Vaccinia Virus Vectors Targeting Peptides for MHC Class II Presentation to CD4⁺ T Cells

CD4⁺ helper T cells play indispensable roles in shaping many aspects of immunity against a wide variety of infections. Following activation, Th CD4⁺ T cells will differentiate into specialized lineages dictated by the expression of individual transcription factors and functionally defined by the cytokines they then produce. These lineages include Th1 (IFN-γ, TNF-α, and IL-2), Th2 (IL-4, IL-5, IL-13), Th17 (IL-17) or T follicular helper (provide help to B cells), and the “upstream” signaling pathways that control the commitment to specific Th-lineages has been rigorously defined in vitro (1–3). Although the primary function of CD4⁺ T cells is traditionally considered to be to provide “help” to B cells, macrophages, and cytotoxic CD8⁺ T cells, activated Th1 CD4⁺ T cells also exhibit direct effector functions and are important for controlling many types of viral, bacterial, and parasitic infections. In fact, several reports implicate Th1-differentiated CD4⁺ T cells as being the critical cell type that orchestrates antiviral immunity (4–6). Th1-committed effector CD4⁺ T cells are also responsible for controlling a number of nonviral intracellular pathogens, particularly those that reside within phagolysosomes, such as Mycobacterium tuberculosis and Leishmania major (7–9). Thus, a better understanding of the complex functions of Ag-specific CD4⁺ T cells activated following diverse types of infections or immunizations may allow for improved vaccine design and development, especially against those pathogens that can successfully avoid the antimicrobial activity of neutralizing Abs and/or cytotoxic CD8⁺ T cells.

In contrast to CD8⁺ T cells, which often generate robust Ag-specific responses against viral infections, Ag-specific CD4⁺ T cells in mice typically undergo less expansion and are often difficult to identify using standard immunological assays. In many cases, extensive enrichment with peptide MHC class II (MHC-II) complexes are necessary to detect endogenous, Ag-specific CD4⁺ T cells by flow cytometry (10), even at the peak of the expansion phase. For example, CD4⁺ T cells are known to be critical for controlling vaccinia virus (VacV) infections in both humans and mice, and a number of MHC-II–presented peptides from poxviruses have been identified (11–14). However, the frequency of CD4⁺ T cells specific for an individual VacV epitope is rather small, with the largest reported response being against I1L_7–21, representing ∼0.15% of CD4⁺ T cells found at the peak of the expansion phase (11). In comparison, the immunodominant CD8⁺ T cell response in C57BL/6 mice (H2-K^b-B8R_20–27) expands to account for ∼10% of the CD8⁺ T cells following a poxvirus infection (15). In addition to the technical challenges of identifying rare, Ag-specific CD4⁺ T cells following infection or immunization, methods of heterologous challenge that are often employed in studies of memory CD8⁺ T cells are far less used to quantify the protective functions, boosting potentials, and Th lineage plasticity of memory CD4⁺ T cells. Furthermore, the development of experimental reagents for generating Ag-specific memory CD4⁺ T cells by viral immunization could result in understanding the quantitative and qualitative features of memory CD4⁺ T cells that are needed to confer protective immunity against important bacterial or parasitic infections such as tuberculosis or leishmaniasis, where attempts to develop durable, effective vaccines have been unsuccessful (16–18).

Because of the potential utility of VacV as a vaccine vector in humans and as a versatile experimental reagent, in this study, we describe the generation of VacV vectors that express known MHC-II–restricted peptides to activate CD4⁺ T cells in mice. Interestingly, we found that VacV expressing only a minimal peptide sequence was not sufficient to activate CD4⁺ T cells in vivo, but rather required incorporating strategies that would target the peptide for more efficient MHC-II presentation. Immunization of mice with VacV expressing MHC-II–targeted peptides resulted in the generation of highly functional effector and memory CD4⁺ T cells that underwent considerable secondary expansion following heterologous challenge. Finally, we demonstrate that VacV expressing an MHC-II–restricted peptide from Leishmania species generates polyfunctional Ag-specific memory CD4⁺ T cells that undergo robust re-expansion following a visceral Leishmania donovani infection. Overall, our findings show that VacV vectors can be used to activate Ag-specific CD4⁺ T cells, but that targeting the peptides for MHC-II presentation is required. These novel viral vectors will be useful not only for studies of Ag-specific CD4⁺ T cell activation and protective functions during poxvirus infections but also as a vaccine strategy to generate Ag-specific, Th1-differentiated memory CD4⁺ T cells against other relevant pathogens, including M. tuberculosis, L. major, and Salmonella enterica.

