Cutting Edge: The Aging Immune System Reveals the Biological Impact of Direct Antigen Presentation on CD8 T Cell Responses

The immune system mobilizes a variety of innate and adaptive immune mechanisms to limit and eliminate infection. In youth, these mechanisms are both robust and overlapping, providing considerable redundancy in protecting against microbial infections. Evaluating the in vivo impact and limits of immune resource redundancy when confronted with microbial immune evasion has been difficult so far. In older age, many mechanisms of protective immunity exhibit defects, allowing us to use old mice as a model of suboptimal immunity, akin to a complex genetic hypomorph for adaptive or innate immunity.

Members of the Orthopoxvirus genus of the Poxviridae family are known to adversely affect individuals with a vulnerable immune system, including older adults (1). Thus, vulnerability to wild-type (wt) ectromelia virus (ECTV) increases with age; antipoxvirus-specific CD8 T cell responses are curtailed both in total number and in function in ECTV-exposed old B6 mice (2), consistent with other models of viral and bacterial infections where CD8 T cell responses are impaired in old mice as compared with their adult counterparts (3–7). By contrast, 14- to 18-mo-old mice infected with poorly pathogenic orthopox viruses, such as vaccinia virus and the mutant strain of ECTV (∆166 ECTV), mounted CD8 T cell responses comparable with adult mice (2). The mechanistic basis for the improved CD8 responses in old mice to attenuated poxviruses remains incompletely understood (8).

Poxviruses use a diverse array of strategies to evade the immune system. At present, it is not known whether and to what extent differences in the expression of viral immune evasion proteins play a role in increased susceptibility of older organisms to wt, but not attenuated, poxviruses (9). Multiple studies have mechanistically dissected cowpox virus (CPXV) immune evasion (10–15). Two viral proteins, CPXV12 and CPXV203, downregulate MHC class I on the surface of infected cells. Consequently, Ag-specific CD8 T cells cannot recognize or exert their effector function on CPXV-infected cells. Importantly, this evasion mechanism does not prevent the priming of a functional CD8 T cell response via cross-presentation (16–18). Indeed, C57BL/6 (B6) mice generate potent CD8 T cell responses to CPXV directed against a conserved immunodominant H-2K^b (K^b)-restricted B8R_20–27 epitope conserved across poxviruses (B8R). However, cross-presentation has been shown to be less effective with aging (17–19). Thus, if direct presentation is blocked by the virus and cross-presentation is crippled with aging, then combined these two deficiencies may explain the reduced CD8 T cell responsiveness with aging (17–19). If this explanation is correct, then restoring direct priming should improve CD8 T cell responses in old mice.

To test this hypothesis, we used a CPXV mutant lacking CPXV12 and CPXV203 (∆12∆203 CPXV) in old mice. We demonstrate that B8R-specific CD8 T cell responses to ∆12∆203 CPXV are significantly improved in both abundance and function, as compared with those primed with wt CPXV. Importantly, repairing direct priming with the mutant virus restored primary CD8 T cell responses in old mice to the same level as in adult mice responding to wt virus, and generated superior memory CD8 T cell responses upon recall in old mice even when compared with adult mice responding to wt CPXV, as judged by clearance of Listeria monocytogenes expressing B8R (Lm-B8R). This demonstrates that direct priming can induce strong effector and memory CD8 T cell responses, which helps explain the evolutionary pressure that led to the generation of CPXV12 and CPXV203 by the virus. Our approach highlights the power of using a vulnerable population with suboptimal immunity as a tool to dissect biological relevance of antimicrobial responses in the face of microbial immune evasion. We conclude that improving direct Ag presentation can be a powerful strategy to induce robust CD8 T cell responses even under conditions of suboptimal immunity (e.g., in the growing elderly segment of the population) and must be considered for vaccines where effective CD8 T cell memory is required to curtail or eliminate infection.

Mouse studies were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Protocols were approved by the Institutional Animal Care and Use Committee at the University of Arizona (IACUC 08-102, PHS Assurance No. A3248-01). Intranasal infections were performed under ketamine/xylazine anesthesia. Footpad (f.p.) injections were performed under isoflurane anesthesia. Euthanasia was performed by isoflurane overdose or cervical dislocation to preserve unobstructed analysis of cellular immunity (20).

