By Phagosome-Lysosome Fusion This Information Is Current As of September 26, 2021

Home , Lysosome, Phagosome

ATP-Mediated Killing of Intracellular Mycobacteria by Macrophages Is a P2X 7 -Dependent Process Inducing Bacterial Death by Phagosome-Lysosome Fusion This information is current as of September 26, 2021. Ian P. Fairbairn, Carmel B. Stober, Dinakantha S. Kumararatne and David A. Lammas J Immunol 2001; 167:3300-3307; ; doi: 10.4049/jimmunol.167.6.3300

http://www.jimmunol.org/content/167/6/3300 Downloaded from

References This article cites 46 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/167/6/3300.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 26, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. ATP-Mediated Killing of Intracellular Mycobacteria by

Macrophages Is a P2X7-Dependent Process Inducing Bacterial Death by Phagosome-Lysosome Fusion1

Ian P. Fairbairn,2 Carmel B. Stober, Dinakantha S. Kumararatne, and David A. Lammas

Mycobacterium tuberculosis survives within host macrophages by actively inhibiting phagosome fusion with lysosomes. Treatment of infected macrophages with ATP induces both cell apoptosis and rapid killing of intracellular mycobacteria. The following studies were undertaken to characterize the effector pathway(s) involved. Macrophages were obtained from p47phox and inducible

NO synthase gene-disrupted mice (which are unable to produce reactive oxygen and nitrogen radicals, respectively) and P2X7 gene-disrupted mice. RAW murine macrophages transfected with either the natural resistance-associated macrophage protein

gene 1 (Nramp1)-resistant or Nramp1-susceptible gene were also used. The cells were infected with bacille Calmette-Gue´rin (BCG), Downloaded from and intracellular mycobacterial trafﬁcking was analyzed using confocal and electron microscopy. P2X7 receptor activation was ؊/؊ essential for effective ATP-induced mycobacterial killing, as its bactericidal activity was radically diminished in P2X7 mac- ,rophages. ATP-mediated killing of BCG within p47phox؊/؊, inducible NO synthase؊/؊, and Nramps cells was unaffected demonstrating that none of these mechanisms have a role in the ATP/P2X7 effector pathway. Following ATP stimulation, BCG- containing phagosomes rapidly coalesce and fuse with lysosomes. Blocking of macrophage phospholipase D activity with butan-

1-ol blocked BCG killing, but not macrophage death. ATP stimulates phagosome-lysosome fusion with concomitant mycobacterial http://www.jimmunol.org/ death via P2X7 receptor activation. Macrophage death and mycobacterial killing induced by the ATP/P2X7 signaling pathway can be uncoupled, and diverge proximal to phospholipase D activation. The Journal of Immunology, 2001, 167: 3300–3307.