Recombinant VacVs were generated by homologous recombination into the thymidine kinase gene using the pSC11 vector, as described previously (19). To generate pSC11-GP_61–80, annealed oligonucleotides (Integrated DNA Technologies) encoding the GP_61–80 epitope and containing BglII- and NotI-compatible overhangs were cloned into the pSC11 vector. To generate pSC11-LAMP1, annealed oligonucleotides encoding the lysosomal-associated membrane protein 1 (LAMP1) (National Center for Biotechnology Information reference sequence: NM_010684.3) signal sequence (aa 1–29) and the lysosomal targeting sequence (aa 368–406) were cloned into the BglII and NotI sites of the pSC11 vector, respectively. Oligonucleotides were designed to maintain functional BglII and NotI endonuclease target sequences at the 3′ end of the LAMP1 signal sequence (BglII) and 5′ end of the lysosome targeting sequence (NotI). Next, oligonucleotides encoding the GP_61–80 or PEPCK_335–351 epitope with BglII- and NotI-compatible overhangs were cloned into the pSC11-LAMP1 vector. To generate pSC11 expressing the MHC-II invariant chain (pSC11-Ii), the MHC-II invariant chain (Ii) (National Center of Biotechnology Information reference sequence: NM_010545.3; aa 1–215) was amplified from cDNA synthesized from mouse splenocytes and cloned into the pSC11 vector using the SalI and NotI sites. Annealed oligonucleotides encoding either GP_61–80 or PEPCK_335–351 with NotI-compatible overhangs were then cloned into pSC11-Ii. To generate recombinant VacV, BSC-40 cells were infected with VacV-Western Reserve (Biodefense and Emerging Infections Resources) and then transfected with the appropriate vector using FuGENE transfection reagent (Promega). The resulting cell lysate was then used to infect 143B thymidine kinase–deficient cells in the presence of BrdU. After three rounds of plaque purification in the presence of BrdU, viruses were screened by DNA sequencing, and successfully recombined viruses were expanded using standard procedures (19).

For in vitro growth curves, confluent monolayers of BSC-40 cells were infected in 12-well plates in a volume of 200 μl at 37°C for 1 h. Following infection, cells were scraped, transferred to microcentrifuge tubes, and subjected to three rounds of freeze-thaw before performing standard plaque assays using BSC-40 cells. For plaque assays of infected skin, ear skin was homogenized in RPMI 1640 medium containing 1% FBS and then subjected to three rounds of freeze-thaw before performing plaque assays using standard protocols. Briefly, serial dilutions were inoculated on BSC-40 cells in a 12-well plate that were then covered with 1% SeaKem agarose in Modified Eagle Medium (Life Technologies). Plaques were visualized 72 h later following overnight incubation with Neutral Red dye.

BSC-40 cells were infected as described in 4In vitro infections and plaque assays, and 90 min postinfection, RNA was purified using the RNeasy Mini Kit (Qiagen) according to manufacturer’s instructions. cDNA was synthesized using the SuperScript First-Strand Synthesis System (Thermo Fisher Scientific) according to manufacturer’s instructions. The resulting cDNA was then used as a template for PCR, using DNA primers specific for the GP_61–80 epitope, the LAMP signal sequence and lysosomal targeting sequence (amplifies across GP_61–80), the invariant chain, or hypoxanthine guanine phosphoribosyl transferase (Hprt) as a positive control.

