NKG2D-Dependent IL-17 Production by Human T Cells in Response to an Intracellular Pathogen¹

The T cells play a vital role in the elimination of various intracellular pathogens, including Mycobacterium tuberculosis, through their effector mechanisms of cytotoxicity and cytokine production. Among various cytokines, IFN-γ is indispensable for resistance to M. tuberculosis in animal models (1, 2), and genetic defects in IFN-γ receptor expression, or in the IL-12 pathway that controls IFN-γ production, result in markedly increased susceptibility to mycobacterial disease (3). Although IFN-γ is essential for host defenses against M. tuberculosis, it is not sufficient, because T cells at the site of disease produce high concentrations of this cytokine. To develop an effective vaccine against M. tuberculosis, it is important to understand factors other than IFN-γ that contribute to antimycobacterial immunity.

IL-17 is a proinflammatory cytokine produced by T cells that plays an important role in the development of autoimmunity (4), enhances graft rejection (5), and protects the host against intracellular pathogens (6, 7). IL-17 induces chemokine production that recruits IFN-γ-producing T cells that inhibit bacterial growth after vaccination in mice infected with M. tuberculosis (6). The frequency of IL-17-producing γδ T cells increases in tuberculosis patients (8), and γδ T cells are the major source for IL-17 in mice infected with M. tuberculosis (9). In contrast, in healthy tuberculin reactors, CD4⁺ central memory cells are the major source of IL-17 (10), so that the major cellular subpopulations that produce IL-17 in response to M. tuberculosis are not well defined. The mechanisms that induce IL-17 production in human M. tuberculosis infection are also unknown.

NKG2D is a surface receptor that is expressed predominantly by NK cells, but also acts as a costimulatory molecule on CD8⁺ cells (11). Previously, we found that NKG2D expressed on CD3⁻CD56⁺ cells is involved in the lysis of M. tuberculosis-infected monocytes and M. tuberculosis-expanded regulatory T cells (12, 13). NKG2D expression on CD4⁺ cells has been described primarily in patients with rheumatoid arthritis and Crohn’s disease (14, 15), and CD4⁺NKG2D⁺ T cells mediate inflammatory responses (14). Interactions between CD40 on APCs and CD40L on T cells can enhance production of inflammatory cytokines (16), and could modulate IL-17 production. In this study, we evaluated the roles of NKG2D and CD40 in regulating IL-17 production in human M. tuberculosis infection. We found that NKG2D expressed on CD4⁺ cells contributes to IL-23-dependent IL-17 production by CD4⁺ cells in human M. tuberculosis infection.

After obtaining informed consent, blood was obtained from 20 healthy tuberculin reactors (persons with a tuberculin skin test diameter of at least 10 mm) and 13 tuberculin-negative donors, and bronchoalveolar lavage fluid was obtained from 5 healthy tuberculin-negative donors. All studies were approved by the Institutional Review Board of the University of Texas Health Center at Tyler.

For flow cytometry, we used FITC anti-CD4, FITC anti-CD14, PE anti-CD62L (all from eBiosciences), and allophycocyanin anti-NKG2D (Biolegend). For neutralization, we used mAbs to IL-1 and IL-6 (both from eBiosciences); CD40, IL-23, and TGF-β (all from R&D Systems); NKG2D (Amgen); and isotype control Abs (eBiosciences). γ-Irradiated M. tuberculosis H37Rv was obtained from J. Belisle, Colorado State University (Fort Collins, CO).

PBMC were isolated by centrifugation over Ficoll-Paque (GE Health). Monocytes and CD4⁺ and CD8⁺ cells were isolated by magnetic beads conjugated to anti-CD14, anti-CD4, or anti-CD8 (all from Miltenyi Biotec), respectively. Cell purity was >95%⁺, as measured by flow cytometry.

Bronchoalveolar lavage was performed by standard methods, and alveolar macrophages were isolated and infected with M. tuberculosis H37Ra, as previously described (17). Cells (1.2 × 10⁶) were adhered to each well of 12-well flat-bottom plates for 2 h. Nonadherent cells were removed by washing each well three times with 4 ml of RPMI 1640, and adherent cells were cultured in RPMI 1640 complete medium.

