Tissue-resident memory T cells (TRMs) have a key role in mediating the host defense against tuberculosis (TB) in mice, but their human counterparts have not been well characterized. In this article, we recruited patients with TB and determined TRM frequency, trafficking, activation marker expression, and cytokine production by flow or mass cytometry at different infection sites, including peripheral blood, pleural fluid, bronchoalveolar lavage fluid, and lung. We found a high frequency of TRMs at all infection sites apart from the peripheral blood. These TRMs exhibited a memory phenotype, were highly activated (based on CD38 and HLA-DR expression), and expressed high levels of trafficking (CCR5 and CXCR6) and exhaustion (PD-1) markers. When stimulated with Mycobacterium tuberculosis, TRMs secreted cytokines, including IFN-γ, TNF-α, and IL-2, and exhibited a multifunctional phenotype. TRMs limited intracellular M. tuberculosis replication in macrophages. These data inform our current understanding of immunosurveillance at different infection sites in patients with TB.

T cells have critical roles in mediating the host defense against Mycobacterium tuberculosis infection but are insufficient to clear the infection (1). No other suitable immune correlates for protective immunity against tuberculosis (TB), however, have been identified, perhaps because immune responses measured ex vivo in the peripheral blood might not faithfully reflect the immune response that occurs at the infection site (2). For example, tissue-resident memory T cells (TRMs) permanently reside in nonlymphoid tissues and cannot be detected in the peripheral blood (3). TRMs exhibit a distinct genetic signature as well as migration, retention, and functional maintenance characteristics that differ from those of peripheral T cells. These cells are typically defined by CD69, CD103, CD49a, and CD44 cell surface expression in both mice and humans (4). Increasing interest has been paid to the relevance of TRMs in disease pathogenesis because of their ability to mount a response against pathogens and accelerate pathogen clearance (5, 6).

Recent studies performed in mice have indicated that TRMs in the lung are critical for TB protection (7); their human counterparts, however, remain relatively uncharacterized in TB. In this article, we aimed to address this knowledge gap by characterizing human TRMs present at infection sites, including the pleural fluid, bronchoalveolar lavage fluid (BALF), and the lung parenchyma from TB patients, in terms of their frequency, phenotype, and function. Based on previous studies, we defined TRMs by the coexpression of CD69 and CD103 (4). We show that TRMs are present at all infection sites in patients with TB. Importantly, TRMs are not detectable in the peripheral blood. These TRMs are activated, expressing a high level of T cell trafficking and exhaustion markers. M. tuberculosis Ag-specific TRMs secrete multiple cytokines and exhibit a multifunctional phenotype. Finally, we show that TRMs can limit intracellular M. tuberculosis replication in macrophages. These data support that TRMs exist in multiple infection sites. Given their absence from the peripheral blood, these findings inform us on the most appropriate sites to detect and study the role of the T cell response in the context of local TB infection. Our data provide a scientific basis for the future design of vaccines and immunotherapies for TB.

Ethical approval for this study was obtained from the Research Ethics Committee of the Shenzhen Third People’s Hospital (Shenzhen, China). All participants provided written informed consent for sample collection and the subsequent analyses. All participants were enrolled at the Shenzhen Third People’s Hospital between January 2017 and July 2019. A TB diagnosis was based on clinical symptoms, chest radiography, microscopy for acid-fast bacilli, M. tuberculosis culture and GeneXpert analysis of sputum and/or BALF and/or pleural effusion, and the response to anti-TB chemotherapy (8). All patients with comorbid HIV were excluded from the study. All patients with TB were undergoing their first round of anti-TB therapy. The cohort demographics and clinical information are provided in Supplemental Table I.

PBMCs were obtained by whole-blood gradient separation. Pleural fluid mononuclear cells (PFMCs) and BALF cells were separated by centrifugation, whereas RBCs were lysed using ACK lysis buffer (Life Technologies). Lung parenchyma tissues (Lung) collected during surgical procedures were incubated with 1 mg/ml collagenase type IV (Roche) and 1 mg/ml DNase I (Roche) for 1 h 37°C before filtering through a 70-mM nylon mesh. The homogenates were resuspended with Percoll (GE Healthcare Systems), and the cells were obtained by gradient separation. All the cells were stored in liquid nitrogen until further use.