pproximately one-third of the world’s population is in- sively unresponsive to further activation by Th1-derived cytokines fected with Mycobacterium tuberculosis, yet 90% of in- as the pathogen undergoes intracellular replication (12). A fected individuals remain healthy (1). Of these, Ͻ10% Innate resistance to intracellular infections is also thought to be will exhibit reactivation disease within their lifetime. Therefore, in genetically determined, influencing the outcome of infection. The the vast majority of individuals, their immune systems can effec- best characterized is the natural resistance-associated macrophage tively control the infection. protein gene 1 (Nramp1)3 (13). This gene codes for a membrane by guest on September 26, 2021 From HIV patients (2) and certain rare immunodeficient indi- transporter protein that acts to inhibit the replication of various viduals (3, 4), we know that human immunity to mycobacterial intracellular pathogens, including bacille Calmette-Gue´rin (BCG), ϩ infections constitutes an interplay between Ag-specific CD4 Th1 by excluding Fe2ϩ (14, 15) and other cations from the pathogen- lymphocytes and host macrophages. The macrophages promote containing phagosomes and promoting its acidification (16). Th1 activity by production of IL-12 (3), IL-18 (5), and IL-23 (6), The macrophage acts as both the major reservoir of infection and the lymphocytes activate macrophages through the release of and the principal effector cell acting to regulate mycobacterial in- ␥ ␣ IFN- (4) and TNF- (7). These two sets of effector cells are then fections in vivo. thought to interact in a positive feedback manner to contain the A major drawback in our understanding of the effector path- infection within T cell-activated macrophages. way(s) operating within human macrophages is the fact that, unlike Mycobacteria, in turn, are thought to survive and replicate in murine model cell systems (17), no mycobactericidal effect is within nonactivated macrophages through a variety of strategies, observed in vitro following stimulation of macrophages with Th1- including the avoidance of generation of reactive oxygen (8) and associated cytokines, including IFN-␥ and TNF-␣ (18). This has nitrogen intermediates (9) and the active inhibition of phagosome been attributed to the inability of such mediators to induce NO maturation (10, 11). The latter strategy preventing exposure to the generation in human cells. However, we know that in vivo mac- antimicrobial peptides and enzymes presents within the acidic ly- rophages are able to contain these organisms in the vast majority sosomal granules. Infected cells are thought to become progres- of infected individuals. This has initiated a search for alternative stimuli that may regulate mycobacteristatic or mycobactericidal activity within infected human cells. Medical Research Council Centre for Immune Regulation, Birmingham Medical We have previously reported that a 15-min exposure to 1–3 mM School, Birmingham University, Edgbaston, Birmingham, United Kingdom ATP results in the apoptosis of BCG-infected monocyte-derived Received for publication October 10, 2000. Accepted for publication July 10, 2001. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 3 Abbreviations used in this paper: Nramp1, natural resistance-associated macrophage 1 I.P.F. is a Wellcome clinical research fellow. protein 1; ADC, albumin dextrose catalase; BCG, bacille Calmette-Gue´rin; BMDM, 2 Address correspondence and reprint requests to Dr. Ian P. Fairbairn, Medical bone marrow-derived macrophage; DAMP, 3-(2,4-dinitroanilino)-3Ј-amino-N-methyl Research Council Centre for Immune Regulation, Birmingham Medical School, dipropylamine; EM, electron microscopy; GFP, green fluorescent protein; iNOS, in- G Birmingham University, Edgbaston, Birmingham, B15 2TT, U.K. E-mail address: ducible NO synthase; L-NMMA, N -monomethyl-L-arginine monoacetate; PLA2, [email protected] phospholipase A2; PLD, phospholipase D; WT, wild type.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00 The Journal of Immunology 3301 human macrophages within 6 h, with a concomitant 50–70% re- protein 60 promoter in a pSMT3 shuttle vector (27) construct. Stock ali- duction in the viability of intracellular mycobacteria. ATP treat- quots, stored in glycerol at Ϫ70°C, were grown to log phase in 7H9 broth ment of a macrophage-free culture of BCG had no effect on bac- supplemented with 10% ADC enrichment medium, 0.2% Tween 80, and 50 ␮g/ml hygromycin (Sigma, St. Louis, MO). terial viability, demonstrating this to be a cell-mediated mechanism (19). Virulent strains of tuberculosis were also found BCG viability assay to be equally as susceptible to this mechanism (20). Macrophages were grown to subconfluence in 200 ␮l/well cultures in 96- 4 The effects of ATP were shown to be mediated by P2X7 recep- well round-bottom microtiter plates. This approximates to 5 ϫ 10 cells in tors, as the cytocidal and bactericidal activity of ATP were blocked 200 ␮l/well. A total of 5 ϫ 105 BCG Evans per well was added. To aid infection, plates were centrifuged at 700 ϫ g for 10 min and incubated at by pretreatment of cells with P2X7 antagonists, oxidized ATP, and KN62, and potentiated by the specific agonist benzoyl-benzyl 37°C for 60 min. Excess BCG were removed by washing. As assessed by CFU, ϳ10% of BCG was taken up, giving a ratio of one viable BCG per ATP (19). macrophage. Where appropriate, the following metabolic inhibitors were G We have shown previously that the mycobactericidal effects of then added for 60 min. N -Monomethyl-L-arginine monoacetate (L- ATP are independent of both reactive oxygen and nitrogen inter- NMMA) (400 ␮M), superoxide dismutase (15 U/ml), catalase (0.1 mg/ml), mediate generation. ATP was found to be equally effective on cells and butan-1-ol (0.05–1%) were all purchased from Sigma. A 30 mM ATP (Boehringer Mannheim, Indianapolis, IN) solution was prepared in culture derived from patients with chronic granulomatous disease as medium, adjusted to pH 7 with sodium bicarbonate, and passed through a against cells from healthy controls. Furthermore, BCG killing oc- 0.22-␮M filter (Millipore, Bedford, MA) to sterilize. It was added to the curred even when such cells were additionally treated in vitro with macrophage cultures to give a final concentration 3 mM, and incubated at NO inhibitors (19). 37°C, 5% CO2. At appropriate time points, the plates were centrifuged at ϫ ␮ Electron microscopy (EM) studies have revealed that ATP-in- 700 g for 10 min. A total of 150 l supernatant was removed and Downloaded from replaced with 130 ␮l 7H9 medium, supplemented as above, plus 20 ␮l2% duced cell death of BCG-infected human macrophages is associ- saponin solution. Cell lysis was promoted by vigorous pipetting and then ated with the formation of cytoplasmic vacuoles into which the incubation at 37°C for 120 min. Fifty-microliter aliquots from each well intracellular bacteria coalesce (21). This was thought to illustrate were serially diluted in 7H9 supplemented with 10% ADC enrichment and the promotion of phagosome-lysosome fusion and/or modulation 0.2% glycerol and plated in 96-well flat-bottom plates. The plates were then incubated at 37°C for 12 days. Microcolonies were counted using an of the phagosome pH within ATP-treated cells, resulting in the inverted microscope (ϫ100 magnification). All experiments were per- destruction of the mycobacteria. formed in quadruplets. http://www.jimmunol.org/ In this study, we report the use of macrophages derived from various gene-disrupted (knockout) mice to further characterize the Lucifer yellow uptake ATP effector mechanism involved and to confirm the selective Pore formation was assessed using a method adapted from that described involvement of the P2X receptor in this mycobactericidal effect. by Steinberg et al. (28). Briefly, macrophages adherent to coverslips were 7 ␮ ␮ Using EM and confocal fluorescent microscopy, we have further incubated with 500 g/ml lucifer yellow dye, 250 M sulfinpyrazone (Sig- ma), and 3 mM ATP in PBS for 5 min. They were washed in 250 ␮M analyzed the multibacillary vacuole formation that occurs follow- sulfinpyrazone in PBS and viewed immediately on a Zeiss confocal laser- ing ATP stimulation. Our findings suggest these play a critical role scanning microscope (Zeiss, Oberkochen, Germany). Cells with pores in mediating the death of mycobacteria as they progressively acid- larger than 463 Da exhibit intracellular staining. ify and fuse with cell lysosomes. Finally, we also demonstrate a by guest on September 26, 2021 Chromium-release assay key role for phospholipase D (PLD) in the P2X7 effector pathway, supporting similar findings reported by Kusner et al. (20). Mouse BMDMs were cultured in 96-well round-bottom microtiter plates, as previously described. Aliquots of 51Cr (ICN Pharmaceuticals, Costa Mesa, CA) were added at a concentration of 2 ␮Ci/well, and the plates Materials and Methods were incubated for 16 h. Cells were washed and incubated with 3 mM ATP Cells and bacteria for a further 16 h. Supernatants were removed, and their activity was counted on a Wallac 1260 Multi Gamma II gamma counter (PerkinElmer The murine macrophage cell line J774a.1 was obtained from the American Wallac, Gaithersburg, MD). The cells themselves were dissolved in 1% Type Culture Collection (Manassas, VA). The cells were cultured in Triton X-100 solution and also counted. Results are expressed as the per- DMEM (Life Technologies, Paisley, U.K.) supplemented with 10% FCS centage of the total activity in the supernatant. and 2 mM glutamine. The murine macrophage cell lines RAW 246.7, wild- type (WT) RAW 7.5R, and RAW 10S were a kind gift from J. M. Black- Confocal fluorescent microscopy well (Cambridge University, Cambridge, U.K.). The 7.5R and 10S lines are stable transfectants, the generation of which has been described previously Macrophages were grown to subconfluence on 13-mm glass coverslips in (22). These were cultured in RPMI (Life Technologies) supplemented with 24-well plates. This approximated to 5 ϫ 105 cells/coverslip. GFP-BCG Ϫ/Ϫ phoxϪ/Ϫ 10% FCS and 2 mM glutamine. P2X7 , p47 , and inducible NO were resuspended in DMEM, and a single cell culture was prepared by synthase (iNOS)Ϫ/Ϫ mice were kind gifts from I. P. Chessell (Glaxo Well- water bath sonication for 5 min. Any remaining clumps were removed by come, Cambridge, U.K.), A. W. Segal (University College of London, slow speed centrifugation (100 ϫ g for 5 min). The bacterial concentration London, U.K.), and F. Y. Liew (University of Glasgow, Glasgow, U.K.), and confirmation of a single cell suspension were determined using a respectively, and have been previously described (23–25). The genetic thoma chamber. The BCG were then added to the macrophages to give a backgrounds of these knockout mice were, respectively, C57/129, C57, and final concentration of 2.5 ϫ 107/coverslip. The plate was centrifuged at ϫ 129. WT C57BL/6 and 129 were used as controls, as appropriate. Bone 700 g for 10 min. The cells were incubated for1hat37°C, 5% CO2, marrow-derived macrophages (BMDMs) were obtained as follows. Femurs washed three times in DMEM to remove any extracellular bacteria, and and tibias were flushed with DMEM, and the bone marrow was washed and incubated for an additional hour at 37°C, 5% CO2. plated out overnight in DMEM supplemented with 10% FCS, 10% GM- LysoTracker (Molecular Probes, Eugene, OR) was added to the cells CSF-containing medium, and 2 mM glutamine. Nonadherent cells were (final concentration of 50 nM) for 1 h before stimulation. ATP was then then removed, washed, and plated at 106/ml in the above medium. The cells added (final concentration of 3 mM), and the cells were incubated for the were cultured for a further 9 days, with changes of the medium on days 3 indicated times. and 6. On day 9, the cells were adherent and morphologically looked like Before fixation, coverslips were washed three times in HBSS (Sigma). mature macrophages. They were then fixed in 3% paraformaldehyde containing 10 mM lysine, M. bovis BCG Evans strain was obtained from Evans Medical (Speke, 140 mM saline, and 40 mM HEPES (pH 7.3) for 15 min at room temper- U.K.) and maintained in log phase growth in 7H9 broth supplemented with ature. After washing in PBS plus 5% dextrose, the coverslip cultures were 10% albumin dextrose catalase (ADC) enrichment medium and 0.2% mounted on glass slides with Perma Fluor (Lipshaw Manufacturing, De- Tween 80. Green fluorescent protein expressing BCG (GFP-BCG) was a troit, MI), and the edges were sealed with nail varnish. kind gift from D. Young (Imperial College of London, London, U.K.). Slides were viewed with a Zeiss confocal laser-scanning microscope. These contained the gene encoding a FACS-optimized GFP protein (26) Twenty fields were examined per slide, and representative pictures were constitutively expressed under the control of the mycobacterial heat shock taken. 3302 ATP INDUCES MYCOBACTERIA DEATH VIA PHAGOSOME-LYSOSOME FUSION

