Although most novel tuberculosis (TB) vaccines are designed for delivery via the muscle or skin for enhanced protection in the lung, it has remained poorly understood whether systemic vaccine-induced memory T cells can readily home to the lung mucosa prior to and shortly after pathogen exposure. We have investigated this issue by using a model of parenteral TB immunization and intravascular immunostaining. We find that systemically induced memory T cells are restricted to the blood vessels in the lung, unable to populate either the lung parenchymal tissue or the airway under homeostatic conditions. We further find that after pulmonary TB infection, it still takes many days before such T cells can enter the lung parenchymal tissue and airway. We have identified the acquisition of CXCR3 expression by circulating T cells to be critical for their entry to these lung mucosal compartments. Our findings offer new insights into mucosal T cell biology and have important implications in vaccine strategies against pulmonary TB and other intracellular infections in the lung.

The lung is an intricate, highly vascularized mucosal organ consisting of two major tissue compartments, the conducting/terminal airways lined with the mucosal tissue and the lung parenchyma (1). Different from other mucosal sites of the body, the respiratory mucosa is in constant contact with the environment for the vital role of the lung in gas exchange critical for the survival of the host. Thus, immune homeostasis in the lung is important for its efficient gas exchange function (1, 2).

Immune regulatory mechanisms are in place to prevent excess damage to the delicate architecture of the lung following the immune responses to infection, vaccination, or airborne allergens. To this end, similar to the other tissue sites, T cell entry to the lung is tightly regulated through chemokine receptor–ligand and integrin–adhesion molecule interactions (35). It has been well established that circulating T cells populate the lung parenchyma and the airway to become mucosal tissue-resident memory T cells (Trm) following the respiratory route of immunologic exposure (4, 6, 7). On the contrary, the T cells activated by the systemic (parenteral) route of immunologic exposure can only populate the lung parenchyma, but not the airway (3, 6, 8). This latter conviction has been based on abundant experimental observations using various approaches, including adoptive T cell transfer and parabiosis (914). Thus, along with the peripheral lymphoid tissues such as the spleen, lymph nodes, and bone marrow, the lung parenchyma was categorized as a common T cell entry site whereas the airway, similar to other mucosal tissue sites, is a restricted site for T cell entry (3, 6). This paradigm contradicts the concept that the lung should remain inflammation-free whenever possible under homeostatic conditions to effectively carry out the gas exchange function. In fact, the intravascular staining protocol has revealed that many cells previously thought to reside within the lung tissue (T cells isolated from the lung homogenates) are not actually in the lung parenchyma, but they reside instead within the lung vasculature (LV). Thus, with this in mind, conclusions from prior studies that assumed all T cells present in lung homogenates actually came from lung tissue need to be reassessed. Of note, hereafter, the term “lung homogenates” is used when describing observations made in total lung without separating it into lung parenchymal tissue (LPT) and LV. Although when using the intravascular staining technique, a large proportion of T cells isolated from the lung homogenates have been shown to confine to LV with limited entry to LPT following systemic immunological exposure (4, 7), the timing and the immune regulatory mechanisms that govern systemic immunization-activated T cell homing to both the LPT and the airway following secondary respiratory infection are still poorly understood.

A clear understanding of the common versus restricted T cell entry sites within the lung and the underlying mechanisms is critical to our knowledge in memory T cell homing and immune protection (15). It is particularly relevant to understanding the mechanisms that govern the homing of parenteral vaccine-activated T cells to LPT and the airway to protect against respiratory infections (16, 17). We and others have consistently shown that respiratory mucosal immunization renders much enhanced protection against respiratory infection or cancer, whereas parenteral immunization fails to do so, paradoxically despite the presence of vaccine-induced T cells in the lung parenchyma (11, 1722). Adoptive airway T cell transfer studies revealed no immune protective defect in such systemically activated T cells but suggested their failure to home to the airway to be the key (11). Thus, for a long time, it has been thought that it is the airway luminal T cells, but not the lung parenchymal counterparts, that are critical to immune protection in the lung (23, 24). Given the practical and wide application of systemic (parenteral) immunization route in human vaccination program, it is imperative to definitively delineate the immunologic underpinnings of ineffective protection against respiratory intracellular infections by parenteral immunization. Such knowledge will aid the development of improved vaccine strategies for human applications (2528).

In this study, using a model of viral vector–based parenteral tuberculosis (TB) immunization and intravascular immunostaining technology, we have investigated lung mucosal homing properties of systemically induced memory T cells and underlying molecular mechanisms before and after pulmonary Mycobacterium tuberculosis infection. By examining T cells in the three lung compartments, that is, LV, LPT, and airway, our study reveals that the LPT, similar to the airway, is also a restricted entry site for systemically induced circulating memory T cells. In contrast to respiratory mucosal-induced counterparts, parenteral immunization-induced T cells remain in the LV, unable to enter either the LPT or the airway under homeostatic conditions. Upon M. tuberculosis infection, it takes many days before Ag-specific T cells can enter these two restricted lung mucosal tissue sites. Acquisition of CXCR3 expression on such T cells in the draining lymph nodes plays a key role in switch-on of the entry of circulating T cells to the restricted lung mucosal sites in systemically immunized hosts. Our study thus shows that lung mucosal-resident memory T cells are not generated following systemic TB immunization and that local inflammation is required for systemically activated T cells to home to lung mucosa, which is mediated by interaction between CXCR3 upregulated in these cells and its ligands IP-10 and MIG. Our findings provide new knowledge in mucosal T cell biology and have important implications in developing effective vaccine strategies against pulmonary TB and other intracellular infections in the lung.

Female BALB/c mice (6–8 wk old) were purchased from Charles River Laboratories (Saint Constant, QC, Canada). CByJ.SJL(B6)-Ptprca/J mice on a BALB/cByJ background (CD45.1) and the CBy.PL(B6)-Thy1a/ScrJ BALB/c congenic strain carrying T lymphocyte–specific Thy1a (Thy1.1) were purchased from The Jackson Laboratory (Bar Harbor, ME). Mice were housed in a specific pathogen-free level B facility at McMaster University. All experiments were carried out in accordance with the guidelines from the Animal Research and Ethics Board at McMaster University.

The construction of a recombinant replication-deficient human type 5 adenovirus expressing an immunodominant M. tuberculosis Ag85A (AdHu5Ag85A) was previously described (21, 28, 29). Vaccine was prepared at 5 × 107 PFU per mouse in PBS. For the respiratory route of immunization, 25 μl of vaccine preparation was instilled intranasally (i.n.). For the parenteral route of immunization, 100 μl of vaccine preparation was injected into the quadriceps muscle of both legs (50 μl per leg).

M. tuberculosis bacilli were grown in Middlebrook 7H9 broth supplemented with Middlebrook oleic acid/albumin/dextrose/catalase enrichment, 0.002% glycerol, and 0.05% Tween 80 for 10–15 d, aliquoted, and stored at −70°C until use (30, 31). Before each in vivo use, M. tuberculosis bacilli were washed with PBS containing 0.05% Tween 80 twice and passed through a 27-gauge needle 10 times to disperse clumps. Pulmonary TB was elicited by intratracheal inoculation of M. tuberculosis bacilli.

FTY720 (Cayman Chemical, Ann Arbor, MI) was prepared in 25% ethanol to inject i.p. in 200 μl at 4 mg/kg body weight to vaccinated mice at selected time points to prevent systemic supply of T cells to lung as well as egress of T cells from lung parenchyma. A single FTY720 injection at the described dose induced deficiency of circulating T cells for up to 5 d as previous reported (32).

Intravascular immunostaining was carried out as previously described (33). Monoclonal anti-mouse CD45.2-PerCP-Cy5.5 (clone) was prepared in 250 μl of PBS at the concentration of 12 μg/ml and injected i.v. via the tail vein. Within 3 min after injection, animals were sacrificed and bled. Blood was collected in heparin and kept at room temperature in the dark. The lung was removed with trachea attached and lavaged to obtain airway luminal cells. The lungs were then processed for total mononuclear cell isolation. All mentioned procedures were completed within 5 min for each animal to prevent the escape of leukocytes from circulation to lung parenchyma. The gating strategy used for discriminating tissue and vasculature cells as follows. Live cells were first gated for singlets and then for lymphocytes followed by CD3+ gating. CD3+ cells were then gated for CD8 and CD45.2. CD8+CD45.2+ cells were defined as the cells in pulmonary vasculature, and CD8+CD45.2 cells were defined as the cells in lung tissue.

Mice were euthanized by exsanguination. Airway luminal cells were collected by exhaustive bronchoalveolar lavage (BAL) as previously described (11, 31). In some experiments conventional BAL samples were saved for chemokine measurements. Subsequently, lungs were cut into small pieces and digested with collagenase type 1 (Sigma-Aldrich, St. Louis, MO) at 37°C in an agitating incubator. A single-cell suspension was obtained by crushing the digested tissue through a 40-μm basket filter. Lymph nodes were teased with a pair of needles and digested in collagenase type IV at 37°C in an agitating incubator. A single-cell suspension was obtained after treating the digested material with EDTA and passing through a 40-μm basket filter. Heparinized blood was mixed with 10× volume of ACK lysing buffer (Life Technologies, Grand Island, NY) and incubated at room temperature for 5 min before removing the lysed RBCs upon centrifugation. A second round of RBC lysis was carried out by resuspending the pellet in 2 ml of ACK lysing buffer. Blood leukocytes were then washed in cold PBS (2–8°C). All isolated cells were resuspended in complete RPMI 1640 medium (RPMI 1640 supplemented with 10% FBS, 1% penicillin/streptomycin, and 1% l-glutamine).