Old (18–22 mo) and adult (12–16 wk) male B6 mice were purchased and/or obtained from the National Institute on Aging Rodent Resource via the Charles River Laboratories (Frederick, MD, and Kingston, NY) and/or The Jackson Laboratory (Bar Harbor, ME). Ten- to sixteen-wk-old male RAG1-KO mice (strain 002216) and Ly5.1 congenic mice (strain 002014) were originally purchased from The Jackson Laboratory. Mice anesthetized i.p. with ketamine and xylazine were infected intranasally (i.n.) with 1 × 10⁶ PFUs CPXV in 10 μl of sterile PBS administered to one naris. Mice anesthetized with isoflurane were infected s.c. in the f.p. with 1 × 10⁶ PFUs of either wt or ∆12∆203 CPXV in 50 μl of sterile PBS. Total T cells or CD8 T cells for adoptive transfer experiments were purified with Miltenyi Biotec (San Diego, CA) reagents and columns using negative magnetic selection. Purity was confirmed at >90% by flow cytometry analysis.

CPXV strain Brighton Red was obtained from the ATCC. The Δ12Δ203 CPXV mutant was kindly provided by Dr. W. Yokoyama and colleagues (10). All viruses were propagated and titered on Vero cells. Titer was determined by plaque assay.

Accutase-treated (eBioscience, San Diego, CA) spleen was disassociated over a 40-μM cell strainer. Blood was hypotonically lysed to eliminate RBCs. Lymphocytes were incubated overnight in B8R tetramer (National Institutes of Health Tetramer Core Facility) and a saturating dose of mAbs including anti-CD3 and -CD8α (eBioscience), stained with Live/Dead Aqua (Life Technologies, Grand Island, NY) and analyzed as described later. Cells stimulated with B8R:K^b peptide (10⁻⁶ M; 21st Century, Marlboro, MA) in the presence of brefeldin A for 5 h were first stained for surface markers and viability as described earlier, then fixed and permeabilized with the BD CytoFix/CytoPerm kit and stained for intracellular IFN-γ, TNF-α (eBioscience), and granzyme B (BD Biosciences, San Jose, CA). Samples were acquired using a BD LSR Fortessa cytometer (BD Biosciences) and analyzed by FlowJo software (Tree Star, Ashland, OR). Cell counts were extrapolated from either a CBC differential collected on a Hemavet LV (Drew Scientific, Waterbury, CT) instrument or by correction during acquisition with Count Beads (Life Technologies, Carlsbad, CA). The two counting methods were confirmed to be consistent.

The protocol was adapted from Jenkins and Moon (21). Spleen and visible lymph nodes (LNs; cervical, axillary, brachial, inguinal, and popliteal) were collected from naive adult (12 wk) and old (18 mo) mice and treated with Accutase (eBioscience) for 30 min at 37°C followed by disassociation over a 40-μM filter. Cells were washed in MACS buffer (PBS, 0.5% BSA, 2 mM EDTA [pH 7.2]) and enriched for total T cells by negative selection (130-095-130; Miltenyi Biotec). Cells were transferred to FACS buffer (PBS + 2% FBS) and maintained cold (4°C) for all remaining steps. Enriched T cells were stained in the presence of Fc block for 1 h with a saturating dose of mAbs for CD3, CD4, and CD8α (eBioscience) and K^b-B8R tetramers (National Institutes of Health Tetramer Core Facility) labeled with two different fluorochromes. Cells were washed twice and stained in a saturating dose of anti-streptavidin PE Ab (BioLegend, San Diego, CA) for 30 min at 4°C. Cells were washed two more times and resuspended in 250 μl of FACS buffer + count beads (Life Technologies). Upon acquisition with BD FACS Diva software, cells were first gated as a CD3⁺ histogram to the axis so that all CD3⁺ events and the count beads would be included. This was selected as the storage gate to minimize the size of the FCS file. The gating strategy for CD8p number is depicted in Supplemental Fig. 1. Final CD8p number was determined with count bead correction for abort rate per manufacturer’s protocol. CD8p from adult and old mice were determined alongside each other, and reduced CD8p numbers with age was consistent with results from previous publications (19, 20). B8R-specific CD8p numbers from adult mice were consistent with the original publication (22).

