Role of B Cells in Mucosal Vaccine–Induced Protective CD8⁺ T Cell Immunity against Pulmonary Tuberculosis

Mycobacterium tuberculosis accounts for 1.5 million deaths annually, making it the largest cause of mortality by a bacterial pathogen (1). Currently, bacillus Calmette–Guérin (BCG) is the only licensed vaccine against tuberculosis (TB) with >3 billion people vaccinated worldwide (2). Although BCG confers protection against severe or disseminated forms of childhood TB, it is not protective against adult pulmonary forms of the disease (2). For this reason, in the past 10 y or so, research has focused on developing heterologous booster vaccines such as recombinant protein–based (plus adjuvant), mycobacterial-based, and viral vector–based systems (2). A number of novel TB vaccines, including H56+IC31, recombinant BCG (VPM1002), Mycobacterium vaccae, MVAAg85A, AERAS402, and adenovirus human serotype 5 Ag85A (AdHu5Ag85A), are currently in the clinical developmental pipeline (2).

M. tuberculosis has evolved to avoid Ab-mediated immunity by primarily residing and growing within macrophages. For this reason, B cells had been largely ignored in the field of M. tuberculosis and TB vaccine research owing to the classical idea that T cell–mediated immunity is crucial to providing protection against intracellular pathogens, whereas B cells play a small role (3, 4). However, recently there has been a growing body of evidence supporting the roles of B cells and Abs in protection against intracellular bacterial pathogens, including M. tuberculosis (4–22). Specifically, by using a number of approaches, varying results from the primary M. tuberculosis challenge model have been obtained (4, 7, 8, 10, 12, 15, 16). For instance, some studies using mice with genetic B cell deficiencies reported unimpaired T cell responses and TB protection compared with wild-type (WT) controls (12, 15). These findings differ from those of Vordermeier et al. (16) who found higher bacterial counts in B cell knockout (KO) mice using high-dose M. tuberculosis challenge. Moreover, several studies using B cell–deficient mice have demonstrated the ability of B cells to modulate T cell responses to infection (5, 8, 17–20). Despite the increasing attention to understanding the role of B cells in host defense against primary M. tuberculosis infection, few studies have investigated the role of B cells in TB vaccine–induced protective immunity (8, 22, 23). Kozakiewicsz et al. (8) represents the only group that has addressed the causal relationship between B cells and the immune responses triggered by the parenteral route of BCG vaccination. To date, the role of B cells in respiratory mucosal TB vaccine–induced T cell immunity still remains unaddressed.

Mounting evidence suggests that inducing immunity at the site of pathogen entry is critical for effective protection. In other words, respiratory mucosal immunization is the most effective way to induce robust protective immunity locally in the lung against pulmonary TB (2, 3). Among all vaccine platforms, viral platforms are particularly amenable for respiratory mucosal immunization because of their natural tropism and potency. The human adenovirus serotype 5 expressing the immunodominant M. tuberculosis–secreted Ag85A (AdHu5Ag85A) is an example of a model viral vector–based booster vaccine that has advanced from bench to human clinical trials (2, 24–30). AdHu5Ag85A, particularly when administered via the respiratory mucosal route, has proven to be highly effective in inducing protective T cell immunity against M. tuberculosis infection in the lung, both as a stand-alone and booster vaccine in a number of animal models (25–27, 29). However, whether respiratory mucosal AdHu5Ag85A vaccination induces Ag-specific B cell responses and whether B cells have a role in influencing T cell responses and protective immunity against M. tuberculosis infection remain unclear.

In the present study, by using murine models and multiple approaches, we have investigated whether 1) respiratory AdHu5Ag85A vaccination induces Ag-specific B cell responses, 2) whether such B cell responses have an immunomodulatory effect on vaccine-induced T cell responses in the lung, and 3) whether B cells contribute to vaccine-induced immune protection against M. tuberculosis infection in the lung. We find that whereas respiratory mucosal vaccination activates B cells, they are not required for vaccine-activated T cell responses and protective immunity against subsequent pulmonary M. tuberculosis challenge. Thus, our findings suggest a minimal role by B cells and highlight a central role of T cells in respiratory mucosal vaccine–induced protective immunity.

