The COX- inhibitor indomethacin reduces Th1 effector and T regulatory cells in vitro in Mycobacterium tuberculosis infection

Abstract

Background: Tuberculosis (TB) causes a major burden on global health with long and cumbersome TB treatment regimens. Host-directed immune modulating therapies have been suggested as adjunctive treatment to TB antibiotics. Upregulated cyclooxygenase-2 (COX-2)-prostaglandin E2 (PGE2) signaling pathway may cause a dysfunctional immune response that favors survival and replication of Mycobacterium tuberculosis (Mtb).

Methods: Blood samples were obtained from patients with latent TB (n = 9) and active TB (n = 33) before initiation of anti-TB chemotherapy. COX-2 expression in monocytes and ESAT-6 and Ag85 specific T cell cytokine responses (TNF-α, IFN-γ, IL-2), proliferation (carboxyfluorescein succinimidyl ester staining) and regulation (FOXP3+ T regulatory cells) were analysed by flow cytometry and the in vitro effects of the COX-1/2 inhibitor indomethacin were measured.

Results: We demonstrate that indomethacin significantly down-regulates the fraction of Mtb specific FOXP3+ T regulatory cells (ESAT-6; p = 0.004 and Ag85; p < 0.001) with a concomitant reduction of Mtb specific cytokine responses and T cell proliferation in active TB. Although active TB tend to have higher levels, there are no significant differences in COX-2 expression between unstimulated monocytes from patients with active TB compared to latent infection. Monocytes in both TB groups respond with a significant upregulation of COX-2 after in vitro stimulation.

Conclusions: Taken together, our in vitro data indicate a modulation of the Th1 effector and T regulatory cells in Mtb infection in response to the COX-1/2 inhibitor indomethacin. The potential role as adjunctive host-directed therapy in TB disease should be further evaluated in both animal studies and in human clinical trials.

Keywords: COX-inhibitors; Cytokines; Host-directed therapy; Monocytes; Regulatory T cells; Tregs; Tuberculosis.

Fig. 1

Flowchart of patient samples. Patients were included at two study locations. Cohort A (latent TB and active TB) denotes patients included at Haukeland University Hospital, whilst cohort B denotes patients included at Oslo University Hospital. The figure also shows the distribution of patient samples in the different assays performed in the study

Fig. 2

COX-2 expression in monocytes. a Gating strategy for identification of true monocytes [34]. b Comparison of COX-2 expression in unstimulated true monocytes in PMBCs from patients with latent TB (LTB, circles) and active TB (ATB, squares). PBMCs from patients with LTB and ATB were left unstimulated (Unstim) or stimulated for 12 h with LPS. P- values were calculated by Wilcoxon matched pairs test for paired samples and Mann- Whitney U test for group wise comparison (n.s = non-significant). Horizontal lines in b represent median values

Fig. 3

Effect of indomethacin on TB antigen induced FOXP3+CD25++ CD4+ Tregs. PBMCs were obtained from patients with active TB (n = 17) prior to treatment and left unstimulated (unstim) or stimulated for 36 h with ESAT-6 or Ag85 without (open boxes) or with (hatched boxes) addition of indomethacin. The figures show percentages of FOXP3+CD25++ (a) and FOXP3 median fluorescence intensity (MFI) (b) in the CD4+ cells. P- values were calculated by Wilcoxon matched pairs test. Plots are shown as box plots together with individual data points with median, IQR and minimum/maximum values. Significant changes between stimulated samples with and without indomethacin are denoted with p-values in figure. Levels of FOXP3+CD25++ CD4+ T cells and FOXP3 MFI were significantly upregulated upon stimulation with both E6 and Ag85 (p < 0.01, values not shown in figure)

Fig. 4

Effect of indomethacin on TB antigen induced intracellular cytokines in CD4+ T cells. Total IFN-у+, IL-2+ and TNF-α+ T cell responses in unstimulated (Unstim), ESAT-6 and Ag85 stimulated PBMCs without (open boxes) and with (hatched boxes) indomethacin after 12 h (a, n = 18) and 36 h (b, n = 17) stimulation. PBMCs were obtained from patients with active TB prior to treatment. P- values calculated by Wilcoxon matched pairs test. Plots are shown as box plots with individual data points with median, IQR and minimum/maximum values. Significant changes between stimulated samples with and without indomethacin are denoted with p-values in figure

Fig. 5

Mtb antigen induced single, duo and polyfunctional cytokine producing CD4+ T cells. PBMCs were obtained from patients with active TB prior to treatment and left unstimulated or stimulated with ESAT- 6 or Ag85 without (open boxes) or with indomethacin (hatched boxes) after 12 h (a) and 36 h (b) stimulation (n = 18 and n = 17 respectively). Boolean gating strategy was used to create cytokine combinations of single-producing, duo-producing and polyfunctional T cells. P- values were calculated by Wilcoxon matched pairs test ( *p <0.05, **p < 0.01). Plots are shown as box plots with median, IQR and minimum/maximum values

Fig. 6

Effect of Indomethacin on CD4+ and CD8+ T cell proliferation. Proliferative responses measured by percentages of CFSE^dim in unstimulated, ESAT-6 and Ag85 stimulated PBMCs without (open boxes) and with (hatched boxes) indomethacin after 6 days stimulation. PBMCs were obtained from patients with active TB prior to treatment. a CD4+ T cells (n = 23). b CD8+ T cells (n = 23). P- values were calculated by Wilcoxon matched pairs test. Plots are shown as box plots with individual data points with median, IQR and minimum/maximum values. Significant changes between stimulated samples with and without indomethacin are denoted with p-values in figure. %CFSEdim cells were significantly upregulated upon stimulation with both E6 and Ag85 (p < 0.001, values not shown in figure)