Durable Mucosal Simian Immunodeficiency Virus-Specific Effector Memory T Lymphocyte Responses Elicited by Recombinant Adenovirus Vectors in Rhesus Monkeys

Abstract

The induction of potent and durable cellular immune responses in both peripheral and mucosal tissues may be important for the development of effective vaccines against human immunodeficiency virus type 1 and other pathogens. In particular, effector responses at mucosal surfaces may be critical to respond rapidly to incoming mucosal pathogens. Here we report that intramuscular injection of nonreplicating recombinant adenovirus (rAd) vectors into rhesus monkeys induced remarkably durable simian immunodeficiency virus (SIV)-specific T lymphocyte responses that persisted for over 2 years in both peripheral blood and multiple mucosal tissues, including colorectal, duodenal, and vaginal biopsy specimens, as well as bronchoalveolar lavage fluid. In peripheral blood, SIV-specific T lymphocytes underwent the expected phenotypic evolution from effector memory T cells (T_EM) to central memory T cells (TCM) following vaccination. In contrast, mucosal SIV-specific T lymphocytes exhibited a persistent and durable T_EM phenotype that did not evolve over time. These data demonstrate that nonreplicating rAd vectors induce durable and widely distributed effector memory mucosal T lymphocyte responses that are phenotypically distinct from peripheral T lymphocyte responses. Vaccine-elicited T_EM responses at mucosal surfaces may prove critical for affording protection against invading pathogens at the mucosal portals of entry.

INTRODUCTION

Mucosal surfaces serve as the primary portal of human immunodeficiency virus type 1 (HIV-1) entry (16), and mucosa-associated lymphoid tissues, particularly in the gastrointestinal tract, experience profound depletion of CD4⁺ T lymphocytes during acute infection (6, 23, 32, 44). The generation of potent and durable mucosal immune responses by vaccination may therefore be important for blocking the transmission event and for containing the early phase of infection (4, 8, 13, 24, 29, 41). Effector responses may prove particularly important because of their ability to respond rapidly to incoming pathogens (17, 39, 40).

Multiple studies have suggested that mucosal T lymphocyte responses may contribute to containing early viral infection in HIV-1-infected individuals and in simian immunodeficiency virus (SIV)-challenged animals (5, 10–12, 14, 15, 20). Previous reports have shown that a variety of vaccine regimens, including peptides, DNA, bacterial vectors, and viral vectors, induce HIV-1- or SIV-specific cellular immune responses at mucosal sites in rhesus monkeys (3, 5, 9, 15, 27, 35, 42, 43). Durable effector responses at mucosal surfaces have been reported for replicating vectors such as rhesus cytomegalovirus (RhCMV) (14, 15) but have not previously been reported for nonreplicating vectors. In particular, nonreplicating vectors have been shown to induce primarily central memory T cell (T_CM) responses rather than effector memory T cell (T_EM) responses in the periphery (14, 15).

We have reported that intramuscular (i.m.) injection of replication-incompetent recombinant adenovirus (rAd) vectors induces antigen-specific mucosal cellular immune responses in mice and rhesus monkeys (19). Here we investigated the durability and phenotypic evolution of SIV-specific T lymphocyte responses in both the periphery and mucosal compartments following the vaccination of rhesus monkeys. Mucosal and peripheral cellular immune responses exhibited similar dynamics, durability, and cytokine secretion profiles, suggesting that these immune compartments are highly coordinated during vaccination. Peripheral SIV-specific T lymphocytes underwent a clear phenotypic evolution from T_EM to T_CM responses following vaccination. In contrast, vaccine-elicited mucosal T lymphocytes demonstrated a distinct phenotype with persistent T_EM responses and without a transition to T_CM responses.

MATERIALS AND METHODS

Rhesus monkeys and immunizations.