Electron microscopy rapidly take up the cell-impermeable dye lucifer yellow following The 3-(2,4-dinitroanilino)-3Ј-amino-N-methyl dipropylamine (DAMP; treatment with ATP for 5 min, indicating P2X7 pore formation had Ϫ/Ϫ Molecular Probes) staining was adapted from Crowle et al. (29). BMDMs occurred (Fig. 2, A and B), whereas cells from P2X7 mice were cultured as described above on 13-mm plastic coverslips and infected failed to do so (Fig. 2, C and D). WT BMDMs were also shown to with BCG Evans at an infection ratio of 5 BCG to 1 macrophage. ATP (3 die in response to ATP stimulation over a 16-h period, as indicated mM final) was then added, and 30 min before the end of the experiment by 51Cr release, whereas macrophages from P2X Ϫ/Ϫ mice re- DAMP (50 ␮M final) was also added. The coverslips were washed three 7 times in PBS and fixed in 2% gluteraldehyde in 0.1 M calcodylate and mained viable (Fig. 3a). 2.5% sucrose (pH 7.35) for 60 min at 4°C. They were washed in 50 mM ammonia chloride, and then progressively dehydrated in ethanol, infiltrated with London Resin white (Agar Scientific, Stanstead, U.K.), and polymer- ATP-mediated killing of intracellular mycobacteria is P2X7 ized with UV light. Ultrathin sections were collected on nickel grids, which dependent were incubated for 10 min with 0.1 M HCl. These were then blocked with Ϫ/Ϫ 0.2% BSA and 0.2% goat serum in PBS for 45 min at room temperature. Having demonstrated that macrophages from P2X7 mice had The primary Ab was a mouse antidinitrophenol monoclonal (NovoCastra, no P2X -associated cytotoxic activity, they were then assessed for Newcastle, U.K.). A mouse IgG1 control Ab (BD PharMingen, San Diego, 7 CA) was also used to exclude nonspecific staining. These were added at a their ability to kill intracellular mycobacteria following ATP stim- concentration of 0.5 ng/ml for 16 h at 4°C. The secondary was a goat ulation. BMDMs were infected with BCG and exposed to ATP (3 antimouse polyclonal Ab conjugated to 5 nM gold particles (British Biocell mM) for 2 h. Compared with WT cells, macrophages from International, Cardiff U.K.). This was used at a dilution of 1/50 and incu- Ϫ/Ϫ P2X7 mice were shown to be ineffective at killing BCG (Fig. bated with the cells for 90 min at room temperature. Grids were counter- 3b). The percentage of BCG killed was reduced from 50% in ATP- stained with urinyl acetate and viewed on a Joel Temscan (Wellwyn Gar- Downloaded from Ͻ Ϫ/Ϫ den City, U.K.) electron microscope. Pictures were taken on Agfa Scientia treated WT cells to 10% in P2X7 cells. Although some re- (Agar Scientific) EM film. The number of gold particles per vacuole was sidual mycobactericidal activity was still occurring, the results counted. Counts were performed for 20 vacuoles, involving at least 10 show that P2X7 is the primary transducer of the intracellular my- macrophages for each experiment. cobactericidal effector pathway activated following ATP stimula- Statistics tion of infected macrophages.

