Patient characteristics

During the study period, 120 individuals were prospectively enrolled, including 80 patients with active TB and 40 non-TB subjects. The demographic characteristics of the enrolled patients are shown in Table 1. The mean age was 62.7 ± 17.6 years in patients with TB and 52.3 ± 17.6 years in the non-TB subjects (p = 0.003). The body mass index (BMI) was 21.4 ± 3.2 in patients with TB and 23.1 ± 3.0 in non-TB subjects (p = 0.007). Among patients with TB, 31.6% had comorbid diabetes, and 10.1% had inactive malignancies. Extrapulmonary TB was present in 14 (17.5%) patients with TB, and sputum smear positivity was observed in 21 (26.3%) of them.

Increased expression of PD-L1 and demethylation-related genes in TB patients

The expression levels of PD-L1, DNMT1, DNMT3A, DNMT3B, TET1, and TET2 in PBMCs obtained from patients with active and no TB are illustrated in Figure 1A. The results revealed that patients with active TB exhibited elevated expression of PD-L1, DNMT3B, TET1, and TET2, while showing decreased expression of DNMT1 compared to that in the non-TB subjects. We also investigated the expression of these genes in patients with active TB before and after the anti-TB treatment. As shown in Figure 1B, after initiating the anti-TB treatment, PD-L1 and TET1 expression was significantly decreased. We also observed a trend towards increased DNMT1 expression after the initiation of anti-TB treatment.

–

FIGURE 1: Comparisons of PD-L1, DNMTs, and TETs expression in peripheral blood mononuclear cells between (A) patients with active TB and non-TB subjects, and in (B) patients with active TB patients upon TB diagnosis and after anti-TB treatment. The data in bar charts are mean values of relative expression compared to GAPDH as reference. Differences between groups were compared using the Mann–Whitney U test. PD-L1 - programmed death-ligand 1; DNMTs - DNA methyltransferases; TETs - ten eleven translocation methylcytosine dioxygenases; 1M - 1-month after anti-TB treatment; 2M - 2-month after anti-TB treatment.

We conducted further investigations to explore the correlation between the expression levels of PD-L1, DNMT1, DNMT3A, DNMT3B, TET1, and TET2, and the clinical characteristics of patients with active TB. The results are shown in Figure 2. Patients with active TB and fever as well as positive sputum smears exhibited significantly higher expression of PD-L1 than in those without fever and with negative sputum smears. Patients with weight loss displayed lower DNMT1 expression. Moreover, patients with extrapulmonary involvement exhibited lower TET1 expression than in those with pulmonary TB. In terms of treatment outcomes, as illustrated in Figure 2, patients who did not achieve smear and culture conversion after one month of treatment showed significantly higher expression of PD-L1 and TET1 than in those who achieved conversion.

–

FIGURE 2: Comparisons of PD-L1, DNMTs, and TETs expression in PBMCs from patients with active TB with various clinical presentations and treatment outcomes. The data in bar charts are mean values. Error bars are standard errors. Differences between groups were compared using the Mann– Whitney U test. PD-L1 - programmed death-ligand 1; DNMTs - DNA methyltransferases; TETs - ten eleven translocation methylcytosine dioxygenases.

Increased expression of PD-L1 and TET1 in lung tissues and macrophages

Lung tissues were surgically resected from two patients with active pulmonary TB. Both lung tissues exhibited granulomatous inflammation and were tested positive for MTB in culture. Immunohistochemistry (IHC) staining revealed co-localization of TET1 and PD-L1 positive cells within inflammatory tissues (Figure 3). These cells were primarily positive for CD68, indicating that they were macrophages.

–

FIGURE 3: Immunohistochemistry (IHC) staining of lung tissues from two patients with pulmonary tuberculosis. IHC staining using anti-PD-L1 and anti-TET1 antibodies showed the presence of PD-L1 and TET1 positive cells. It also demonstrated the co-localization of PD-L1 and TET1. IHC staining using an anti-CD68 antibody suggested that these PD-L1/TET1 positive cells were primarily macrophages. PD-L1 - programmed death-ligand 1; TET1 - ten eleven translocation methylcytosine dioxygenases.

We conducted immunofluorescence (IF) analysis of THP-1 cells stimulated with MTB-related proteins. Upon treatment with MTB whole cell lysate (H37Rv, 10 μg/ml) and EsxA protein (1 μg/ml), TET1 and PD-L1 expression was increased in THP-1 cells for both stimulations (Figure 4). In the merged images, immunoreactivity towards the TET1 antibody overlapped with the PD-L1 staining. These findings suggest that human macrophages with high PD-L1 expression also exhibit high TET1 expression following MTB-related stimulation.