Age- and sex-matched 6- to 8-wk-old female BALB/c and B cell KO mice (Igh J^tm1Dhu N?+N2) were purchased from Taconic (31). Age- and sex-matched 6- to 8-wk-old female BALB/c SCID mice were purchased from The Jackson Laboratory. Mice were housed in a specific pathogen-free facility at McMaster University in accordance with the Animal Research Ethics Board.

Mice were vaccinated through the intranasal (i.n.) route with recombinant replication-deficient Ad-based TB vaccine (AdHu5Ag85A). Development of the vaccine has been previously described (25). Each mouse received a dose of 5 × 10⁷ PFU in 25 μl PBS. After a duration of 4 wk, the animal was sacrificed.

M. tuberculosis bacilli preparation was performed as previously described (29, 32). The challenge experiments were carried out using the virulent strain M. tuberculosis H₃₇Rv (ATCC 27294). Mice were challenged i.n. with a dose of 10,000 CFU M. tuberculosis in 25 μl PBS per mouse.

The level of bacterial burden was determined by plating serial dilutions of the lung homogenates and left to incubate at 37°C for 21 d. Colonies were counted and calculated and displayed as log₁₀ CFU per organ.

B cells were depleted using 5D2 anti-mouse CD20 Ab (anti-CD20; Genentech, South San Francisco, CA). Mice were given an i.p. injection at an initial dose of 250 μg per mouse 2 d prior to vaccination and subsequently 200 μg per mouse every other week for maintenance of B cell depletion.

Ninety-six–well flat-bottom Immuno plates (Thermo Fisher Scientific) were coated with 0.01 μg/μl Ag85 complex proteins and left overnight at 4°C. Plates were washed with dilution buffer and samples were then added at various dilutions and incubated for 1 h at 37°C. Plates were washed again and anti-mouse IgG2a and IgA Abs (R&D Systems) were added at a dilution of 1:500 for 1 h at 37°C. Plates were again washed and strepavidin–alkaline phosphatase was added and incubated for 30 min at 37°C. Plates were washed and substrate was added (p-nitrophenyl phosphate tablet in diethanolamine) incubated at room temperature for 15–30 min. Lastly, 2 N NaOH stop solution was added and plates were read on a spectrophotometer at 405 nm.

Bronchoalveolar lavage (BAL) was carried out on lungs to collect airway luminal cells as well as the conventional BAL supernatants (25). Mononuclear cells were isolated from different tissues as previously described (27).

Approximately 2 million mononuclear cells from BAL, lung, and spleen were cultured for 5 h in the presence of GolgiPlug (5 μg/ml; BD Pharmingen, Mississauga, ON, Canada) with or without Ag85A-specific CD8 (MPVGGQSSF) T cell peptide at a concentration of 1 μg per well for 5–6 h. Cells were first incubated for 15 min on ice with the CD16/CD32 Fc block Ab (BD Pharmingen). The cells were then immunostained using a tetramer specific for Ag85A CD8 T cell peptide bound to BALB/c MHC class I allele H-2L^d and PE fluorochrome (National Institutes of Health Tetramer Core) for 1 h at room temperature and then washed. Extracellular staining was then carried out with mAbs, washed, permeabilized, and stained intracellularly. The following fluorochrome-conjugated mAbs were used: CD3-v450, CD8a-PE-Cy7, CD4-allophycocyanin-Cy7, and IFN-γ–allophycocyanin (BD Pharmingen). B cells were stained with B220-PE and CD19-PerCP-Cy5.5 (BD Pharmingen).

B cells were purified from naive WT spleens (>95% purity) using the EasySep mouse pan B cell isolation kit (StemCell Technologies, Vancouver, BC, Canada). T cells were purified from 4-wk AdHu5Ag85A (i.n.) vaccinated mouse lungs (>95% purity) using the mouse pan T cell isolation kit II (Miltenyi Biotec, Auburn, CA).

Lungs were fixed in 10% formalin and kept in 5 ml 10% formalin for at least 72 h. Tissue sections were stained with H&E for histological examination.