Adult outbred rhesus monkeys (Macaca mulatta) of Indian ancestry that were specific pathogen free for B virus, simian T-lymphotropic virus type I, simian type D retroviruses, and SIV were housed at the New England Primate Research Center (Southborough, MA). All studies were approved by the Harvard Medical School Institutional Animal Care and Use Committee. The presence of the Mamu-A*01 allele was determined by PCR and sequencing (33). The immunization schedules of the different groups of monkeys are summarized in Table 1. Monkeys were injected i.m. with 10¹¹ viral particles (vp) of rAd vectors of various serotypes expressing SIVmac239 Gag in 1 ml of sterile phosphate-buffered saline divided equally between the two quadriceps muscles. Peripheral blood was collected to determine peripheral SIV-specific T lymphocyte responses. Bronchoalveolar lavage (BAL) fluid and pinch biopsy specimens of the colorectal, duodenal, and vaginal mucosae were collected to evaluate mucosal SIV-specific T cell responses. Healthy Mamu-A*01⁺ monkeys were used to determine the background SIV-specific immune responses, cell activation status, and memory phenotype of T lymphocytes in both peripheral and mucosal biopsy specimens.

Table 1.

Immunization schedules used in this study

Group (no. of monkeys)	Priming (time point [wk])	Boosting (time point [wk])	Manu-A*01 statusa
			Positive	Negative
Single immunization, rAd26 (5)	rAd26 Gag (0)	None	5	0
Homologous prime- boost, rAd26- rAd26 (6)	rAd26 Gag (0)	rAd26 Gag (26)	3	3
Heterologous prime- boost, rAd5HVR48- rAd26 (3)	rAd5HVR48 Gag (0, 16)	rAd26 Gag (88)	2	1
Heterologous prime- boost, rAd26/ rAd5 (6)	rAd26 Gag (0)	rAd5 Gag (32)	0	6

^aNumber of monkeys with or without the Mamu-A*01 allele.

Lymphocyte isolation.

Lymphocytes were isolated from peripheral blood by Ficoll density gradient sedimentation. BAL fluid lymphocytes were collected by centrifuging lavage fluid at 1,500 rpm for 5 min. Mucosal lymphocytes were isolated from tissues essentially as previously described (28). Briefly, biopsy specimens were incubated in RPMI 1640 medium containing 10% fetal bovine serum supplemented with 200 U/ml type IV collagenase (Sigma-Aldrich) and 30 U/ml DNase I (Sigma-Aldrich) at 37°C with rocking for 30 min. The digested biopsy specimen tissues were then homogenized, and the solution was strained with a 70-μm cell strainer (BD Falcon). The cell suspension was then centrifuged at 1,800 rpm for 25 min on a 35% Percoll (Sigma-Aldrich) gradient. Pellets containing lymphocytes were collected and processed for assays.

Tetramer staining.

Multiparameter tetramer staining assays were performed essentially as previously described (37, 43). All monoclonal antibodies (MAbs) were purchased from BD Biosciences unless otherwise indicated. Peripheral blood mononuclear cells (PBMC; 1 × 10⁶) or mucosal lymphocytes were stained with Mamu-A*01 tetramers labeled with phycoerythrin (PE) and folded around the immunodominant SIV Gag epitope CM9 (CTPYDINQM; also known as p11c) (2) in conjunction with predetermined titers of MAbs against CD3 (SP34.2; Alexa 700), CD4 (L200; AmCyan), CD8 (SK1; allophycocyanin-cyanine 7 [APC-Cy7]), CD69 {TP1.55.3; PE-Texas Red [energy-coupled dye; (ECD)]; Beckman Coulter}, CCR7 (150503; fluorescein isothiocyanate [FITC]; R&D Systems), CD95 (DX2; APC), and CD28 (L293; peridinin chlorophyll A-cyanine 5.5 [PerCP-Cy5.5]) at room temperature for 30 min. Stained samples were fixed with 1.0% paraformaldehyde solution and evaluated with an LSR II (BD Biosciences). Flow data were analyzed using FlowJo software (Tree Star). Approximately 300,000 to 500,000 events were collected per sample. The background for tetramer staining was typically below 0.01% of the gated CD8⁺ T lymphocytes from peripheral blood and below 0.05% of the gated CD8⁺ T lymphocytes for biopsy samples.