–

FIGURE 4: Immunofluorescence (IF) staining of THP-1 cells treated with MTB-related materials. THP-1 cells were treated with H37Rv whole cell lysate (10 µg/ml) and EsxA protein (1 µg/ml) for 24 hours. IF staining demonstrated increased expression of PD-L1 (red) overlapped with TET1 (green) in THP-1 cells. PD-L1, programmed death-ligand 1; TET1, ten eleven translocation methylcytosine dioxygenase 1.

PD-L1 production in macrophages correlated with methylation modulation

We treated THP-1 cells with MTB whole cell lysate (H37Rv, 10 and 20 μg/ml) and EsxA protein (1 and 2 μg/ml) and measured the expression of PD-L1 and methylation-related genes. Increased expression of PD-L1 and TET1, and decreased expression of DNMT1, DNMT3A, and DNMT3B, was noted in THP-1 cells after MTB-related stimulation (Supplementary Figure 1).

To elucidate how DNA methylation of the PD-L1 promoter contributes to the downregulation of PD-L1 gene expression in vivo, we employed treatment with 5’Aza to inhibit DNMT1. THP-1 cells were pre-treated with different concentrations of 5’Aza (1 μM and 5 μM), followed by treatment with MTB whole cell lysate (H37Rv, 10 μg/ml and 20 μg/ml) and EsxA protein (5 μM and 10 μM). As shown in Figure 5A, a discernible increase in PD-L1 protein expression was observed upon the treatment of THP-1 cells with MTB whole-cell lysate and EsxA protein. Notably, PD-L1 protein expression showed a further dose-dependent increase when THP-1 cells were pre-treated with 5’Aza. Consistently, PD-L1 gene expression was also increased after pre-treatment with 5’AZA in THP-1 cells (Figure 5B).

–

FIGURE 5: Effect of DNA methylation inhibitor and TET1 gene knockdown on PD-L1 gene expression in THP-1 cells with (A, B) pre-treatment of 5’AZA, and (C) shRNA-mediated TET1 knockdown. THP-1 cells were treated with H37Rv whole cell lysate (10 µg/ml) and EsxA protein (2 µg/ml). A) Production of PD-L1 protein was evaluated using a western blot assay with densitogram. Data are presented as the mean ± standard deviation (SD) for each group. PD-L1, programmed death-ligand 1. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

The effect of promoter methylation on PD-L1 expression was further confirmed by shRNA-mediated TET1 knockdown. PD-L1 expression was significantly decreased in THP-1 cells with TET1 knockdown, both in the presence and absence of treatment with MTB whole cell lysate and EsxA (Figure 5C). These findings provide compelling evidence for the role of promoter methylation in modulating the expression of PD-L1 in macrophages.

Decreased PD-L1 promoter methylation in active TB patients

The methylation status of the PD-L1 promoter was assessed using bisulfite sequencing, and the results are shown in Figure 6. The analysis revealed a significantly reduced methylation in the cg19724470 region, among patients diagnosed with active TB compared with that in non-TB subjects. Conversely, both active TB and non-TB subjects displayed low methylation levels in the cg14305799 region, with no statistically significant differences observed between the two groups.

–

FIGURE 6: PD-L1 promoter methylation analysis in PBMCs from patients with active TB and non-TB subjects. CpG loci cg19724470 and cg14305799 were analyzed. The data in bar charts are mean values. Error bars are standard deviations. The data were compared using the Mann-Whitney U test. TB, tuberculosis.

Patients and settings

This prospective observational study was conducted at a tertiary medical center in northern Taiwan. Eligible participants were newly diagnosed tuberculosis patients with culture-proven cases between June 2020 and July 2022. To establish a comparison group, individuals without TB were also enrolled during the same study period. Demographic characteristics, clinical presentations, and radiological presentations were collected and reviewed during the enrollment interviews. The study protocol was approved by the Institutional Review Board of Taipei Veterans General Hospital (IRB number 2020-06-014BC, 2020-12-001CC). Written informed consent was obtained from each participant or their authorized representative(s) prior to enrollment. Details are shown in Supplementary material.

RNA isolation and quantitative reverse transcription PCR

PBMCs were isolated from peripheral blood samples collected from the participants on the day of enrollment. Total RNA was extracted from PBMCs using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Quantitative reverse transcription PCR was performed in duplicate using Power SYBR Green Master Mix (Roche, Mannheim, Germany) and analyzed using a Stratagene Mx3000P real-time PCR system. The mRNAs evaluated were PD-L1, DNMT1, DNMT3A, DNMT3B, TET1, and TET2. The relative expressions of target mRNAs were compared to GAPDH as reference. The primer sequences used for real-time PCR are provided in Supplementary Table 1.

