Background
Tuberculosis (TB) is a communicable disease caused by
Mycobacterium tuberculosis (
M. tb), which infected about a quarter of world’s population [
1]. Until the pandemic of COVID-19, TB remains the leading cause of death worldwide from a single infectious pathogen. In 2020, approximately 5.8 million people were newly diagnosed with TB, among which over 157 thousand were detected with drug-resistant TB [
1]. The continuous high disease burden worldwide urges the development of new strategies for TB treatment and prevention. Host-directed therapy is one of the promising approaches, given the outcome of human infection with
M. tb is largely dependent on host immune status. Most people infected with
M. tb can be asymptomatic during their lifetime, with only about 5% fail to contain
M. tb and eventually develop TB [
2]. Research into the interplay between host immune system and pathogen is vital for the development of novel host-directed therapy for TB.
Immune checkpoint proteins can regulate the immune response in malignancies and infectious diseases via numerous types of activating and inhibitory signals between antigen-presenting cells (APCs) and T cells [
3,
4]. Receptor-ligand interaction is required for the transduction of second signal, following the first signal conveyed by the interaction of MHC molecules on APCs and T cell receptors on effector T cells loaded with cognate antigens [
3]. Co-stimulatory receptor-ligand interactions that help amplify effector T cell responses include CD28–CD80, 4-1BB (also known as CD137)-4-1BB ligand, CD27–CD70. To avoid over-reactivation, effector T cell response also require suppression signals generated upon co-inhibitory receptor–ligand interactions, including B- and T-lymphocyte attenuator (BTLA)-herpes virus entry mediator (HVEM), glucocorticoid-induced TNFR-related (GITR)- GITR ligand, lymphocyte-activation gene 3 (LAG-3)- MHC, programmed cell death protein 1 (PD-1)- programmed cell death 1 ligand 1 (PD-L1)/programmed cell death 1 ligand 2 (PD-L2), T-cell immunoglobulin and mucin domain-3 (TIM-3)- Galectin 9, cytotoxic T-lymphocyte associated antigen 4 (CTLA-4)- CD80. Intracellular molecules, such as indoleamine 2,3-dioxygenase (IDO) were also found to exert immune checkpoint functions in T cells [
3]. Researches have investigated the impact of immune checkpoints on the immune responses against a range of chronic infections, including malaria, hepatitis B virus and human immunodeficiency virus infection [
3,
5]. Since a well-balanced host immune response against
M. tb is required to clear or contain the infection, while preventing potentially damaging inflammatory response [
6], further insight into the role of immune checkpoints in anti-TB immunity is required for the development of immune therapy for TB.
In this study, we aimed to provide preliminary data on the characteristic profile of soluble immune checkpoints (sICs) in patients with distinct M. tb infection status, and their association with immune response at lesion sites, and to assess their dynamic change during anti-TB treatment.
Methods
Study population
In total, 44 individuals were enrolled in this study, including 10 individuals with latent tuberculosis infection (LTBI), and 10 healthy controls (HC), and 24 patients with pulmonary tuberculosis (PTB), among which 10 patients were diagnosed with tuberculous pleurisy (TBP). All the TB patients were recruited from Wuxi Fifth People’s Hospital between February 2019 and May 2021. Individuals with LTBI and HC were recruited from the relatives of TB patients and the volunteers of Huashan Hospital during the same period.
PTB patients were diagnosed based on chest radiological evidence, supplemented by a positive result of culture for M. tb and/or Xpert MTB/RIF test in the sputum or bronchoalveolar lavage fluid; and individuals clinically diagnosed based on chest radiological evidence, supplemented by clinical symptoms of TB, or a positive tuberculin skin test or interferon-γ release assay (IGRA) result were also included. Patients with TBP were defined as TB patients showed chest radiological evidence of pleural involvement, and met one of the following criteria: a positive result of culture for M. tb and/or Xpert MTB/RIF test in the pleural effusion; histologically confirmed M. tb infection by pleural biopsy; or an elevated concentration of adenosine deaminase (ADA) in pleural effusion along with a positive tuberculin skin test or IGRA result. Individuals with LTBI and HC were IGRA-positive and -negative, respectively. In addition, they had no evidence of active tuberculosis (ATB). All enrolled participants were free of HIV infection, autoimmune disease or other chronic infections, and not undergoing immune-modulating treatment.
Peripheral blood was drawn from all participants, while pleural effusion was only collected from TBP patients. Plasma and tuberculous pleural effusion (TPE) supernatant were separated and stored at – 80 °C until further assessment. Demographic and clinical characteristics, including age, gender, bacillus Calmette-Guerin (BCG) vaccination history, results of diagnostic tests for TB, and laboratory test results of TPE were also collected.