The Student’s t test was used for all statistical analysis. http://www.jimmunol.org/ PLD activity is essential for ATP/P2X -mediated BCG killing, Results 7 but not macrophage death Macrophage BCG killing kinetics post-ATP stimulation PLD is involved in a number of signaling cascades via the pro- ATP stimulation of BCG-infected J774 cells resulted in ϳ50% duction of phosphatidic acid from phosphorylcholine. If cells are loss of mycobacterial CFU within 60 min posttreatment (Fig. 1). pretreated with butan-1-ol, PLD preferentially uses the primary Loss of mycobacterial viability was detectable within 15 min and alcohol as a substrate producing the nonsignaling phosphatidyl bu- maximal at 2 h posttreatment. Subsequently, all viability studies tanol (30). Treatment with butan-1-ol after BCG infection, but be- were performed at 2 h post-ATP stimulation. fore ATP stimulation, showed a concentration-dependent inhibi- by guest on September 26, 2021 Similar BCG killing kinetics was observed in another murine tion of BCG killing post-ATP (Fig. 4a), but had no effect on macrophage cell line (RAW), and with the human monocyte cell macrophage death, as measured by chromium release (Fig. 4b). line (THP-1), and in murine BMDMs from various mouse strains (C57BL/6, BALB/c, and 129) (results not shown). The ATP/P2X7 effector mechanism was conﬁrmed as a cell- Radical oxygen intermediates and radical nitrogen intermediates mediated function, as ATP had no effect on BCG grown in mac- are not required for BCG killing post-ATP/P2X stimulation rophage-free cultures (results not shown). 7 Experiments were performed to investigate whether the ATP-in-

ATP-mediated pore formation and macrophage death are P2X7 duced mycobactericidal effects observed in murine macrophages dependent were attributable to the generation of reactive oxygen or reactive Ϫ/Ϫ nitrogen intermediates. BMDMs from P2X7 mice were examined for their responses following ATP stimulation. Cells from WT mice were shown to The high levels of NO required for killing mycobacteria within murine macrophages are generated by the enzyme iNOS after appropriate priming and stimulation of the cells. Macrophages from iNOSϪ/Ϫ mice are therefore unable to up-regulate nitrogen radical production (25). p47phox is a key enzyme in the NADPH reductase pathway, and mice with a targeted disruption of this gene are unable to produce reactive oxygen radicals (24). BMDMs from iNOSϪ/Ϫ and p47phoxϪ/Ϫ mice and WT littermates were infected with BCG and treated with ATP (3 mM) for 2 h. The results (Fig. 5) reveal that ATP was equally effective in inducing bacterial death within iNOSϪ/Ϫ and p47phoxϪ/Ϫ macrophages as in WT cells. FIGURE 1. BCG viability post-ATP stimulation of J774 macrophages. To conﬁrm that production of neither oxygen or nitrogen radi- Macrophages of the murine cell line J774 were infected with BCG for 60 cals was required for the ATP effect, nitrogen radical production min. Extracellular BCG were washed off, and the macrophages were then was blocked in the p47phoxϪ/Ϫ macrophages using the NO inhib- stimulated with 3 mM ATP. After the indicated time, they were lysed and itor L-NMMA, and oxygen radical production was blocked in the assayed for BCG CFU. The time course shows rapid cell death within 15 Ϫ/Ϫ min, with a maximum response within 60 min. SDs for quadruplets are iNOS macrophages using superoxide dismutase and catalase shown, and the results are representative of three experiments. Similar (Fig. 5). ATP killing of BCG was shown to be equally as effective results were found with human monocyte-derived macrophages, RAW in all cell treatments. The data reveal that oxygen and nitrogen macrophages, and BMDMs from BALB/c, C57BL/6, and 129 mice. radicals do not have a role in the ATP effector mechanism. The Journal of Immunology 3303

Ϫ/Ϫ FIGURE 2. Macrophage membrane permeabilization following ATP stimulation. Macrophages were cultured from the bone marrow of P2X7 mice and WT littermates. They were stimulated with 3 mM ATP for 5 min in the presence of the cell-impermeable dye lucifer yellow. The cells were then washed and immediately viewed with confocal ﬂuorescent (A and C) and transmission (B and D) microscopy. A and B demonstrate stimulated WT macrophages Ϫ/Ϫ that are permeable to the dye, whereas C and D demonstrate stimulated macrophages from P2X7 mice whose membranes have remained impermeable. The pictures are representative of numerous ﬁelds from three experiments.

Killing post-P2X7 stimulation is not affected by Nramp1 10S), were used. ATP treatment of BCG-infected 7.5R and 10S genotype cells was shown to be equally as effective in inducing killing of The susceptibility of inbred mice strains to BCG is associated with BCG, as was observed in the WT RAW cells (Fig. 6). This indicates that expression of Nramp1 does not affect the ability of RAW a single point mutation of Nramp1 (31). Both the RAW and J774 Downloaded from macrophage cell lines are derived from BALB/c mice, which carry cells to kill BCG in response to ATP stimulation. the susceptible mutation of Nramp1 (31). Further support for this ﬁnding is provided by the observation To examine the possible inﬂuence of Nramp1 on the ATP-bac- that BMDMs from BALB/c, C57BL/6, and 129 mice were all terial killing effect, RAW cells, stably transfected with either the equally effective at killing intracellular mycobacteria following resistant Nramp1 (RAW 7.5R) or the susceptible Nramp1 (RAW ATP stimulation. BALB/c and C57BL/6 mice are known to be susceptible to BCG infection and carry the susceptible Nramp1 mutation, whereas 129 mice are a resistant strain and carry the http://www.jimmunol.org/ by guest on September 26, 2021