PFMCs (1 × 106) were suspended in RPMI 1640 containing 10% FBS and then incubated with 10 μg/ml M. tuberculosis H37Rv lysate at 37°C for 24 h. During the final 4 h of the incubation, 10 μg/ml brefeldin A (Sigma-Aldrich) was added to the culture media. The cultured cells were then harvested for staining with Abs suitable for flow cytometry or cytometry by time-of-flight (CyTOF).

For cell surface labeling, 1 × 106 cells were blocked with Fc-block reagent (BD Biosciences). Then, the following Abs were added and incubated for 30 min: CD8 (RPA-T8; BioLegend), CD69 (FN50; BioLegend), CD4 (RPA-T4; BD Biosciences), CD45RA (HI100; BD Biosciences), CCR7 (G043H7; BioLegend), and CD103 (Ber-ACT8; BD Biosciences). After incubation, the samples were washed and reconstituted in PBS for flow cytometric analysis on an FACSCanto II Flow Cytometer. For intracellular staining, the cells were fixed and permeabilized using Fix/Perm Buffer (BD Biosciences). Then, the following Abs were added to the cells for intracellular labeling: TNF-α (MAb11; BioLegend), IFN-γ (B27; BD Biosciences) and IL-2 (MQ1-17H12; BioLegend). The Abs were incubated for 60 min before the cells were washed and fixed with 1.6% paraformaldehyde. Data analysis was performed using FlowJo software version 10.

The Abs for CyTOF are listed in Supplemental Table II. All CyTOF staining was performed at room temperature, as previously described (9). In brief, the cells were incubated in 1 M cisplatin (Fluidigm) for 5 min. Fc-block reagent was then added to the cells for 10 min. Next, the cells were stained with Abs for 30 min for surface labeling and then washed once in CyTOF staining buffer (Fluidigm). For intracellular labeling, the cells were fixed and permeabilized using Fix/Perm Buffer (Fluidigm) prior to incubation with the appropriate Abs for 30 min. Then, the cells were washed twice by centrifugation with PBS, resuspended in 1.6% paraformaldehyde (Fluidigm) in the presence of 125 nM iridium intercalator (Fluidigm), and incubated overnight at 4°C. After three further washes, the cells were resuspended in PBS containing EQ calibration beads (Fluidigm) before analysis on a Helios mass cytometer. Data analysis was conducted using Cytobank (Cytobank, Mountain View, CA) and FlowJo (Tree Star, Ashland, OR).

Both M. tuberculosis infection and intracellular M. tuberculosis growth assays were performed as previously described (10). Briefly, monocytes (5 × 104 per well) from patients with TB were cultured in RPMI 1640 medium supplemented with 10% FBS, human rIL-4 (BD Biosciences), and GM-CSF (Sigma-Aldrich) for 7 d. The H37Ra inoculum was added at a multiplicity of infection = 5. After overnight infection at 37°C, the extracellular H37Ra in the medium was removed. The memory (CD45RACCR7) CD4+CD69, CD4+CD69+, CD8+CD69­, and CD8+CD69 T cells were enriched from PFMCs by flow cytometry, and then 5 × 105 cells per well were incubated with H37Ra-infected autologous macrophages. After incubation in 5% CO2 at 37°C for 3 d, the culture supernatants were aspirated, and the cells were lysed with lysis buffer (0.067% SDS in Middlebrook 7H9) prior to plating for CFU.

GraphPad Prism 6.0 software (GraphPad) was used for all statistical calculations. Heatmaps were displayed in R using the heatmap.2 function from the ggplot package. A p < 0.05 was considered statistically significant with a 95% confidence interval. The statistical analyses performed are indicated in the figure legends.

CD69 and CD103 are two key markers used to identify TRMs (11, 12). We first determined the distribution of TRMs between different infection sites by flow cytometric analysis of CD4+ and CD8+ T cells based on CD69 and CD103 coexpression (Fig. 1A). We found that CD69+CD103+ T cells were significantly more frequent at sites of TB infection, (i.e., in the PFMCs [2–5%], BALF [13–30%], and lung [15–28%]) compared with in the PBMCs (0.05–0.3%) (Fig. 1B), suggesting that TRMs are present at different TB infection sites.