Mononuclear cells from BAL, lungs, lymph nodes, and blood were plated in U-bottom 96-well plates at a concentration of 20 million cells/ml. For tetramer immunostaining, a tetramer for the immunodominant CD8 T cell peptide (MPVGGQSSF) of Ag85A bound to the BALB/c MHC class I allele H-2Ld (National Institutes of Health Tetramer Core) was used. Cells were stained with the tetramer for 1 h in the dark at room temperature and washed and stained with surface Abs. In some experiments cells were stained for memory surface markers CD62L, CD127, and CD103 or for chemokine receptors CXCR3, CXCR6, CCR5, and CD11a. The mAbs used include CD3-V450, CD8-PE-Cy7, tetramer-PE, CD62L-PerCP-Cy5.5, CD127-FITC, and CD103-Qdot. For intracellular cytokine staining cells were cultured in the presence of GolgiPlug (5 mg/ml brefeldin A; BD Pharmingen) with or without stimulation for 5–6 h with an Ag85A-specific CD8 T cell–specific peptide (MPVGGQSSF) as previously described (11, 31, 32) at a concentration of 1 μg per well. In some experiments lung mononuclear cells were stimulated with crude bacillus Calmette–Guérin and M. tuberculosis culture filtrate Ags at a concentration of 1 μg per well of each stimuli (34). The cytotoxicity of CD8 T cells was evaluated by adding CD107a-FITC during the period of Ag stimulation (32). After incubation, cells were washed and blocked with CD16/CD32 in 0.5% BSA-PBS for 15 min on ice and then stained with the appropriate fluorochrome-labeled mAbs. The mAbs used included CD8a-PE-Cy7, CD4-allophycocyanin-Cy7, IFN-γ–allophycocyanin, and CD3-V450. All mAbs were bought from BD Biosciences. Immunostained cells were processed according to the BD Biosciences instructions for flow cytometry and run on an LSR II flow cytometer. Data were analyzed using FlowJo software (version 9; Tree Star, Ashland, OR).

CD8 T cells were purified from single-cell suspension of lung mononuclear cells. The EasySep mouse CD8+ T cell isolation kit (Stemcell Technologies, Vancouver, BC, Canada) was used as per the manufacturer’s instruction. Purified cells were resuspended in PBS for adoptive transfer.

Chemokine protein contents in BAL fluids were quantified using Luminex multianalyte technology (Luminex Molecular Diagnostics, Toronto, ON, Canada) according to the manufacturer’s protocols.

To block CXCR3, mice were treated with anti-CXCR3 blocking mAb (clone CXCR3-173; Bio X Cell) or isotype control at the concentration of 500 μg in total volume 200 μl i.p. per mouse at designated times. Blocking capability of this Ab is well documented in previously published studies (35, 36). To neutralize IP-10 or both IP-10 and MIG, polyclonal rabbit anti-murine IP-10 or MIG serum (200 μl per injection) or normal rabbit serum as control (200 μl per injection) was injected i.p into immunized mice at designated times following mycobacterial infection. Polyclonal rabbit anti-murine IP-10 or MIG serum and normal rabbit serum were provided by Dr. S.L. Kunkel (37).

RNA from purified CD8 T cells was isolated using an RNeasy mini kit (Qiagen, Germantown, MD). Disrupted cells in RLT buffer were processed according to the manufacturer’s instructions. RNA was eluted using 30–50 μl of RNase-free water. Samples were stored at −70°C until use. Quality of RNA and subsequent microarray was carried out by the Center for Applied Genomics of The Hospital for Sick Children (Toronto, ON, Canada). The quality of RNA samples was analyzed using the Agilent 2100 bioanalyzer, which uses RNA 6000 Nano LabChip platform (Agilent Technologies Canada, Mississauga, ON, Canada). On average, the RNA integrity number for all samples included in the study was >7. RNA samples were then subjected to RNA microarray expression analysis using the Affymetrix Mouse Gene 2.0 ST array. This array contains protein-coding regions.

Genome-wide gene expression analysis was performed using an Affymetrix Mouse Gene 2.0 ST microarray (Microarray Facility, Center for Applied Genomics of The Hospital for Sick Children, Toronto, ON, Canada). For each time point, RNA samples obtained from three independent experiments were analyzed independently to generate triplicate sets of data per time point. The data discussed in this study have been deposited in the National Center for Biotechnology Information’s Gene Expression Omnibus (38) and are accessible through accession number GSE100288 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE100288). Pairwise comparisons were performed for each time point to determine statistical differences (p ≤ 0.05) in fold changes. Genes up- or downregulated >2-fold were used in further analyses. The lists of differentially expressed genes in pairwise comparisons were generated using Gene Ontology. The Panther DB Web site was used to determine whether the changes in the genes mapped to a particular biological process (39). Of the genes categorized according to biological processes, genes mapped to immune system process and biological adhesion were subjected to further analysis to group the genes that had similar relative changes. Some genes that were not categorized by Panther but known to have a role in immune system processes were also included in the relative change analysis.

A p value <0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001). A two-tailed Student t test was performed for pairwise comparisons. One-way ANOVA followed by a Tukey post hoc test was performed to compare more than two groups. All analyses were performed by using GraphPad Prism software (GraphPad Software, La Jolla, CA).

For microarray data analysis, all statistical analyses were performed in R version 3.2.3 and various Bioconductor packages (40). Principal component analysis was performed using the rgl package (https://cran.r-project.org/web/packages/rgl/index.html); hierarchical clustering was performed using the gplots package (https://cran.r-project.org/web/packages/gplots/index.html); differential expression analysis, including fold change calculations and Venn diagrams, was performed using the limma package (41) with p values corrected by using the Benjamini–Hochberg procedure (42).

We have previously reported that similar to respiratory mucosal (i.n.) TB immunization, parenteral (i.m.) TB immunization induces T cells detectable in lung homogenates, but different from the i.n. route, i.m. immunization fails to induce lung airway T cells (11, 21). Lack of airway T cells by i.m. immunization is associated with the lack of protection. These observations at the time were in line with the immunologic dogma that the lung parenchyma was a common T cell entry site whereas the airway is a restricted T cell entry site (3, 6). However, immunologically it has remained difficult to reconcile the presence of T cells in lung homogenates with almost a complete lack of immune protection. To begin solving this enigma, we wanted to better understand the T cell responses to secondary infection in i.m. immunized hosts by comparing them with those i.n. immunized hosts.

Using AdHu5Ag85A (21, 29), we first examined how i.m. induced T cells and i.n. induced T cells respond to respiratory M. tuberculosis infection and their relationship to airway T cells. To this end, mice were immunized either i.m. or i.n. with AdHu5Ag85A and in the memory phase of T cell responses (4 wk) were challenged with (AdHu5Ag85A/M. tuberculosis) or without (AdHu5Ag85A) M. tuberculosis. Ag-specific CD8+tetramer+ (Ag85A tet+) T cell responses were examined in the lung homogenate–derived mononuclear cells and BAL fluids at 4 d after pulmonary M. tuberculosis infection (Fig. 1A). As expected, both i.m. and i.n. immunization induced T cells present in lung homogenates (Fig. 1B), but such lung homogenate T cells increased further significantly upon M. tuberculosis infection only in i.m., not in i.n., immunized hosts (Fig. 1B). Alternatively, i.m. immunization led to no detectable airway T cells before and after M. tuberculosis challenge (Fig. 1C). In contrast, i.n. immunization led to the significant presence and increases of airway T cells before and after M. tuberculosis infection, respectively (Fig. 1C). Taken together, these observations suggest that i.m. and i.n. TB vaccine–induced lung homogenate T cells respond to pulmonary M. tuberculosis infection differently and that increased lung homogenate T cells in i.m. vaccinated hosts in response to infection are associated with lack of airway T cells, whereas unaltered lung homogenate T cells in i.n. vaccinated hosts are associated with increased airway T cells.

FIGURE 1.

Failure of parenteral immunization-induced memory T cells to move into the airway upon respiratory pathogen exposure. (A) Experimental schema. Numbers of Ag-specific (tet+) CD8+tet+ cells in the lung homogenates (B) and airway (C) of i.m. and i.n. immunized animals before and after M. tuberculosis challenge. (D) Experimental schema showing FTY720 injection introduced to the experimental protocol described in (A) but only with i.n. mucosal AdHu5Ag85A immunization. Numbers of Ag-specific CD8+tet+ cells in the lung homogenates (E) and airway (F) are shown. Data are expressed as the mean ± SEM of T cell numbers of three to four mice per group, representative of two independent experiments. *p < 0.05, compared with the paired control.

FIGURE 1.

Failure of parenteral immunization-induced memory T cells to move into the airway upon respiratory pathogen exposure. (A) Experimental schema. Numbers of Ag-specific (tet+) CD8+tet+ cells in the lung homogenates (B) and airway (C) of i.m. and i.n. immunized animals before and after M. tuberculosis challenge. (D) Experimental schema showing FTY720 injection introduced to the experimental protocol described in (A) but only with i.n. mucosal AdHu5Ag85A immunization. Numbers of Ag-specific CD8+tet+ cells in the lung homogenates (E) and airway (F) are shown. Data are expressed as the mean ± SEM of T cell numbers of three to four mice per group, representative of two independent experiments. *p < 0.05, compared with the paired control.

Close modal

The above data suggest a temporal relationship between lung homogenate T cells and airway T cells. To investigate this further, we used FTY720 1 d before M. tuberculosis infection in i.n. immunized hosts (Fig. 1D) to cut off the systemic T cell supply and its transmigration between the lung parenchyma and the airway, as previously described (32). Indeed, when T cell transmigration between the lung parenchyma and the airway was blocked, the number of airway T cells declined sharply upon M. tuberculosis infection (AdHu5Ag85A/M. tuberculosis/FTY), which was accompanied by increased T cells detected in lung homogenates (Fig. 1E, 1F). These data suggest that increased airway T cells in i.n. immunized animals upon M. tuberculosis infection are due to the transmigrated T cells from lung parenchyma.

Collectively, the above data suggest that the T cells present in lung homogenates induced by parenteral TB immunization behave differently from those induced by respiratory mucosal TB immunization in that they fail to move into the airway upon infection, raising the question of whether they represented authentic lung parenchymal T cells.

To further interrogate the authenticity of parenteral immunization-induced T cells present in lung homogenates, we examined the immunologic properties associated with mucosal Trm in 4 wk i.m. and i.n. immunized animals. Although similar to i.n. immunization-induced lung homogenate T cells, i.m. immunization-induced lung homogenate T cells produced IFN-γ and underwent degranulation (CD107+) following ex vivo Ag stimulation, and the frequencies of these cells were significantly lower (Fig. 2A). Of importance, the Ag-specific (tet+) lung homogenate T cells by i.m. immunization lacked the expression of CD103, an integrin molecular marker for Trm (4, 6, 7), in contrast with high CD103 expression on the lung homogenate T cells induced by i.n. immunization (Fig. 2B). Furthermore, the lung homogenate T cells by i.m. immunization displayed a significant resting memory (CD127+/CD62L) or a central memory (CD127+/CD62+) phenotype, whereas, as expected, most (90%) of the Trm by i.n. immunization were of an effector memory phenotype (CD127/CD62) (30) (Fig. 2C). The lung homogenate T cells by i.m. immunization also declined sharply by day 90 postimmunization, contrasting to the lasting Trm population induced by i.n. immunization (Fig. 2D). These data indicate that parenteral TB immunization–induced T cells present in lung homogenates are not bona fide mucosal Trm.

