Lower IL-7 Receptor Expression of Monocytes Impairs Antimycobacterial Effector Functions in Patients with Tuberculosis

Tuberculosis (TB) is a chronic mycobacterial disease caused by bacteria of the Mycobacterium tuberculosis complex. Host response against M. tuberculosis infection is characterized by manifold immune mechanisms leading to asymptomatic latent infection in the vast majority of individuals. Effector cells of the adaptive immunity, especially T cells, and also innate immune cells contribute to immune surveillance that is crucial for protection against TB disease progression (1). A well-balanced immune response is associated with latent M. tuberculosis infection, whereas aberrant features characterize immune pathology of acute TB (2). Previous studies showed that immune pathognomonic features of TB (e.g., changes of lymphocyte and monocyte proportions) can be detected in the peripheral blood of patients with TB (2, 3).

Several studies investigated altered subsets and functions of monocytes in TB (reviewed in Ref. 4). Peripheral blood monocytes are a heterogeneous population that forms the reservoir for distinct phagocyte lineages, like macrophages and dendritic cell subsets. The majority of studies distinguished subsets of monocytes by the expression of CD14 and CD16 on MHC class II–positive leukocytes. CD14⁺/CD16^– cells were termed “classical” (M1) monocytes and CD16⁺/CD14^– cells were termed “alternative” (M2) monocytes. A third subset of monocytes that expresses CD14 and CD16 concomitantly (termed intermediate or M1/2) was previously found to be enriched in patients with acute TB (5–8). M1/2 cells were also termed “inflammatory” monocytes and differed functionally from the other subsets of monocytes (9, 10). In TB, higher M1/2 proportions were hypothesized to impair monocyte proliferation, circulation, and dendritic cell maturation (11–15). How changes in monocyte phenotype and function contribute to TB immunopathology remains hardly defined. Besides the mentioned effects on monocyte proliferation, differentiation, and survival, interaction of derived macrophages and dendritic cells with T cells is potentially affected (16). In this regard, the programmed cell death (PD) pathway (mediated by PD-1/PD-L1 interaction) has been found to affect T cell functions in TB (17, 18). The relevance of this pathway has recently been highlighted by the fact that immune modulation targeting the check point inhibitor PD-1 in cancer patients led to reactivation of latent M. tuberculosis infection (19–22).

IL-7 is an important homeostatic cytokine during lymphocyte development and maturation. Especially T cells depend on IL-7 availability for proliferation, survival, and memory generation. In this regard, fine-balanced IL-7Rα expression is crucial to avoid immune pathology (e.g., in autoimmune diseases or transplantation) (23). An important role of IL-7 for host response against M. tuberculosis was suggested by Maeurer et al., who showed increased protection of M. tuberculosis–infected mice when treated with recombinant IL-7 (24). In accordance, we and others showed IL-7 promoting effects on M. tuberculosis–specific T cell cytokine expression in individuals with latent M. tuberculosis infection (25, 26). Our own previous studies indicated a role of IL-7 serum levels and IL-7Rα expression in TB (26, 27). Aberrant low-soluble IL-7Rα serum concentrations and decreased T cell membrane IL-7Rα expression was found in acute patients with TB (26). Concomitant high IL-7 concentrations in patients suggested lower IL-7 sensitivity and impaired IL-7 promoted T cell functions in acute TB (26).

Only sparse information about IL-7 effects on monocytes is available. Studies from the last century described antimicrobial and antitumor cytotoxic effects of monocytes/macrophages treated with IL-7 (28, 29). One study described IL-7–mediated killing effects against monocyte-derived macrophages (MDMs) infected with Mycobacterium avium (30). The biological relevance of these finding, however, remains elusive because IL-7Rα expression on monocytes is low and indirect effects [e.g., by IL-7 sensitive T cells (31)] may bias results. Recently, Al-Mossawi et. al. (32) demonstrated LPS-induced IL-7Rα upregulation on monocytes. They described rapid IL-7Rα upregulation on monocytes in vitro and TNF-α was identified as a crucial factor (32). IL-7 effects on the PD-1/PD-L1 pathway have been demonstrated in animal models of type 1 diabetes (33, 34). In these studies, IL-7 induced decreased PD-1 expression of effector T cells, and IL-7Rα blocking Abs enhanced PD-1 expression and reverted established T1D disease (33, 34).

