive a “second hit” that promotes their survival (51, 52). Hence, it is conceivable to postulate that different DC subsets cooperatively regulate different aspects of CD8⁺ T cell effector responses such as early programming in DLNs and secondary differentiation in the lung tissue (15). Future studies using Notch-2^−/− CD11c-Cre mice (53, 54) or IRF4^−/− mice (55) that lack CD103⁻ CD11b⁺ in the lungs should inform us on whether the differentiation in the inflamed lungs is controlled by yet another APC population.

Collectively, our work adds to the growing body of literature that defines the nature of DC subsets required for induction of antiviral CD8⁺ T cell effector responses. We showed a prominent role for Batf3-dependent DCs in regulating CD8⁺ T cell effector responses with consequences on host resistance against respiratory VacV infection. Targeting Batf3-DCs could potentially optimize CD8⁺ T cell-based vaccine efficacy and could possibly guide immunization strategies that utilize intranasal delivery of viral antigens.

MATERIALS AND METHODS

Mice.

Eight- to 12-week-old female C57BL/6 (CD45.2) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). B6.129S(C)-Batf3^tm1kmm/J (catalog no. 013755) or Batf3^−/− mice were also purchased from The Jackson Laboratory (Bar Harbor, ME) and maintained at the University of Florida animal facility. All animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Florida (OLAW Assurance no. A3377-01).

Viruses and infections.

Recombinant vaccinia virus (VacV-WR) and virus expressing full-length ovalbumin (rVacV-WR-OVA) were grown in HeLa cells, and their titers were subsequently determined on Vero E6 cells as described previously (56). Mice were intranasally infected with either 1.5 × 10⁴ PFU of VacV-WR or 2 × 10⁴ PFU of VacV-WR-OVA in a volume of 10 μl.

VacV titer assay.

Following VacV-WR infection, ovaries from individual mice were homogenized and sonicated for 1 min with a pause every 10s using an ultrasonic cleaner (1210 Branson). Serial dilutions were made and viral titers were determined by plaque assays on confluent VeroE6 cells as described previously (10, 29, 56).

CD8⁺ T cell adoptive transfer.

For adoptive transfer experiments, 5 × 10⁴ naive WT OT-I CD8⁺ T cells were purified from spleens with MACS Technology (Miltenyi Biotec), FACS sorted, and transferred into WT C57BL/6 mice via the intravenous route as described previously (31, 57). One day later, mice were infected with rVacV-WR-OVA as described above.

Flow cytometry.

Preparation of cells, extracellular/intracellular staining, data acquisition, and data analysis were performed as described previously (12, 29, 32, 33).

Statistical analysis.

Data were analyzed using GraphPad Prism version 5.0 software (GraphPad, San Diego, CA). Statistical analyses were performed using two-tailed, unpaired Student's t test with 95% confidence intervals unless otherwise indicated. Two-way analysis of variance (ANOVA) was used to determine differences in weight loss profiles, and the Mantel-Cox test was utilized for survival analysis. Unless otherwise indicated, data represent the mean ± SEM; P < 0.05 was considered statistically significant.