C57BL/6N and C57BL/6J mice (6–10 wk of age, female) were purchased from Charles River/National Cancer Institute and The Jackson Laboratory, respectively. SMARTA and P14 TCR-tg mice have been described previously and were maintained by sibling-with-sibling mating (20, 21). For adoptive transfers, cells were isolated from the spleen and transferred i.v. in a volume of 200 μl of PBS and mice were infected the following day. VacV and VacV expressing GP_33–41 (VacV-GP_33–41) have been described previously and were generated using standard protocols (19, 22). VacV-GP_33–41 was independently sequenced in our laboratory to confirm only expression of the minimal epitope. VacV skin infections were performed by pipetting 10 μl of PBS containing 5 × 10⁶ PFU onto the ventral ear skin and poking the ear skin 25–30 times with a 16.5-gauge needle. Lymphocytic choriomeningitis virus (LCMV)–Armstrong infections were performed by injecting 2 × 10⁵ PFU i.p. in PBS. ActA-deficient Listeria monocytogenes expressing the GP_33–41 and GP_61–80 epitopes from LCMV (23) was grown in tryptic soy broth (Sigma-Aldrich) supplemented with 50 μg/ml streptomycin (Sigma-Aldrich) at 37°C until OD600 = 0.1 (10⁸ CFU/ml), and 5 × 10⁶ CFU were injected i.v. Leishmania parasites (L. donovani BOB) were routinely cultured at 26°C in RPMI medium supplemented with 25 mM HEPES (pH 6.9), 10 mM glutamine, 7.6 mM hemin, 0.1 mM adenosine, 10% (v/v) heat-inactivated FCS, and 25 μg/ml G418. The initial density was constantly maintained at 2.5 × 10⁵ cell/ml for no more than six passages. Parasites were grown to the stationary phase, and 5 × 10⁶ parasites were washed twice with PBS and injected i.v.

For limiting dilution assays to quantify parasite loads in mice, serial 4-fold dilutions of liver and spleen cell lysates were cultivated in 96-well plates in a Leishmania culture medium to which 10% FBS and 100 μM hypoxanthine were added, as described previously (24). Growth in individual wells was monitored after 2 wk by visual inspection.

The following Abs were used in this study: CD4 (RM4-5; BioLegend), CD8⍺ (53-6.7; BioLegend), Thy1.1 (OX-7; BioLegend), CD44 (1M7; BioLegend), CD69 (H1.2F3; BioLegend), CD103 (2E7; BioLegend), V⍺2 (B20.1; BioLegend), IFN-γ (XMG1.2; BioLegend), TNF-⍺ (MP6-XT22; BioLegend), and IL-2 (JES6-5H4; BioLegend). MHC-II tetramers were obtained from the National Institutes of Health Tetramer Core Facility. Tetramer staining was performed for 1 h at 37°C, and a tetramer loaded with a human CLIP peptide was used as a negative staining control. All other staining was performed for 15–30 min at 4°C. Intravascular labeling was performed as described previously (25). Briefly, 3 μg of anti-V⍺2 was injected i.v. in 200 μl of PBS, and tissues were harvested 3 min later. Data were acquired using an LSRII Flow Cytometer (BD Biosciences) in the Oregon Health & Science University Flow Cytometry Core Facility. Flow cytometry data were analyzed using FlowJo Software (BD Biosciences) Version 9.9.6 or 10.5.3.

Naive Thy1.1⁺ SMARTA CD4⁺ T cells or P14 CD8⁺ T cells were isolated from the spleen and washed twice with PBS before labeling with 1 μM CFSE at 37°C for 15 min. Cells were then washed twice in RPMI 1640 medium supplemented with 10% FBS and transferred into naive Thy1.2⁺ recipients. The following day, mice were infected on the left ear skin with 5 × 10⁶ PFU of the indicated VacV strain, and proliferation was analyzed by CFSE dilution using FlowJo 9.9.6. The percentage divided and expansion index were calculated using the FlowJo proliferation platform as described previously (26). Briefly, the percentage of cells divided was determined by calculating the average number of divisions of responding cells divided by the average number of divisions of all cells to calculate the proportion of cells that underwent division. The expansion index was calculated by calculating the fold-increase of the number of cells.