In the current study, we investigated the role of IL-7Rα expression and IL-7–mediated effects in monocyte subpopulations from patients with TB, asymptomatic contacts of indexed TB patients (termed “contacts” throughout), and healthy controls. IL-7–mediated effects on monocytes were further characterized in vitro by cytokine receptor signaling and by using a flow cytometry–based antimycobacterial kill assay.

We recruited patients with TB (n = 33), asymptomatic contacts of indexed TB patients (contacts; n = 30), and healthy controls (n = 20) from September 2019 to November 2019 in Ghana. TB patients with HIV coinfection (n = 2) were excluded from the study. Age and gender characteristics of study groups are summarized in Table I. TB patients and contacts were recruited at the Agogo Presbyterian Hospital, the St. Mathias Catholic Hospital, the Atebubu District Hospital, and the Sene West District Hospital. Diagnosis for active TB was based on patient history, clinical evaluation, chest x-ray, and sputum smear test. GeneXpert (Cepheid) analyses were done for all TB patients. TB patients were included prior to initiation of treatment, and chemotherapy was initiated immediately thereafter. Contacts showed no symptoms of TB but were close relatives living in the same household with indexed TB patients according to self-report and direct observation. We showed recently that limited sensitivity of immune tests for detection of M. tuberculosis infection constrains classification of contacts as latently M. tuberculosis–infected individuals in Ghana (35–37). In this study, we found that 18 of 30 contacts (60%) were positive according to in vitro tests for M. tuberculosis Ags (i.e., ESAT6/CFP10) (35). This was only moderately lower as compared with confirmed TB patients (70% positive). Our own previous studies suggested a high likelihood of asymptomatic M. tuberculosis infection for contacts on the basis of the applied criteria (35, 36).

Healthy controls were recruited from the Kumasi Center of Collaborative Research for Tropical Medicine staff. High prevalence of TB in Ghana and bacillus Calmette–Guérin (BCG) vaccination at birth render identification of community controls negative for M. tuberculosis–specific immune responses difficult. Hence, we included staff members without recent contact to TB patients according to self-reports. Notably, 17 of 20 community controls (85%) were positive for M. tuberculosis Ag-specific immune tests, again demonstrating limited capacity of immune tests to distinguish between recent and remote infection.

The present study received approval from the Committee on Human Research, Publication and Ethics (CHRPE/AP/023/18) at the School of Medical Sciences at the Kwame Nkrumah University of Science and Technology in Kumasi, Ghana. All study subjects gave written informed consent prior to recruitment. For in vitro monocyte and MDM experiments, buffy coat cells retrieved from the transfusion medicine department at the Heinrich-Heine-University in Duesseldorf were used. The local ethics committee approved this study (ID: 5445).

Heparinized whole blood (10 ml) was collected from all study participants. PBMCs were purified from blood and buffy coats by density gradient centrifugation (Histopaque-1077; Sigma-Aldrich) following manufacturers’ guidelines. PBMCs not used for ex vivo phenotypic analyses were cryopreserved in FBS containing 10% DMSO. For buffy coats, monocytes were isolated from PBMC (1 × 10⁸) by immunomagnetic noncontact enrichment (EasySept Monocytes Negative Selection Kit; StemCell Technology). The purity of monocytes after enrichment was determined by flow cytometry using HLA-DR positivity as a marker and ranged from 82 to 92%.

PBMCs (2 × 10⁵) from all study subjects were directly stained with fluorescence-labeled Abs against human PD-L1 (BV421, clone MIH3; BioLegend), CD14 (BV650, clone M5E2; BioLegend), CD16 (allophycocyanin, clone 3G8; BioLegend), HLA-DR (BV750, clone L243; BioLegend), CD11b (PE-Cy7, clone ICRF; BioLegend), IL-7Rα (synonymous: CD127) (PE, clone A019D5; BioLegend) and viability dye (eFluor780; eBiosciences) as described before (38). In brief, PBMCs were incubated with 10% FCS/PBS on ice for 10 min and were stained thereafter for 30 min on ice in the dark. Isotype controls for CD127 (PE, IgG1κ, BioLegend) and PD-L1 (Bv421, IgG1κ, BioLegend) and fluorescence minus one controls for monocyte-staining panel were included. After the staining, cells were washed and measured using a CytoFlex S (Beckman Coulter) flow cytometer. At least 9 × 10⁴ live cells were recorded. FlowJo software (Version 10, FlowJo LLC) was used to analyze the data.