Intracellular cytokine staining (ICS) assays.

ICS assays were performed essentially as previously described (25, 26). PBMC (1.5 × 10⁶) or mucosal lymphocytes were incubated for 6 h at 37°C with medium, 10 pg/ml phorbol myristate acetate, and 1 μg/ml ionomycin (Sigma-Aldrich) or 2-μg/ml SIVmac239 Gag peptide pools. Cultures contained monensin (GolgiStop; BD Biosciences) and 1 μg/ml MAb against human CD49d (clone 9F10; BD Biosciences). The cells were then stained with predetermined titers of MAbs against CD3 (SP34; Alexa 700), CD4 (L200; AmCyan), CD8 (SK1; APC-Cy7), CD28 (L293; PerCP-Cy5.5), CD95 (DX2; PE), CD69 (TP1.55.3; PE-Texas Red [ECD]; Beckman Coulter), gamma interferon (IFN-γ; B27; PE-cyanine 7 [PE-Cy7]), interleukin-2 (IL-2; MQ1-17H12; APC), and tumor necrosis factor alpha (TNF-α; Mab11; FITC). In certain studies, MAbs against TNF-α (Mab11; Pacific Blue; eBioscience) and CCR7 (150503; FITC; R&D Systems) were used to facilitate the evaluation of T lymphocyte subsets and phenotypes. Background responses were typically below 0.01% of the gated CD4⁺ or CD8⁺ peripheral T lymphocytes and below 0.05% of the gated CD4⁺ or CD8⁺ mucosal lymphocytes. ICS values were analyzed following subtraction of the background. Polyfunctionality of CD8⁺ and CD4⁺ T lymphocytes was analyzed using Boolean gating of FlowJo software.

Statistical analysis.

Comparisons of immunologic data among groups of monkeys were performed by analysis of variance (ANOVA) with Turkey posttest adjustments utilizing GraphPad Prism 4.

RESULTS

rAd vectors induce widely distributed and durable SIV-specific cellular immune responses in rhesus monkeys.

We initiated studies by investigating the anatomic distribution and durability of mucosal Gag-specific CD8⁺ T lymphocyte responses in rhesus monkeys following (i) single-shot, (ii) homologous prime-boost, and (iii) heterologous prime-boost immunizations with rAd vectors expressing SIV Gag. We first evaluated the ability of a single immunization with rAd26 expressing SIV Gag (1) to elicit mucosal cellular immune responses. Five Mamu-A*01-positive monkeys were immunized i.m. with 10¹¹ vp rAd26 SIV Gag, and CD8⁺ T lymphocyte responses specific for the Mamu-A*01-restricted Gag epitope CM9 (CTPYDINQM) (2) were determined by multiparameter tetramer-binding assays at 2 weeks postvaccination. All five monkeys developed CM9-specific CD8⁺ T lymphocyte responses not only in their peripheral blood but also in their colorectal and duodenal mucosae (Fig. 1A). Naïve Mamu-A*01⁺ monkeys exhibited background tetramer-binding responses of <0.05% (Fig. 1A). The magnitude of CM9-specific CD8⁺ T lymphocyte responses in the colorectal and duodenal mucosae was 11 to 83% (average, 40%) of the magnitude of the responses in peripheral blood.

Fig. 1.