FIGURE 2.

Lack of mucosal Trm property by parenteral immunization-induced T cells present in lung homogenates. (A) Representative dot plots showing frequencies of Ag-specific CD8+IFN-γ+ and CD8+CD107+ T cells out of a total of CD8 T cells in the Ag85A ex vivo–stimulated lung mononuclear cells analyzed at 30 d after i.m. and i.n. AdHu5Ag85A immunization. (B) Representative dot plots depicting frequencies of CD103-expressing CD8+tet+ T cells gated on total CD8+ T cells of lung mononuclear cells without Ag stimulation. The bar graph shows the numbers of CD103+CD8+tet+ T cells in the lung parenchyma. (C) Representative dot plots depicting memory subsets of Ag-specific CD8+tet+ T cells in lung mononuclear cells without Ag stimulation. Bar graph compares the relative frequencies of memory T cell subsets between i.m. and i.n. immunizations. (D) Kinetic changes in Ag-specific CD8+tet+ T cells in the lung homogenates following i.m. and i.n. immunization analyzed at 30 or 90 d postimmunization. Bar graph shows the numbers of Ag-specific CD8+tet+ T cells in the lung homogenates. The graphic data are expressed as the mean ± SEM of T cell frequencies or absolute numbers of three to four mice per group, representative of two to three independent experiments. *p < 0.05 compared with i.n. route of immunization.

FIGURE 2.

Lack of mucosal Trm property by parenteral immunization-induced T cells present in lung homogenates. (A) Representative dot plots showing frequencies of Ag-specific CD8+IFN-γ+ and CD8+CD107+ T cells out of a total of CD8 T cells in the Ag85A ex vivo–stimulated lung mononuclear cells analyzed at 30 d after i.m. and i.n. AdHu5Ag85A immunization. (B) Representative dot plots depicting frequencies of CD103-expressing CD8+tet+ T cells gated on total CD8+ T cells of lung mononuclear cells without Ag stimulation. The bar graph shows the numbers of CD103+CD8+tet+ T cells in the lung parenchyma. (C) Representative dot plots depicting memory subsets of Ag-specific CD8+tet+ T cells in lung mononuclear cells without Ag stimulation. Bar graph compares the relative frequencies of memory T cell subsets between i.m. and i.n. immunizations. (D) Kinetic changes in Ag-specific CD8+tet+ T cells in the lung homogenates following i.m. and i.n. immunization analyzed at 30 or 90 d postimmunization. Bar graph shows the numbers of Ag-specific CD8+tet+ T cells in the lung homogenates. The graphic data are expressed as the mean ± SEM of T cell frequencies or absolute numbers of three to four mice per group, representative of two to three independent experiments. *p < 0.05 compared with i.n. route of immunization.

Close modal

Collectively, our data thus far have revealed that parenterally induced memory T cells detectable in lung homogenates do not possess the immune properties of bona fide mucosal Trm and are unable to get mobilized into the airway lumen upon pulmonary M. tuberculosis infection. These observations raised the question about the geographical localization of these cells within the lung. Recent findings by others have suggested that the conventional techniques long and widely being used to isolate lung tissue mononuclear cells are ineffective in ridding the cells residing within the LV (43). In other words, very likely the T cells present in lung homogenates identified up to now by using such techniques are the cells located either within both the LPT and LV or within the LV (4, 7). From this point on, the T cells identified to be located truly within lung parenchyma by intravascular immunostaining are called LPT cells.

Intravascular immunostaining allows a reliable discrimination of LPT cells from those in the LV (33). To this end, animals were i.v. injected with an anti-mouse CD45.2 mAb conjugated to PerCP-Cy5.5 fluorochrome and subsequently sacrificed within 3 min (Fig. 3A). We validated this technique by showing that almost all lung parenchymal T cells in naive mice were restricted to the LV (Supplemental Fig. 1) and that >95% of peripheral blood T cells were labeled with anti-CD45.2 (CD45.2 PerCP-Cy5.5+), and the vast majority of T cells in the airway were negative in i.n. immunized mice (Fig. 3B). When these criteria were met, the i.v. mAb-labeled lung cells were identified to be the cells located within LV whereas the unlabeled ones were those in LPT (Fig. 3B). Thus, using this technique we have examined Ag-specific T cell distribution in LV and LPT at 4 wk following parenteral and respiratory mucosal routes of immunization (Fig. 3A). We found that most of the Ag-specific (tet+) T cells induced by i.m. immunization and detectable in lung homogenates were in fact located in LV (Fig. 3C), in keeping with much higher frequencies of Ag-specific T cells in the peripheral blood of i.m. immunized mice than in i.n. immunized mice (Fig. 3D). Alternatively, most of the Ag-specific T cells induced by i.n. immunization were located in the LPT (Fig. 3C). These data indicate that the Ag-specific T cells previously identified to be lung parenchymal T cells in the hosts parenterally immunized with TB vaccine are in fact primarily located in the LV and are not bona fide mucosal Trm.

FIGURE 3.

Parenteral immunization-induced T cells present in lung homogenates are restricted to LV. (A) Experimental schema outlining the procedure of intravascular immunostaining. (B) Representative dot plots showing the expected distribution of CD3+CD45.2+ or CD3+CD45.2 total T cell populations in peripheral blood (PB), airway, LPT, and LV of i.n. immunized animals for validation of intravascular immunostaining techniques. (C) Representative dot plots comparing the relative distribution (frequencies) of Ag-specific CD8+tet+ T cells in LPT and LV of i.m. and i.n. immunized animals. Bar graph shows the numbers of Ag-specific CD8+tet+ T cells in LPT and LV. (D) Representative dot plots showing frequencies of circulating Ag-specific CD8+tet+ T cells gated on CD3+ cells in the peripheral blood of i.m. and i.n. immunized animals. The graphic data are expressed as the mean ± SEM of T cell frequencies or numbers of three to four mice per group, representative of two to three independent experiments.

FIGURE 3.

Parenteral immunization-induced T cells present in lung homogenates are restricted to LV. (A) Experimental schema outlining the procedure of intravascular immunostaining. (B) Representative dot plots showing the expected distribution of CD3+CD45.2+ or CD3+CD45.2 total T cell populations in peripheral blood (PB), airway, LPT, and LV of i.n. immunized animals for validation of intravascular immunostaining techniques. (C) Representative dot plots comparing the relative distribution (frequencies) of Ag-specific CD8+tet+ T cells in LPT and LV of i.m. and i.n. immunized animals. Bar graph shows the numbers of Ag-specific CD8+tet+ T cells in LPT and LV. (D) Representative dot plots showing frequencies of circulating Ag-specific CD8+tet+ T cells gated on CD3+ cells in the peripheral blood of i.m. and i.n. immunized animals. The graphic data are expressed as the mean ± SEM of T cell frequencies or numbers of three to four mice per group, representative of two to three independent experiments.

Close modal

We have thus far shown that parenterally induced Ag-specific memory T cells are confined to LV (Fig. 3) and are unable to home to the airway shortly (4 d) after pulmonary M. tuberculosis infection, which is in sharp contrast to the rapid increase in Ag-specific T cell responses in i.n. immunized mice (Fig. 1B). The question remained then whether such systemically induced T cells in circulation could eventually home to the lung mucosal tissue following pulmonary M. tuberculosis infection. To this end, mice were immunized i.m., infected with M. tuberculosis at 4 wk and sacrificed at days 0, 7, 14, and 21 postinfection (Fig. 4A). Before sacrifice, intravascular immunostaining was performed and immune responses in the three lung compartments were analyzed as depicted in Fig. 3A. Although total CD8 T cells increased in LV at 7 d postinfection, the increases in LPT and the airway were minimal (Fig. 4B). Significant increases were seen at 14 d. Further examination of Ag-specific CD8 T cells (tet+) revealed that markedly increased Ag-specific LPT cells were not seen until 14 d postinfection (Fig. 4C). Accompanied with increased LPT Ag-specific T cells were markedly increased airway Ag-specific T cells at day 14 (Fig. 4C). In sharp contrast, Ag-specific CD8 T cells significantly further increased in the LPT of i.n. immunized animals by 7 d postinfection (Fig. 4D). These data indicate that not only the parenteral TB immunization–induced T cells are largely confined to LV, but there is also a significant delay for them to move into LPT and airway lumen in response to pulmonary M. tuberculosis infection.

FIGURE 4.

Much delayed movement of parenteral immunization-induced memory T cells from LV into both the LPT and airway after respiratory pathogen exposure. (A) Experimental schema. Representative dot plots show time-dependent changes in relative distribution (frequencies) of total CD3+CD8+ T cells (B) and (frequencies) of Ag-specific CD8+tet+ T cells (C) in LPT and LV compartments at days 0, 7, 14, and 21 after M. tuberculosis challenge. Bar graphs show the numbers of total CD3+CD8+ T cells (B) and Ag-specific CD8+tet+ T cells (C) in airway, LPT, and LV at days 0, 7, and 14 after M. tuberculosis challenge. (D) Bar graph shows the number of total Ag-specific CD8+tet+ T cells in airway, LPT, and LV on i.n. immunized mice before (day 0) and at 7 d after M. tuberculosis infection. The graphic data are expressed as the mean ± SEM of T cell frequencies or numbers of three to four mice per group, representative of two independent experiments. ***p < 0.001.

FIGURE 4.

Much delayed movement of parenteral immunization-induced memory T cells from LV into both the LPT and airway after respiratory pathogen exposure. (A) Experimental schema. Representative dot plots show time-dependent changes in relative distribution (frequencies) of total CD3+CD8+ T cells (B) and (frequencies) of Ag-specific CD8+tet+ T cells (C) in LPT and LV compartments at days 0, 7, 14, and 21 after M. tuberculosis challenge. Bar graphs show the numbers of total CD3+CD8+ T cells (B) and Ag-specific CD8+tet+ T cells (C) in airway, LPT, and LV at days 0, 7, and 14 after M. tuberculosis challenge. (D) Bar graph shows the number of total Ag-specific CD8+tet+ T cells in airway, LPT, and LV on i.n. immunized mice before (day 0) and at 7 d after M. tuberculosis infection. The graphic data are expressed as the mean ± SEM of T cell frequencies or numbers of three to four mice per group, representative of two independent experiments. ***p < 0.001.