Freshly isolated CD4⁺ cells or CD8⁺ cells were cultured in 12-well plates at 2 × 10⁶ cells/well in RPMI 1640 containing penicillin (Life Technologies) and 10% heat-inactivated human serum, with or without 2 × 10⁵ autologous monocytes/well. CD4⁺ cells and monocytes were cultured in the presence or absence of γ-irradiated M. tuberculosis H37Rv (10 μg/ml) at 37°C in a humidified 5% CO₂ atmosphere. After 3 days, cell-free culture supernatants were collected, aliquoted, and stored at −70°C until cytokine concentrations were measured. For some experiments, 10 μg/ml neutralizing Abs to IL-1, IL-6, IL-23, or TGF-β were added.

CD4⁺ cells and autologous monocytes were cultured in the presence or absence of γ-irradiated M. tuberculosis H37Rv, and after 48 h, CD4⁺CD62L⁺ and CD4⁺CD62L⁻ cells were isolated, using positive selection with CD62L microbeads. Aliquots of CD4⁺CD62L⁺ and CD4⁺CD62L⁻ cells were placed on ELISPOT plates, and the number of cells that produce IL-17 was detected (e-Biosciences). In some experiments, monocytes and either freshly isolated CD4⁺CD62L⁺ or CD4⁺CD62L⁻ cells were cultured in the presence or absence of γ-irradiated M. tuberculosis H37Rv. After 48 h, cells were placed on ELISPOT plates, and the number of cells that produce IL-17 was detected.

To measure intracellular IL-17 in cocultures of CD4⁺ cells and γ-irradiated M. tuberculosis H37Rv-stimulated monocytes, we used the intracellular staining kit from eBiosciences. For triple staining, allophycocyanin anti-NKG2D and FITC anti-CD4 were first added. After washing with PBS and 2% FCS, cells were fixed in Cytofix/Cytoperm and washed twice in 1× permeabilization/wash solution. PE anti-IL-17 was then added in staining buffer and 1× permeabilization/wash solution. After incubation at 4°C for 40 min, cells were washed in PBS and 2% FCS, and analyzed by flow cytometry.

Supernatants were stored at −70°C until cytokine concentrations were measured. IL-23, IL-17, and IFN-γ levels were measured by ELISA (eBiosciences).

Results are shown as the mean ± SE. For data that were normally distributed, comparisons between groups were performed by Student’s paired or unpaired t test, as appropriate. For data that were not normally distributed, the nonparametric Mann-Whitney U test was performed.

IL-17 is generally considered to be produced by CD4⁺ cells, and one recent study showed that IL-17 is produced by CD4⁺ cells, but not CD8⁺ or γδ T cells that are stimulated with Mycobacterium bovis bacillus Calmette-Guerin (10). However, CD8⁺ cells also produce IL-17 in response to Klebsiella pneumoniae (18). To determine the cellular source of IL-17 in response to M. tuberculosis, we isolated CD4⁺ and CD8⁺ cells from PBMC of three healthy tuberculin reactors, and cultured them in the presence or absence of irradiated M. tuberculosis for 24–96 h. M. tuberculosis induced IL-17 production by CD4⁺ cells at all time points, with maximum concentrations 72 h after stimulation (Fig. 1 A). In contrast, M. tuberculosis did not elicit IL-17 production by CD8⁺ cells.

To determine whether IL-17 production depends on prior Ag exposure, we isolated CD4⁺ cells from PBMC of 15 tuberculin reactors and 13 tuberculin-negative persons, and cultured them in the presence or absence of M. tuberculosis for 72 h. IL-17 production by M. tuberculosis-stimulated CD4⁺ cells from healthy tuberculin reactors increased markedly from 35 ± 15 to 887 ± 195 pg/ml (p < 0.001, Mann-Whitney U test; Fig. 1,B). In contrast, M. tuberculosis stimulation did not elicit IL-17 production by CD4⁺ cells from 13 tuberculin-negative donors (4.7 ± 0.8 vs 87 ± 51 pg/ml, p = 0.13, Mann-Whitney U test; Fig. 1 B).