All statistical analyses were performed using GraphPad Prism software. A two-tailed Student t test was used for pairwise comparisons whereas a one-way ANOVA using a Tukey post hoc test was carried out when comparing multiple groups.

It has been previously shown that respiratory mucosal AdHu5Ag85A vaccine-induced Ag85A-specific T cell immunity is critical for protection (25, 27, 30). To begin evaluating the role of B cells in AdHu5Ag85A-induced T cell immunity, we first examined whether B cells were activated to produce Ag85A-specific Ab responses in the circulation and on the surface of the respiratory mucosa following respiratory mucosal AdHu5Ag85A vaccination. Indeed, by using ELISA, Ag85A-specific IgG2a Ab responses were detectable in the serum starting from 14 d postvaccination with titers markedly increased by 5-fold at 28 d postvaccination (Fig. 1A). In the BAL fluids, there were only minimally detectable levels of Ag85A-specific IgG2a and IgA at 14 d postvaccination whereas they increased remarkably at 28 d postvaccination (Fig. 1B, 1C). These data indicate that respiratory mucosal AdHu5Ag85A vaccination leads to the induction of mycobacterial Ag-specific B cell activation and Ab production both systemically and locally in the lung.

Having established that respiratory mucosal AdHu5Ag85A vaccination leads to Ag-specific B cell responses, we next examined the role of B cells in vaccine-induced Ag-specific T cell responses in the lung. We first addressed this issue by employing genetic B cell–deficient mice (Jh^−/− KO). To this end, both WT and Jh^−/− KO mice were vaccinated i.n. with AdHu5Ag85A and at 4 wk postvaccination, and Ag85A tetramer (tet⁺) and intracellular IFN-γ–producing CD8⁺ T cell responses were evaluated using immunostaining in the BAL and lung interstitium. In comparison with WT mice, the number of tet⁺CD8⁺ T cells in the BAL of Jh^−/− KO mice was reduced by ∼40% (Fig. 2A). Similarly, the number of CD8⁺IFN-γ⁺ cells in the BAL was also reduced by ∼50% (Fig. 2C). However, there were comparable numbers of tet⁺CD8⁺ and CD8⁺IFN-γ⁺ T cells in the lung interstitium of WT and Jh^−/− KO mice (Fig. 2B, 2D). Moreover, the number of CD4⁺IFN-γ⁺ cells was reduced within the BAL and lungs of Jh^−/− KO mice when compared with WT animals (Fig. 2E, 2F). These data suggest that genetic B cell deficiency leads to reduced CD8⁺ T cell responses only in the airway but not in the lung interstitium, whereas CD4⁺ T cell responses are altered in both compartments following respiratory mucosal AdHu5Ag85A vaccination.

Although genetic B cell–deficient mice have often been used for investigation by others, there could be inherent issues. For example, Jh^−/− KO mice were reported to have reduced T cells due to the absence of the chemokines required for the normal formation of T cell zones and abnormalities in dendritic cell populations (33–35). Given the limitations to using Jh^−/− KO mice, we next used an anti-mouse CD20 Ab (anti-CD20), also known as rituximab (licensed therapeutic) (36), to deplete B cells in WT mice. CD20 is a surface marker present on B cells but not on plasma cells (37). To this end, anti-CD20 or a control IgG was injected i.p. to WT mice 2 d prior to vaccination (Supplemental Fig. 1A). A second dose of anti-CD20 was administered at 2 wk after i.n. AdHu5Ag85A vaccination for the maintenance of B cell depletion (7, 8, 37, 38). This B cell depletion regimen effectively wiped out B cells in the lymph nodes, peripheral blood, and lung before vaccination (data not shown) as verified by immunostaining for CD19 and B220 surface markers expressed on developing and mature B cells. B cell depletion was effectively maintained in the lymph nodes, peripheral blood, and lung at 4 wk postvaccination (Supplemental Fig. 1B–D). As a result, different from control IgG-treated animals, such B cell–depleted animals were unable to produce anti-Ag85A IgG2a and IgA in the peripheral blood and lung following i.n. AdHu5Ag85A vaccination (Supplemental Fig. 1E).