Ears from infected mice were removed, and dorsal and ventral sides of the ear pinna were separated and incubated in 1 ml HBSS (Life Technologies) containing CaCl₂ and MgCl₂ supplemented with 125 U/ml collagenase II (Invitrogen) and DNAse-I (Sigma-Aldrich) at 37°C for 1 h. Whole tissue suspensions were then generated by forcing the digested skin through a wire mesh screen. Leukocytes were then purified by resuspending the cells in 10 ml of 35% (v/v) Percoll (GE Healthcare) in HBSS in 50-ml conical tubes followed by centrifugation for 10 min at room temperature.

Spleens from mice were harvested and single cell suspensions were generated by gently forcing the spleen through a mesh screen. RBCs were lysed by resuspending cell pellets in 150 mM NH4Cl, 10 mM KHCO3, and 0.1 mM Na-EDTA. Cells were then washed in RPMI 1640 medium containing 5% FBS and incubated at 37°C in the presence of 2 μM of the indicated peptide and brefeldin A (BioLegend) for 5–6 h. Intracellular cytokine staining was performed using the CytoFix/CytoPerm Kit (BD Biosciences) according to the manufacturer’s instructions. Briefly, surface Ags were stained as described in 8Flow cytometry and Abs, followed by permeabilization with 100 μl of CytoFix/CytoPerm for 15 min at 4°C. Staining for cytokines was performed in Perm/Wash Buffer for 20 min at 4°C.

Statistics were calculated using Prism Software (GraphPad), using the unpaired Student t test or ANOVA with Tukey correction for multiple comparisons.

To determine whether VacV vectors that express MHC-II–restricted peptides would activate CD4⁺ helper T cells of a desired Ag specificity, we generated recombinant VacV that expresses GP_61–80 from LCMV, which includes the minimal amino acid sequence required for binding to I-A^b, GP_66–77 (Fig. 1A; see 2Materials and Methods). In addition, we also generated two additional viruses, incorporating strategies that have been reported to increase the efficiency of MHC-II presentation of Ags in vivo. In one strategy, we generated recombinant VacV expressing GP_61–80 flanked by two protein sequences of LAMP1, where the N-terminal portion of the protein serves as the signal sequence and the C-terminal sequence targets the chimeric protein to lysosomes (27). As an alternative strategy to direct peptides for MHC-II presentation, we also generated VacV where Ii was directly conjugated to the N terminus of GP_61–80 (28). All viruses (VacV-GP_61–80, VacV-LAMP1-GP_61–80, VacV-Ii-GP_61–80) exhibited similar growth in BSC-40 cells compared with the thymidine kinase–VacV control virus (Fig. 1B). To next verify mRNA expression of the recombinant sequences, we infected BSC-40 cells with VacV expressing GP_61–80, LAMP1-GP_61–80, or Ii-GP_61–80 and analyzed expression of the cloned inserts using RT-PCR. Amplification of cDNA synthesized from VacV-infected cells using GP_61–80–specific primers demonstrated that all three viruses expressed the GP_61–80 sequence (Fig. 1C). To next confirm expression of the GP_61–80 chimeric proteins, we also used primers that specifically amplified LAMP1-GP_61–80 or Ii-GP_61–80. Thus, these data demonstrate the successful generation of recombinant VacV expressing GP_61–80 alone or conjugated to other protein sequences that have been reported to target Ags for MHC-II presentation.

