Samples were acquired on a BD LSRFortessa using FACSDiva software (version 6.2, BD Biosciences) and analyzed using FlowJo software (version 9.5.3, Treestar, Ashland, OR, USA). CD8+ cells were enriched by positive selection using magnetic beads (MACS, Miltenyi Biotec). Cells were fluorescence-activated cell sorted (FACS) by BD FACSAriaIII cell sorter using CD39-PE (Biolegend). Purity of all cell sorts was ≥97% as assessed by flow cytometry. Cell lines were tested for their capacity to inhibit proliferation of a Th1

responder clone (Rp15 1–1) and its cognate M. tuberculosis hsp65 p3–13 peptide, presented by HLA-DR3 positive, irradiated (20 Gy) PBMCs as APCs in a coculture assay that has been previously reported [8, 34]. Proliferation was measured

after 3 days of coculture by addition of 0.5 μCi/well and (3H)thymidine incorporation was assessed after 18 h. Values represent means from triplicate HM781-36B wells. For the CFSE-labeling assay, the Rp15 1–1 Th1-responder clone was labeled with 0.005 μM of CFSE and the irrelevant, isogenic T-cell clone (R2F10), with different peptide specificity and HLA-DR2 restriction, with 0.5 μM of CFSE, similar in design to previously described [13]. After 16 h of coculture with 5 × 104 CD8+CD39+ T cells, the p3–13 peptide (50 ng/mL) and HLA-DR3 positive KU-60019 in vivo APCs, cells were harvested and stained for CD3, CD4, and CD8. CFSE intensity was measured on a BD LSRFortessa using FACSDiva software and analyzed using FlowJo software. ARL 67156 trisodium salt hydrate (Sigma-Aldrich) was added to the well in 150 μM and daily during the 3 days of coculture. Anti-CD39 monoclonal antibody BY40/OREG-103 (Orega Biotech, Ecully, France) was added to the well at the first day of coculture at a final concentration of 10 μg/mL, as was the IgG1

isotype control (R&D Systems). Values represent mean ± SE from triplicate wells. Suppressive capacity of CD8+CD39+ Oxymatrine T cells was independent of original proliferation of the Th1 clone, as tested by reducing the cognate peptide concentration in the coculture assays. Reversal of suppression was calculated in proportion to original clone proliferation in the absence of Treg cells, since ARL and anti-CD39 monoclonal antibody interfered directly with Th1 clone proliferation signals in the CD39 pathway, as demonstrated by reduced (3H)thymidine incorporation after 3 days. Percentage blocking was calculated after natural logarithmic transformation, and inhibition of proliferation in the presence and absence of blocking agents was calculated and expressed as percentage [8]. Raw data can be provided per request. Mann–Whitney tests and Wilcoxon signed-ranks tests were performed using GraphPad Prism (version 5, GraphPad Software, San Diego, CA, USA) and SPSS statistical software (version 20, SPSS IBM, Armonk, NY, USA). We acknowledge EC FP6 TBVAC contract no. LSHP-CT-2003–503367, EC FP7 NEWTBVAC contract no. HEALTH-F3–2009—241745, and EC FP7 ADITEC contract no. HEALTH.2011.1.