A representative example of the monocyte gating procedure is provided as Fig. 1B. After excluding cell debris, viability dye–positive, and HLA-DR–negative cells, CD14/CD16 expression was used to categorize CD14⁺/CD16^– (M1), CD14⁺/CD16⁺ (M1/2), and CD14^–/CD16⁺ (M2) monocytes. CD14/CD16–double-negative cells were excluded from further analyses.

Cryopreserved PBMC were thawed and washed in X-vivo15 medium (Lonza) supplemented with L-glutamine (2 mM) and penicillin/streptomycin (50 U/ml) at a concentration of 2.5 × 10⁵ cells per well. STAT phosphorylation was performed as described before (39). In brief, prewarmed medium with or without human recombinant IL-7 (10 ng/ml) was added and cells were incubated at 37°C for 15 min. Thereafter cells were fixed in 100 µl of true nuclear fixation buffer (BioLegend) for 15 min, permeabilized in ice-cold, pure methanol (130 µl), and washed before staining with Abs against human pSTAT5 (PE, clone: SRBCZX; eBioscience), pSTAT1 (AF647, clone A17012A), CD14, CD16 (all BioLegend) for 30 min on ice. Measurement was performed using an LSR-Fortessa Flow Cytometer (BD Bioscience) and analysis was done with FlowJo software as described above.

Monocytes and MDMs were used for these experiments. For generation of MDM, monocytes were incubated in macrophage-complete medium (RPMI 1640 containing FCS [10%], HEPES [1%], and L-glutamine [1%]) for 4 d (37°C, 5% CO₂). Monocytes or MDMs (5 × 10⁴ cells per well) were incubated for 24 h in complete medium (total volume, 200 μl). The following stimuli were added: BCG [multiplicity of infection 0.5, viable or dead; preparation as described (40)], LPS, and mycobacterial components derived from the M. tuberculosis H37Rv strain (i.e., mycobacterial soluble cell wall proteins, lipoarabinomannan [LAM] and lipomannan [LM]; all provided by Biodefense and Emerging Infections Research Resources) using indicated concentrations. Afterwards the supernatants were removed, and monocytes/MDM samples were incubated in PBS containing 10 mM EDTA and 0.5 BSA on ice for 15 min to detach adherent cells. Ab staining for HLA-DR, IL-7Rα, and viability dye was done as described above including an isotype control for the IL-7Rα. Measurement was performed using an LSR-Fortessa Flow Cytometer (BD Bioscience) and analysis was done with FlowJo software.

A live/dead (LD) reporter plasmid containing BCG reporter strain was used to analyze mycobacterial infection of MDMs and viability as described before (40). In brief, MDM were incubated in the presence or absence of IL-7 for 4 d. MDM were then infected with log phase grown LD-BCG (OD_600nm between 0.8 and 0.9) treated with anhydrotetracycline (ATC; Sigma-Aldrich) for induction of GFP in viable mycobacteria at a multiplicity of infection of 1:1 for 4 h (37°C, 5% CO₂). Non-ATC–treated LD-BCG were used as a negative control. IL-7 (10 ng/ml) was also added to the LD-BCG containing medium for samples treated with IL-7 on day 0. Following infection, extracellular bacteria were removed by washing and prewarmed complete medium for incubation of 48 h. Thereafter MDM were detached as described above and either used for CFU (see below) or flow cytometry–based analyses. For flow cytometry, MDM were stained with Abs against CD11b and the viability dye for 30 min on ice in the dark. MDM were then fixed (Perm/Fix; BioLegend) according to manufacturers’ guidelines. Before measurement, count beads (1 × 10⁴, 123-count eBeads; Biosciences) were added to each sample to normalize acquired samples for MDM counts. Measurement was performed using an LSR-Fortessa Flow Cytometer and analysis was done with FlowJo software.