Widely distributed and durable mucosal CD8⁺ T lymphocyte responses elicited by rAd vectors in rhesus monkeys. Rhesus monkeys were subjected to a single-immunization, a homologous prime-boost, or a heterologous prime-boost rAd vaccination regimen. SIV-specific CD8⁺ T lymphocyte responses from peripheral blood and mucosal specimens specific for the Mamu-A*01-restricted Gag epitope CM9 (CTPYDINQM) were determined by multiparameter tetramer-binding assays. The percentage of CM9-specific CD8⁺ T lymphocytes (%Tet⁺) is depicted. (A) SIV-specific CD8⁺ T cell responses in multiple anatomic compartments in naïve monkeys (left graph) and 2 weeks following a single vaccination with rAd26 Gag (right graph). (B) SIV-specific CD8⁺ T lymphocyte responses following rAd26 Gag priming at week 0 and homologous rAd26 Gag boosting at week 26. (C) Representative CM9-specific tetramer binding of gated CD3⁺ CD8⁺ T lymphocytes at 2 weeks postboost following rAd5HVR48 Gag priming at weeks 0 and 16 and rAd26 boosting at week 88. One Mamu-A*01-negative monkey (no. 153) was included as a negative control. (D) SIV-specific CD8⁺ T lymphocyte responses following rAd5HVR48 Gag priming and rAd26 Gag boosting at weeks 90, 110, and 224, corresponding to weeks 2, 22, and 136 postboost.

We next evaluated the ability of homologous and heterologous rAd prime-boost regimens to induce mucosal Gag-specific CD8⁺ T lymphocyte responses. Three Mamu-A*01-positive monkeys were immunized with rAd26 Gag at week 0 and boosted with the homologous vector rAd26 Gag at week 26. Peripheral and mucosal CD8⁺ T lymphocyte responses following two homologous rAd26 Gag immunizations were not substantially higher than those induced by a single rAd26 immunization, presumably as a result of the development of vector-specific neutralizing antibodies generated by the priming immunization (Fig. 1B). In contrast, two Mamu-A*01-positive monkeys primed at weeks 0 and 16 with rAd5HVR48 Gag (38) and boosted at week 88 with the heterologous vector rAd26 Gag developed greater magnitude CD8⁺ T lymphocyte responses in multiple mucosal compartments at 2 weeks following the boost immunization, including in colorectal, duodenal, and vaginal mucosae, as well as in BAL fluid, with magnitudes comparable (average of 132% for colorectal, duodenal, and vaginal responses) with those observed in peripheral blood (Fig. 1C). As a negative control, a similarly vaccinated Mamu-A*01-negative monkey (no. 153) exhibited no tetramer-binding responses as expected. Mucosal and peripheral tetramer-binding CD8⁺ T lymphocyte responses showed similar kinetics, with a peak at 2 weeks postboost (week 90) and a 65 to 70% decline in magnitude by 22 weeks postboost (week 110), but then showed remarkable durability, with only an additional 5 to 10% decline over the next 2 years (week 224). Colorectal, duodenal, vaginal, and BAL fluid responses all exhibited long-term durability in these monkeys (Fig. 1D).

Mucosal vaccine-elicited SIV-specific T lymphocytes secrete multiple cytokines.

To evaluate the cytokine secretion profiles of SIV-specific CD8⁺ and CD4⁺ T lymphocytes elicited by the heterologous rAd5HVR48-rAd26 prime-boost regimen, multiparameter ICS assays were performed using both peripheral and mucosal T lymphocytes. Figure 2A depicts a representative ICS assay showing IFN-γ expression from CD8⁺ T lymphocytes from peripheral blood, as well as from colorectal, duodenal, and vaginal biopsy specimens. TNF-α responses of CD8⁺ T lymphocytes were comparable to IFN-γ responses, whereas IL-2 responses were typically severalfold lower. Following the boost immunization, all three monkeys in this study exhibited multiple subpopulations of Gag-specific CD8⁺ and CD4⁺ T lymphocytes that secreted IFN-γ, TNF-α, and IL-2 both alone and in combination. The overall magnitude, cytokine secretion profiles, and polyfunctionality appeared comparable in the systemic and mucosal compartments. CD4⁺ T lymphocyte responses were severalfold lower in magnitude than CD8⁺ T lymphocyte responses in all of the compartments (Fig. 2B).

Fig. 2.