Close modal

To understand whether the parenteral TB vaccine–induced Ag-specific resting memory T cells underwent phenotypic changes at some point after M. tuberculosis infection, we further characterized and compared the central memory T cell (TCM), effector memory T cell (TEM), and effector T cell (TE) phenotypes of tet+ CD8 T cells in the mediastinal lymph node (MLN), LV, LPT, and airway (BAL) before (day 0) and after (day 7 and day 21) M. tuberculosis infection based on experimental design described in Fig. 4A. We find that consistent with the information presented in Fig. 4, before (day 0) and at day 7 after M. tuberculosis infection, there was a lack of any Ag-specific T cells in the lung (LPT and BAL) whereas they increased, particularly TCM, in MLN by day 7 after M. tuberculosis infection (Supplemental Fig. 2). Of note, the relative numbers of TCM, TEM, and TE in LV remained small and comparable in proportions between day 0 and day 7. In contrast, by day 21 after M. tuberculosis infection, numbers of Ag-specific CD8 T cells that homed to the lung (LPT and airway) were markedly increased and, of note, the relative magnitude of secondary TE, including both TEM and TE, became much higher than TCM in LV, LPT, and BAL. These data further suggest that following M. tuberculosis infection, parenteral TB vaccine–induced resting memory T cells remain largely confined to LV for quite some time, and they then undergo a phenotypic change, becoming the T cells of effector phenotypes, before and after homing to the lung (LPT and airway lumen).

The above model of parenteral TB immunization and subsequent respiratory M. tuberculosis challenge thus provides the opportunity to study the molecular pathways involved in the lack and switch-on of lung mucosal homing of systemically induced memory T cells. To begin understanding the mechanisms, we first compared the parenteral TB immunization–induced T cells with those by mucosal TB immunization in their intrinsic capabilities to home to the same-shared (inflamed) lung microenvironment by using a dual adoptive T cell transfer approach. To this end, the congenic CD8 T cells purified from the lung homogenates of i.m. (Thy1.2+CD45.2+) and i.n. (Thy1.2+CD45.1+) immunized mice were mixed in 1:1 ratio and i.v. transferred into the Thy1.1+ recipient mice. The mice were then infected with M. tuberculosis. At 6 d after pathogen exposure, some animals were sacrificed and T cells from the spleen and lung were immunostained ex vivo for Thy1.2, whereas other animals were injected i.v with allophycocyanin fluorochrome–labeled anti-Thy1.2 mAb to assess the relative portions of LV Thy1.2+ T cells (Fig. 5A). Ex vivo Thy1.2 staining revealed that a much smaller portion (1:3 ratio or 20.3/74.9%) of transferred i.m. T cells (Thy1.2+CD45.2+) was detectable in the lung homogenates compared with i.n. T cells (Thy1.2+CD45.1+) (Fig. 5B), accompanied by a reversed distribution pattern (3:1 ratio or 62.1/34.6%) in the spleen (Fig. 5C) and LV (68.4/27.3%) (Fig. 5D). Taken together, these data on the one hand suggest a lack of lung mucosal-homing property in parenteral TB immunization–induced T cells as a mechanism for their inability to home to the LPT at earlier stages of pulmonary M. tuberculosis infection. On the other hand, they suggest the acquisition of lung mucosal-homing property by these T cells at later stages of M. tuberculosis infection (around day 14).

FIGURE 5.

Poor lung mucosal-homing capability of parenteral immunization-induced memory T cells following adoptive transfer. (A) Experimental schema outlining the procedure of a dual adoptive T cell transfer approach whereby congenic CD8 T cells from the lungs of i.m. (Thy1.2+CD45.2+) and i.n. (Thy1.2+CD45.1+) immunized animals were mixed in 1:1 ratio (dot plot) and i.v. transferred into the Thy1.1+ recipient animals that were then infected via the respiratory route with M. tuberculosis for 6 d before analysis. (B and C) Representative dot plots and bar graphs displayed at bottom show the relative frequencies of adoptively transferred i.m. Thy1.2+CD45.2+ and i.n. Thy1.2+CD45.1+ T cells in lung homogenates and spleen, respectively, following ex vivo Thy1.2 immunostaining of mononuclear cells harvested from the lung and spleen. (D) Representative dot plots and bar graphs show the relative frequencies of adoptively transferred i.m. Thy1.2+CD45.2+ and i.n. Thy1.2+CD45.1+ T cells in LV analyzed following i.v. injection of Thy1.2-allophycocyanin mAb. The graphic data are expressed as the mean ± SEM of T cell frequencies of three to four mice per group, representative of two independent experiments. ***p < 0.001.

FIGURE 5.

Poor lung mucosal-homing capability of parenteral immunization-induced memory T cells following adoptive transfer. (A) Experimental schema outlining the procedure of a dual adoptive T cell transfer approach whereby congenic CD8 T cells from the lungs of i.m. (Thy1.2+CD45.2+) and i.n. (Thy1.2+CD45.1+) immunized animals were mixed in 1:1 ratio (dot plot) and i.v. transferred into the Thy1.1+ recipient animals that were then infected via the respiratory route with M. tuberculosis for 6 d before analysis. (B and C) Representative dot plots and bar graphs displayed at bottom show the relative frequencies of adoptively transferred i.m. Thy1.2+CD45.2+ and i.n. Thy1.2+CD45.1+ T cells in lung homogenates and spleen, respectively, following ex vivo Thy1.2 immunostaining of mononuclear cells harvested from the lung and spleen. (D) Representative dot plots and bar graphs show the relative frequencies of adoptively transferred i.m. Thy1.2+CD45.2+ and i.n. Thy1.2+CD45.1+ T cells in LV analyzed following i.v. injection of Thy1.2-allophycocyanin mAb. The graphic data are expressed as the mean ± SEM of T cell frequencies of three to four mice per group, representative of two independent experiments. ***p < 0.001.

Close modal

To identify the potential molecular determinants of lung mucosal homing of systemically induced T cells, we analyzed the difference in transcriptional signature of parenteral TB immunization–induced Ag-specific CD8 T cells collected when they were largely in LV and after they homed to lung mucosal tissue following pulmonary M. tuberculosis infection, by using the Affymetrix Mouse Gene 2.0ST microarray. We elected to focus on day 0, 7, and 21 time points postinfection to compare the T cells located in LV (day 0 and day 7, largely resting memory T cells) with those in lung mucosal tissue (day 21, largely secondary TE) (Fig. 6A). The gene expression profiles of these cells were analyzed in a pairwise comparison of day 0 versus day 7, day 0 versus day 21, and day 7 versus day 21 (Fig. 6A). A total of 106 genes were up- or downregulated by at least 2-fold in the cells from 7-d animals compared with those from uninfected controls (day 0 versus day 7). Alternatively, 1790 genes were up- or downregulated by at least 2-fold in the cells from 21 d–infected mice compared with cells from 7 d–infected mice (day 7 versus day 21) (Fig. 6B). Hierarchical clustering analysis and differential gene expression analysis revealed a similarity between day 0 and day 7 profile and a much greater dissimilarity between day 7 and day 21 or day 0 and day 21 (Fig. 6B, Supplemental Fig. 3A).

FIGURE 6.

Distinctive transcriptional gene expression profiles in parenteral immunization-induced T cell populations residing within LV and lung mucosal tissue. (A) Experimental schema outlining step-by-step procedure of transcriptional gene profiling by microarray. (B) Hierarchical clustering of the nine samples displaying gene expression profile of triplicates per time point. Clustering was performed by using complete linkage and Euclidean distance. Venn diagrams show the numbers of up- or downregulated genes across the three pairwise comparisons of different time points. (CE) Fold changes in chemokine receptor/ligand and integrin gene expression profiles by pairwise comparison between days 0 and 7 or days 7 and 21. Data presented were obtained from triplicate sets of lung cell–derived RNA samples per time point (8 mice per set; 24 mice per time point).

FIGURE 6.

Distinctive transcriptional gene expression profiles in parenteral immunization-induced T cell populations residing within LV and lung mucosal tissue. (A) Experimental schema outlining step-by-step procedure of transcriptional gene profiling by microarray. (B) Hierarchical clustering of the nine samples displaying gene expression profile of triplicates per time point. Clustering was performed by using complete linkage and Euclidean distance. Venn diagrams show the numbers of up- or downregulated genes across the three pairwise comparisons of different time points. (CE) Fold changes in chemokine receptor/ligand and integrin gene expression profiles by pairwise comparison between days 0 and 7 or days 7 and 21. Data presented were obtained from triplicate sets of lung cell–derived RNA samples per time point (8 mice per set; 24 mice per time point).

Close modal

To further understand the relative changes in the profile of the genes involved in immune responses, Panther analysis was performed. From the list generated from Panther analysis, genes that code for activation and cytokine production, chemokines, chemokine receptors, integrin, and metalloproteinase critical for T cell homing and T cell activation over time upon infection (3, 5) were further grouped according to the patterns of relative changes over time as shown in Supplemental Fig. 3B. The immune genes included in the analysis, the fold changes in pairwise comparison, and the groupings of genes are listed in Table I. Most of the chemokine receptor, chemokine ligand, and integrin gene expression levels continued to increase from day 0 to day 21. Among those genes, chemokine ligands, but not chemokine receptor genes, were altered on CD8 T cell samples of day 7 compared with day 0 (Fig. 6C). However, expression of the genes encoding chemokine receptors such as CXCR6, CCR5, and CXCR3 was significantly upregulated on the T cells residing in lung mucosal tissue (day 21) compared with those located in LV (Fig. 6D, Supplemental Fig. 3B, Table I, group C genes). Similarly, genes encoding for heterodimer integrin molecules of LFA-1 (CD11a or ITGAL), α4β7, and αEβ7 were also upregulated (Fig. 6E). Alternatively, the genes categorized in group D (Supplemental Fig. 3B, Table I) were downregulated. The above data indicate a striking difference in global expression of immune regulatory genes in T cell populations outside and inside of the lung mucosal tissue and, in particular, an upregulation of the cell transmigration genes in the T cells residing in the lung mucosal tissue.