Proportions and absolute numbers of live or dead BCG-infected MDMs were deduced from flow cytometric analyses and stringent gating strategies were applied using the uninfected and ATC-treated MDMs. Equal sample volumes were acquired by adjusting for count beads. This allows comparison of absolute MDM numbers from culture to estimate potential experimental effects on host cell viability (as described in Ref. 40). Gating procedures of representative samples are provided as Fig. 5A.

For CFU, PBS containing 0.5% Tween80 (120 μl, Sigma-Aldrich) was added to lyse LD-BCG–infected MDMs and incubated for 30 min on ice. Three dilution steps in PBS (1:10, 1:100, 1:1000) were prepared and inoculated on 7H11 agar plates supplemented with Middelbrook OADC enrichment (BD Biosciences) and hygromycin (50 μg/ml; Gentaur Biotech). LD-BCG colonies were counted after incubation (37°C) for 3 wk.

All statistical analyses were performed using GraphPad Prism v8 software (GraphPad Software, La Jolla, CA). Study group comparisons were performed by the nonparametric Mann–Whitney U test. The Wilcoxon sign rank test was used to test for IL-7–specific effects on STAT5 phosphorylation. Pearson correlation coefficient was applied to evaluate the relation between individual IL-7Rα and PD-L1 expression of monocyte subsets. The Student paired t test was applied for IL-7Rα induction by mycobacteria and components in monocytes and MDMs. A p value < 0.05 was considered statistically significant.

Previous studies found aberrant distributions of monocyte subpopulations in patients with TB (16). Initially we compared proportions of HLA-DR positive monocytic cells in the blood of patients with TB, asymptomatic contacts of indexed TB patients (termed “contacts” throughout) and healthy control donors from Ghana (Table I). Patients with TB had generally higher proportions of monocytes as compared with contacts and control donors (Fig. 1A, left graph). Moderate differences were detected for CD14 or CD16 expressing monocyte proportions and only contacts had higher proportions of CD16 positive monocytes as compared with controls (Fig. 1A, right graphs). To further characterize differences, monocyte subpopulations were classified according to expression of CD14 and CD16 into CD14⁺/CD16⁻ monocytes (M1), CD14⁻/CD16⁺ monocytes (M2), and CD14⁺/CD16⁺ monocytes (M1/2) (for gating procedure see Fig. 1B). Monocyte subset classification revealed significantly lower proportions of M1 in patients with TB and contacts as compared with controls (Fig. 1C). In contrast, proportions of M1/2 were higher in TB patients and contacts as compared with controls (Fig. 1C). Contacts differed from patients with TB by increased M2 proportions as well as significantly lower M1/2 proportions (Fig. 1C).

Differences in the subsets of monocytes from patients with TB was previously found to be accompanied by changes in monocyte phenotype. Aberrant monocyte PD-L1 expression has been described for TB (8) and we detected markedly higher PD-L1 expression especially in monocytes from patients with TB (Fig. 1D). These differences were most evident for M1/2 and M2 subsets as compared with contacts and controls (Fig. 1D). No differences for PD-L1 expression were seen between contacts and controls. These results suggested distinct effects of M. tuberculosis infection and TB disease on monocyte subsets and phenotype.

Own previous studies found lower IL-7Rα expression on T cells from patients with TB (26). Therefore, we next compared monocyte IL-7Rα expression between the study groups. Moderate IL-7Rα expression was detected on monocytes ex vivo and only a small proportion of monocytes expressed higher IL-7Rα levels (Fig. 2A). Notably, patients with TB had lower proportions of IL-7Rα_high monocytes and also showed lower mean IL-7Rα expression as compared with both, contacts and controls (Fig. 2B). Comparisons of monocyte subpopulations detected lower IL-7Rα_high monocytes in patients with TB for all subsets as compared with controls but only for M1/2 and M1 when compared with contacts (Fig. 2C). Contacts showed lower IL-7Rα_high M2 and M1/2 proportions as compared with controls (Fig. 2C).