With the B cell depletion model fully characterized, we evaluated the role of B cells in vaccine-induced Ag85A-specific T cell responses at 4 wk postvaccination (Supplemental Fig. 1A). We found that both the frequencies and the number of Ag85A tetramer-specific and IFN-γ–producing CD8⁺ T cell responses in the BAL of B cell–depleted animals were similar to those in control IgG-treated, B cell–competent animals (Fig. 3A, 3B). Likewise, similar frequencies and numbers of Ag85A tetramer-specific and IFN-γ–producing CD8⁺ T cells were seen in the lung interstitium of B cell–depleted and B cell–competent animals (Fig. 3C, 3D). Taken together, these data suggest that B cells play a negligible role in respiratory mucosal AdHu5Ag85A-induced Ag specific T cell responses in the lung.

Having examined the role of B cells in mucosal vaccine–induced Ag85A-specific T cell responses, we next evaluated the level of anti-TB protection in the lung of animals lacking B cells and compared it with B cell–competent counterparts following i.n. AdHu5Ag85A vaccination and pulmonary M. tuberculosis challenge. Thus, WT control mice, genetic Jh^−/− KO mice, and the WT mice depleted of B cells (as depicted in Supplemental Fig. 1A) were vaccinated i.n. with AdHu5Ag85A. These groups of mice were then challenged via the airway with virulent M. tuberculosis H37Rv at 4 wk postvaccination. Unvaccinated WT and Jh^−/− KO mice were also challenged with M. tuberculosis as a comparison. The levels of M. tuberculosis infection in the lung were assessed in all groups at 2 wk postchallenge. Consistent with previously published studies using B cell KO mice (16), unvaccinated Jh^−/− KO mice had increased M. tuberculosis burden in the lung compared with unvaccinated WT controls (Fig. 4A). Alternatively, compared with unvaccinated WT controls, i.n. AdHu5Ag85A-vaccinated WT mice had significantly reduced levels of infection in the lung (Fig. 4A). Compared to unvaccinated Jh^−/− KO mice, vaccinated Jh^−/− KO mice also demonstrated significantly reduced infection, although not to the same extent as in vaccinated WT animals (Fig. 4A). Furthermore, similar to vaccinated WT B cell–competent mice, the vaccinated WT mice depleted of B cells (anti-CD20 treated) had significantly reduced levels of infection in the lung (Fig. 4A).

Because the extent of histopathology in the lung serves as an independent sensitive indicator of anti-TB immune protection, we went on to assess histopathological changes in the lungs of these animals. In agreement with their increased bacterial infection in the lung, unvaccinated Jh^−/− KO mice displayed much worsened granulomatous inflammation over that in the lung of unvaccinated WT mice (Fig. 4B, top two rows). Alternatively, consistent with their significantly reduced infection in the lung, vaccinated WT B cell–competent mice, vaccinated B cell–depleted WT mice, and vaccinated Jh^−/− KO mice all had reduced lung histopathology compared with their respective unvaccinated counterpart (Fig. 4B, bottom three rows). Taken together, the above data suggest that B cells are not required for effective protection mediated by respiratory mucosal AdHu5Ag85A vaccination.

Although respiratory mucosal vaccination significantly reduced bacterial infection in the lung of Jh^−/− KO mice, the extent of enhanced protection was less than that seen in vaccinated B cell–depleted WT mice (Fig. 4). These data suggest that the changes other than B cell deficiency in Jh^−/− KO animals have contributed to the altered protective immunity (33–35). To investigate this further, we undertook an adoptive B cell transfer approach. To this end, Jh^−/− KO mice were reconstituted i.v. with B cells purified from naive WT animals. As a control, a group of Jh^−/− KO mice were left without reconstitution. After 2 d after B cell transfer, mice were vaccinated i.n. with AdHu5Ag85A. Four weeks postvaccination, mice were challenged i.n. with M. tuberculosis H₃₇Rv and bacterial burden was accessed in the lungs at 2 wk postchallenge (Fig. 5A). We found that compared with the Jh^−/− KO control group without B cell reconstitution, B cell reconstitution in Jh^−/−KO mice did not enhance protection against pulmonary M. tuberculosis infection (Fig. 5B). These data suggest that the suboptimal protection against pulmonary M. tuberculosis infection mediated by respiratory mucosal AdHu5Ag85A vaccination in Jh^−/− KO mice may not be due to the lack of B cells.