The above findings suggest that CD4⁺ cells are the major source of M. tuberculosis-induced IL-17. To identify the CD4⁺ cell subpopulation that produces IL-17, we cultured CD4⁺ cells from PBMC of 10 healthy tuberculin reactors with autologous monocytes, in the presence or absence of M. tuberculosis. After 48 h, CD62L⁺ and CD62L⁻ cells were isolated by magnetic selection, and ELISPOT was performed to identify IL-17-producing cells. M. tuberculosis increased the frequency of IL-17⁺ cells from 7.2 ± 2.5 per 10⁶ to 114 ± 23 per 10⁶ CD4⁺CD62L⁻ cells, compared with 8.9 ± 2.3 per 10⁶ CD4⁺CD62L⁺ cells (p < 0.001; Fig. 1 C), indicating that Ag-experienced, CD4⁺CD62L⁻ cells are the major source for IL-17.

Because some CD62L⁺ cells can lose CD62L expression during culture, we also determined the capacity of freshly isolated CD4⁺CD62L⁺ and CD4⁺CD62L⁻ cells to produce IL-17 upon culturing with M. tuberculosis. Monocytes from six healthy tuberculin reactors were cultured with freshly isolated CD4⁺CD62L⁺ or CD4⁺CD62L⁻ cells, in the presence or absence of γ-irradiated M. tuberculosis H37Rv. After 48 h, cells were placed on ELISPOT plates, and the number of cells that produce IL-17 was detected. M. tuberculosis increased the frequency of IL-17⁺ cells from 0 to 86 ± 28 per 10⁶ CD4⁺CD62L⁻ cells, compared with 38 ± 12 per 10⁶ CD4⁺CD62L⁺ cells (p = 0.05).

Costimulatory molecules play an important role in eliciting cytokine production by T cells. Therefore, we studied the role of the costimulatory molecules, NKG2D and CD40, in IL-17 production. We cultured CD4⁺ cells and autologous monocytes from PBMC of healthy tuberculin reactors with M. tuberculosis, in the presence of anti-NKG2D or anti-CD40 or isotype control Ab (10 μg/ml). In eight healthy tuberculin reactors, anti-NKG2D reduced IL-17 levels by ∼60% (1016 ± 284 vs 423 ± 186 pg/ml, p < 0.001, Mann-Whitney U test; Fig. 2,A), compared with isotype control Ab (Fig. 2 A). In seven healthy tuberculin reactors, anti-CD40 Ab had no effect on IL-17 production (398 ± 60 vs 403 ± 58 pg/ml).

We next asked whether NKG2D regulates IL-17 production by Ag-experienced CD4⁺ CD62L⁻ cells. We cultured CD4⁺ cells and autologous monocytes from PBMC of healthy tuberculin reactors with M. tuberculosis, in the presence of anti-NKG2D or isotype control Ab (10 μg/ml). In eight healthy tuberculin reactors, anti-NKG2D reduced the frequency of CD4⁺CD62⁻ IL-17⁺ cells from 128 ± 26 per 10⁶ cells to 32 ± 8 per 10⁶ cells (p = 0.006; Fig. 2 B), whereas in four healthy tuberculin reactors, isotype control Ab had no effect (81 ± 20 per 10⁶ cells vs 59 ± 18 per 10⁶ cells, p = 0.19).

The above findings suggest that NKG2D expressed on CD4⁺ cells is essential for optimal IL-17 production. To determine whether CD4⁺NKG2D⁺ cells are the source of IL-17, we cultured CD4⁺ cells from PBMC of six healthy tuberculin reactors with autologous monocytes, in the presence or absence of M. tuberculosis. After 48 h, cells were stained for CD4, NKG2D, and IL-17. M. tuberculosis increased the number of CD4⁺NKG2D⁺ cells from 1.7 ± 0.3% to 3.7 ± 0.7% (p = 0.02). Similarly, M. tuberculosis increased the number of CD4⁺IL-17⁺ cells from 0.5 ± 0.3% to 1.2 ± 0.3% (p < 0.001). As shown in the left panels of Fig. 2,C, IL-17-producing CD4⁺ cells are not NKG2D⁺ cells. Expression of NKG2D was not due to contaminating NK cells in the CD4⁺ cell preparations, which only contained 0.6% CD56⁺ cells (Fig. 2 C, top right panel).