Because PD-L1 also showed TB disease–specific differences for subsets of monocytes, we next compared IL-7Rα and PD-L1 expression between the study groups. Notably, we found an inverse correlation between IL-7Rα and PD-L1 expression on M1 and M1/2 for patients with TB but not for contacts or controls (Fig. 2D). In contrast, IL-7Rα expression was positively correlated with PD-L1 on M2 in controls but not in patients with TB or contacts. This indicated monocyte subset–specific IL-7Rα differences between the study groups and potential association with differential monocyte function indicated by PD-L1.

Lower IL-7Rα monocyte expression of patients with TB suggested impaired monocyte IL-7 sensitivity. Hence, we next compared IL-7 signaling of monocytes via STAT1 and STAT5 pathways between the study groups. Ex vivo STAT1 analyses showed moderately lower pSTAT1 levels in contacts as compared with patients with TB and similar levels to controls (Fig. 3A). Interestingly, constitutive STAT5 phosphorylation was markedly lower in both, TB patients and contacts, as compared with controls (Fig. 3B). Next, we compared IL-7–induced pSTAT5 of monocytes in vitro between the study groups. Controls showed significantly increased pSTAT5 levels of monocytes stimulated with IL-7 and contacts also had moderately increased IL-7–induced pSTAT5. Notably, monocytes from patients with TB showed no pSTAT5 differences with or without IL-7 (Fig. 3C). To compare IL-7 effects between the study groups, we calculated pSTAT5 differences (i.e., IL-7 versus without) and detected lower IL-7 induced pSTAT5 in both, patients with TB and contacts, as compared with controls (Fig. 3D). These results confirmed functional effects of lower IL-7Rα expression on monocytes from patients with TB and suggested impaired monocyte sensitivity to IL-7 as compared with controls.

Only limited information on IL-7Rα regulation of monocytes is available. Recently, Al-Mossawi et al. (32) showed LPS-induced upregulation of the IL-7R α-chain in monocytes. We performed cell culture experiments using enriched monocyte populations to compare IL-7Rα expression during culture and after stimulation. In vitro culture induced partial maturation of monocytes (MDM) and this was indicated by increased HLA-DR expression after 96 h (Fig. 4A). Interestingly, IL-7Rα expression was moderately increased after 24 h and also showed marked upregulation in MDM after 96 h (Fig. 4A). Next, we analyzed monocyte markers CD14 and CD16 during in vitro culture (Fig. 4B). Especially CD16 was markedly upregulated, indicating culture-dependent effects on monocyte phenotype. Against this background, we excluded monocyte subtyping from in vitro analysis and determined IL-7Rα expression on all monocytes as well as MDM after in stimulation with LPS, Mycobacterium bovis BCG mycobacteria, and M. tuberculosis components.

As described, LPS significantly induced IL-7Rα expression in monocytes after 24 h (Fig. 4C). Interestingly, viable BCG mycobacteria induced IL-7Rα expression to a similar level as LPS (∼60% IL-7Rα_high monocytes), whereas dead BCG mycobacteria only induced moderate IL-7Rα_high proportions of monocytes (Fig. 4C). M. tuberculosis H37Rv components LAM, LM, and purified protein derivatives (PPDs) did not induce detectable IL-7Rα expression but cell wall extracts of H37Rv induced moderate proportions of IL-7Rα_high monocytes (Fig. 4C). Human peripheral blood monocytes hardly phagocytose mycobacteria and, therefore, we generated in vitro cultured MDM (96 h) to measure induction of IL-7Rα. LPS was less potent inducer of IL-7Rα in MDMs, whereas both viable and dead BCG as well as H37Rv cell wall extracts markedly induced IL-7Rα expression (Fig. 4D). These results demonstrated that mycobacteria as well as mycobacterial cell wall components induce IL-7Rα expression in monocytes and MDMs.