FIGURE 1.

TRMs are highly frequent at infection sites in patients with TB. (A) CD69 and CD103 expression in PBMCs, PFMCs, BALF, and lung parenchyma isolated from patients with TB, as assessed by flow cytometry. (B) The frequency of CD69+CD103+ cells in PBMCs (n = 19), PFMCs (n = 15), BALF (n = 30), and lung (n = 5). Unpaired Student t test compared with PBMCs. *p ≤ 0.05, **p ≤ 0.01. (C) The proportions of the four CD45RA/CCR7 subsets in CD69+CD103+ cells in the indicated compartments.

FIGURE 1.

TRMs are highly frequent at infection sites in patients with TB. (A) CD69 and CD103 expression in PBMCs, PFMCs, BALF, and lung parenchyma isolated from patients with TB, as assessed by flow cytometry. (B) The frequency of CD69+CD103+ cells in PBMCs (n = 19), PFMCs (n = 15), BALF (n = 30), and lung (n = 5). Unpaired Student t test compared with PBMCs. *p ≤ 0.05, **p ≤ 0.01. (C) The proportions of the four CD45RA/CCR7 subsets in CD69+CD103+ cells in the indicated compartments.

Close modal

T cells are typically subdivided into naive, central memory, effector memory (TEM) and late TEM cells based CD45RA and CCR7 expression (13). We found that both CD4+ and CD8+ TRMs in infection sites expressed low CCR7 and CD45RA levels. In addition, most TRMs (range 77.09–98.16%) exhibited a TEM phenotype (CD45RACCR7) (Fig. 1C). These data indicate that a high frequency of TRMs in the CD4+ and CD8+ T cell population is present at infection sites in patients with TB. These TRMs exhibit a TEM-like phenotype.

To comprehensively analyze the TRM immune landscape, we analyzed 28 markers for activation, trafficking, and exhaustion simultaneously by CyTOF (Supplemental Table II) (14). We determined the gating thresholds for each parameter by plotting each marker against a second marker known to express that protein (Supplemental Fig. 1A). We divided CD4+ and CD8+ T cells into four subsets based on CD69 and CD103 expression (CD69CD103, CD69CD103+, CD69+CD103, and CD69+CD103+) and then compared the frequencies of the different markers between the four subsets (Supplemental Fig. 1B). In this article, we found similar protein expression patterns presented by TRMs from different infection sites (Fig. 2A, 2B). For example, we found that CD69+CD103+ T cells presented a TEM phenotype (CD45RACCR7) (Supplemental Fig. 1C, 1D). Moreover, CCR5 and CXCR6 expression by CD69+CD103+ subsets was significantly higher in CD8+ T cells from PFMCs, BALF, and lung compared with the other three subsets (Fig. 2C, 2D).

FIGURE 2.

TRMs in infection sites express high levels of T cell activation, tracking, and exhaustion markers. Heatmaps showing the average frequency of positive expression for the different markers within the indicated cell subsets found in PFMCs (n = 6), BALF (n = 6), and lung (n = 3) for CD4+ (A) and CD8+ (B) T cells. The frequencies of the indicated different marker-expressing CD4+ (C) and CD8+ (D) T cells found at found in PFMCs, BALF, and lung [related to (A)]. One-way ANOVA with Tukey multiple comparison. *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 2.

TRMs in infection sites express high levels of T cell activation, tracking, and exhaustion markers. Heatmaps showing the average frequency of positive expression for the different markers within the indicated cell subsets found in PFMCs (n = 6), BALF (n = 6), and lung (n = 3) for CD4+ (A) and CD8+ (B) T cells. The frequencies of the indicated different marker-expressing CD4+ (C) and CD8+ (D) T cells found at found in PFMCs, BALF, and lung [related to (A)]. One-way ANOVA with Tukey multiple comparison. *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal

Some markers did, however, show tissue-type and cell-type specificity. For example, the level of CD38 and HLA-DR expression in PFMCs by CD4+ and CD8+CD69+ subsets was higher than by CD69 subsets, and their frequencies in PFMCs were higher than in the BALF and lung. In CD4+T cells, both CCR5 and CXCR6 were expressed at significantly higher levels by CD69+ subsets than CD69 subsets in PFMCs and BALF but not in the lung. These findings might, however, be due to the small lung tissue sample size (n = 3) used in our study. Notably, we detected that more CD8+CD69+ subsets from PFMCs and BALF expressed significantly higher levels of CD161 and PD-1 than CD69 subsets (Fig. 2C, 2D). Taken together, the high frequency of CD38, HLA-DR, CCR5, CXCR6, CD161, and PD-1 expression by TRMs suggests that these cells at TB infection sites are activated.