Ϫ/Ϫ FIGURE 3. a, Chromium release post-ATP stimulation of P2X7 Ϫ/Ϫ BMDMs. b, Intracellular BCG killing by ATP-stimulated P2X7 Ϫ/Ϫ BMDMs. BMDMs were cultured from P2X7 mice and WT littermates. a, Macrophages were labeled overnight with 51Cr. Excess chromium was removed. Where indicated, they were then stimulated for 16 h with 3 mM ATP. ␥ Activity was counted in both the supernatant and cells. Results are FIGURE 4. a, Effect of PLD inhibition on BCG killing by J774 mac- expressed as percentage released over total. SDs for quadruplets are shown, rophages post-ATP stimulation. J774 macrophages were infected with and the results are representative of three experiments. After stimulation, BCG and then treated for 60 min with the indicated concentrations of the amount of chromium released by the WT macrophages increases, in- butan-1-ol to inhibit PLD activity. They were then stimulated with 3 mM dicating that cell death had occurred. There was no response in the ATP for 2 h. BCG viability was assessed and expressed as percentage of Ϫ/Ϫ P2X7 macrophages. b, The macrophages were infected with BCG for BCG killed. Inhibition of PLD activity completely inhibits BCG killing. 60 min, and extracellular BCG were washed off. Where appropriate, mac- SDs for quadruplets are shown, and the results are representative of three rophages were stimulated with 3 mM ATP for 120 min. They were then experiments. b, Effect of PLD inhibition on macrophage death post-ATP lysed and assayed for BCG CFU. The percentage of intracellular BCG stimulation. 51Cr-labeled macrophages were pretreated with the indicated killed is calculated by comparing the CFU from stimulated with the CFU concentrations of butan-1-ol for 60 min before 16-h stimulation with 3 mM from unstimulated macrophages. SDs for four experiments, each done in ATP. Inhibition of PLD activity has no effect on macrophage death. SDs Ϫ/Ϫ quadruplet, are shown. WT, but not P2X7 , macrophages are effective at for quadruplets are shown, and the results are representative of three killing BCG. experiments. 3304 ATP INDUCES MYCOBACTERIA DEATH VIA PHAGOSOME-LYSOSOME FUSION

BCG in non-ATP-treated cells (Fig. 7, A and B) appeared green, indicating that they were residing in nonacidic vesicles separate from the red-staining lysosomes. However, in ATP-treated cells (Fig. 7, C and D), an increasing percentage of the BCG appeared yellow (due to colocalization with the red LysoTracker), indicating that they were residing within acidic organelles that had fused with lysosomes. To quantify the colocalization of the BCG, serial ob- servations were expressed as the percentage of colocalized BCG over total intracellular BCG (Fig. 8). The results illustrate the pro- gressive acidiﬁcation of the BCG-containing phagosomes. These experiments show that BCG-containing phagosomes rapidly fuse with lysosomes following ATP stimulation. Experiments using UTP, GTP, and AMP failed to show phagosome-lysosome fusion (results not shown), suggesting that ATP

was acting via the P2X7 receptor. In addition, phagosome-lyso- FIGURE 5. Intracellular BCG killing by BMDMs does not require ox- Ϫ/Ϫ some fusion was absent in ATP-stimulated P2X7 macrophages ygen or nitrogen radicals. BMDMs were cultured from p47phoxϪ/Ϫ, Ϫ/Ϫ (Fig. 7, E and F). The data demonstrate that the phagosome-lyso- iNOS , and WT littermates. They were infected with BCG before being some fusion that occurs following ATP stimulation is dependent

stimulated with 3 mM ATP for 2 h. BCG viability was assessed by CFU. Downloaded from on and is a function of P2X activation. Results are expressed as percentage of BCG killed. Where indicated, 400 7 ␮M L-NMMA, 15 U/ml superoxide dismutase, or 0.1 mg/ml catalase was added 60 min before stimulation. SDs for quadruplets are shown, and the BCG-containing phagosomes fuse to form acidified results are representative of two to three experiments. In all scenarios, the multibacillary vacuoles post-P2X7 stimulation: EM studies macrophages remained effective at killing BCG, indicating that neither High resolution transmission EM was used to investigate the mor- nitrogen nor oxygen radicals are required. phological changes accompanying the fusion of BCG-containing http://www.jimmunol.org/ phagosomes with lysosomes following ATP stimulation. ATP-in- resistant gene (31). In total, the data suggest that Nramp1 does not duced acidification of BCG-containing phagosomes was confirmed have a role in ATP-mediated killing of intracellular mycobacteria. by the prestaining of cells with DAMP, an acidophilic dye that has a preferential affinity for acidic organelles such as lysosomes.

BCG-containing phagosomes fuse with lysosomes post-P2X7 DAMP staining was then detected using a gold-labeled anti- stimulation: confocal fluorescent microscopy studies dinitrophenol Ab. Fig. 9, A and B, illustrates a non-ATP-stimulated Mycobacteria survive within macrophages by inhibiting the mat- macrophage in which the BCG-containing phagosome has little to uration of their phagosomes, thus avoiding fusion with lysosomes no gold labeling, and is therefore shown not to be acidic. This (10). Lysosomes provide a potentially toxic environment for my- supports previous findings (10, 11, 32) that BCG normally reside by guest on September 26, 2021 cobacteria containing various hydrolases and cationic proteins in within individual nonacidic phagosomes that do not fuse with ly- an acidic environment. The promotion of phagosome-lysosome fu- sosomes. Within 30 min of ATP stimulation, marked changes sion by ATP was therefore investigated as a potential mycobacte- could be observed within infected macrophages (Fig. 9, C and D). ϫ rial-killing mechanism. At low power ( 5000), the BCG were observed to coalesce into multibacillary vacuoles (Fig. 9C). These are heavily labeled with Confocal fluorescent microscopy was used to show that P2X7 stimulation of infected BMDMs resulted in the rapid fusion of the DAMP (Fig. 9D), indicative of vacuole acidification, in contrast to BCG-containing phagosomes with lysosomes (Fig. 7). The acido- the non-DAMP-labeled individual BCG-containing phagosomes philic dye LysoTracker (Molecular Probes) was used to prelabel seen in unstimulated cells. the cell lysosomes red, and GFP-BCG were used to infect the By counting the number of gold particles per vacuole, the BMDMs. When analyzed by confocal fluorescent microscopy, the amount of DAMP staining in the various phagosomes/vacuoles was quantified (Fig. 10). Overall, the results confirm and illustrate the rapid acidification of phagosomes that occurs post-ATP stimulation. In addition, the EM studies illustrate that this involves the fusion of individual BCG-containing phagosomes both with each other and with lysosomes. To confirm that the formation of acidified multibacillary vacu-

oles was speciﬁcally due to ATP stimulation of P2X7 receptors, the Ϫ/Ϫ EM study was repeated using cells derived from P2X7 mice. Fig. 9, E and F, illustrates that at 2 h post-ATP stimulation of Ϫ/Ϫ P2X7 BMDMs, the BCG-containing phagosomes have neither acidiﬁed nor coalesced.

