Potential role for ESAT6 in dissemination of M. tuberculosis via human lung epithelial cells

Abstract

ESAT6 has recently been demonstrated to cause haemolysis and macrophage lysis. Our studies demonstrate that ESAT6 causes cytolysis of type 1 and type 2 pneumocytes. Both types of pneumocytes express membrane laminin, and ESAT6 exhibits dose-dependent binding to both cell types and to purified human laminin. While minimal ESAT6 was detected on the surface of Mycobacterium tuberculosis grown in vitro, exogenously provided ESAT6 specifically associated with the bacterial cell surface, and the bacterium-associated ESAT6 retained its cytolytic ability. esat6 transcripts were upregulated approximately 4- to approximately 13-fold in bacteria replicating in type 1 cells, and approximately 3- to approximately 5 fold in type 2 cells. In vivo, laminin is primarily concentrated at the basolateral surface of pneumocytes where they rest on the basement membrane, which is composed primarily of laminin and collagen. The upregulation of esat6 transcripts in bacteria replicating in pneumocytes, the specific association of ESAT6 with the bacterial surface, the binding of ESAT6 to laminin and the lysis of pneumocytes by free and bacterium-associated ESAT6 together suggest a scenario wherein Mycobacterium tuberculosis replicating in pneumocytes may utilize surface ESAT6 to anchor onto the basolateral laminin-expressing surface of the pneumocytes, and damage the cells and the basement membrane to directly disseminate through the alveolar wall.

Fig. 1

ESAT6 causes dose- and time-dependent cytolysis of type 1 and 2 pneumocytes. A and B. (A) Type 1 pneumocytes (WI26 cells) and (B) type 2 pneumocytes (A549 cells) were incubated with several concentrations of ESAT6 or CFP10 for 24, 48 and 72 h at which time MTT assay was performed to quantify lysis. Cells exposed to medium alone were used as controls; each assay condition was tested in six wells. Per cent cytolysis was calculated by using the formula: (Mean OD₅₃₀ of 6 control wells – Mean OD₅₃₀ of 6 test wells)/(Mean OD₅₃₀ of 6 control wells) × 100. For CFP10, only results obtained at 72 h are shown. These assays were done atleast 3 times and a single representative assay with each cell line is shown. C. Type 1 pneumocytes are significantly more sensitive to ESAT6-mediated cytolysis than type 2 pneumocytes at 6 μg ml^–1 (*P = 0.0098) and 10 μg ml^–1 (**P = 0.0005) of ESAT6. Per cent cytolysis obtained after 72 h of incubation of the two cell lines with different concentrations of ESAT6 is plotted. Values are mean percentage cytolysis ± SD from at least three independent experiments. D. Microscopic visualization of cytolysis of type 1 and 2 pneumocytes when exposed to 10 μg ml^–1 M. tb proteins for 72 h. Cells were visualized using a Nikon microscope at 100× magnification. Photographs shown are digitally magnified to show the monolayers.

Fig. 2

ESAT6 causes cytolysis of type 1 and 2 pneumocytes when complexed with CFP10. A. Confirmation of complex formation: Various dilutions of a 1:1 molar mixture of ESAT6 and CFP10 (containing 1, 0.2, 0.1 and 0.05 μg ml^–1 of ESAT6) were added to anti-CFP10 IgG coated wells, as were the same concentrations of ESAT6 or CFP10 alone and detected with anti-ESAT6 IgG. The Delta OD₄₉₀ (mean OD₄₉₀ with anti-ESAT6 IgG at each concentration of the complex or the individual protein minus the mean OD₄₉₀ in wells exposed to anti-ESAT6 IgG) is plotted. Immune complexes captured by anti-CFP10 and detected by anti-ESAT6 antibodies were present at all dilutions of the ESAT6 and CFP10 mixture. B. Cytolysis of type 1 and type 2 pneumocytes with ESAT6:CFP10 complexes (black bars) or by ESAT6 alone (grey bars). Both experiments were done twice with 3–6 replicates providing similar results and one representative experiment is shown.

Fig. 3

Binding of ESAT6 and CFP10 to type 1 and type 2 pneumocytes. (A) WI26 cells or (B) A549 cells were incubated with ESAT6 or CFP10 (10 or 20 μg ml^–1) or buffer (in triplicates) followed by exposure to anti-His antibodies and FITC-conjugated secondary antibodies were analysed by FACS. Only ESAT6 showed binding to the two cell types. The per cent FITC-labelled cells obtained when cells were incubated with either protein minus per cent FITC labelled cells exposed to anti-His antibodies and FITC-conjugated secondary antibody from a representative experiment is shown. The binding of ESAT6 to type 1 pneumocytes was significantly higher than type 2 cells at 10 μg ml^–1 (*P = 0.032) and similar at 20 μg ml^–1.