To identify M. tuberculosis Ag-specific TRMs at infection sites in patients with TB, we isolated and stimulated PFMCs with M. tuberculosis. Then, we monitored cytokine and chemokine production by the four CD69/CD103 T cell subsets (Fig. 3A). First, we found that the percentage of CD69+ cells in PFMCs was similar before and after M. tuberculosis stimulation. In CD4+ T cells, both CD69+ subsets in the PFMCs expressed high IFN-γ (∼30%), TNF-α (∼40%), IL-2 (∼15%), MIP1β (∼18%), GM-CSF (∼25%), and Tim-3 (∼20%) levels after in vitro stimulation. By contrast, the CD69CD103 subset showed low to no cytokine and chemokine expression after M. tuberculosis stimulation (Fig. 3A). We verified these results by conventional flow cytometry in an independent cohort. In this article, we consistently detected IFN-γ, TNF-α and IL-2 expression by both CD69+ subsets (Supplemental Fig. 2A). However, IL-4, IL-5, IL-6, IL-10, IL-17, and IL-22 cytokine production was very low in all four subsets, regardless of M. tuberculosis stimulation (Fig. 3A).

FIGURE 3.

TRMs from the PFMCs express multiple cytokines in response to M. tuberculosis stimulation. PFMCs from patients with TB (n = 7) were stained with a panel of intracellular markers after stimulation or not with M. tuberculosis. (A) The frequencies of the indicated markers in four CD69/CD103 subsets for CD4+ and CD8+ T cells, as assessed by CyTOF. (B) The frequencies of the indicated cytokines (IFN-γ, IL-2, and/or TNF-α expressed alone or simultaneously) after M. tuberculosis stimulation across the different CD4+ and CD8+ T cell subsets. Paired Student t test compared with unstimulated cells (“−”). *p < 0.05, **p < 0.01. (C) The proportion of CD69+CD103+ T cells in PFMCs producing one (1+, gray), two (2+, blue), or three (3+, orange) cytokines (IFN-γ, IL-2, and/or TNF-α) after M. tuberculosis simulation. (D) Monocyte-derived macrophages isolated from untreated patients with active TB (n = 5) were infected with H37Ra overnight, and autologous CD69 and CD69+ T cells were purified and cocultured. The cultured cells were then lysed for CFU. Paired Student t test compared with mock (without purified cells added to autologous macrophages). *p < 0.05.

FIGURE 3.

TRMs from the PFMCs express multiple cytokines in response to M. tuberculosis stimulation. PFMCs from patients with TB (n = 7) were stained with a panel of intracellular markers after stimulation or not with M. tuberculosis. (A) The frequencies of the indicated markers in four CD69/CD103 subsets for CD4+ and CD8+ T cells, as assessed by CyTOF. (B) The frequencies of the indicated cytokines (IFN-γ, IL-2, and/or TNF-α expressed alone or simultaneously) after M. tuberculosis stimulation across the different CD4+ and CD8+ T cell subsets. Paired Student t test compared with unstimulated cells (“−”). *p < 0.05, **p < 0.01. (C) The proportion of CD69+CD103+ T cells in PFMCs producing one (1+, gray), two (2+, blue), or three (3+, orange) cytokines (IFN-γ, IL-2, and/or TNF-α) after M. tuberculosis simulation. (D) Monocyte-derived macrophages isolated from untreated patients with active TB (n = 5) were infected with H37Ra overnight, and autologous CD69 and CD69+ T cells were purified and cocultured. The cultured cells were then lysed for CFU. Paired Student t test compared with mock (without purified cells added to autologous macrophages). *p < 0.05.

