ATP-induced autophagy is associated with rapid killing of intracellular mycobacteria within human monocytes/macrophages

Abstract

Background: We have previously reported that ATP treatment of M bovis-BCG infected human macrophages induces P2X7 receptor-dependent killing of intracellular mycobacteria. The mechanism mediating this bactericidal effect has not been full characterized but is known to be Ca2+-dependent and to promote the maturation and acidification of mycobacteria-containing phagosomes. In this study we demonstrate that the ATP/P2X7-mediated, mycobactericidal effect also involves the induction of cell autophagy.

Results: We report that 3 mM ATP induces rapid cell autophagy in THP1 cells and monocyte-derived macrophages within 30 minutes post-treatment, as revealed by the expression of LC3-II bands on western blot analysis. Using Ca2+-free media and selective P2X7 agonists and antagonists, ATP-induced cell autophagy was shown to be Ca2+ and P2X7 receptor-dependent. Electron microscopy of ATP-treated, BCG-infected MDMs revealed the presence of the bacteria within characteristic double-membraned autophagosomes. Confocal analysis further confirmed that pharmacological inhibition of autophagy by wortmannin or pre-treatment of macrophages with anti-P2X7 antibody blocked ATP-induced phago-lysosomal fusion. Induction of cell autophagy with ATP was also temporally associated with a fall in intracellular mycobacterial viability, which was suppressed by treatment with wortmannin or the selective P2X7 antagonist, oxidized ATP (oATP).

Conclusion: We provide the first evidence that ATP/P2X7-mediated killing of intracellular mycobacteria involves the induction of cell autophagy. The findings support the hypothesis that autophagy plays a key role in the control of mycobacterial infections.

Figure 1

Induction of autophagy in THP1 cells and MDMs, following treatment with ATP, is both calcium and P2X₇ – dependent. A. THP1 cells undergo autophagy (lane 1) following 30 minutes exposure to ATP (3 mM), as revealed by the presence of LC3-II (16 kDa) by Western blot. Lane 2 represents untreated control THP-1 cells. B. ATP (3 mM) treated MDMs also express LC3-II (lane 2), which is absent in non-treated control cells (lane 1). C. THP1 cells treated with 3 mM ATP for 30 minutes cultured in calcium-free medium (lane 1) and calcium-replete medium (lane 2). LC3-II expression was observed in ATP-treated cells cultured in calcium-replete medium (lane 2) but was absent in cells cultured in calcium-free media (lane 1). D. THP1 cells were treated with different P2 agonists and antagonists. Lane 1 represents untreated control cells, while lanes 2 and 3 represent cells pre-treated with oATP (0.3 mM) and anti-P2X₇ antibody (3 μg/ml) for 2 hours and 1 hour respectively, prior to treatment with ATP (3 mM for 30 minutes). Lanes 4, 5 and 6 represent cells treated with ATP (3 mM), UTP (3 mM) and bzATP (3 mM) for 30 minutes. The presence of an LC3-II band, indicating cell autophagy, was observed only in lanes 4 and 6.

Figure 2

ATP-induced autophagy triggers phagosome-lysosome fusion. A. Confocal images of MDMs treated with 3 mM ATP for 30 minutes and immuno stained for intracellular LC3. B. Confocal images of live MDMs infected with GFP-BCG (green) and pre-pulsed with Lysotracker (red) to stain acidic lysosomes. Cells were pre-incubated with anti-P2X7 (3 μg/ml) or wortmannin (100 nM) for 1 hr and then treated with 3 mM ATP for 30 minutes. Note scale bars = 10 um. C. Graph showing the percentage of Lysotracker positive, GFP-BCG containing phagosomes. Histogram shows means ± s.e.m. (n = 25–60 phagosomes). *** = P < 0.01.

Figure 3

Ultrastructural appearance of infected macrophages following exposure to ATP. A. Electron micrograph of a non-ATP treated, BCG-infected macrophage showing the presence of the bacteria within phagosomes (BP). B. Exposure to ATP is associated with the appearance of numerous double-membrane autophagic vacuoles (AV) within the cell cytoplasm. C. A BCG-infected human macrophage (MDM) following treatment with 3 mM ATP for 30 minutes, the bacteria localize within autophagic vacuoles (AV) with clearly demonstrable inner (double black arrows) and outer (double white arrows) membranes. D. A higher magnification image of a number of mycobacteria contained within the inner and outer membranes of an autophagic vacuole.

Figure 4

The mycobactericidal effect of ATP is reversed by wortmannin and oATP. MDMs, cultured for 5 days, were infected overnight with BCG (MOI of 5:1) and pulsed with ATP (3 mM) for 30 minutes, with or without pre-treatment with wortmannin (100 nM/1 hour) or oATP (0.3 mM/2 hours). The kinetics of BCG viability was determined by ³H-uridine incorporation, monitored at various times post treatment. The figure illustrates the pattern of results obtained from three separate experiments performed in triplicate. The symbols indicate mean values, and vertical bars indicate the standard error. The significance of the various treatment relative to the untreated control are illustrated by: ** = P < 0.005 * = P < 0.05. The results from one ³H-uridine uptake assay are tabulated as mean counts per minute (CPM) data +/- SEM at each time point examined.