The above data have together suggested that B cells play a minimal role in respiratory mucosal vaccine–induced, T cell–mediated protection and that the mucosal vaccine–induced suboptimal protection seen in genetic B cell–deficient Jh^−/− KO animals is not due to B cell deficiency per se. To further investigate, we next employed an approach involving SCID mice and adoptive T and B cell transfer. The T and B cells would be purified from respiratory mucosal vaccine–primed animals. This approach allowed the exclusive assessment of the functionality of the adoptively transferred, vaccine-primed T and B cells in a naive host in the absence of endogenous lymphocytes. To this end, both WT and Jh^−/− KO mice were vaccinated i.n. for 4 wk and the vaccine-primed T and B cells were then purified from these animals. As a control, the naive T and B cells were also purified from unvaccinated WT mice. Purified T cells alone or both purified T and B cells were adoptively transferred to groups of naive SCID mice. The SCID mice were then subsequently challenged with M. tuberculosis H37Rv and the levels of bacterial burden in the lungs were assessed at 2 wk postchallenge (Fig. 6A). As expected, SCID mice that received the naive T and B cells from unvaccinated WT mice had high levels of M. tuberculosis infection in the lung whereas those receiving the vaccine-primed T cells from vaccinated WT mice had significantly reduced M. tuberculosis infection in the lung (Fig. 6B). Importantly, adoptive transfer of vaccine-primed B cells in conjunction with transfer of vaccine-primed T cells to SCID mice did not further improve protection or reduce infection (Fig. 6B). These data further support our conclusion that B cells do not contribute significantly to respiratory mucosal vaccine–induced, T cell–mediated protection. Alternatively, further compared with the SCID mice receiving the vaccine-primed T cells from vaccinated WT mice, the mice receiving the vaccine-primed T cells isolated from vaccinated Jh^−/− KO mice were not as well protected, having significanly higher levels of M. tuberculosis infection in the lung (Fig. 6B). This finding, in conjunction with the data presented in Fig. 5, suggests that the suboptimal protection against pulmonary M. tuberculosis infection in mucosal-vaccinated Jh^−/− KO mice is due to suboptimal T cell immunity resulting from other immune alterations inherent to these mice than the lack of B cells.

It has been well established that T cell–mediated immunity plays an important role in immune protection against intracellular pathogens, including M. tuberculosis (24, 39). However, recent revisits to the role of B cells and Ab responses in protection against primary pulmonary M. tuberculosis infection have revealed a potential role for B cells in anti-TB host defense (4, 7, 8, 10, 12, 15, 16), although there is a difference in opinion in the field (4, 40). Despite such progress, the role of TB vaccine–activated B cells in T cell activation and mucosal protective immunity still remains poorly understood. In this study, by using a model viral-vectored respiratory mucosal TB vaccine (AdHu5Ag85A) and a number of experimental approaches we have addressed this question. Our findings indicate that B cells are not required for respiratory mucosal vaccine–induced protective immunity against pulmonary TB and highlight a central role for T cells in this process.