Cytokines secretion profiles of vaccine-elicited mucosal T lymphocytes. The cytokine secretion profiles of SIV-specific CD8⁺ and CD4⁺ T lymphocytes elicited by the heterologous rAd5HVR48-rAd26 prime-boost regimen were determined by multiparameter ICS assays. Polyfunctionality was analyzed with Boolean gating using FlowJo software. (A) Representative data on IFN-γ secretion from peripheral and mucosal CD8⁺ T lymphocytes from animal 210 at 6 weeks postboost. (B) Summary of CD8⁺ and CD4⁺ T lymphocyte cytokine secretion profiles in peripheral and mucosal compartments at 6 weeks postboost.

Persistent activation of vaccine-elicited SIV-specific T lymphocytes at mucosal sites.

Mucosal lymphoid tissues represent effector sites that frequently confront invading pathogens, and thus mucosal T lymphocytes typically demonstrate an activated phenotype characterized by CD69 expression (Fig. 3A) (21, 22, 31, 43). However, the kinetics of cellular immune activation in the periphery and in the mucosa following vaccination remains to be determined. In peripheral blood, the total CD8⁺ and CD4⁺ T lymphocytes from the three rAd5HVR48-rAd26 vaccinated monkeys exhibited transient and low levels of CD69 upregulation following vaccination, with CD69 expression decreasing to baseline levels within 4 to 8 weeks (Fig. 3B). In contrast, the total CD8⁺ and CD4⁺ T lymphocytes isolated from colorectal, duodenal, and vaginal biopsy specimens showed persistent high levels of CD69 expression (Fig. 3C). SIV-specific CD8⁺ T lymphocytes in peripheral blood similarly exhibited transient CD69 expression shortly after the week 88 rAd26 boost (week 90), which largely resolved by week 110, whereas SIV-specific CD8⁺ T lymphocytes at mucosal surfaces exhibited persistent CD69 expression, similar to resident mucosal CD8⁺ T lymphocytes (Fig. 3D). These findings were confirmed in eight additional Mamu-A*01-positive rhesus monkeys that received one or two immunizations with rAd26 Gag (Fig. 3E).

Fig. 3.

T cell activation of vaccine-elicited peripheral and mucosal T lymphocytes. (A) Expression of CD69 on peripheral and mucosal CD8⁺ and CD4⁺ T lymphocytes in 12 healthy rhesus monkeys. **, P < 0.001 (ANOVA comparing the percentages of cells expressing CD69). (B) CD69 expression on total peripheral T lymphocytes following rAd5HVR48 priming at weeks 0 and 16 and rAd26 boosting at week 88. (C) CD69 expression on total mucosal T lymphocytes following rAd5HVR48 priming at weeks 0 and 16 and rAd26 boosting at week 88. (D) CD69 expression on tetramer-positive, Gag-specific CD8⁺ T lymphocytes following rAd5HVR48-rAd26 immunization. ND, not done. (E) CD69 expression on tetramer-positive, Gag-specific CD8⁺ T lymphocytes following rAd26 immunization in eight additional rhesus monkeys.

Phenotypic evolution of vaccine-elicited SIV-specific T lymphocytes in peripheral blood.

We next analyzed the phenotypic evolution of SIV-specific T lymphocyte responses in the peripheral blood of monkeys immunized with the heterologous rAd5HVR48-rAd26 prime-boost regimen as described in Fig. 1C andD. Peripheral CD8⁺ and CD4⁺ T lymphocytes expressing the memory marker CD95 were divided into three subpopulations according to the expression of the lymph node homing receptor CCR7 and the costimulatory molecule CD28. CD95⁺ CCR7⁻ CD28⁺ defined transitional effector memory T_EM1 cells, CD95⁺ CCR7⁻ CD28⁻ defined fully differentiated effector memory T_EM2 cells, and CD95⁺ CCR7⁺ CD28⁺ defined central memory T_CM cells, as previously reported (36).