Discussion FIGURE 6. ATP-stimulated killing of BCG by RAW macrophages In this study, loss of intracellular BCG viability was shown to transfected with resistant or susceptible Nramp1. The 7.5R and 10S RAW occur within 15 min following ATP stimulation of infected murine macrophages, which have been respectively transfected with the resistant and susceptible Nramp1 genes, were infected with BCG and stimulated macrophages. The rapidity of the effect excluded a role for either with 3 mM ATP for 2 h. BCG viability was assessed by CFU and expressed nutrient and/or iron deprivation as the antimycobacterial mechanism as the percentage of BCG killed. Ability to kill post-ATP is independent of responsible. The speed of the bactericidal activity indicates the which Nramp1 gene has been transfected. SDs for quadruplets are shown, involvement of a preformed toxic mediator that does not require de and the results are representative of three experiments. novo synthesis. This is in agreement with our previous ﬁndings, in The Journal of Immunology 3305

FIGURE 7. Confocal fluorescent microscopy images showing BCG and lysosome localization. Confocal fluorescent (A, C, and E) and transmission (B, D, and F) microscopy of macrophages infected with GFP-BCG and incubated with LysoTracker to stain lysosomes red. A and B, An unstimulated macrophage containing a green BCG separate from the red lysosomes. C and D, A macrophage that has been stimulated for 240 min with 3 mM ATP. The Downloaded from intracellular BCG now appear yellow, indicating that colocalization and phagosome-lysosome fusion have occurred. The extracellular BCG remains green. Ϫ/Ϫ E and F, A P2X7 macrophage that has been stimulated with ATP. No colocalization has occurred. Representative pictures from at least three experiments are shown. which we showed that the protein synthesis inhibitor cyclohexi- involvement of an additional purinergic receptor that may contrib- http://www.jimmunol.org/ mide was unable to block ATP-induced BCG killing (19). ute to the loss of bacterial viability seen in ATP-treated cells. We The kinetics of ATP-induced BCG killing was more rapid than have provisionally characterized this additional purine receptor as induction of cell death. However, although ATP-stimulated cell P2Y11 (35). death and BCG killing were both shown to be specifically P2X7 Crossover studies using inhibitors to block both nitrogen radical dependent, their effector signaling pathways were found to diverge production in p47phoxϪ/Ϫ BMDMs and oxygen radical production Ϫ/Ϫ downstream of P2X7 activation. Blocking of cellular PLD activity in iNOS BMDMs also demonstrated unequivocally that neither with butan-1-ol then inhibited BCG killing, but not macrophage oxygen nor nitrogen radicals have a role in ATP killing of intra- death in response to ATP stimulation. cellular BCG. This confirms our previous findings, which showed The P2X7 receptor (formally known as P2Z) is highly expressed that monocyte-derived macrophages from patients with chronic by guest on September 26, 2021 on cells of the immune system, predominately cells of monocytic granulomatous disease are able to effectively kill BCG following lineage (33). However, macrophages are thought to express a num- ATP stimulation, and that the addition of the nitrogen radical in- ber of additional purinergic receptors, including: P2Y 1,2,4,5,6,10,11 hibitor L-NMMA was also unable to inhibit ATP-mediated BCG and P2X (34), some of which can respond to ATP with vary- 1,4,5,7 killing in these cells (19). ing sensitivity. Ϫ Ϫ Nramp1 is an important mechanism in murine defenses against The fact that P2X / macrophages are unable to effectively kill 7 various intracellular infections including BCG, and has been re- BCG in response to ATP demonstrates unequivocally that P2X7 is the major receptor involved in this specific effector pathway. How- ported to be associated with human resistance to tuberculosis (36). ever, we also observed a small, but nonsignificant level of resid- Nramp1 is also known to be expressed on phagosome membranes Ϫ/Ϫ (37), and has been proposed to play a role in both iron transport ual BCG killing within ATP-treated P2X7 cells. This is in agreement with Kusner et al. (20), who noted that ATP was and phagosome acidification (16). In this study, Nramp1 was not found to be involved in the ATP-induced mycobactericidal effector slightly more effective than the more specific P2X7 agonist benzoyl-benzyl ATP. Taken together, these findings indicate the possible process, as macrophages with either the susceptible or resistant form of the Nramp1 allele were shown to be equally capable of killing intracellular BCG following ATP stimulation. The molecular probes LysoTracker and DAMP are weak bases that are used to study phagosome acidification because they are selectively taken up within cells by acidic organelles. In this study, both LysoTracker and DAMP preferentially targeted BCG-containing phagosomes following ATP stimulation of infected macrophages, indicating that phagosome-lysosome fusion had occurred. EM studies using DAMP labeling also illustrated that the BCG-containing phagosomes both acidified and fused with each other following ATP stimulation. This is in agreement with Molloy FIGURE 8. Colocalization between intracellular, BCG, and lysosomes et al. (21), who also reported similar fusion of mycobacteria-con- following ATP stimulation. The results from experiments similar to those taining phagosomes following ATP treatment. that gave the images shown in Fig. 7 were quantified. BCG that appeared Importantly, the phagosome acidification and phagosome fusion yellow were said to be colocalized and expressed as a percentage of total intracellular BCG. Over 200 BCG from at least 30 macrophages and three induced by ATP were not seen in BMDMs from P2X7 gene-dis- different experiments were assessed. SDs are shown. Increasing colocal- rupted mice. This illustrates that, like BCG killing, these fusion ization after ATP stimulation is demonstrated. processes occur downstream of P2X7 activation. 3306 ATP INDUCES MYCOBACTERIA DEATH VIA PHAGOSOME-LYSOSOME FUSION