Because our in vitro analysis demonstrated that all recombinant viruses exhibited similar growth and expressed the correct variations of the GP_61–80 peptide sequence, we next tested whether they would be presented by MHC-II to activate CD4⁺ T cells in vivo. We labeled naive Thy1.1⁺ SMARTA TCR-tg CD4⁺ T cells (specific for I-A^b-GP_66–77 (20)) with CFSE and transferred them into naive B6 mice. The left ear skin was then infected with VacV or each of the VacV strains expressing GP_61–80. On day 3 postinfection, all viruses caused similarly high levels of infection in the skin (Fig. 2A). Surprisingly, VacV expressing only the GP_61–80 peptide did not stimulate any detectable proliferation of naive SMARTA CD4⁺ T cells in the draining lymph node, suggesting the peptide sequence was not being efficiently presented on MHC-II to CD4⁺ T cells (Fig. 2B), whereas VacV expressing GP_61–80 conjugated to either LAMP1 or Ii caused significant levels of proliferation (Fig. 2B–D). Activation of naive CD4⁺ T cells was site specific, as proliferation was not detected in contralateral nondraining lymph nodes. Notably, VacV-Ii-GP_61–80 infection consistently caused more expansion of SMARTA CD4⁺ T cells compared with the LAMP1-conjugated sequence (Fig. 2D), which could be due to less MHC-II presentation using this strategy or because it has also been reported that cathepsin D cleaves the GP_61–80 sequence (between aa 74 and 75) within lysosomes (29). In contrast, VacV expressing only the MHC class I (MHC-I)–restricted peptide GP_33–41 from LCMV caused proliferation of naive, Ag-specific P14 TCR-tg CD8⁺ T cells (Fig. 2E–G), demonstrating that VacV expression of minimal MHC-I–restricted peptides is sufficient to activate naive CD8⁺ T cells. Therefore, these data show that targeting peptides for MHC-II presentation is necessary to activate Ag-specific CD4⁺ T cells during VacV skin infection.

The previous data demonstrated that VacV expressing GP_61–80 conjugated to Ii was the more efficient strategy to activate naive CD4⁺ T cells in vivo. Therefore, we next tested whether viral immunization would stimulate a systemic response that resulted in the generation of memory CD4⁺ T cells. We again transferred naive SMARTA CD4⁺ T cells into B6 mice and infected them with VacV-Ii-GP_61–80 by either i.p. injection or scarification of the skin to determine if route of infection would influence the magnitude of expansion or the formation of memory cells. Both routes of immunization caused significant clonal expansion of SMARTA CD4⁺ T cells compared with a control VacV infection, and at 30 d postinfection, memory cells could be detected in the circulation (Fig. 3A–C). SMARTA CD4⁺ T cells could also be detected in the skin following infection by scarification (but not i.p. infection) after clearance of the viral infection (Fig. 3D), which occurs within 15 d postinfection (30). Most of the SMARTA CD4⁺ T cells in the previously infected skin expressed CD69 (Fig. 3E, 3F), suggesting the development of tissue residency. Interestingly, in contrast to what we and others have shown for tissue-resident CD8⁺ T cells that form following VacV infection (30, 31), the majority of Ag-specific CD4⁺ T cells in the skin did not express CD103 (Fig. 3E, 3F). Despite their lack of CD103 expression, SMARTA CD4⁺ T cells in the skin were protected from i.v. labeling, demonstrating that these T cells reside within the skin parenchyma (Fig. 3G, 3H). Thus, these data demonstrate that immunization with VacV expressing peptides targeted for MHC-II presentation results in the formation of memory CD4⁺ T cells.

Using monoclonal TCR-tg SMARTA CD4⁺ T cells as an experimental readout, the previous data demonstrated that peptides required targeting to the MHC-II presentation pathway to stimulate the activation of Ag-specific CD4⁺ T cells in vivo. Therefore, we next determined whether infection with VacV expressing GP_61–80 would also cause the activation and expansion of the rare, Ag-specific CD4⁺ T cells found within the endogenous repertoire. In agreement with our previous data, i.p. immunization with VacV-Ii-GP_61–80, but not VacV-GP_61–80 or VacV-LAMP-GP_61–80, caused expansion of I-A^b-GP_66–77–specific CD4⁺ T cells on day 7 postinfection (Fig. 4A, 4B). Most of the Ag-specific CD4⁺ T cells produced IFN-γ (suggesting Th1-differentation), and approximately half of the cells exhibited hallmarks of polyfunctionality (e.g., production of TNF-α and IL-2) (Fig. 4C, 4D). Therefore, these data demonstrate that immunization with VacV-Ii-GP_61–80 activates and expands Ag-specific CD4⁺ T cells that exhibit features of Th1 differentiation.