Table I.
Differential expression of genes of interest found by Panther analysis
GenesDay 7 versus Day 0Day 21 versus Day 0Day 21 versus Day 7Group
Ccl9 1.82 2.68 1.47 
Itgae 1.29 6.75 5.23 
Cxcl11 2.13 3.75 1.76 
IL1B 2.06 4.67 2.27 
Cxcl9 2.93 8.02 2.73 
Cxcl13 1.82 2.37  
Cxcl3 3.01 6.48  
Fasl −1.35 1.93 2.61 a 
Cxcl5 2.57 1.64 −1.56 a 
IL2RA  1.92 2.05 
Itgb2  2.14 1.86 
TNF  2.25 1.8 
Timd4  2.17 2.03 
IFNZ  2.23 2.92 
Gzmb  2.27 2.45 
Itga4  2.29 2.06 
Itgb7  2.36 2.04 
Icos  2.45 2.68 
Pdcd1  2.46 2.49 
Gzmk  2.48 2.82 
Ccl4  2.53 2.74 
Itgal  2.57 2.11 
IRF4  2.6 2.79 
Ccr5  2.69 2.82 
Cxcr3  2.72 2.54 
Ccl17  2.81 2.12 
Nos2  3.04 2.6 
IFNG  3.08 3.55 
Cxcr6  3.24 3.97 
CD8A  3.43 3.31 
Xcr1  3.52 3.17 
Itgax  3.58 2.4 
Ccl22  3.78 3.07 
Xcl1  4.08 3.07 
IL12B  4.63 3.77 
Ctla4  5.36 5.54 
Ptgis  −1.92 −2.14 
TGFBR3  −1.78 −2.14 
Itgb5  −1.63 −2.04 
IL6  −4.14 −3.18 
S1pr3  −3.22 −3.53 
Ccl2  −2.84 −2.8 
Itga8  −2.76 −2.85 
Ccl11  −2.64 −2.17 
Socs2  −2.63 −2.63 
Dll1  −2.52 −2.37 
Sgpp2  −2.52 −2.59 
Pim3  −2.36 −2.38 
Itgbl1  −2.31 −2.03 
Cxcl14  −2.28 −2.56 
IL1R1  −2.26 −2.02 
Socs5  −2.18 −2.08 
IL33 −1.38 −2.73 −1.97 
Epcam −1.34 −2.78 −2.08 
IL5 −1.72 −3.8 −2.21 
GenesDay 7 versus Day 0Day 21 versus Day 0Day 21 versus Day 7Group
Ccl9 1.82 2.68 1.47 
Itgae 1.29 6.75 5.23 
Cxcl11 2.13 3.75 1.76 
IL1B 2.06 4.67 2.27 
Cxcl9 2.93 8.02 2.73 
Cxcl13 1.82 2.37  
Cxcl3 3.01 6.48  
Fasl −1.35 1.93 2.61 a 
Cxcl5 2.57 1.64 −1.56 a 
IL2RA  1.92 2.05 
Itgb2  2.14 1.86 
TNF  2.25 1.8 
Timd4  2.17 2.03 
IFNZ  2.23 2.92 
Gzmb  2.27 2.45 
Itga4  2.29 2.06 
Itgb7  2.36 2.04 
Icos  2.45 2.68 
Pdcd1  2.46 2.49 
Gzmk  2.48 2.82 
Ccl4  2.53 2.74 
Itgal  2.57 2.11 
IRF4  2.6 2.79 
Ccr5  2.69 2.82 
Cxcr3  2.72 2.54 
Ccl17  2.81 2.12 
Nos2  3.04 2.6 
IFNG  3.08 3.55 
Cxcr6  3.24 3.97 
CD8A  3.43 3.31 
Xcr1  3.52 3.17 
Itgax  3.58 2.4 
Ccl22  3.78 3.07 
Xcl1  4.08 3.07 
IL12B  4.63 3.77 
Ctla4  5.36 5.54 
Ptgis  −1.92 −2.14 
TGFBR3  −1.78 −2.14 
Itgb5  −1.63 −2.04 
IL6  −4.14 −3.18 
S1pr3  −3.22 −3.53 
Ccl2  −2.84 −2.8 
Itga8  −2.76 −2.85 
Ccl11  −2.64 −2.17 
Socs2  −2.63 −2.63 
Dll1  −2.52 −2.37 
Sgpp2  −2.52 −2.59 
Pim3  −2.36 −2.38 
Itgbl1  −2.31 −2.03 
Cxcl14  −2.28 −2.56 
IL1R1  −2.26 −2.02 
Socs5  −2.18 −2.08 
IL33 −1.38 −2.73 −1.97 
Epcam −1.34 −2.78 −2.08 
IL5 −1.72 −3.8 −2.21 

The list contains the genes coding for cytokines, chemokines, chemokine receptors, integrins, and metalloproteinases critical for T cell homing and activation over time upon infection. Only the genes that exhibited a fold change of at least 2 in at least one of the pairwise comparisons (corrected p value <0.05) are present in the table. The group column provides the identifiers of the gene groups plotted in Supplemental Fig. 3B.

a

Groups containing only a single gene were not assigned an identifier.

Having established the distinctive immune gene expression profiles between LV– and lung mucosal tissue–restricted T cell populations in parenterally immunized animals, we elected to further examine T cell surface protein expression of CXCR3, CXCR6, CCR5, and CD11a (ITGAL) by flow cytometry, a set of molecules that have been implicated in T cell homing to lung mucosal tissue (4, 14, 4446). Inconsistent with its transcriptional profiles, prior to pulmonary M. tuberculosis infection (day 0) the LV Ag-specific T cells expressed marginal levels of CXCR3, whereas the lung mucosal tissue–associated T cells after M. tuberculosis infection (day 21) showed marked levels of CXCR3 surface protein expression (Fig. 7A). Alternatively, contrary to their transcriptional profiles, the lung mucosal T cells showed only a minimally increased CXCR6 and CCR5 expression (Fig. 7A). Compared to CXCR3 expression, CD11a expression was already high before pathogen exposure (day 0) on LV T cells and underwent further induction on lung mucosal-associated T cells after M. tuberculosis infection (day 21) (Fig. 7A).

FIGURE 7.

Upregulation of CXCR3 protein expression on parenteral immunization-induced memory T cells residing in draining lymph nodes and lung mucosal tissue. Mice parenterally immunized for 4 wk were infected with or without M. tuberculosis and sacrificed at selected time points. Lung mononuclear cells without stimulation were subjected to immunostaining. (A) Representative dot plots and bar graphs show the relative frequencies and number of Ag-specific CD8+tet+ T cells expressing CXCR3, CXCR6, CCR5, and CD11a in the lung right before (day 0) and at 21 d postinfection, respectively. (B) Bar graph shows the frequencies of Ag-specific CD8+tet+ T cells expressing the above chemokine receptor/integrin molecules in the MLN and lung homogenates of parenteral-immunized mice at 14 d postinfection. (C) Representative dot plots comparing the frequencies of CXCR3-expressing Ag-specific CD8+tet+ T cells with CD11a-expressing counterparts distributed in the MLN, blood, lung homogenates, and airway at 14 d postinfection. Graphic data are expressed as the mean ± SEM of T cell numbers or frequencies of three to four mice per group, representative of two independent experiments. *p < 0.05, ***p < 0.001.

FIGURE 7.

Upregulation of CXCR3 protein expression on parenteral immunization-induced memory T cells residing in draining lymph nodes and lung mucosal tissue. Mice parenterally immunized for 4 wk were infected with or without M. tuberculosis and sacrificed at selected time points. Lung mononuclear cells without stimulation were subjected to immunostaining. (A) Representative dot plots and bar graphs show the relative frequencies and number of Ag-specific CD8+tet+ T cells expressing CXCR3, CXCR6, CCR5, and CD11a in the lung right before (day 0) and at 21 d postinfection, respectively. (B) Bar graph shows the frequencies of Ag-specific CD8+tet+ T cells expressing the above chemokine receptor/integrin molecules in the MLN and lung homogenates of parenteral-immunized mice at 14 d postinfection. (C) Representative dot plots comparing the frequencies of CXCR3-expressing Ag-specific CD8+tet+ T cells with CD11a-expressing counterparts distributed in the MLN, blood, lung homogenates, and airway at 14 d postinfection. Graphic data are expressed as the mean ± SEM of T cell numbers or frequencies of three to four mice per group, representative of two independent experiments. *p < 0.05, ***p < 0.001.

Close modal

Further analysis of geographical acquisition carried out at an earlier time point (day 14) after M. tuberculosis infection revealed an even greater extent of CXCR3 expression on the T cells present in the MLN and blood before they moved into the lung mucosal tissue, whereas CD11a expression was lower on MLN T cells compared with those in the lung tissue (Fig. 7B, 7C). Indeed, comparison of the relative expression of CXCR3 and CD11a in various compartments shows that CXCR3 expression progressively declined as the T cells moved from the MLN and blood to lung parenchyma and airway whereas CD11a expression progressively increased as they entered and passed through these sites (Fig. 7C). The differential expression levels of CXCR3 and CD11a are not restricted only to Ag-specific CD8 T cells, as this was also the case with total CD8 T cells (Supplemental Fig. 4). Taken together, the above data strongly imply a role for CXCR3, but not CD11a (or CXCR6 and CCR5), in lung mucosal homing of parenterally activated T cells.

Because our findings thus far suggested a role of CXCR3 in lung mucosal homing of parenterally induced Ag-specific T cells, we next examined the protein content of the chemokine ligand IP-10 for CXCR3 in the lung at days 0 and 14 after M. tuberculosis infection and compared it with chemokine ligands MCP-1 and MIP-1α for CCR5 (5). Consistent with the overall CXCR3 and CCR5 expression profiles on T cells, there was little production of their respective chemokine ligands in the lung prior to M. tuberculosis infection or prior to T cell lung homing (day 0) (Fig. 8A). However, at day 14 postinfection, only IP-10, but not MCP-1 or MIP-1α, was significantly increased in the lung of parenterally immunized hosts (Fig. 8A).

FIGURE 8.

Requirement of CXCR3 signaling pathway for lung mucosal homing of systemically activated memory T cells. (A) Chemokine protein levels in the airway (BAL fluids) of parenteral-immunized animals before (day 0) and at 14 d after M. tuberculosis infection. (B) Experimental schema outlining the timelines of CXCR3 blockade or IP-10 neutralization or the control Ab treatment following i.m. immunization and respiratory M. tuberculosis challenge. (C) Representative dot plots and bar graph showing frequencies and numbers of Ag-specific CD8+tet+ T cells, respectively, located within LPT and LV. (D) Representative dot plots and bar graph showing frequencies and numbers of Ag-specific CD8+tet+ T cells, respectively, located within the airway. (E) Bar graph showing numbers of Ag-specific CD8+tet+ T cells located within LPT and LV of the control Ab-treated mice and those treated with both anti–IP-10 and anti-MIG Abs. Graphic data are expressed as the mean ± SEM of T cell numbers from three to four mice per group, representative of two to three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 between indicated comparisons.