To study the phenotypic evolution of vaccine-elicited CD8⁺ T lymphocytes in peripheral blood, multiparameter tetramer binding assays and multiparameter ICS assays were performed longitudinally with these three vaccinated monkeys over a period of 4 years (224 weeks) (Fig. 4A). Gag-specific CD8⁺ T lymphocytes in PBMC were predominantly T_EM1 and T_EM2 shortly after vaccination. These responses evolved into a clear T_CM phenotype, with 75 to 80% of the peripheral Gag-specific CD8⁺ T lymphocytes developing T_CM markers (CD95⁺ CD28⁺ CCR7⁺) over time (Fig. 4A and B). A similar phenotypic shift from a T_EM to a T_CM response was observed in ICS assays for both CD8⁺ and CD4⁺ Gag-specific T lymphocytes (Fig. 4A).

Fig. 4.

Phenotypic evolution of vaccine-elicited T lymphocytes in peripheral blood. (A) Rhesus monkeys were primed with rAd5HVR48 Gag at weeks 0 and 16 (thin arrows at the bottom) and boosted with rAd26 Gag at week 88 (thick arrow at the bottom). The magnitudes and phenotypes of SIV-specific T lymphocyte responses were determined by multiparameter tetramer-binding assays (top graphs) and CD8⁺ and CD4⁺ T lymphocyte ICS assays (middle and bottom graphs, respectively). Gag-specific, tetramer-binding, or IFN-γ-secreting T lymphocytes were divided into transitional effector memory T_EM1, fully differentiated effector memory T_EM2, and central memory T_CM cells according to the expression of CD95, CCR7, and CD28. (B) Expanded time frame postimmunization of monkey 184. (C) Magnitude of IFN-γ secretion from CD8⁺ and CD4⁺ T lymphocytes elicited by the rAd26-rAd5 prime-boost regimen in six additional rhesus monkeys. (D) Phenotypic evolution of Gag-specific, IFN-γ⁺ T lymphocytes elicited by the rAd26-rAd5 prime-boost regimen at weeks 2, 16, and 24 postboost. *, P < 0.05 (ANOVA comparing the percentages of T_CM cells at weeks 16 and 24 with those at week 2 postboost).

To confirm the generalizability of these observations in a larger number of monkeys, we assessed the phenotypic evolution of vaccine-elicited Gag-specific T lymphocytes in a separate cohort of six Mamu-A*01-negative monkeys immunized with a different heterologous rAd26 Gag prime-rAd5 Gag boost regimen (26) using multiparameter ICS assays. Peripheral blood was collected at weeks 2, 16, and 24 following the boost immunization. Animals developed robust Gag-specific CD8⁺ and CD4⁺ T lymphocyte responses at 2 weeks postboost, followed by a typical contraction (Fig. 4C). As we observed in the previous experiment, both CD8⁺ and CD4⁺ T lymphocytes demonstrated a phenotypic shift from T_EM to T_CM following vaccination (P < 0.05 [ANOVA with Turkey posttest]) (Fig. 4D).

Persistent effector memory SIV-specific T lymphocytes in mucosal tissues.

Mucosal microenvironments are phenotypically distinct from the periphery and may substantially impact the phenotype of vaccine-elicited T lymphocytes (19, 22, 30, 31, 34). The total mucosal T lymphocytes isolated from colorectal, duodenal, and vaginal biopsy specimens, as well as from BAL fluid, were >80% CD95⁺ memory T lymphocytes (Fig. 5A), whereas the total peripheral T lymphocytes isolated from blood and lymph nodes included substantial numbers of CD95-naïve lymphocytes, consistent with previous results (36, 43). Mucosal CD8⁺ CD95⁺ and CD4⁺ CD95⁺ T lymphocytes were also >90% T_EM1 and T_EM2, while peripheral T lymphocytes contained balanced numbers of T_EM1, T_EM2, and T_CM subpopulations (Fig. 5B).

Fig. 5.