One hallmark of cellular immunological memory is the capacity to undergo more robust re-expansion following heterologous challenge compared with the primary infection. Therefore, we infected mice that had been previously immunized with VacV or VacV-Ii-GP_61–80 with attenuated L. monocytogenes expressing the GP_61–80 epitope (23). Upon rechallenge, GP_66–77–specific CD4⁺ T cells underwent robust secondary expansion in mice that had been previously infected with VacV-Ii-GP_61–80, but not VacV (Fig. 4E–G). In fact, the frequency of boosted GP_66–77–specific CD4⁺ T cells in the circulation was ∼10 times higher compared with mice receiving control VacV immunization. Overall, these data demonstrate that immunization with VacV-Ii-GP_61–80 generates functional effector and memory CD4⁺ T cells that are able to rapidly expand following secondary challenge.

Because our data demonstrated that VacV-Ii-GP_61–80 caused the activation of naive Ag-specific CD4⁺ T cells, we next tested whether MHC-II–targeted peptides would reactivate an established memory CD4⁺ T cell population. To test this, we infected naive B6 mice with LCMV, which generates I-A^b-GP_66–77–specific memory CD4⁺ T cells that can be identified in the circulation (Fig. 5A). We then challenged LCMV-immune mice or naive controls on the left ear skin with VacV-Ii-GP_61–80 (Fig. 5B). On day 7 postinfection, GP_66–77–specific memory CD4⁺ T cells in LCMV-immune mice had undergone an ∼5-fold re-expansion in the draining lymph node (compared with contralateral nondraining lymph node) (Fig. 5C, 5D). Notably, VacV-Ii-GP_61–80 viral load in the skin was significantly lower in LCMV-immune mice (Fig. 5E), suggesting that, like memory CD8⁺ T cells (32, 33), memory CD4⁺ T cells are able to provide protection against poxvirus infection. To test whether memory CD4⁺ T cell–mediated protection was Ag-specific, we infected the ear skin of LCMV-immune mice with VacV, VacV-GP_61–80, or VacV-Ii-GP_61–80. LCMV-immune mice were only protected against VacV-Ii-GP_61–80, (Fig. 5F) demonstrating that protection is both Ag-specific and requires the targeting of antigenic peptides to the MHC-II pathway. Thus, these data reveal that I-A^b-GP_66–77–specific memory CD4⁺ T cells generated by LCMV infection become reactivated following VacV-Ii-GP_61-80 skin infection and also provide protective immunity.

Using the LCMV model Ag GP_61–80, the previous set of experiments provided strong evidence that expression of MHC-II–restricted peptides by VacV required additional targeting strategies to activate CD4⁺ T cells in vivo. Therefore, we next determined whether this strategy would activate CD4⁺ T cells that are specific for a relevant human pathogen, such as intracellular parasites from the genus Leishmania. Notably, in contrast to most viral infections that can also be controlled by cytotoxic CD8⁺ T cells, protective immunity against cutaneous and visceral leishmaniasis is largely mediated by IFN-γ–producing Th1-committed CD4⁺ T cells (34, 35). Because an I-A^b–restricted epitope from Leishmania has now been identified within the glycosomal phosphoenolpyruvate carboxykinase (PEPCK) gene, aa 335–351 (36), this presented an opportunity to determine whether Ag-specific memory CD4⁺ T cells could be generated with VacV immunization that would subsequently respond to a Leishmania infection. Notably, PEPCK is expressed during both the promastigote and amastigote stage of the parasite life cycle and its sequence is conserved across Leishmania species (36), suggesting it could be a relevant target for rational vaccine design against leishmaniasis. To determine whether memory PEPCK_335–351–specific CD4⁺ T cells formed following VacV immunization, we generated VacV-Ii-PEPCK_335–351 and VacV-LAMP1-PEPCK_335–351 using the strategies described in Fig. 1A. Naive B6 mice were first immunized with VacV-Ii-PEPCK_335–351, and at 30 d postinfection, PEPCK_335–351–specific CD4⁺ T cells were analyzed from the spleen. As predicted, PEPCK_335–351–specific memory CD4⁺ T cells exhibited a strong Th1-biased lineage commitment and produced IFN-γ, TNF-α, and IL-2 following stimulation with PEPCK_335–351 peptide (Fig. 6A–C). Thus, immunization with VacV expressing PEPCK targeted for MHC-II presentation generates Ag-specific, Th1-differentiated memory CD4⁺ T cells.