FIGURE 8.

Requirement of CXCR3 signaling pathway for lung mucosal homing of systemically activated memory T cells. (A) Chemokine protein levels in the airway (BAL fluids) of parenteral-immunized animals before (day 0) and at 14 d after M. tuberculosis infection. (B) Experimental schema outlining the timelines of CXCR3 blockade or IP-10 neutralization or the control Ab treatment following i.m. immunization and respiratory M. tuberculosis challenge. (C) Representative dot plots and bar graph showing frequencies and numbers of Ag-specific CD8+tet+ T cells, respectively, located within LPT and LV. (D) Representative dot plots and bar graph showing frequencies and numbers of Ag-specific CD8+tet+ T cells, respectively, located within the airway. (E) Bar graph showing numbers of Ag-specific CD8+tet+ T cells located within LPT and LV of the control Ab-treated mice and those treated with both anti–IP-10 and anti-MIG Abs. Graphic data are expressed as the mean ± SEM of T cell numbers from three to four mice per group, representative of two to three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 between indicated comparisons.

Close modal

The data further support the involvement of the CXCR3 ligand signaling pathway in lung mucosal homing of parenterally induced Ag-specific T cells. To investigate this further, we used an Ab-mediated neutralizing approach to block CXCR3 or IP-10. To this end, mice were first immunized i.m. for 4 wk to allow memory T cell development and then infected with M. tuberculosis via the respiratory route. The animals were then subjected to anti-CXCR3 or anti–IP-10 or control Ab treatment at designated time points postinfection (Fig. 8B). Mice were sacrificed 13 d postinfection, and 3–4 min before sacrifice they were subjected to intravascular staining. Upon examination of Ag-specific T cells in LV, LPT, and airway, we found that blockade of CXCR3 most significantly diminished T cell appearance in LPT (Fig. 8C) and airway (Fig. 8D). Neutralization of IP-10 also decreased such T cells in the LPT, albeit to a lesser extent (Fig. 8C), and airway (Fig. 8D). Because IP-10 neutralization only partially reduced T cell homing to lung parenchyma, we next blocked both IP-10 and MIG, another CXCR3 ligand (Fig. 8E). Indeed, when both IP-10 and MIG chemokines were neutralized, a further reduced T cell lung mucosal tissue homing was observed (Fig. 8E). Collectively, the above data indicate that the CXCR3 signaling pathway represents a critical switch-on mechanism for the ultimate homing of systemically TB vaccine–induced Ag-specific T cells to the restricted mucosal entry sites, that is, the LPT and airway lumen, after pulmonary M. tuberculosis infection.

T lymphocytes are critically required for host defense against intracellular infection and cancer in the lung. However, for T cells to be effective, they have to be at, or home to, the battleground in time (15). Given the importance of their geographical location to the benefit of the host’s survival, certain pathogens such as M. tuberculosis even have evolved sophisticated mechanisms to delay the arrival of the primed circulating Ag-specific T cells at lung mucosal sites of infection (47, 48). Thus, this represents a unique challenge to develop vaccine strategies that can provide robust protection in the lung via eliciting the lasting Trm within, or a type of circulating T cells capable of rapid entry to the lung mucosa upon pathogen exposure. However, until recently it was thought that the lung parenchyma was a common T cell entry site, whereas the airway was a restricted T cell entry site following systemic or parenteral T cell activation (3, 6). This immunologic dogma posed two issues. First, under homeostatic conditions, the lung shall remain as inflammation-free as possible to ensure efficient gas exchange (1, 2). Second, it can hardly explain why the presence of memory T cells detectable in lung homogenate-derived mononuclear cells almost completely fails to provide protection in parenterally immunized hosts (11, 17, 18, 22, 24). However, this conflicted view has recently become resolvable by the understanding that the lung parenchyma is not a single entity as thought before; rather, it is contaminated with cells retained in the LV (33, 43). However, the underlying immune molecular events that govern systemic immunization–activated T cell homing to the lung have awaited understanding. In the present study, by using a model of parenteral TB immunization and examining T cells in the LV, lung parenchyma tissue, and airway before and after respiratory pathogen exposure, we have shown that systemically induced resting memory T cells by parenteral TB immunization with a viral-based TB vaccine are restricted to the LV and are unable to populate either the LPT or the airway under homeostatic conditions. We further found that upon pulmonary M. tuberculosis infection, it takes as many as 10–14 d before such T cells concurrently enter the LPT and airway lumen. Acquisition of CXCR3 expression on circulating T cells is required for their entry to these respiratory mucosal tissue compartments in parenterally immunized hosts.

The understanding of entry of circulating Ag-specific T cells to the lung mucosa was misled by earlier studies because of the knowledge that the lung consists of only two major compartments, the conducting airways and the lung parenchyma (1). However, the recent appreciation that the lung homogenate-derived mononuclear cells are contaminated with cells retained in the LV and that such cells are different from the cells residing within the LPT has led to the new paradigm that the T cell entry to the LPT is tightly controlled under homeostatic conditions, ensuring efficient gas exchange in the lung (12, 49). Our findings further support this new paradigm that systemically induced memory T cells by parenteral TB immunization are restricted to the LV, unable to populate either the LPT or the airway under homeostatic conditions. Additionally, our findings provide a clear mechanistic explanation as to why systemically activated memory T cells are poorly immune protective of the lung. This is particularly relevant to the intracellular pathogens against which T cell immunity, but not B cell/Ab-based immunity, is critically required. It is known that some of such intracellular pathogens, in particular M. tuberculosis, have evolved with the mechanisms to slow down T cell priming and the appearance of T cell immunity at the site of infection (27, 47, 48). Indeed, we find that it takes many days after pulmonary M. tuberculosis exposure before systemically induced resting memory T cells begin to assume effector phenotypes, upregulate CXCR3 expression, and gain entry to the LPT and airway. Although the time needed for circulating T cells to enter the lung mucosal sites may vary depending on the nature of lung immunological encounter, it has been well established that parenteral immunization is inferior to respiratory mucosal immunization in lung protection in a number of scenarios evaluating various vaccine platforms against M. tuberculosis, viruses, and cancer (8, 11, 12, 1722, 49).

Our study holds important implications in the hunt for novel vaccination strategies. Indeed, the respiratory mucosal route of immunization represents a robust means to establish protective Trm in the lung prior to pathogen encounter (4, 16, 17). However, developing both effective and safe respiratory mucosal vaccines for human application has been a daunting challenge due to limited choices of vaccine backbone and adjuvant, as well as higher safety expectations. In part, this is the reason why the vast majority of human vaccines currently in use are administered parenterally. Our current findings suggest that to significantly improve the efficacy of parenteral vaccination in mucosal protection requires new strategies to overcome the T cell mucosal homing barrier. One way is to design immune strategies that will systemically upregulate the expression on vaccine-induced T cells of lung mucosal homing molecules such as CXCR3. Such T cells may more quickly get mobilized into the site of infection in response to increased local chemokine gradient. However, as we have shown in the present study and in a separate report (30), the establishment of such a chemokine gradient is also retarded by active M. tuberculosis infection in the lung. This situation may, however, differ somewhat under homeostatic conditions in human lungs that are constantly exposed to environmentally borne agents (4, 50). Notwithstanding, a recent human vaccine study has found that parenteral (intradermal) vaccine-activated T cells display only a very limited capability to home to human airways under homeostatic conditions, contrary to aerosol vaccine-induced counterparts (51). We have previously shown that prior respiratory mucosal viral infection can signal systemically activated T cells to home to the lung mucosa (34). It still remains to be understood whether the prior respiratory viral infection–mediated signaling is due to resident memory T cells induced by prior viral infection or simply due to altered lung microenvironment. Findings in the present study suggest that resident memory T cell–derived signals alone are not sufficient enough for circulating memory T cells to home to the lung because in our competitive T cell homing assay, despite the preferential homing of i.n. induced Trm to the lung, i.m. induced T cells still fail to home to the lung. Another strategy to draw the circulating Ag-specific T cells to the lung for enhanced protection is to deliberately apply soluble (and safe) proimmune or antigenic agents to the respiratory tract (18, 34, 52), likely via upregulated expression of lung mucosal homing receptors such as CXCR3 on circulating T cells. Recruited T cells by antigenic agents may self-sustain locally via a specific Ag-dependent mechanism (32). Admittedly, our present study was not designed to investigate the difference in lung-homing mechanisms between parenteral vaccine-induced and i.n. mucosal vaccine-induced circulating resting memory T cells before M. tuberculosis infection. Such resting memory T cell homing mechanisms may differ from those required for lung homing of circulating effector/TEM following infection. The knowledge from future studies in this regard will further help design improved vaccine strategies.

A number of previous studies have studied the homing molecules involved in T cell migration to the lung following primary or secondary respiratory viral infections (14, 4446, 53). Lee et al. (54) and we compared the transcriptional profile of lung CD8 T cells induced by respiratory mucosal and parenteral immunization and identified unique chemokine receptor expression by respiratory mucosal immunization-induced cells. However, until now, the molecular switch underpinning the homing of circulating T cells to lung mucosal sites in systemically infected or immunized hosts has remained poorly understood. Using the gene microarray analysis and loss of function approaches, we have identified the CXCR3–ligand interaction to be a key signaling pathway involved in the entry of parenteral vaccine-induced T cells from LV into lung mucosal sites. Although different from the skin and gut, the lung-specific T cell homing molecules have remained elusive (47, 15), and our findings in conjunction with previous studies (4446) suggest an indispensable role for CXCR3 signaling in T cell lung homing. CXCR3 signaling has also recently been shown to be critical for antitumor T cell trafficking across tumor vasculature (35). Our data further indicate the acquisition of CXCR3 by systemically induced T cells to be in the lymph nodes draining the lung infection site (Supplemental Fig. 4), supporting the current view that dendritic cells in the draining lymphoid tissues play an important role in imprinting T cells with mucosal homing signature (6, 53). We also observed that the T cells upon the lung entry downregulate CXCR3 expression probably due to its ligand binding. In contrast, influenza infection–induced memory CD4 T cells in the lung continued to express CXCR3 (46). To this end, it has been shown that pathogen specificity can have a profound effect on CXCR3 receptor expression of lung memory CD8 T cells due to differential inflammatory cytokines induced by specific pathogens (44). Other chemokine receptors, that is, CCR5, CCR3, and CCR4, have also been implicated in T cell homing to the lung following respiratory viral infections and in response to allergic asthma (14, 53, 55, 56), and lung dendritic cells play a critical role in upregulation of CCR4 on lung-homing T cells (53). In this study, we demonstrate that CXCR3 upregulation on systemic vaccine-induced CD8 T cells is required for their entry to lung mucosa following lung infection. A recent study examining trafficking of T cells to the lung of parenteral protein-based vaccine-immunized mice also demonstrated CD4 T cells expressing CXCR3 to efficiently migrate into the M. tuberculosis–infected lung parenchyma (57).