Because mycobacteria induce IL-7Rα expression of monocytes and MDMs, we next characterized the influence of IL-7 on MDM function. We performed a previously described mycobacterial kill assay using an LD-BCG reporter strain for fluorescence-based flow cytometry analysis (40). MDM were infected with BCG constitutively expressing mCherry to allow identification of infected MDM (Fig. 5A). Only in the presence of tetracyclin (ATC), viable BCG coexpress GFP, and this was used to discriminate live and dead BCG in infected MDM (Fig. 5A). Concomitant measure of count beads allows to deduce absolute cell numbers as well as proportions of MDMs infected with live or dead BCG. About 80% of MDM were infected with BCG and 50% of MDMs contained viable BCG (Fig. 5B). In the presence of IL-7, however, the proportion of MDM containing viable BCG decreased significantly (mean 39%) (Fig. 5B). Because infection efficacy and MDM numbers in culture were not different between IL-7–treated and –nontreated samples (data not shown), the results indicated increased mycobacterial killing. To analyze if MDM viability was affected, we determined absolute cell counts. In the presence of IL-7 the number of viable BCG containing MDM decreased (Fig. 5C). This excluded a possible bias due to IL-7 viability effects and confirmed increased mycobacterial killing of IL-7–treated MDM.

We also performed analyses of CFUs using LD-BCG–infected MDM with and without IL-7. Again, we detected a significant decrease in CFU when culturing BCG from IL-7–treated MDMs (Fig. 5D). Calculation of proportional decrease “kill rate” of IL-7–treated MDM (versus non–IL-7–treated MDM) showed a mean IL-7 effect of ∼27% on antimycobacterial MDM cytotoxicity (Fig. 5E). We concluded that lower IL-7Rα expression on monocytes from TB patients may affect monocyte and MDM function including impaired capacity to kill mycobacteria in macrophages.

The role of IL-7 in lymphocyte development and memory generation is well established and initial results indicated impaired T cell IL-7 sensitivity in acute TB (26). In this study, we demonstrate that lower IL-7Rα expression and lower IL-7 sensitivity is a monocyte feature of patients with TB. (M1) CD14⁺ and (M1/2) CD14⁺/CD16⁺ monocytes showed decreased expression of the IL-7Rα and impaired IL-7 signaling ex vivo. M. tuberculosis infection was previously shown to affect peripheral blood monocytes with potential implications on monocyte function and derived phagocyte populations (4). Increased monocyte proportions (also measured as monocyte lymphocyte ratios) as well as upregulation of CD16 on M1 are pathognomonic features of acute TB (3, 6–8, 41). The present study confirmed generally increased proportions of monocytes as well as higher frequencies of M1 and M1/2 subsets in patients with TB. In addition, contacts of indexed TB patients had higher proportions of M1/2 as compared with healthy community controls. This is of special interest because differences in proportions of monocyte subsets were shown to predict TB disease progression of M. tuberculosis–infected contacts (42). Prospective follow-up studies on contacts of patients with TB may decipher the role of monocyte phenotype changes during TB disease progression.

Especially M1/2 were previously shown to be increased in acute TB with potential implications on fate and functions of monocyte. M1/2 differ from M1 in several aspects, including phagocytosis, Ag presentation, and cytokine expression (9). Indeed, upregulation of CD16 on M1 (as the marker of M1/2) was discussed as an immune evasion mechanism of M. tuberculosis infection (9). Described consequences may include increased monocyte cell death (5, 43), increased immune modulatory “tolerogenic” monocyte functions (6–8), impaired APC functions (44), and impaired development of dendritic cells (11, 14, 45). IL-7 may be involved in several of these aspects, and prevention of cell death as well as increased survival are well-known functions of IL-7 in T cells (46). Because low IL-7Rα expression of M1 and M1/2 was associated by increased PD-L1, an important inducer of apoptosis, one can assume that decreased survival of monocyte during acute TB is a consequence of lower IL-7Rα expression. In accordance, we found impaired monocyte IL-7 sensitivity, measured by STAT5 phosphorylation, in patients with TB. Whether low sensitivity for IL-7 and low expression of IL-7Rα in monocytes is indicating a kind of exhaustion [as found for impaired IL-7Rα expression in T cells from chronic infections (47)] will have to be addressed by future studies.