Stimulation of human macrophages

For in vitro studies, THP-1 human leukemia cells (TIB202; ATCC, Manassas, VA, USA) were cultured in RPMI 1640 medium at a density of 1 × 10⁶ cells/ml. The cells were differentiated into macrophages by treatment with 20 ng of phorbol 12-myristate 13-acetate (PMA, P8139; Sigma-Aldrich; St. Louis, MO, USA) for 24 h, followed by a 72-hour resting period before treatment with MTB-related stimulants. PD-L1, DNMT1, DNMT3A, DNMT3B, TET1, and TET2 expression levels were measured before and after stimulation.

MTB-related materials and in vitro stimulation

The MTB-related materials used in this study included MTB whole cell lysates (M. tuberculosis, Strain H37Rv, Whole Cell Lysate, NR-14822) and recombinant EsxA protein (Recombinant Protein Reference Standard, NR-49424; both obtained from BEI Resources; Manassas, VA, USA). THP-1 cells were treated with MTB whole-cell lysates at concentrations of 10 and 20 µg/ml and with EsxA protein at concentrations of 1 and 2 µg/mL. The treatment durations were 6 h for RNA isolation and 24 h for western blotting. PMA-treated THP-1 cells were plated in 24-well microculture plates at a density of 1 × 10⁶ cells/well before in vitro stimulation. Details are shown in Supplementary material.

Immunohistochemistry (IHC) and immunofluorescence (IF) analysis

To identify the cells expressing PD-L1 and TET1, lung specimens obtained from two patients with active pulmonary TB were subjected to IHC analysis. Briefly, 4 µm sections of lung tissues were pretreated using an antigen retrieval solution and then incubated with anti-PD-L1 antibody (R&D Systems) and anti-TET1 antibody (GeneTex, Irvine, CA, USA).The bound antibodies were detected using a horseradish peroxidase-conjugated compact polymer system. IHC staining was also performed using an anti-CD68 antibody (Dako; Glostrup, Denmark) to identify the macrophages in lung tissues.

For IF analysis, THP-1 cells with and without MTB-related stimulation were stained using primary antibodies against PD-L1 (GeneTex; Irvine, CA, USA) and TET1 (GeneTex; Irvine, CA, USA). Nuclear staining was achieved with 4′,6-diamidino-2-phenylindole (DAPI). Fluorescence images were captured using a 3D HISTECH PANNORAMIC SCAN system and were analyzed using panoramic viewer software. Details are shown in Supplementary material.

DNMTs blockade with 5’Aza and western blot assay

To investigate the role of DNMTs in PD-L1 expression, THP-1 cells were treated with the DNMTs inhibitor, 5-Aza-2’-deoxycytidine (5-Aza) (Sigma-Aldrich), at concentrations of 1 μM and 5 μM for 96 hours prior to stimulation with MTB-related materials. For western blotting, proteins harvested from THP-1 cells were electrotransferred onto polyvinylidene fluoride membranes. The membranes were incubated with primary antibodies against PD-L1 (Cell Signaling Technology, Danvers, MA, USA) for 3 h and then with anti-rabbit IgG horseradish peroxidase-conjugated secondary antibodies for 1 h. The blots were visualized using an enhanced chemiluminescence western blot detection system. Details are shown in Supplementary material.

TET1 knockdown with shRNA

TET1 knockdown THP-1 cell lines were generated by stably expressing short hairpin RNAs (shRNAs) targeting TET1 mRNA. THP-1 cells were transduced with TET1 shRNA (TRCN0000075026) or copGFP control lentiviral particles (C6-4-2) obtained from the National RNAi Core Facility at Academia Sinica, Taiwan. The sequences of TET1 shRNAs are listed in Supplementary Table 1.

Details are shown in Supplementary material.

Promoter methylation analysis by bisulfite sequencing

CpG methylation analyses were conducted by bisulfite sequencing as previously described 29. Genomic DNA extracted from the PBMCs was treated with bisulfite. PCR was performed on bisulfite-treated DNA to amplify the promoter regions of PD-L1. Considering the previously reported high methylation status of cg19724470 and cg14305799 30, methylation-specific primers were designed to cover these two CpG loci and used for further investigation. Details are shown in Supplementary material.

Statistical analysis

Intergroup differences were analyzed using a non-parametric, two-tailed Mann-Whitney U test for numerical variables and Fisher’s exact test for categorical variables. Statistical significance was defined as p < 0.05. All statistical analyses were performed using SPSS version 25.0 (Chicago, IL, USA) and GraphPad Prism 7.0 (San Diego, CA, USA).

Data availability statement

The data presented in this study are available on request from the corresponding author.