Persistent transitional memory phenotype of vaccine-elicited T lymphocytes in mucosal compartments. (A) CD95 expression on total CD8⁺ and CD4⁺ T lymphocytes from multiple systemic and mucosal compartments in rhesus monkeys. (B) Total CD8⁺ CD95⁺ and CD4⁺ CD95⁺ T lymphocytes were divided into T_EM1, T_EM2, and T_CM subpopulations by the expression of CCR7 and CD28 in both vaccinated and naïve monkeys. (C) Representative data showing T_EM1, T_EM2, and T_CM subpopulations of peripheral and mucosal tetramer-positive, Gag-specific CD8⁺ T lymphocytes from animal 210 at week 90 (2 weeks postboost) and week 110 (22 weeks postboost). (D) T_EM1, T_EM2, and T_CM subpopulations of peripheral and mucosal tetramer-positive, Gag-specific CD8⁺ T lymphocytes following rAd5HVR48-rAd26 immunization. ND, not done. (E) T_EM1, T_EM2, and T_CM subpopulations of tetramer-positive, Gag-specific CD8⁺ T lymphocytes at weeks 2, 28, and 40 following rAd26 immunization in eight additional rhesus monkeys. *, P < 0.05 (ANOVA comparing the percentages of T_CM cells at weeks 28 and 40 with those at week 2 postimmunization).

We next evaluated the phenotypic evolution of mucosal Gag-specific CD8⁺ T lymphocytes in the three monkeys immunized with the heterologous rAd5HVR48-rAd26 prime-boost regimen. While peripheral Gag-specific CD8⁺ T lymphocytes showed the characteristic transition from T_EM to T_CM following the boost immunization (Fig. 4 and 5C and D), the majority of the mucosal Gag-specific CD8⁺ T lymphocytes isolated from colorectal, duodenal, or vaginal biopsy specimens exhibited a persistent T_EM phenotype (dominated by transitional effector memory T_EM1 cells) that did not substantially change from week 90 to week 224 (week 2 to week 136 postboost) (Fig. 5C and D). Thus, vaccine-elicited, Gag-specific CD8⁺ T lymphocytes at mucosal surfaces exhibited a durable and stable T_EM1 phenotype for over 2 years.

To confirm these results with eight additional rhesus monkeys, the memory phenotypes of mucosal Gag-specific CD8⁺ T lymphocytes elicited by a single rAd26 immunization or a homologous rAd26 prime-boost regimen were assessed. While Gag-specific CD8⁺ T lymphocytes in peripheral blood exhibited the characteristic shift from T_EM to T_CM from week 2 to week 40 following immunization (P < 0.05 [ANOVA with Turkey posttest]), mucosal Gag-specific CD8⁺ T lymphocytes in colorectal and duodenal biopsy specimens exhibited a persistent T_EM1 phenotype (Fig. 5E), similar to the results of the previous experiment. These data demonstrate that mucosal Gag-specific CD8⁺ T lymphocytes elicited by rAd vectors exhibited a persistent T_EM1 phenotype at mucosal surfaces.

DISCUSSION

The induction of potent and durable virus-specific cellular immune responses at mucosal sites may be critical for the development of an effective HIV/AIDS vaccine. Previous reports have shown that a variety of vaccine regimens can induce mucosal SIV-specific cellular immune responses in rhesus monkeys (3, 5, 9, 14, 15, 27, 35, 42, 43), but the phenotype of these mucosal cellular immune responses has remained unclear. It has been assumed that nonreplicating vaccine vectors elicit primarily T_CM responses, which would result in a delay in responding to antigenic challenges. Here we demonstrate that nonreplicating rAd vectors induced widely distributed and remarkably durable Gag-specific T lymphocyte responses that persisted for more than 2 years at multiple mucosal sites in rhesus monkeys. In the periphery, antigen-specific T lymphocytes exhibited a clear phenotypic evolution from T_EM to T_CM. In contrast, antigen-specific T lymphocytes at mucosal surfaces exhibited a persistent T_EM phenotype following vaccination. The mechanism accounting for these strikingly different phenotypes remains unclear but likely involves signals in the local mucosal microenvironment on vaccine-elicited T lymphocytes (19, 21, 30). We speculate that durable T_EM responses at mucosal surfaces elicited by vaccination may be capable of responding rapidly to invading pathogens at the mucosal portals of entry.