Ex vivo pSTAT5 levels were constitutively lower in monocytes from patients with TB and contacts as compared with controls. This suggested in vivo effects on pSTAT5 signaling pathways in monocytes from recently M. tuberculosis–infected individuals. However, because the majority of contacts are related to index TB patients and/or live in the same household also affects independent factors (e.g., genetic, environmental, behavioral) and cannot be excluded. Own ongoing studies to investigate immune responses during M. tuberculosis–specific immune conversion in contacts from patients with TB will address this question. Potential functional implications of both lower constitutive and impaired IL-7–induced monocyte pSTAT5 levels will be addressed by future studies.

IL-7 functional effects in monocytes need to be further characterized. In this regard, we demonstrated increased MDM killing of BCG mycobacteria in the presence of IL-7. These results were in accordance with a previous study that showed IL-7 effects on anti–M. avium monocyte cytotoxicity (30). We excluded IL-7 confounding effects on MDM phagocytosis efficiency as well as IL-7 effects on monocyte survival to strengthen our conclusion that IL-7 promotes MDM mediated antimycobacterial effector functions. The underlying mechanisms remain elusive. However, our own initial studies on IL-7 in vitro effects showed increased HLA-DR expression of monocytes in the presence of IL-7 (P. Hehenkamp, M. Hoffmann, S. Kummer, C. Reinauer, C. Döing, K. Förtsch, E. Mayatepek, T. Meissner, M. Jacobsen, and J. Seyfarth, submitted for publication). Therefore, increased Ag presentation may contribute to improved in vitro monocyte effector mechanisms. This finding fits well with described impaired APC and dendritic cell functions in TB (11, 14, 44). Impaired IL-7–induced HLA-DR expression may also contribute to low Ag presentation capacity of tuberculous granuloma and less efficient T cell responses in TB (48).

IL-7 is secreted by different cell types including epithelial stroma cells and dendritic cells (46). Besides a role of IL-7 in the induction of memory generation in T cells, one may speculate about autologous IL-7 effects on monocytes and during development of dendritic cells. Specifically, initial evidence from in vitro studies showed that IL-7 affects monocyte-derived dendritic cell generation (49, 50). The present study showed an association of lower IL-7Rα expression on PD-L1_high monocytes. If direct effects of impaired IL-7 monocyte response in TB patients are causative for increased PD-L1 expression needs to be investigated by future studies. However, the biological significance of IL-7Rα expression on monocyte subsets and potential effects on the interaction with T cells via the PD-1/PD-L1 pathway should intensify the efforts to further characterize the role of IL-7 in TB.

IL-7Rα regulation mechanisms of monocytes are hardly defined. Recently, Al-Mossawi et al. (32) showed that LPS-induced IL-7Rα expression on monocytes. We confirmed LPS-induced IL-7Rα expression and provided evidence that viable BCG mycobacteria and M. tuberculosis cell wall components induced monocyte IL-7Rα expression in vitro. Dead BCG hardly induced IL-7Rα expression on monocytes, whereas MDMs showed IL-7Rα upregulation also in the presence of dead BCG. Moderate IL-7Rα receptor upregulation was also seen in the absence of in vitro stimulation. The mechanisms underlying regulation of IL-7Rα expression in monocytes need further investigation. Al-Mossawi et al. (32) provided evidence that TNF-α is involved in LPS-induced monocyte IL-7Rα expression, but we detected only minor effects when blocking TNF-α (data not shown). Further analyses are necessary to unravel the underlying mechanisms.

Our study provides the first evidence, to our knowledge, for a potential role of IL-7 on monocyte function in TB pathogenesis and generally initiates novel questions about the influence of IL-7 on monocyte and dendritic cell populations. Beyond the field of TB, this topic is of interest for IL-7–dependent immune pathologies like rheumatoid arthritis. Finally, ongoing IL-7–interventional trials for treatment of sepsis patients (reviewed in Ref. 51) may consider potential impaired IL-7 sensitivity of monocytes during disease pathology. In this study, IL-7Rα expression differences on monocytes qualify as candidate biomarkers for treatment efficacy.

We thank all study participants, study nurses, and physicians.

The authors have no conflicts of interest.