Close modal

We also found that CD8+CD69+ T cell subsets produced high cytokine levels following M. tuberculosis stimulation. The frequencies of these cytokines in CD8+ T cells were, however, lower than those in CD4+ T cells. Specifically, the frequency of IFN-γ–producing cells was 12.11% in CD8+ TRMs versus 37.27% in CD4+ TRMs (Fig. 3A). Interestingly, we detected high levels of granzyme B and perforin in both CD8+CD69+ subsets, regardless of M. tuberculosis stimulation (Fig. 3C). However, we found that only the CD4+CD69+CD103+ subset expressed granzyme B and perforin regardless of M. tuberculosis stimulation (Fig. 3A).

We also found that the frequency of CD69+CD103+ T cells was very low in CD4+ (0.27 ± 0.05%) and CD8+ (0.98 ± 0.1183%) T cells from non-TB patients (Supplemental Fig. 2B). However, the TRM cytokine profile in non-TB patients significantly differed from that in patients with TB. In particular, the frequencies of cytokine-producing cells (such as IFN-γ, TNF-α, IL-2, GM-CSF, and MIP1β) in CD69+CD103+ cells were extremely low in PFMCs from non-TB patients, regardless of M. tuberculosis stimulation. In CD4+ T cells, the four subsets in the PFMCs expressed a low level of IFN-γ (∼1.6%), TNF-α (∼2.62%), IL-2 (∼0.77%), MIP1β (∼0.7%), GM-CSF (∼4.43%), and Tim-3 (∼1.50%) in response to M. tuberculosis stimulation. Similar data were found in CD8+ T cells (Supplemental Fig. 2C). These data support that TRMs found in the pleural fluid in patients with TB exhibit an M. tuberculosis Ag-specific Th1-cell response.

Th1 cells can gain the capacity to secrete several cytokines; such “multifunctional” Th1 cells mainly produce IFN-γ, TNF-α, and IL-2 (15). The frequency of IFN-γ+TNF-α+IL-2+CD4+ T cells increases in patients with active TB and normalizes after anti-TB treatment (16, 17). Some studies performed in humans have shown that CD4+CD103+ T cells isolated from the human lung have the potential to elicit a rapid effector response to a secondary challenge and upregulate the genes required for multifunctional cytokine production (6, 18, 19). We found that in CD4+ T cells, the frequency of CD69+CD103+T cells producing two or three cytokines (IFN-γ, TNF-α, and/or IL-2) increased upon M. tuberculosis stimulation (Supplemental Fig. 2D). Specifically, we found that 13% CD69+CD103+T cells produced three cytokines compared with CD4+CD69 subsets (both were <1%) (Fig. 3B). In CD8+CD69+CD103+ T cells, we detected a low frequency (∼2.5%) of cells that produced three cytokines (Fig. 3B). We thus classified the M. tuberculosis–responsive cells in the four subsets in terms of their ability to produce three (3+), two (2+) or one (1+) cytokine(s) according to the IFN-γ, TNF-α, and IL-2 expression profile. In CD4+ T cells, the relative proportion of 3+ (IFN-γ+TNF-α+IL-2+) CD69+CD103+T cells within the total population of cytokine-producing cells was ∼28.71%; this value was much higher than the frequency of 3+ cells in the other three subsets (4.83–19.28%). Together, we classified 58.86% CD4+CD69+CD103+ T cells as 2+ and 3+ (Fig. 3C). More interestingly, we also classified∼40% of CD8+CD69+CD103+ T cells as 2+ and 3+ (Fig. 3C).

Murine studies have shown that tissue-resident T cells likely have a protective role in TB (7, 20). We thus wondered whether TRMs can provide protective immunity in TB patients. Because macrophages are the predominant cell type to control M. tuberculosis, we asked whether TRMs could limit intracellular M. tuberculosis replication in human macrophages. Because of the limited number of CD69+CD103+ cells present in the pleural fluid, we isolated memory (CD45RACCR7) CD69+ or CD69+ cells from CD4+ and CD8+ T cells in PFMCs by flow cytometry and cocultured these cells with M. tuberculosis–infected autologous macrophages. The CFUs of macrophages in coculture with CD4+CD69+ or CD8+CD69+ cells were significantly lower than those cultured alone. These data suggest that macrophages cocultured with TRMs from PFMCs showed a stronger ability to control intracellular M. tuberculosis replication. Interestingly, CD69+ cells from both CD4+ and CD8+ T cells seemed to limit intracellular M. tuberculosis growth (Fig. 3D). Thus, we conclude that CD69+ T cells in patients with active TB exhibit anti–M. tuberculosis effector function in terms of limiting intracellular M. tuberculosis replication in cultured macrophages.