Previously, we found that among the different NKG2D ligands, only ULBP1 expression was up-regulated on M. tuberculosis-infected monocytes (12). To determine whether NKG2D and ULBP1 interaction is responsible for IL-17 production by CD4⁺ cells, we cultured CD4⁺ cells and autologous monocytes from PBMC of healthy tuberculin reactors with M. tuberculosis, in the presence of anti-ULBP1 or isotype control Ab (10 μg/ml). In six healthy tuberculin reactors, anti-ULBP1 significantly inhibited IL-17 production by CD4⁺ cells from 589 ± 157 to 145 ± 33 pg/ml (p = 0.03; Fig. 2,D), and isotype control Ab had no effect (Fig. 2 D).

Previous studies have demonstrated that the initial development of Th17 cells from naive T cells requires TGF-β and IL-6, produced by APCs, and that subsequent expansion of Th17 cells depends on IL-23 (19, 20). To determine whether specific monokines contribute to M. tuberculosis-induced IL-17 production, we cultured M. tuberculosis-activated monocytes from healthy tuberculin reactors with autologous CD4⁺ cells, in the presence or absence of neutralizing Abs to IL-23, TGF-β, IL-1, and IL-6. In 10 healthy tuberculin reactors, anti-IL-23 significantly abrogated IL-17 production by CD4 cells (1041 ± 149 to 280 ± 89 pg/ml, p < 0.001; Fig. 3,A), whereas Abs to TGF-β, IL-1, IL-6, and isotype control IgG2 had no effect (Fig. 3 A).

Our results above suggest that IL-23 contributes to M. tuberculosis-induced IL-17 production. Because APCs are believed to be the major cellular source of IL-23 (21, 22), we isolated CD14⁺ monocytes from PBMC of 14 healthy tuberculin reactors and cultured them in the presence or absence of M. tuberculosis for 48 h. IL-23 concentrations increased 3-fold upon M. tuberculosis stimulation (23.8 ± 6.5 vs 62.7 ± 12 pg/ml, p = 0.01; Fig. 3 B).

Because M. tuberculosis first encounters the immune system in the lung, we next evaluated the capacity of alveolar macrophages from five healthy donors to secrete IL-23 in response to M. tuberculosis. Infection with M. tuberculosis H37Ra significantly up-regulated IL-23 production (6.4 ± 3.7 vs 32.7 ± 7.9 pg/ml, p = 0.01; Fig. 3 B).

We hypothesized that NKG2D, expressed by activated CD4⁺ cells, increased production of IL-23 by M. tuberculosis-stimulated monocytes, which in turn increased IL-17 production by CD4⁺ T cells. To test this hypothesis, we cultured M. tuberculosis-stimulated monocytes with autologous CD4⁺ cells from healthy tuberculin reactors. CD4⁺ cells enhanced IL-23 production by M. tuberculosis-stimulated monocytes, and this effect was markedly reduced by anti-NKG2D (451 ± 69 pg/ml vs 204 ± 70 pg/ml, p = 0.03; Fig. 4). Isotype control Abs did not affect IL-23 production (Fig. 4).

Recent studies in animal models have demonstrated that IL-17 is essential for development of a protective response to vaccination against M. tuberculosis (6), but limited information is available on the mechanisms that control IL-17 production in the human response to M. tuberculosis infection. In the current study, we found that Ag-experienced CD4⁺ cells are the major source for IL-17, and that IL-23, but not TGF-β or IL-6, contributes to IL-17 production. Furthermore, addition of CD4⁺ cells to M. tuberculosis-stimulated monocytes increased IL-23 production, and this was mediated through NKG2D. Anti-NKG2D and anti-ULBP1 inhibited IL-17 production, suggesting that NKG2D on CD4⁺ cells interacts with ULBP1 on monocytes to induce IL-23-dependent IL-17 production in human M. tuberculosis infection.