Mice were i.v. infected with 10⁵ CFUs of Lm-B8R kindly provided by Dr. S.S. Way. Forty-eight hours postchallenge, spleens were harvested, weighed, and placed into sterile PBS. Tissues were homogenized mechanically using a Tissue-Tearor electric homogenizer (BioSpec Products, Bartlesville, OK). Serial dilutions were made in sterile PBS and plated onto brain heart infusion agar. Plates were incubated overnight at 37°C. The log10 CFU/g of tissue was calculated as: log10 [(CFU/dilution factor) × ([organ weight + homogenate volume]/organ weight)] (23).

Multiple models of infection have demonstrated an age-associated decrease in primary Ag-specific CD8 T cell responses (2–5, 8). We infected adult (12–16 wk) and old (18–22 mo) mice with wt CPXV and evaluated the B8R-specific CD8 T cell response (Fig. 1). Old mice exhibited reduced absolute numbers of, and diminished polyfunctional cytokine production among, the B8R-specific effector CD8 T cells on days 7–8 postinfection (p.i.) (Fig. 1A, 1B). Moreover, as previously shown, many tetramer-specific CD8 T cells from old mice failed to produce IFN-γ (Fig. 1C). Similar age-related differences were found with both the f.p. (Fig. 1) and intranasal (data not shown) administration of the virus.

Because CD8 T lymphocyte precursor (CD8p) numbers decline with age (19, 24, 25), we next asked whether that reduction curtailed primary CD8 T cell response in old mice to CPXV. We confirmed that B8R-specific CD8p numbers were reduced with age (12 wk old = 1603 ± 124.2; 18 mo old = 810.0 ± 115.9, Fig. 2A; flow cytometry gating strategy, see Supplemental Fig. 1). Because the environment of the old mouse is reported to impair T cell responses, we compared the responsiveness of adult and old CD8 T cells in the adult environment. We transferred total adult or old splenic and LN T cells into Rag1^−/− mice and challenged them with wt CPXV (Fig. 2B). Numbers of total T cells, CD8 T cells, B8R-specific CD8 T cells, and IFN-γ–producing CD8 T cells did not significantly differ with age on day 7 p.i. with i.n. administered CPXV (Fig. 2C–F). Therefore, a 2-fold reduction of B8R CD8 precursor numbers in 18-mo-old mice was not sufficient to impair the response of old CD8 T cells in a young adult environment. This result further suggests that in the context of appropriate environmental cues, old CD8 T cells may be able to produce robust primary responses.

CPXV12 and CPXV203 work together to prevent MHC I Ag presentation on the surface of infected cells, blocking their detection by CD8 T cells (18). Although cross-presentation allows for priming a CD8 T cell response against CPXV in young mice, cross-presentation is impaired with age (26–29). We hypothesized that restoring direct presentation during priming may improve CD8 T cell responses in old mice and provide insight into how the basic biology of Ag presentation may impact immunity in a vulnerable population with suboptimal cellular immunity.

We confirmed that CPXV titers were equivalent in adult and old mice infected with either wt CPXV or CPXV∆12∆203 (Supplemental Fig. 2A). We next compared the CD8 T cell responses against CPXV and CPXV∆12∆203, both among the endogenous old or adult T cells, and, to control for possible intrinsic CD8 T cell defects in old mice, also among adult Ly5.1 congenic CD8 T cells transferred into adult and old mice (Fig. 3A). In adult mice, we confirmed (18) that there were no significant differences in the CD8 T cell response between the two viruses as measured by total number of CD8 T cells (Supplemental Fig. 2B), the number of B8R-specific CD8 T cells (Fig. 3B), or the frequency of IFN-γ–producing CD8 T cells (Fig. 3C). By contrast, old mice mounted an improved response (significantly higher number of B8R-specific and higher frequency of IFN-γ–producing CD8 T cells) in response to CPXV∆12∆203 (Fig. 3B, 3C). Total numbers of CD8 T cells remained comparable for both viruses (Supplemental Fig. 2B). The magnitude of the endogenous response in old mice remained 3- to 4-fold lower compared with adult mice.