The study was approved by the institutional ethics review board of Huashan Hospital, Fudan University. Verbal informed consent was obtained for all the investigations, and written informed consent was obtained from each participant in the study prior to enrollment. The study was performed in accordance with the guidelines of the Declaration of Helsinki and relevant regulations.
Quantification of sICs
Plasma samples of PTB and TBP patients were collected at baseline prior to anti-TB treatment and during treatment. Baseline TPE samples were additionally collected in TBP patients. The ProcartaPlex™ Human Immuno-Oncology Checkpoint Panel Immunoassay Kit (Thermo Fisher, Waltham, MA) was used to measure the concentrations of fourteen sICs, including soluble BTLA (sBTLA), sHVEM, sGITR, sIDO, sLAG-3, sPD-1, sPD-L1, sPD-L2, sTIM-3, sCD28, sCD80, s4-1BB, sCD27 and sCTLA-4. According to the manufacturer’s protocol, all samples were assayed in duplicate, using 25 μL of sample per well. The concentration of the samples was calculated by plotting the expected concentration of the standards against the fluorescent signals generated by each standard. Detection and analysis of assayed samples were performed using the Luminex 200™ System (Thermo Fisher, Waltham, MA, USA).
Statistical analysis
Statistical analysis was performed using GraphPad Prism 8 (GraphPad Software, Inc. La Jolla, CA, USA). Continuous variables were compared between independent groups using Mann–Whitney test (two groups) and Kruskal–Wallis test (multiple groups). Categorical variables were compared using Fisher's exact test. Paired data were analyzed using the paired Wilcoxon rank test. The correlation between laboratory test markers of pleural effusion, including specific gravity, protein, ADA, lactate dehydrogenase (LDH) and lymphocyte percentage, and the levels of sICs was assessed with Spearman correlation. The results with a P value of < 0.05 were considered significant.
Discussion
In this study, we performed comprehensive assessment of the circulating levels multiple sICs in M. tb infection for the first time, using four distinct populations, including patients with PTB, TBP, individuals with LTBI, and healthy individuals. We revealed a screwed balance between the co-inhibitory and co-stimulatory molecules in different status of M. tb infection.
M. tb manipulates host immunity to persist in latent form or develop chronic infection by impairing the bactericidal machinery of APCs and interfering T-cell mediated immunity [
7]. In untreated
M. tb infection, T cells upregulate the display of multiple co-inhibitory molecules, such as PD-1, TIM-3, LAG-3, reduce production of IFN-γ, secrete suppressive cytokines, and proliferate less, suggesting a suppressive immune status in ATB [
8]. The expression level of co-inhibitory receptors, PD-1, CTLA-4 and TIM-3, on both lymphocytes and monocytes was also found associated with
M. tb infection status and changes during LTBI treatment [
9]. Similar changes of circulating sICs levels were found in our ATB and LTBI patients. However, such change in plasma sTIM-3 level did not recur in subgroup analysis where culture-negative ATB patients with a clinical diagnosis were excluded from the ATB group, neither did that in plasma sLAG-3 level. One possible explanation could be the correlation between the sICs levels with disease severity, as Wang et al. found the TIM-3 expression on CD8 + T cells in smear-positive PTB patients was significantly higher than that in smear-negative patients [
10]. And an elevated sLAG-3 level in PTB patients was also found correlated to with cavity sizes > 4 mm on chest radiography [
11].
We have also observed increased sCD28, sCD27 and s4-1BB in
M. tb infection. Researches showed activated T cells also upregulate numerous co-stimulatory molecules, such as CD28:B7, ICOS:ICOS-L, 4-1BB:4-1BBL, during persisting infections to mount efficient T-cell response [
12]. Studies of gene modified murine models and TB patients suggested a positive role of co-stimulatory molecules in anti-TB immunity. A reduced cytokine production and killing capacity of
M. tb-specific effector CD8 + T cell was found
M. tb-infected mice with ICOS deficiency during late-stage
M. tb infection [
13]. Together, our results showed a skewed balance of co-inhibitory and co-stimulatory molecules in individuals with ATB and LTBI, suggesting a suppressive immune status in