This study has characterized the TRMs profile (based on CD69 and CD103 coexpression) at different sites of infection in patients with TB. Our key findings derived from patients with TB are as follows: 1) CD4+ and CD8+ TRMs are highly expanded in nonlymphoid tissues but not in the peripheral blood; 2) TRMs found at TB infection sites exhibit memory and an activated phenotype; 3) CD4+ TRMs exhibit a cytotoxic phenotype similar to CD8+ T cells; 4) M. tuberculosis–specific TRMs exhibit a multifunctional phenotype based on multiple cytokine expression; and 5) TRMs limit intracellular M. tuberculosis replication in macrophages.

We identified TRMs by CD69 and CD103 expression. Numerous other cell surface markers have been used to study TRMs in previous studies, including CD49a, CXCR6, CD101, PD-1, CD62L, KLRG1, and CX3CR1 (6, 21). Whereas CD69 is expressed by most TRMs, CD103 is primarily expressed by CD8+ T cells and not CD4+ T cells (4). Because of the wide acceptance of CD103 as a CD8+ TRM marker (5, 6, 21), we used CD103 expression to compare CD8+ TRMs with CD4+ TRMs. Others have used CD103 as an indicator of protective immunity after murine vaccination with bacillus Calmette-Guérin vaccination (22). The protective immunity elicited after a novel mucosal TB vaccination seems to be associated with the expression of CD4+ and CD8+CD103+ T cells in the lung (23). Some studies performed in the human context have found that CD4+ CD103+ T cells isolated from the human lung have the potential to elicit a rapid effector response to a secondary challenge and can upregulate the genes required for multifunctional cytokine production (6, 18, 19). Although these findings support that CD103 is a valid marker of TRMs, further studies are required to identify accurate molecular markers for TB-specific CD8+ and CD4+ TRMs.

There is a growing body of evidence supporting that TRMs have a protective role during different pathogenic infections (21, 24). Although relatively few studies have directly investigated the protective role of TRMs in TB, one early study showed that parenchymal T cells exhibit greater control of M. tuberculosis infection than their intravascular counterparts (7). In addition, the protective role of TRMs in the lung was recently established after adoptive transfer of TRMs collected from mice that received the mucosal bacillus Calmette-Guérin vaccination. Notably, the CD69+CD103+ cells expanded in the vaccinated mice and promoted a high level of IFN-γ or IFN-γ/TNF-α/IL-2–producing TRMs-mediated cell responses, which seemed to act as surrogates for protective immunity against TB (22, 23). This earlier study, however, did not fully characterize the function of TRMs in human TB.

Although more work is needed, our data strongly suggest that TRMs present in the infectious site of patients with TB have a protective role in the immune response to M. tuberculosis. At the infectious site, we found that TRMs exhibit memory, with an activated and multifunctional phenotype. CD4+ TRMs in the pleural fluid also expressed high levels of granzyme B and perforin, implicating that CD4+ TRMs exhibit a cytotoxic phenotype similar to CD8+ T cells in TB. Furthermore, both CD4+ and CD8+M. tuberculosis Ag-specific TRMs exhibited a profile reminiscent of multifunctional Th1 cells, as they simultaneously expressed IFN-γ, TNF-α, and IL-2. Finally, although we had access to only a small number of samples in this study, these TRMs showed a strong protective immune response in TB.

Because CD69 is also upregulated on recently activated T cells, we cannot definitively establish that all CD69+ cells analyzed are TRMs. Indeed, a key limitation of this study is that only tissues from patients with active TB were investigated because of the ethical and logistical challenges of performing the invasive procedures on asymptomatic latently infected individuals or the patients who have been cured from TB. It would be informative to examine in future whether the heightened susceptibility of HIV patients with low CD4+ T cell counts is associated with the reduction of the tissue-resident T cells described in this article.