Because we found that Ag-specific memory CD4⁺ T cells were generated following infection with VacV-Ii-PEPCK_335–351, we next tested whether this memory CD4⁺ T cell population would undergo re-expansion during a Leishmania infection. Interestingly, we found that mice immunized with either the VacV-Ii-PEPCK_335–351 or VacV-LAMP1-PEPCK_335–351 caused expansion of endogenous PEPCK_335–351–specific CD4⁺ T cells in the circulation, demonstrating that, in contrast to LCMV GP_61–80, both strategies efficiently target this peptide for MHC-II presentation in vivo (Fig. 6D). At 25 d postimmunization, we then infected immunized mice with L. donovani promastigotes by i.v. injection, which infects the liver and spleen of mice and is used as an experimental model of visceral leishmaniasis (37). Following challenge with L. donovani, PEPCK_335–351–specific CD4⁺ T cells in mice that had been immunized with the two VacV-PEPCK vectors underwent robust secondary expansion in the circulation, spleen, and liver compared with VacV-immunized controls (Fig. 6D–G). However, despite this robust secondary expansion, parasite burden was not significantly reduced in VacV-PEPCK immunized animals (Fig. 6H, 6I), underscoring the ability of L. donovani to evade elimination by the adaptive immune system and highlighting the challenges of developing an effective vaccine against leishmaniasis. Nevertheless, these data show that VacV vectors targeting a Leishmania-specific peptide for MHC-II presentation generates functional memory CD4⁺ T cells that become reactivated during a visceral Leishmania infection.

It is well established that CD4⁺ T cells play critical roles in protective immunity against a number of intracellular pathogens, but the relevant effector functions and cell types that coordinate memory Th1 CD4⁺ T cell–mediated protective immunity is less clear (38–40). Control of intracellular pathogens, such as M. tuberculosis and Leishmania, requires Th1-produced IFN-γ and TNF-⍺ to act directly on infected cells. Alternatively, Th1 cytokines can also provide protection by promoting the formation of highly inflammatory tissue microenvironments through the upregulation of chemokines, such as CXCL9 and CXCL10 (41). In contrast to this cytokine-mediated protection, during some viral infections (including poxviruses, CMV, and dengue virus), activated CD4⁺ T cells have been demonstrated to have direct cytotoxic capabilities (42–44). Finally, Ag-specific memory CD4⁺ T cells can also cause more robust maturation of professional APC during viral infection, resulting in stronger priming and expansion of antiviral cytotoxic CD8⁺ T cells (5). Thus, it appears that effector/memory CD4⁺ T cells possess a variety of mechanisms that could contribute to antiviral immunity, but the importance of any individual effector function for controlling diverse types of infections may be highly pathogen- and/or tissue-specific.