FIGURE 9. EM studies of BMDMs infected with BCG and stimulated with ATP. BMDMs were cul- Ϫ/Ϫ tured from P2X7 mice and WT littermates. They were infected with BCG and, where indicated, stimulated with ATP. Before fixation, cells were incubated with DAMP that localizes to acidic organelles. The DAMP was then gold labeled. A, C, and E, Magnifi- cation, ϫ5,000; B and F, ϫ50,000; and D, ϫ40,000. Downloaded from A and B, An unstimulated macrophage. BCG reside in individual phagosomes that are not gold labeled, indicating they are not acidic. C and D, A macrophage that has been stimulated with 3 mM ATP for 30 min. The BCG-containing phagosomes have coalesced, and the resulting vacuole shown in D is heavily gold labeled, indicating it is acidic. E and F, A macrophage http://www.jimmunol.org/ Ϫ/Ϫ from a P2X7 mouse after ATP stimulation for 30 min. The BCG-containing phagosomes have neither coalesced nor acidified. Representative pictures from at least three experiments are shown. by guest on September 26, 2021

There was an observed temporal relationship between ATP-in- large acidified vacuoles formed by coinfection with other intracel- duced phagosome acidification and BCG killing that supports the lular organisms (39). Fusion with lysosomes would then expose hypothesis that ATP-induced phagosome-lysosome fusion directly the BCG to the detrimental effects of low pH and to the toxic results in killing of the intracellular BCG. It is known that myco- actions of lysosomal proteases and antibacterial peptides. The bacteria grow poorly in acidic environments (38) and the growth rapid killing effect of ATP observed in this study would suggest rate of intracellular M. tuberculosis is impaired when they enter that exposure to these lysosomal mediators results in swift mycobactericidal activity. In confirmation of the findings reported by Kusner et al. (20), we also demonstrated that PLD activation is required for ATP-induced BCG killing. It is therefore of interest that PLD has been impli- cated in a number of organelle trafficking functions, e.g., degranulation by mast cells (40) and endosome fusion (41). PLD activation produces increased levels of phosphatidic acid, a known fusogen to endosome membranes (41). PLD may also activate ad-