Collectively, our study has established the lung parenchyma tissue, similar to the airway, to be a restricted homing site for systemically induced resting memory T cells under homeostatic conditions. The entry of such T cells from the circulation to these respiratory mucosal sites is tightly regulated, involving the engagement of CXCR3 signaling pathway, which can be slow developing and requires respiratory microbial exposure. Our findings hold important implications in the development of novel effective immunization strategies against TB and other respiratory infectious diseases.

We are grateful for technical assistance from Anna Zganiacz and Xueya Feng. We also acknowledge the provision of MHC class I tetramers from the National Institutes of Health Tetramer Core.

This work was supported by funds from the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council of Canada.

The microarray data presented in this article have been submitted to the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE100288.

The online version of this article contains supplemental material.

Abbreviations used in this article:

     
  • AdHu5Ag85A

    recombinant replication-deficient human type 5 adenovirus expressing an immunodominant M. tuberculosis Ag85A

  •  
  • BAL

    bronchoalveolar lavage

  •  
  • i.n.

    intranasal(ly)

  •  
  • LPT

    lung parenchymal tissue

  •  
  • LV

    lung vasculature

  •  
  • MLN

    mediastinal lymph node

  •  
  • TB

    tuberculosis

  •  
  • TCM

    central memory T cell

  •  
  • TE

    effector T cell

  •  
  • TEM

    effector memory T cell

  •  
  • tet

    tetramer

  •  
  • Trm

    tissue-resident memory T cell.