How effector and/or memory CD4⁺ T cells recognize intracellular pathogens that result in activation and/or protection remains an active field of research. In the most classic model of MHC-II Ag presentation (45), proteins from the extracellular environment are acquired by phagocytosis or endocytosis that are then delivered into specialized late endosomal/lysosomal Ag-processing compartments that contain a variety of proteases. Following proteolytic cleavage, these processed peptides displace CLIP on recycled MHC-II molecules, which are then transported to the cell surface for presentation to CD4⁺ T cells. However, considerable evidence has accumulated, suggesting that alternative pathways exist that allows cells to directly present cytosolic proteins on MHC-II (46). For example, a recent report found that dendritic cells infected with modified VacV Ankara presented MHC-II–restricted peptides from an endogenous presentation pathway to CD4⁺ T cells, a process that required functional proteasomes and autophagy (47). In agreement with this finding, UV-inactivated ectromelia virus (the causative agent of mousepox) does not activate CD4⁺ T cells in vivo (48), suggesting that a live replicating infectious virus is required for optimal Ag processing and subsequent display of viral Ags on MHC-II. Alternative pathways of viral Ag presentation is not likely a feature of only poxviruses, as CD4⁺ T cell responses to an alternatively presented Ag have been described for several different viruses, including influenza, CMV, and HIV (49–51). Our study suggests that in VacV-infected cells, cytosolic peptides are not efficiently presented by an alternative endogenous pathway but that fusion to the invariant chain is an effective strategy to target peptides for MHC-II presentation. Finally, it is becoming evident that expression of MHC-II is not restricted to only professional APCs and B cells but can be induced on some nonhematopoietic cells in response to inflammatory cytokines (52–54). Thus, expression of MHC-II on nonhematopoietic cells in combination with nonclassical, endogenous MHC-II Ag presentation could potentially be an important mechanism for effector/memory CD4⁺ T cells to identify and eliminate virally infected cells in nonlymphoid tissues, such as the skin, by either direct cytolytic activity or through the production of antiviral cytokines.

Although memory CD4⁺ T cells clearly play important roles in providing host defense and protective immunity against a variety of pathogens, studies analyzing their formation and differentiation in vivo are far more limited compared with studies of memory CD8⁺ T cells. In particular, the extent of Th lineage commitment and plasticity of Ag-specific memory CD4⁺ T cells following reactivation remains relatively ill-defined. However, recent studies have found that during LCMV or L. monocytogenes infections, the majority of activated CD4⁺ T cells exhibit features primarily of the Th1 or T follicular helper lineages and that these commitments appear to be somewhat maintained in the ensuing memory population (55–57). One potential caveat to these studies is that CD4⁺ T cells play little role in protection against these two pathogens, thereby making it difficult to evaluate whether specific T helper lineages of memory CD4⁺ T cells are more protective against reinfection versus another. In fact, it has recently been shown that Ag-specific memory CD4⁺ T cells elicited following immunization become highly pathogenic rather than protective during chronic LCMV infection (58). This contrasts our finding in this study, demonstrating that memory CD4⁺ T cells protect against poxvirus infection, suggesting that the amount of protection provided by memory Th1 CD4⁺ T cells may be highly virus specific (20). Finally, at least in mice, memory CD4⁺ T cell populations seem to be less numerically stable compared with memory CD8⁺ T cells (59), suggesting that frequent “booster” immunization may be necessary to maintain protective levels of memory CD4⁺ T cells (23), but how multiple rounds of Ag stimulation influences T helper lineage commitment, effector functions, or trafficking potentials of memory CD4⁺ T cells remains largely unknown.

In summary, we report the generation of VacV vectors that efficiently target MHC-II–restricted peptides that activate Ag-specific CD4⁺ T cells. Unlike MHC-I–restricted peptides, our findings demonstrate that more complex targeting strategies are required for viral vectors to express Ags that will be presented to CD4⁺ T cells in vivo. Importantly, we show that the Ag-specific memory CD4⁺ T cells generated by viral immunization rapidly respond to heterologous challenge with pathogens expressing the common Ag, demonstrating that this experimental approach may be useful in understanding the extent of protection (or immunopathology) that is provided by memory CD4⁺ T cells against a wide variety of infections. Overall, these viral vectors provide an opportunity to further investigate the activation, lineage commitment, and effector mechanisms employed by memory CD4⁺ T cells, which could ultimately result in understanding how to improve vaccine strategies against diseases that rely on the protective functions of CD4⁺ T cells.

We thank Dr. Ann Hill for critical review of the manuscript and also Dr. Jon Yewdell and Dr. James Gibbs for providing the pSC11 plasmid. We would also like to acknowledge the Oregon Health & Science University Flow Cytometry Core Facility.

The authors have no financial conflicts of interest.