Fig. 4

Presence of laminin on the surface of type 1 and type 2 pneumocytes. A. Cells were exposed to anti-laminin antibodies or normal rabbit antibodies followed by FITC-conjugated anti-rabbit antibodies. Type 1 pneumocytes have more laminin/cell than type 2 pneumocytes. Cells were exposed to each antibody in triplicates and the experiment was done two times. B. Detection of laminin in membrane preparations of cells. Different concentrations of the membrane preparations were tested in triplicates and the experiment was performed twice with similar results. The OD values (mean ± SD) from one representative experiment are plotted. C. ESAT6 binds to purified laminin: increasing concentrations of ESAT6 were added to wells coated with laminin or BSA at 1 μg ml^–1 and the binding of ESAT6 detected with anti-ESAT6 IgG. Each condition was tested in triplicate and the experiment was performed twice with similar results. Values (mean ± SD) from one representative experiment are plotted. As compared with BSA the binding of ESAT6 to laminin was significantly higher at all concentrations tested (*P < 0.0001). D. Specific binding of ESAT6 to laminin: various concentrations of His-tagged M. tb proteins (ESAT6, MS and SodC) were incubated with laminin (1 μg ml^–1) coated wells and binding of M. tb proteins was detected with anti-His mAbs. Compared with SodC, the binding of ESAT6 to laminin was significantly higher at all concentrations tested (*P < 0.0001). E. Inhibition of binding of ESAT6 to laminin by anti-ESAT6 IgG. ESAT6 preincubated with anti-ESAT6 IgG (striped bars) or pre-immune IgG (dotted bars) was added to laminin-coated wells (in triplicates) and the laminin-bound ESAT6 was detected by anti-His IgG. Per cent inhibition of binding of ESAT6 to laminin from one representative experiment is plotted. Inhibition of ESAT6–laminin interaction by anti-ESAT6 IgG was significantly higher compared with inhibition by pre-immune IgG at both dilutions tested (*P = 0.009 at 1:500 and **P = 0.037 at 1:1000).

Fig. 5

Paucity of ESAT6 in and on the M. tb cell wall. A. M. tb cell wall preparations were tested for presence of ESAT6 (grey bars) or CFP10 (black bars) by ELISA with respective antibodies. B. γ-Irradiated M. tb H₃₇Rv, H₃₇Rv:ΔESAT6 and H₃₇Rv:ΔRD1 were coated at various concentrations in ELISA plates and the respective antibodies used to detect ESAT6 or CFP10 on the surface of the intact bacterial cells. ΔOD₄₉₀ = Mean OD₄₉₀ with anti-protein IgG at any protein/bacterial concentration – OD₄₉₀ with pre-immune IgG at the same concentration. C–F. Immunoelectron microscopy of ultrathin sections of γ-irradiated M. tb probed with anti-ESAT6 IgG (C), pre-immune IgG from the ESAT-6 immunized animal (D), anti-CFP10 IgG (E) and pre-immune IgG (F) from the CFP10 immunized animal.

Fig. 6

Association of exogenous ESAT6 with M. tb cell surface. A. Re-association of exogenously added His-tagged ESAT6 or CFP10 (2 μg ml^–1) with M. tb H₃₇Rv surface as detected by anti-His Abs. B–D. Re-association of ESAT6 with surface of M. tb H₃₇Rv (B), M. tb H₃₇Rv:Δcfp10 (C) or M. tb H₃₇Rv:ΔRD1 (D) as detected by anti-ESAT6 IgG. E. Cell surface associated ESAT6 is cytolytic. Type 1 pneumocytes and type 2 pneumocytes incubated with various dilutions of γ-irradiated M. tb H₃₇Rv:ΔRD1 carrying reassociated ESAT6 for 72 h were tested by MTT assay for cytolysis and mean OD₅₃₀ from 3 replicates at each concentration were used to calculate the per cent cytolysis.

Fig. 7

esat6 transcripts are upregulated in M. tb replicating in type 1 and 2 pneumocytes. The fold change (ratios of copies of esat6 or 23S rRNA normalized to 16S rRNA copies in bacteria from type 1 or 2 pneumocytes to the normalized copies of respective genes in broth grown bacteria) ± standard deviation (error bars) from triplicates of a representative of two experiments are shown.