Based on our preliminary data obtained from TB patients thus far, we propose that CD69 and CD103 markers can be used to identify and isolate TRMs from human tissues. Studies that transfer these cells to humanized mice might now be established to study the function of TRMs in more detail. Overall, TRMs are not detectable in peripheral blood; therefore, analyses of the immune response using peripheral blood cells might miss this crucial cellular component and an arm of the immune response against M. tuberculosis infection. Ultimately, our data highlight the complexity of TRMs in human TB and give us a more-comprehensive and updated understanding of local immunity in patients with TB.

We thank Dr. Jessica Tamanini (Shenzhen University Health Science Center, ETediting) for editing the manuscript prior to submission.

This work was supported by grants from the National Science and Technology Major Project (2017ZX10201301-001 and 2017ZX10201301-008), the Natural Science Foundation of China (81671984, 81772150 and 81571951), the Science and Technology Project of Shenzhen (JCYJ20160427184123851), and the Jin Qi team of the Sanming Project of Medicine in Shenzhen (SZSM201412001).

The online version of this article contains supplemental material.

Abbreviations used in this article:

BALF

bronchoalveolar lavage fluid

CyTOF

cytometry by time-of-flight

PFMC

pleural fluid mononuclear cell

TB

tuberculosis

TEM

effector memory T cell

TRM

tissue-resident memory T cell.