B cells are not only Ab-producing immune cells but may also be involved in Ag presentation and other immune effector activities (9, 19). Despite the presence of conflicting data, there has been increasing evidence to suggest a role of B cells, including their production of anti–M. tuberculosis Abs and cytokines such as IL-10 in modulating the responses of T cells and other innate immune cells in the course of primary pulmonary M. tuberculosis infection (4, 7, 8, 10, 12, 15, 16). In the present study, we find that respiratory mucosal AdHu5Ag85A vaccination activates B cells to produce Ag85A-specific Abs both in the airways and circulation. However, our overall data argue against a significant role of B cells in mucosal vaccine–induced T cell responses and protective immunity. This conclusion is supported by the evidence generated by using three major approaches including the use of genetic B cell–deficient Jh^−/− KO mice, anti-CD20–mediated B cell depletion in WT mice, and adoptive T and B cell transfer to SCID and Jh^−/− KO mice. Taking our present findings and the previous studies by others into consideration, it is likely that B cells may play a more prominent role in the regulation of protective T cell immunity in the course of primary M. tuberculosis infection where T cell priming relies on Ag presentation by M. tuberculosis–infected or M. tuberculosis Ag–bearing APCs. In the case of mucosal TB vaccination, the relative involvement of B cells in T cell immunity perhaps depends on the nature of vaccine backbone. The viral backbone as used in the present study represents a robust T cell activator that may function independent of B cell functionality (2). Furthermore, because the Ag85A expressed by AdHu5Ag85A vaccine is a single secreted M. tuberculosis protein, Abs against this Ag have only a limited role in protection. Indeed, previous studies have demonstrated a protective role only for the type of Abs that are directed against multiple surface Ags of M. tuberculosis (11, 17, 19). Because we have found little evidence in supporting a major role of B cells and Abs in protective T cell immunity, we did not elect to use passive serum transfer approach in the present study. In this regard, a previous study has shown serum transfer to be inconsequential to TB protection in B cell KO animals (7). Our findings together help clarify the contentious role of B cells in the field of TB research in general and suggest B cells to be differentially required in the immunologic process following primary M. tuberculosis infection and in mucosal vaccine–induced immune activation and protection. Specifically, our evidence indicates that different from the setting of host defense against primary M. tuberculosis infection, B cells are minimally required for respiratory mucosal TB vaccine–induced T cell immunity.

The importance of airway luminal T cells in providing protection against pulmonary M. tuberculosis infection has been well established by our laboratory (24). In the present study, although we have shown that respiratory mucosal TB vaccination can still go on to activate Ag-specific T cells both in the airway and lung interstitium of the genetic B cell–deficient animals (Jh^−/− KO), the numbers of airway luminal T cells are lower than in WT control animals, which appears to be associated with suboptimal protection seen in Jh^−/− KO mice. However, by using a B cell depletion approach, we have observed comparable levels of vaccine-activated CD8⁺ T cell responses both in the airway and lung interstitium associated with uncompromised protection. In our previous studies, we have repeatedly shown that CD8⁺ T cells are the primary immune cells involved in providing protection in our mucosal AdHu5Ag85A vaccine model (27). We have also provided evidence that CD4⁺ T cells are not required for the generation of immune protective CD8⁺ T cell responses following respiratory mucosal AdHu5Ag85A vaccination (30). This latter observation prompted us to investigate further the possibility that the suboptimal protective immunity in Jh^−/− KO mice was not a result of lacking B cell functionality per se but rather was due to suboptimal CD8⁺ T cell activation resulting from other immunologic changes inherent to this type of transgenic mice, including the abnormal structure of T cell zones and altered dendritic cell populations in the lymphoid tissues (33–35). In this regard, indeed we have observed significantly reduced Ag85A-specific T cell priming in the lung-draining lymph node following respiratory mucosal vaccination (data not shown). Additionally, because we have observed suboptimal CD4⁺ T cell responses in the Jh^−/− KO mice, we cannot rule out that this may also have contributed to suboptimal CD8⁺ T cell activation. These considerations provide a plausible explanation for our findings that reconstitution of B cells in vaccinated Jh^−/− KO mice was unable to improve immune protection on the one hand and that the Jh^−/− KO mouse–derived T cells transferred to SCID hosts were not as immune protective as those isolated from WT mice on the other hand. Collectively, the findings from our present investigation cautions against the sole use of genetic B cell–deficient mice and supports the relevance of including other complementary approaches, including Ab-mediated B cell depletion and adoptive B cell transfer in investigating the role of B cells in host defense against TB.

We thank Dr. John Chan for advice, Genentech, Inc. for providing the anti-CD20 Abs, and the National Institutes of Health Tetramer Core for providing Ag85A MHC class I tetramers.

The authors have no financial conflicts of interest.