IL-17 is an inflammatory T cell cytokine that causes autoimmune disease (4), but can also contribute to resistance against intracellular pathogens, including viruses, fungi, parasites, and mycobacteria (6, 8, 9, 23, 24, 25). IL-17 is not essential for resistance to M. tuberculosis (26), but is critical for vaccine-induced protection against challenge with M. tuberculosis (6). γδ T cells are the major source for IL-17 in murine infection with several intracellular pathogens, including M. tuberculosis (9, 27, 28). However, we found that Ag-experienced CD4⁺CD62L⁻ cells are the major source for IL-17 in human M. tuberculosis infection, confirming recent findings that IL-17 is produced by CD4⁺, but not CD8⁺ or γδ T cells, from persons with latent tuberculosis infection (10).

Differentiation of naive T cells to Th17 cells requires TGF-β and IL-6, but expansion of Th17 cells depends on IL-23 (19, 29). In animal models, IL-23 is essential for IL-17 production during infection by Helicobacter pylori, Pneumocystis carinii, and K. pneumoniae (18, 30, 31). However, for mycobacterial infection, one study showed that IL-23 was required for the IL-17 response to tuberculosis (26), but another demonstrated that IL-23 was dispensable for IL-17 production during infection with M. bovis bacillus Calmette-Guerin (32). We found that IL-23, but not IL-6 or TGF-β, is required for T cells from persons with M. tuberculosis infection to produce IL-17 in response to mycobacterial Ags. IL-23 expands CD4⁺ memory cells to IL-17-producing cells (33), and our results suggest that Th17 cells are present in individuals with latent infection, and are expanded by IL-23. Although we found no role for IL-1, IL-6, and TGF-β in IL-17 production, similar to findings of others (34, 35), our experimental system evaluated Ag-experienced cells, and it is possible that IL-1, IL-6, and TGF-β are important for differentiation of Th17 cells from naive cells.

Our most surprising finding was that NKG2D enhanced production of IL-23 by APCs, and contributed to IL-17 production by CD4⁺ cells in response to M. tuberculosis. NKG2D expression on CD4⁺ cells has been described primarily in patients with rheumatoid arthritis and Crohn’s disease (14, 15). CMV can also expand NKG2D⁺CD4⁺ cells from CMV-seropositive individuals (36). In these settings, TNF-α and IL-15 induced expression of NKG2D (14, 15), which appeared to act as an alternative costimulatory molecule, because NKG2D⁺ cells were CD28⁻ (15, 36).

We found that NKG2D expression on CD4⁺ cells was enhanced by exposure to monocytes and M. tuberculosis, and that production of both IL-23 and IL-17 was inhibited by neutralization of either NKG2D or its ligand on APCs, ULBP1. We have previously shown that, of the multiple ligands for NKG2D, only ULBP1 is up-regulated on the surface of mononuclear phagocytes during M. tuberculosis infection (12). During the early stages of mycobacterial infection, mononuclear phagocytes produce soluble mediators, including IL-15 (37), TNF-α (38), and IL-23 (the current study). We speculate that IL-15 and TNF-α enhance expression of NKG2D on CD4⁺ cells, which interacts with ULBP1 to increase IL-23 production (Fig. 4), and thus IL-17 production by CD4⁺CD62L⁻ cells. IL-23 production was significantly increased by the addition of only 4 T cells per 10 monocytes (Fig. 4), a ratio that may mimic the limited number of activated T cells that are present in the lung during the early stages of infection.

In summary, we found that NKG2D expressed on CD4⁺ cells interacts with ULBP1 on M. tuberculosis-infected mononuclear phagocytes to enhance IL-23 secretion by mononuclear phagocytes, which in turn favors IL-17 production by CD4⁺CD62L⁻ cells. This suggests a novel role for NKG2D in protection against intracellular bacteria such as M. tuberculosis.

The authors have no financial conflict of interest.