This pointed to additional age-related defects in the endogenous CD8 T cell responses, which we substantiated by examination of responding transferred adult CD8 T cells in the adult or old environments. We saw improved priming with CPXV∆12∆203 in both adult and old environments, as evidenced by more Ly5.1 congenic B8R-specific CD8 T cells and increased percentages of IFN-γ⁺ cells responding to B8R peptide (Fig. 3D, 3E). Indeed, adult CD8 T cells primed with CPXV∆12∆203 in the old environment expanded to the same numbers like adult CD8 T cells responding to wt CPXV in the adult environment (Fig. 3D). The total number of donor Ly5.1 CD8 T cells recovered was comparable between groups (Supplemental Fig. 2C). Therefore, the restoration of direct Ag presentation improved primary CD8 T cell responses. Of note, previous studies examining CPXV∆12∆203 in young mice, where robust CD8 T cell responses exist in the absence of direct presentation, were unable to resolve the impact of immune evasion on CD8 T cell responses. The positive contribution of direct Ag presentation was below the limit of resolution for the assay and became fully evident only when we examined a vulnerable old population with suboptimal immunity.

Efficacious and protective long-term memory responses are key to immune defense. We, therefore, tested the efficacy of memory CD8 T cell responses elicited by CPXV∆12∆203. The acute CD8 T cell response measured in the blood on day 7 p.i. recapitulated previous data from the spleen, with superior responses to both infections in adult versus old mice and an improved response in old mice to ∆12∆203 as compared with wt CPXV (Fig. 4A). In the resting memory phase, at day 45 p.i., Ag-specific CD8 T cell frequencies were comparable between groups, albeit those in old mice immunized with the ∆12∆203 trended higher (Fig. 4B). Most importantly, upon challenge with Lm-B8R, mice vaccinated with CPXV∆12∆203 exhibited improved clearance as measured by decreased bacterial burden in the spleen (Fig. 4C). This clearance was similarly improved in both adult and old mice, and CD8 memory T cell responses in old mice vaccinated with CPXV∆12∆203 were as potent as their adult counterparts in clearing the bacterium.

The data presented in this article underscore the significant contribution of direct Ag presentation to the priming of functional CD8 T cell memory, as one of the key mechanisms to long-term resistance to infections. Our results stress the extent to which these mechanisms function in a vulnerable population to provide superior CD8 T cell memory, over and above what would have been predicted based on primary responses. Specifically, by circumventing immune evasion mechanisms that downregulate class I expression and prevent the participation of infected DCs in CD8 T cell priming, we were able to generate strongly protective memory CD8 T cell responses against an intracellular bacterium.

These results confirm our prior data in the West Nile virus model (8) that in old mice weak primary CD8 T cell responses do not predict strength of CD8 T cell protection upon recall. They also underscore the importance of strong direct Ag presentation in the course of priming of functional CD8 T cell memory in a vulnerable population and highlight the unappreciated interplay between immune evasion and reduced immune resources in determining long-term immune protection.

Our studies also highlight the importance of using vulnerable populations with suboptimal immunity as a tool to dissect biological relevance of microbial immune evasion in the context of antimicrobial and antivaccine responses. Indeed, CD8 responses after direct Ag presentation were not significantly improved when assessed in adult mice that possess strong, competent, cross-presentation mechanisms (26–28). The contribution of direct presentation could only be appreciated after adoptive transfer of adult CD8 T cells into old mice and by studies of memory populations, neither of which would have been undertaken in the absence of the improved endogenous response measured in old mice.

The earlier basic findings have substantial and immediate relevance to vaccine design. The contribution of poor CD8 T cell responses to increased susceptibility to infection with aging is pathogen specific (3, 4, 6, 30). Nonetheless, our study highlights that ensuring robust direct presentation is a promising vaccine-engineering strategy in vulnerable populations against strictly intracellular microbial pathogens that do not lend themselves to Ab neutralization. Such infections include, but are not limited to, Listeria and Salmonella and viruses whose control requires robust CD8 T cell responses, including respiratory syncytial virus.

K^b-restricted B8R tetramer was obtained through the National Institutes of Health Tetramer Core Facility.

The authors have no financial conflicts of interest.