This study confirms and extends prior reports (3, 19, 43) showing that i.m. injection of rAd vectors can induce widely distributed cellular immune responses in multiple mucosal tissues, including the gastrointestinal tract, the vaginal tract, and the respiratory tract in rhesus monkeys. We have previously shown that priming within mucosal tissues is not required for T lymphocytes to traffic to mucosal tissues and that i.m. injection of rAd vectors appears to overcome the traditional compartmentalization of immune responses into peripheral and mucosal components (19). Thus, we speculate that antigen-specific T lymphocytes were primed in the periphery and then efficiently trafficked to mucosal surfaces, where they acquired the phenotypic characteristics of typical mucosal T lymphocytes as a result of signals in the mucosal microenvironments. We cannot exclude the alternative possibility that replication-incompetent rAd vectors trafficked directly to mucosal surfaces, but in vivo imaging of rAd vectors expressing luciferase did not demonstrate vector trafficking to the mucosa after the i.m. injection of mice (18).

In the present study, we evaluated vaccine-elicited mucosal immune responses longitudinally for >2 years following vaccination, allowing an unprecedented degree of characterization of the kinetics, durability, and phenotypic evolution of mucosal immune responses compared with prior studies. We observed that peripheral SIV-specific T lymphocytes elicited by rAd vectors exhibited a dramatic and reproducible phenotypic shift from T_EM to T_CM following vaccination (Fig. 4). In contrast, mucosal SIV-specific T lymphocytes elicited by rAd vectors exhibited a persistent T_EM phenotype (dominated by transitional effector memory T_EM1 cells) without evidence of evolution for over 2 years following vaccination (Fig. 5C to E).

The relative importance of and functional differences between the T_EM1 and T_EM2 cells remain to be determined, but both of these effector memory subpopulations have been shown to respond more rapidly to antigen stimuli than T_CM cells (40). Replicating RhCMV vectors have been shown to induce durable and persistent effector responses comprising both T_EM1 and T_EM2 cells in rhesus monkeys both in the periphery and at mucosal surfaces (15). It has therefore been assumed that replicating viral vectors would be required to elicit durable T_EM responses (14, 15). In this report, we show that nonreplicating rAd vectors induced persistent T_EM1 responses at multiple mucosal surfaces. Whether T_EM1 mucosal effector memory T lymphocyte responses are sufficient to protect against mucosal virus challenges remains to be determined. The implications of these data with respect to the findings from the phase 2b STEP study evaluating a rAd5 Gag/Pol/Nef vaccine (7) also require further investigation, given the substantial differences in the cellular immune phenotypes elicited by the different serotype rAd vectors (25).

In summary, our data demonstrate that i.m. injection of nonreplicating rAd vectors induced widely distributed and remarkably durable mucosal SIV-specific cellular responses that persisted for over 2 years in rhesus monkeys. Although we did not compare the i.m. and mucosal routes of immunization, previous experiments with mice have not suggested additional benefits of using the mucosal immunization route for rAd vectors (18). Vaccine-elicited mucosal SIV-specific T lymphocytes exhibited phenotypic features profoundly different from those of peripheral SIV-specific T lymphocytes. In particular, vaccine-elicited mucosal T lymphocytes exhibited a persistently activated T_EM1 phenotype without the characteristic shift to T_CM cells that was readily observed in the periphery. These durable and activated T_EM responses at the mucosal portal of entry may prove useful in containing the early events following mucosal viral challenges. Future studies should therefore directly evaluate the protective efficacy of these mucosal immune responses against viral challenges in rhesus monkeys.