1
World Health Organization
.
2018
.
Global Tuberculosis Report 2018.
WHO Press
,
Geneva, Switzerland
.
2
Seder
,
R. A.
,
P. A.
Darrah
,
M.
Roederer
.
2008
.
T-cell quality in memory and protection: implications for vaccine design. [Published erratum appears in 2008 Nat. Rev. Immunol. 8: 486.]
Nat. Rev. Immunol.
8
:
247
258
.
3
Topham
,
D. J.
,
E. C.
Reilly
.
2018
.
Tissue-resident memory CD8+ T cells: from phenotype to function.
Front. Immunol.
9
:
515
.
4
Kumar
,
B. V.
,
W.
Ma
,
M.
Miron
,
T.
Granot
,
R. S.
Guyer
,
D. J.
Carpenter
,
T.
Senda
,
X.
Sun
,
S. H.
Ho
,
H.
Lerner
, et al
.
2017
.
Human tissue-resident memory T cells are defined by core transcriptional and functional signatures in lymphoid and mucosal sites.
Cell Rep.
20
:
2921
2934
.
5
Thome
,
J. J.
,
D. L.
Farber
.
2015
.
Emerging concepts in tissue-resident T cells: lessons from humans.
Trends Immunol.
36
:
428
435
.
6
Mueller
,
S. N.
,
L. K.
Mackay
.
2016
.
Tissue-resident memory T cells: local specialists in immune defence.
Nat. Rev. Immunol.
16
:
79
89
.
7
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
.
8
Chen
,
X.
,
Q.
Yang
,
M.
Zhang
,
M.
Graner
,
X.
Zhu
,
N.
Larmonier
,
M.
Liao
,
W.
Yu
,
Q.
Deng
,
B.
Zhou
.
2009
.
Diagnosis of active tuberculosis in China using an in-house gamma interferon enzyme-linked immunospot assay.
Clin. Vaccine Immunol.
16
:
879
884
.
9
Amir
,
E. D.
,
X. V.
Guo
,
O.
Mayovska
,
A. H.
Rahman
.
2018
.
Average overlap frequency: a simple metric to evaluate staining quality and community identification in high dimensional mass cytometry experiments.
J. Immunol. Methods
453
:
20
29
.
10
Qiu
,
Y.
,
J.
Chen
,
H.
Liao
,
Y.
Zhang
,
H.
Wang
,
S.
Li
,
Y.
Luo
,
D.
Fang
,
G.
Li
,
B.
Zhou
, et al
.
2012
.
Tim-3-expressing CD4+ and CD8+ T cells in human tuberculosis (TB) exhibit polarized effector memory phenotypes and stronger anti-TB effector functions.
PLoS Pathog.
8
: e1002984.
11
Cibrián
,
D.
,
F.
Sánchez-Madrid
.
2017
.
CD69: from activation marker to metabolic gatekeeper.
Eur. J. Immunol.
47
:
946
953
.
12
Wong
,
M. T.
,
D. E.
Ong
,
F. S.
Lim
,
K. W.
Teng
,
N.
McGovern
,
S.
Narayanan
,
W. Q.
Ho
,
D.
Cerny
,
H. K.
Tan
,
R.
Anicete
, et al
.
2016
.
A high-dimensional atlas of human T cell diversity reveals tissue-specific trafficking and cytokine signatures.
Immunity
45
:
442
456
.
13
Sallusto
,
F.
,
D.
Lenig
,
R.
Förster
,
M.
Lipp
,
A.
Lanzavecchia
.
1999
.
Two subsets of memory T lymphocytes with distinct homing potentials and effector functions.
Nature
401
:
708
712
.
14
Matos
,
T. R.
,
H.
Liu
,
J.
Ritz
.
2017
.
Research techniques made simple: mass cytometry analysis tools for decrypting the complexity of biological systems.
J. Invest. Dermatol.
137
:
e43
e51
.
15
Counoupas
,
C.
,
J. A.
Triccas
.
2019
.
The generation of T-cell memory to protect against tuberculosis.
Immunol. Cell Biol.
97
:
656
663
.
16
Qiu
,
Z.
,
M.
Zhang
,
Y.
Zhu
,
F.
Zheng
,
P.
Lu
,
H.
Liu
,
M. W.
Graner
,
B.
Zhou
,
X.
Chen
.
2012
.
Multifunctional CD4 T cell responses in patients with active tuberculosis.
Sci. Rep.
2
:
216
.
17
Caccamo
,
N.
,
G.
Guggino
,
S. A.
Joosten
,
G.
Gelsomino
,
P.
Di Carlo
,
L.
Titone
,
D.
Galati
,
M.
Bocchino
,
A.
Matarese
,
A.
Salerno
, et al
.
2010
.
Multifunctional CD4(+) T cells correlate with active Mycobacterium tuberculosis infection.
Eur. J. Immunol.
40
:
2211
2220
.
18
Oja
,
A. E.
,
B.
Piet
,
C.
Helbig
,
R.
Stark
,
D.
van der Zwan
,
H.
Blaauwgeers
,
E. B. M.
Remmerswaal
,
D.
Amsen
,
R. E.
Jonkers
,
P. D.
Moerland
, et al
.
2018
.
Trigger-happy resident memory CD4+ T cells inhabit the human lungs.
Mucosal Immunol.
11
:
654
667
.
19
Wilk
,
M. M.
,
A.
Misiak
,
R. M.
McManus
,
A. C.
Allen
,
M. A.
Lynch
,
K. H. G.
Mills
.
2017
.
Lung CD4 tissue-resident memory T cells mediate adaptive immunity induced by previous infection of mice with Bordetella pertussis.
J. Immunol.
199
:
233
243
.
20
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
.
21
Ogongo
,
P.
,
J. Z.
Porterfield
,
A.
Leslie
.
2019
.
Lung tissue resident memory T-cells in the immune response to Mycobacterium tuberculosis.
Front. Immunol.
10
:
992
.
22
Perdomo
,
C.
,
U.
Zedler
,
A. A.
Kühl
,
L.
Lozza
,
P.
Saikali
,
L. E.
Sander
,
A.
Vogelzang
,
S. H.
Kaufmann
,
A.
Kupz
.
2016
.
Mucosal BCG vaccination induces protective lung-resident memory T cell populations against tuberculosis.
MBio
7
: e01686-16.
23
Hart
,
P.
,
A.
Copland
,
G. R.
Diogo
,
S.
Harris
,
R.
Spallek
,
W.
Oehlmann
,
M.
Singh
,
J.
Basile
,
M.
Rottenberg
,
M. J.
Paul
,
R.
Reljic
.
2018
.
Nanoparticle-fusion protein complexes protect against Mycobacterium tuberculosis infection.
Mol. Ther.
26
:
822
833
.
24
Snyder
,
M. E.
,
D. L.
Farber
.
2019
.
Human lung tissue resident memory T cells in health and disease.
Curr. Opin. Immunol.
59
:
101
108
.

The authors have no financial conflicts of interest.

This article is distributed under The American Association of Immunologists, Inc., Reuse Terms and Conditions for Author Choice articles.

Supplementary data