M. tb infection.
The sICs levels in plasma and TPE of TBP patients were compared in this study, and most sICs were upregulated, suggesting a more activated and potent anti-TB immune response might exist in pathological lesion sites. The upregulation of co-inhibitory receptor might help prevent overactivated immune response during TBP. As shown in preclinical models, both
M. tb-infected PD-1 knockout mice and anti-PD-1 antibody-treated rhesus macaques developed detrimental inflammation and exacerbated disease, suggesting checkpoint-mediated co-inhibition is involved in harnessing T-cell-driven immune pathology and the control of
M. tb infection [
14,
15]. There have been several studies on the PD-1/PD-Ls pathways in
M. tb infection, which all showed an upregulated expression of PD-1 and PD-L1 on both pleural and peripheral CD4 + T cells in ATB patients [
16,
17]. Yin et al. also found an increasing sPD-1 level but a comparable level of sPD-L1 level in TPE [
17], Partially consistent with the findings of Yin et al., we revealed a higher level of sPD-1 but a significantly lower level of sPD-L1 in TPE, which might result from the repressed release of PD-L1 from pleural mesothelial cells or lymphocytes.
Studies of TBP patients identified the pleural T cells with an effector phenotype of CD27-CD45RA-CCR7-CD62L- mediates local
M. tb-specific immune response by simultaneously production of multiple cytokines [
18,
19]. Consistent with previous finding, we found a low sCD27 in TPE were correlated negatively with the specific gravity and protein level of TPE, which indicates severer chronic inflammation at lesion sites. However, several researches consider the lower expression of CD27 on
M. tb-specific CD4 + T cells is associated with persistent ATB [
20,
21]. Therefore, further investigation on the role of CD27 in
M. tb infection is needed to draw a conclusion.
LAG-3 expression is significantly induced in the lungs of macaques with ATB, and correlates with diminished responses and increased bacterial burden [
22]. Silencing LAG-3 signaling in macaque lung enhanced killing of
M. tb in CD4 + T cells, by interfering with the mitochondrial apoptosis pathway and increasing IFN-γ expression [
23]. We here found that in TPE the sLAG-3 level was positively correlated with the ADA level. As ADA is a highly sensitive and specific marker for TBP diagnosis [
24], we surmised sLAG-3 might be a candidate biomarker for TPE.
Follow-up assessment revealed the circulating level of sCTLA4 and sCD28 might be correlated with favorable anti-TB treatment outcome. CTLA4 and CD28 are paired immune checkpoint and co-stimulatory molecules that compete for shared ligands CD80 and CD86 [
12]. Studies on chronic viral infection and in vitro blockade showed CTLA4 contribute to the impaired immune response, induction of T cell exhaustion and the failure of immunological control of the persisting pathogens by pathogen-specific T cells [
12,
25]. CTLA4 attenuates T cell activation by inhibiting co-stimulatory signal via CD28 and transmitting inhibitory signals to T cells [
25]. Our results suggested a recovering balance between co-inhibitory and co-stimulatory signals as the clearance of bacteria load in
M. tb infection.
More attention should be given to BTLA, since sBTLA was found not only correlated with
M. tb-specific marker ADA in TPE, but also steadily decline in response to anti-TB treatment. Intense research in the field of cancer, autoimmune diseases and various infection has revealed the interaction of BTLA with its ligand HVEM has a potent inhibitory effect on adaptive, majorly T-cell mediated, immune response [
26]. However, the effect and mechanism of BTLA-HVEM signaling in
M. tb infection still remain poorly understood. Recent studies on
M. tb infection have found that, in comparison with HC, ATB patients exhibited enhanced expression of BTLA on myeloid and plasmacytoid dendritic cells (DCs) from peripheral blood and TPE, different from the high level of BTLA expression on T cells observed in virus infection [
27‐
29]. TB-driven BTLA upregulation on DCs impaired the expression of DCs mature marker CD83 and activation marker HLA-DR, and suppress the ability of DCs to induce Th17 and Th22 response, while promoting Th2 response [
28,
29]. Additionally, BTLA + DCs can promote Foxp3 expression in naïve CD4 + T cells through the upregulation of CD5 and subsequent inhibition of PI3K/mTOR activation, therefore inducing extrathymic Treg differentiation [
30]. Together, these findings indicate the activation of BTLA-HVEM pathway is involved in the pathogenesis of TB. Further insight is required to understand the mechanism of BTLA-mediated regulation on anti-TB immunity.
Some limitations in the present study must be noted. First, the number of individuals involved in our study was small. And a number of clinically diagnosed ATB patients who lacks positive bacteriological test results were included. More bacteriologically confirmed ATB patients should be enrolled in further study to provide more power. Second, ATB patients were followed up only once during their anti-TB therapy, resulting from the retrospective study design. Serial follow-up at different time points should be done to depict the kinetics of sICs in a more detailed way. To develop novel biomarkers of TB, further investigation on the sICs that showed correlation with M. tb infection status or treatment response in this study is required.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.