1
Holt
,
P. G.
,
D. H.
Strickland
,
M. E.
Wikström
,
F. L.
Jahnsen
.
2008
.
Regulation of immunological homeostasis in the respiratory tract.
Nat. Rev. Immunol.
8
:
142
152
.
2
Wissinger
,
E.
,
J.
Goulding
,
T.
Hussell
.
2009
.
Immune homeostasis in the respiratory tract and its impact on heterologous infection.
Semin. Immunol.
21
:
147
155
.
3
Woodland
,
D. L.
,
J. E.
Kohlmeier
.
2009
.
Migration, maintenance and recall of memory T cells in peripheral tissues.
Nat. Rev. Immunol.
9
:
153
161
.
4
Shane
,
H. L.
,
K. D.
Klonowski
.
2014
.
Every breath you take: the impact of environment on resident memory CD8 T cells in the lung.
Front. Immunol.
5
:
320
.
5
Griffith
,
J. W.
,
C. L.
Sokol
,
A. D.
Luster
.
2014
.
Chemokines and chemokine receptors: positioning cells for host defense and immunity.
Annu. Rev. Immunol.
32
:
659
702
.
6
Shin
,
H.
,
A.
Iwasaki
.
2013
.
Tissue-resident memory T cells.
Immunol. Rev.
255
:
165
181
.
7
Schenkel
,
J. M.
,
D.
Masopust
.
2014
.
Tissue-resident memory T cells.
Immunity
41
:
886
897
.
8
Beverley
,
P. C.
,
S.
Sridhar
,
A.
Lalvani
,
E. Z.
Tchilian
.
2014
.
Harnessing local and systemic immunity for vaccines against tuberculosis.
Mucosal Immunol.
7
:
20
26
.
9
Klonowski
,
K. D.
,
K. J.
Williams
,
A. L.
Marzo
,
D. A.
Blair
,
E. G.
Lingenheld
,
L.
Lefrançois
.
2004
.
Dynamics of blood-borne CD8 memory T cell migration in vivo.
Immunity
20
:
551
562
.
10
Masopust
,
D.
,
V.
Vezys
,
A. L.
Marzo
,
L.
Lefrançois
.
2001
.
Preferential localization of effector memory cells in nonlymphoid tissue.
Science
291
:
2413
2417
.
11
Santosuosso
,
M.
,
X.
Zhang
,
S.
McCormick
,
J.
Wang
,
M.
Hitt
,
Z.
Xing
.
2005
.
Mechanisms of mucosal and parenteral tuberculosis vaccinations: adenoviral-based mucosal immunization preferentially elicits sustained accumulation of immune protective CD4 and CD8 T cells within the airway lumen.
J. Immunol.
174
:
7986
7994
.
12
Wu
,
T.
,
Y.
Hu
,
Y. T.
Lee
,
K. R.
Bouchard
,
A.
Benechet
,
K.
Khanna
,
L. S.
Cauley
.
2014
.
Lung-resident memory CD8 T cells (TRM) are indispensable for optimal cross-protection against pulmonary virus infection.
J. Leukoc. Biol.
95
:
215
224
.
13
Galkina
,
E.
,
J.
Thatte
,
V.
Dabak
,
M. B.
Williams
,
K.
Ley
,
T. J.
Braciale
.
2005
.
Preferential migration of effector CD8+ T cells into the interstitium of the normal lung.
J. Clin. Invest.
115
:
3473
3483
.
14
Kohlmeier
,
J. E.
,
S. C.
Miller
,
J.
Smith
,
B.
Lu
,
C.
Gerard
,
T.
Cookenham
,
A. D.
Roberts
,
D. L.
Woodland
.
2008
.
The chemokine receptor CCR5 plays a key role in the early memory CD8+ T cell response to respiratory virus infections.
Immunity
29
:
101
113
.
15
Baaten
,
B. J.
,
A. M.
Cooper
,
S. L.
Swain
,
L. M.
Bradley
.
2013
.
Location, location, location: the impact of migratory heterogeneity on T cell function.
Front. Immunol.
4
:
311
.
16
Lycke
,
N.
2012
.
Recent progress in mucosal vaccine development: potential and limitations.
Nat. Rev. Immunol.
12
:
592
605
.
17
Belyakov
,
I. M.
,
J. D.
Ahlers
.
2009
.
What role does the route of immunization play in the generation of protective immunity against mucosal pathogens?
J. Immunol.
183
:
6883
6892
.
18
Belyakov
,
I. M.
,
D.
Isakov
,
Q.
Zhu
,
A.
Dzutsev
,
D.
Klinman
,
J. A.
Berzofsky
.
2006
.
Enhancement of CD8+ T cell immunity in the lung by CpG oligodeoxynucleotides increases protective efficacy of a modified vaccinia Ankara vaccine against lethal poxvirus infection even in a CD4-deficient host.
J. Immunol.
177
:
6336
6343
.
19
Patel
,
A.
,
Y.
Zhang
,
M.
Croyle
,
K.
Tran
,
M.
Gray
,
J.
Strong
,
H.
Feldmann
,
J. M.
Wilson
,
G. P.
Kobinger
.
2007
.
Mucosal delivery of adenovirus-based vaccine protects against Ebola virus infection in mice.
J. Infect. Dis.
196
(
Suppl. 2
):
S413
S420
.
20
Sandoval
,
F.
,
M.
Terme
,
M.
Nizard
,
C.
Badoual
,
M. F.
Bureau
,
L.
Freyburger
,
O.
Clement
,
E.
Marcheteau
,
A.
Gey
,
G.
Fraisse
, et al
.
2013
.
Mucosal imprinting of vaccine-induced CD8+ T cells is crucial to inhibit the growth of mucosal tumors. [Published erratum appears in 2013 Sci. Transl. Med. 5: 178er2.]
Sci. Transl. Med.
5
:
172ra20
.
21
Wang
,
J.
,
L.
Thorson
,
R. W.
Stokes
,
M.
Santosuosso
,
K.
Huygen
,
A.
Zganiacz
,
M.
Hitt
,
Z.
Xing
.
2004
.
Single mucosal, but not parenteral, immunization with recombinant adenoviral-based vaccine provides potent protection from pulmonary tuberculosis.
J. Immunol.
173
:
6357
6365
.
22
Forbes
,
E. K.
,
C.
Sander
,
E. O.
Ronan
,
H.
McShane
,
A. V.
Hill
,
P. C.
Beverley
,
E. Z.
Tchilian
.
2008
.
Multifunctional, high-level cytokine-producing Th1 cells in the lung, but not spleen, correlate with protection against Mycobacterium tuberculosis aerosol challenge in mice.
J. Immunol.
181
:
4955
4964
.
23
Kohlmeier
,
J. E.
,
D. L.
Woodland
.
2006
.
Memory T cell recruitment to the lung airways.
Curr. Opin. Immunol.
18
:
357
362
.
24
Jeyanathan
,
M.
,
A.
Heriazon
,
Z.
Xing
.
2010
.
Airway luminal T cells: a newcomer on the stage of TB vaccination strategies.
Trends Immunol.
31
:
247
252
.
25
Berzofsky
,
J. A.
,
J. D.
Ahlers
,
I. M.
Belyakov
.
2001
.
Strategies for designing and optimizing new generation vaccines.
Nat. Rev. Immunol.
1
:
209
219
.
26
Rollier
,
C. S.
,
A.
Reyes-Sandoval
,
M. G.
Cottingham
,
K.
Ewer
,
A. V.
Hill
.
2011
.
Viral vectors as vaccine platforms: deployment in sight.
Curr. Opin. Immunol.
23
:
377
382
.
27
Griffiths
,
K. L.
,
S. A.
Khader
.
2014
.
Novel vaccine approaches for protection against intracellular pathogens.
Curr. Opin. Immunol.
28
:
58
63
.
28
Afkhami
,
S.
,
Y.
Yao
,
Z.
Xing
.
2016
.
Methods and clinical development of adenovirus-vectored vaccines against mucosal pathogens.
Mol. Ther. Methods Clin. Dev.
3
:
16030
.
29
Smaill
,
F.
,
M.
Jeyanathan
,
M.
Smieja
,
M. F.
Medina
,
N.
Thanthrige-Don
,
A.
Zganiacz
,
C.
Yin
,
A.
Heriazon
,
D.
Damjanovic
,
L.
Puri
, et al
.
2013
.
A human type 5 adenovirus-based tuberculosis vaccine induces robust T cell responses in humans despite preexisting anti-adenovirus immunity.
Sci. Transl. Med.
5
:
205ra134
.
30
Lai
,
R.
,
M.
Jeyanathan
,
C. R.
Shaler
,
D.
Damjanovic
,
A.
Khera
,
C.
Horvath
,
A. A.
Ashkar
,
Z.
Xing
.
2014
.
Restoration of innate immune activation accelerates Th1-cell priming and protection following pulmonary mycobacterial infection.
Eur. J. Immunol.
44
:
1375
1386
.
31
Jeyanathan
,
M.
,
D.
Damjanovic
,
C. R.
Shaler
,
R.
Lai
,
M.
Wortzman
,
C.
Yin
,
A.
Zganiacz
,
B. D.
Lichty
,
Z.
Xing
.
2013
.
Differentially imprinted innate immunity by mucosal boost vaccination determines antituberculosis immune protective outcomes, independent of T-cell immunity.
Mucosal Immunol.
6
:
612
625
.
32
Jeyanathan
,
M.
,
J.
Mu
,
S.
McCormick
,
D.
Damjanovic
,
C. L.
Small
,
C. R.
Shaler
,
K.
Kugathasan
,
Z.
Xing
.
2010
.
Murine airway luminal antituberculosis memory CD8 T cells by mucosal immunization are maintained via antigen-driven in situ proliferation, independent of peripheral T cell recruitment.
Am. J. Respir. Crit. Care Med.
181
:
862
872
.
33
Anderson
,
K. G.
,
K.
Mayer-Barber
,
H.
Sung
,
L.
Beura
,
B. R.
James
,
J. J.
Taylor
,
L.
Qunaj
,
T. S.
Griffith
,
V.
Vezys
,
D. L.
Barber
,
D.
Masopust
.
2014
.
Intravascular staining for discrimination of vascular and tissue leukocytes.
Nat. Protoc.
9
:
209
222
.
34
Jeyanathan
,
M.
,
N.
Thanthrige-Don
,
S.
Afkhami
,
R.
Lai
,
D.
Damjanovic
,
A.
Zganiacz
,
X.
Feng
,
X. D.
Yao
,
K. L.
Rosenthal
,
M. F.
Medina
, et al
.
2015
.
Novel chimpanzee adenovirus-vectored respiratory mucosal tuberculosis vaccine: overcoming local anti-human adenovirus immunity for potent TB protection.
Mucosal Immunol.
8
:
1373
1387
.
35
Mikucki
,
M. E.
,
D. T.
Fisher
,
J.
Matsuzaki
,
J. J.
Skitzki
,
N. B.
Gaulin
,
J. B.
Muhitch
,
A. W.
Ku
,
J. G.
Frelinger
,
K.
Odunsi
,
T. F.
Gajewski
, et al
.
2015
.
Non-redundant requirement for CXCR3 signalling during tumoricidal T-cell trafficking across tumour vascular checkpoints.
Nat. Commun.
6
:
7458
.
36
Chaturvedi
,
V.
,
J. M.
Ertelt
,
T. T.
Jiang
,
J. M.
Kinder
,
L.
Xin
,
K. J.
Owens
,
H. N.
Jones
,
S. S.
Way
.
2015
.
CXCR3 blockade protects against Listeria monocytogenes infection-induced fetal wastage.
J. Clin. Invest.
125
:
1713
1725
.
37
Zeng
,
X.
,
T. A.
Moore
,
M. W.
Newstead
,
J. C.
Deng
,
S. L.
Kunkel
,
A. D.
Luster
,
T. J.
Standiford
.
2005
.
Interferon-inducible protein 10, but not monokine induced by gamma interferon, promotes protective type 1 immunity in murine Klebsiella pneumoniae pneumonia.
Infect. Immun.
73
:
8226
8236
.
38
Edgar
,
R.
,
M.
Domrachev
,
A. E.
Lash
.
2002
.
Gene expression omnibus: NCBI gene expression and hybridization array data repository.
Nucleic Acids Res.
30
:
207
210
.
39
Mi
,
H.
,
Q.
Dong
,
A.
Muruganujan
,
P.
Gaudet
,
S.
Lewis
,
P. D.
Thomas
.
2010
.
PANTHER version 7: improved phylogenetic trees, orthologs and collaboration with the Gene Ontology Consortium.
Nucleic Acids Res.
38
:
D204
D210
.
40
Gentleman
,
R. C.
,
V. J.
Carey
,
D. M.
Bates
,
B.
Bolstad
,
M.
Dettling
,
S.
Dudoit
,
B.
Ellis
,
L.
Gautier
,
Y.
Ge
,
J.
Gentry
, et al
.
2004
.
Bioconductor: open software development for computational biology and bioinformatics.
Genome Biol.
5
:
R80
.
41
Smyth
,
G. K.
2004
.
Linear models and empirical bayes methods for assessing differential expression in microarray experiments.
Stat. Appl. Genet. Mol. Biol.
3
:
Article3
doi:10.2202/1544-6115.1027
.
42
Benjamini
,
Y.
,
D.
Drai
,
G.
Elmer
,
N.
Kafkafi
,
I.
Golani
.
2001
.
Controlling the false discovery rate in behavior genetics research.
Behav. Brain Res.
125
:
279
284
.
43
Anderson
,
K. G.
,
H.
Sung
,
C. N.
Skon
,
L.
Lefrancois
,
A.
Deisinger
,
V.
Vezys
,
D.
Masopust
.
2012
.
Cutting edge: intravascular staining redefines lung CD8 T cell responses.
J. Immunol.
189
:
2702
2706
.
44
Slütter
,
B.
,
L. L.
Pewe
,
S. M.
Kaech
,
J. T.
Harty
.
2013
.
Lung airway-surveilling CXCR3hi memory CD8+ T cells are critical for protection against influenza a virus.
Immunity
39
:
939
948
.
45
Sakai
,
S.
,
K. D.
Kauffman
,
J. M.
Schenkel
,
C. C.
McBerry
,
K. D.
Mayer-Barber
,
D.
Masopust
,
D. L.
Barber
.
2014
.
Cutting edge: control of Mycobacterium tuberculosis infection by a subset of lung parenchyma-homing CD4 T cells.
J. Immunol.
192
:
2965
2969
.
46
Kohlmeier
,
J. E.
,
T.
Cookenham
,
S. C.
Miller
,
A. D.
Roberts
,
J. P.
Christensen
,
A. R.
Thomsen
,
D. L.
Woodland
.
2009
.
CXCR3 directs antigen-specific effector CD4+ T cell migration to the lung during parainfluenza virus infection.
J. Immunol.
183
:
4378
4384
.
47
Cooper
,
A. M.
2009
.
Cell-mediated immune responses in tuberculosis.
Annu. Rev. Immunol.
27
:
393
422
.
48
Jeyanathan
,
M.
,
S.
McCormick
,
R.
Lai
,
S.
Afkhami
,
C. R.
Shaler
,
C. N.
Horvath
,
D.
Damjanovic
,
A.
Zganiacz
,
N.
Barra
,
A.
Ashkar
, et al
.
2014
.
Pulmonary M. tuberculosis infection delays Th1 immunity via immunoadaptor DAP12-regulated IRAK-M and IL-10 expression in antigen-presenting cells.
Mucosal Immunol.
7
:
670
683
.
49
Morabito
,
K. M.
,
T. R.
Ruckwardt
,
A. J.
Redwood
,
S. M.
Moin
,
D. A.
Price
,
B. S.
Graham
.
2017
.
Intranasal administration of RSV antigen-expressing MCMV elicits robust tissue-resident effector and effector memory CD8+ T cells in the lung.
Mucosal Immunol.
10
:
545
554
.
50
Welsh
,
R. M.
,
L. K.
Selin
.
2002
.
No one is naive: the significance of heterologous T-cell immunity.
Nat. Rev. Immunol.
2
:
417
426
.
51
Satti
,
I.
,
J.
Meyer
,
S. A.
Harris
,
Z. R.
Manjaly Thomas
,
K.
Griffiths
,
R. D.
Antrobus
,
R.
Rowland
,
R. L.
Ramon
,
M.
Smith
,
S.
Sheehan
, et al
.
2014
.
Safety and immunogenicity of a candidate tuberculosis vaccine MVA85A delivered by aerosol in BCG-vaccinated healthy adults: a phase 1, double-blind, randomised controlled trial.
Lancet Infect. Dis.
14
:
939
946
.
52
Santosuosso
,
M.
,
S.
McCormick
,
E.
Roediger
,
X.
Zhang
,
A.
Zganiacz
,
B. D.
Lichty
,
Z.
Xing
.
2007
.
Mucosal luminal manipulation of T cell geography switches on protective efficacy by otherwise ineffective parenteral genetic immunization.
J. Immunol.
178
:
2387
2395
.
53
Mikhak
,
Z.
,
J. P.
Strassner
,
A. D.
Luster
.
2013
.
Lung dendritic cells imprint T cell lung homing and promote lung immunity through the chemokine receptor CCR4.
J. Exp. Med.
210
:
1855
1869
.
54
Lee
,
L. N.
,
D.
Baban
,
E. O.
Ronan
,
J.
Ragoussis
,
P. C.
Beverley
,
E. Z.
Tchilian
.
2010
.
Chemokine gene expression in lung CD8 T cells correlates with protective immunity in mice immunized intra-nasally with Adenovirus-85A.
BMC Med. Genomics
3
:
46
.
55
Lloyd
,
C. M.
,
T.
Delaney
,
T.
Nguyen
,
J.
Tian
,
C.
Martinez-A
,
A. J.
Coyle
,
J. C.
Gutierrez-Ramos
.
2000
.
CC chemokine receptor (CCR)3/eotaxin is followed by CCR4/monocyte-derived chemokine in mediating pulmonary T helper lymphocyte type 2 recruitment after serial antigen challenge in vivo.
J. Exp. Med.
191
:
265
274
.
56
Mikhak
,
Z.
,
M.
Fukui
,
A.
Farsidjani
,
B. D.
Medoff
,
A. M.
Tager
,
A. D.
Luster
.
2009
.
Contribution of CCR4 and CCR8 to antigen-specific TH2 cell trafficking in allergic pulmonary inflammation.
J. Allergy. Clin. Immunol.
123
:
67
73.e63
.
57
Woodworth
,
J. S.
,
S. B.
Cohen
,
A. O.
Moguche
,
C. R.
Plumlee
,
E. M.
Agger
,
K. B.
Urdahl
,
P.
Andersen
.
2017
.
Subunit vaccine H56/CAF01 induces a population of circulating CD4 T cells that traffic into the Mycobacterium tuberculosis-infected lung.
Mucosal Immunol.
10
:
555
564
.

The authors have no financial conflicts of interest.

Supplementary data