ditional phospholipases such as phospholipase A2 (PLA2) (42–

45). Interestingly, PLA2 is known to be induced within ATP- treated macrophages (46). This in turn may then promote FIGURE 10. Acidity of BCG-containing vacuoles in ATP-stimulated J774 macrophages. Electron micrographs were prepared as in Fig. 9. BCG- endosome fusion via the generation of arachidonic acid (47). infected J774 macrophages were stimulated with 3 mM ATP for the indi- We hypothesize that BCG actively inhibit the fusigenicity of cated times. The number of gold particles per BCG-containing phagosome their phagosomes, thereby preventing phagosome maturation, i.e., or vacuole is proportional to its acidity. lysosome fusion and acidification. One mechanism by which they The Journal of Immunology 3307 may achieve this effect is by altering the lipid composition of their 22. Barton, C. H., S. H. Whitehead, and J. M. Blackwell. 1995. Nramp transfection phagosome membranes. We propose that ATP might act to over- transfers Ity/Lsh/Bcg-related pleiotropic effects on macrophage activation: influ- ence on oxidative burst and nitric oxide pathways. Mol. Med. 1:267. come this inhibition via the generation of fusinogenic mediators 23. Sikora, A., J. Liu, C. Brosnan, G. Buell, I. Chessel, and B. R. Bloom. 1999. Cutting edge: purinergic signaling regulates radical-mediated bacterial killing such as PLD and/or its downstream mediators, e.g., PLA2 and mechanisms in macrophages through a P2X7-independent mechanism. J. Immu- arachidonic acid that act to promote organelle fusion. nol. 163:558. Further investigation of this intracellular bacterial killing mech- 24. Tkalcevic, J., M. Novelli, M. Phylactides, J. P. Iredale, A. W. Segal, and J. Roes. anism may help to elucidate a key macrophage-mediated antimy- 2000. Impaired immunity and enhanced resistance to endotoxin in the absence of neutrophil elastase and cathepsin G. Immunity 12:201. cobacterial effector process. In addition, the signaling pathway(s) 25. Wei, X. Q., I. G. Charles, A. Smith, J. Ure, G. J. Feng, F. P. Huang, D. Xu, involved may prove to be amenable to modulation via targeted W. Muller, S. Moncada, and F. Y. Liew. 1995. Altered immune responses in mice drug design for the treatment of multidrug-resistant tuberculosis. lacking inducible nitric oxide synthase. Nature 375:408. 26. Cormack, B. P., R. H. Valdivia, and S. Falkow. 1996. FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33. 27. Snewin, V. A., M. P. Gares, P. O. Gaora, Z. Hasan, I. N. Brown, and References D. B. Young. 1999. Assessment of immunity to mycobacterial infection with 1. Smith, P. E., and A. R. Moss. 1994. Epidemiology of tuberculosis. In Tubercu- luciferase reporter constructs. Infect. Immun. 67:4586. losis: Pathogenesis, Protection, and Control. B. R. Bloom, ed. Am. Soc. Micro- 28. Steinberg, T. H., J. A. Swanson, and S. C. Silverstein. 1988. A prelysosomal biol., Washington, DC, p. 47. compartment sequesters membrane-impermeant fluorescent dyes from the cyto- 2. Pedersen, C., and J. O. Nielsen. 1987. Tuberculosis in homosexual men with HIV plasmic matrix of J774 macrophages. J. Cell Biol. 107:887. disease. Scand. J. Infect. Dis. 19:289. 29. Crowle, A. J., R. Dahl, E. Ross, and M. H. May. 1991. Evidence that vesicles 3. Altare, F., D. Lammas, P. Revy, E. Jouanguy, R. Doffinger, S. Lamhamedi, containing living, virulent Mycobacterium tuberculosis or Mycobacterium avium P. Drysdale, D. Scheel-Toellner, J. Girdlestone, P. Darbyshire, et al. 1998. In- in cultured human macrophages are not acidic. Infect. Immun. 59:1823. herited interleukin 12 deficiency in a child with bacille Calmette-Gue´rin and 30. Yang, S. F., S. Freer, and A. A. Benson. 1967. Transphosphatidylation by phos- Downloaded from Salmonella enteritidis disseminated infection. J. Clin. Invest. 102:2035. pholipase D. J. Biol. Chem. 242:477. 4. Jouanguy, E., F. Altare, S. Lamhamedi, P. Revy, J. F. Emile, M. Newport, 31. Malo, D., K. Vogan, S. Vidal, J. Hu, M. Cellier, E. Schurr, A. Fuks, N. Bumstead, M. Levin, S. Blanche, E. Seboun, A. Fischer, and J. L. Casanova. 1996. Inter- K. Morgan, and P. Gros. 1994. Haplotype mapping and sequence analysis of the feron-␥-receptor deficiency in an infant with fatal bacille Calmette-Gue´rin infec- mouse Nramp gene predict susceptibility to infection with intracellular parasites. tion. N. Engl. J. Med. 335:1956. Genomics 23:51. 5. Sugawara, I., H. Yamada, H. Kaneko, S. Mizuno, K. Takeda, and S. Akira. 1999. 32. Sturgill-Koszycki, S., P. H. Schlesinger, P. Chakraborty, P. L. Haddix, Role of interleukin-18 (IL-18) in mycobacterial infection in IL-18-gene-disrupted H. L. Collins, A. K. Fok, R. D. Allen, S. L. Gluck, J. Heuser, and D. G. Russell. mice. Infect. Immun. 67:2585. 1994. Lack of acidification in Mycobacterium phagosomes produced by exclu- 6. Oppmann, B., R. Lesley, B. Blom, J. C. Timans, Y. Xu, B. Hunte, F. Vega, sion of the vesicular proton-ATPase. [Published erratum appears in 1994 Science http://www.jimmunol.org/ N. Yu, J. Wang, K. Singh, et al. 2000. Novel p19 protein engages IL-12p40 to 263:1359.] Science 263:678. form a cytokine, IL-23, with biological activities similar as well as distinct from 33. Collo, G., S. Neidhart, E. Kawashima, M. Kosco-Vilbois, R. A. North, and IL-12. Immunity 13:715. G. Buell. 1997. Tissue distribution of the P2X7 receptor. Neuropharmacology 7. Flynn, J. L., M. M. Goldstein, J. Chan, K. J. Triebold, K. Pfeffer, 36:1277. C. J. Lowenstein, R. Schreiber, T. W. Mak, and B. R. Bloom. 1995. Tumor 34. Berchtold, S., A. L. Ogilvie, C. Bogdan, P. Muhl-Zurbes, A. Ogilvie, G. Schuler, necrosis factor-␣ is required in the protective immune response against Myco- and A. Steinkasserer. 1999. Human monocyte derived dendritic cells express bacterium tuberculosis in mice. Immunity 2:561. functional P2X and P2Y receptors as well as ecto-nucleotidases. FEBS Lett. 8. Manca, C., S. Paul, C. E. Barry III, V. H. Freedman, and G. Kaplan. 1999. 458:424. Mycobacterium tuberculosis catalase and peroxidase activities and resistance to 35. Stober, C. B., D. A. Lammas, D. S. Kumararatne, S. L. Lightman, and oxidative killing in human monocytes in vitro. Infect. Immun. 67:74. C. A. McArdle. 2001. ATP-mediated killing of Mycobacterium bovis bacille 9. Yu, K., C. Mitchell, Y. Xing, R. S. Magliozzo, B. R. Bloom, and J. Chan. 1999. Calmette Gue´rin (BCG) within human macrophages is calcium-dependent and Toxicity of nitrogen oxides and related oxidants on mycobacteria: M. tubercu- associated with the acidification of mycobacteria-containing phagosomes. J. Im- by guest on September 26, 2021 losis is resistant to peroxynitrite anion. Tuber. Lung Dis. 79:191. munol. 166:6276. 10. Hart, P. D., M. R. Young, A. H. Gordon, and K. H. Sullivan. 1987. Inhibition of 36. Bellamy, R., C. Ruwende, T. Corrah, K. P. McAdam, H. C. Whittle, and phagosome-lysosome fusion in macrophages by certain mycobacteria can be ex- A. V. Hill. 1998. Variations in the NRAMP1 gene and susceptibility to tubercu- plained by inhibition of lysosomal movements observed after phagocytosis. losis in West Africans. N. Engl. J. Med. 338:640. J. Exp. Med. 166:933. 37. Gruenheid, S., E. Pinner, M. Desjardins, and P. Gros. 1997. Natural resistance to infection with intracellular pathogens: the Nramp1 protein is recruited to the 11. Clemens, D. L., and M. A. Horwitz. 1995. Characterization of the Mycobacterium membrane of the phagosome. J. Exp. Med. 185:717. tuberculosis phagosome and evidence that phagosomal maturation is inhibited. 38. Piddington, D. L., A. Kashkouli, and N. A. Buchmeier. 2000. Growth of Myco- J. Exp. Med. 181:257. bacterium tuberculosis inadefined medium is very restricted by acid pH and 12. Reiner, N. E. 1994. Altered cell signaling and mononuclear phagocyte deactiva- ϩ Mg2 levels. Infect. Immun. 68:4518. tion during intracellular infection. Immunol. Today 15:374. 39. Gomes, M. S., S. Paul, A. L. Moreira, R. Appelberg, M. Rabinovitch, and 13. Vidal, S., M. L. Tremblay, G. Govoni, S. Gauthier, G. Sebastiani, D. Malo, G. Kaplan. 1999. Survival of Mycobacterium avium and Mycobacterium tuber- E. Skamene, M. Olivier, S. Jothy, and P. Gros. 1995. The Ity/Lsh/Bcg locus: culosis in acidified vacuoles of murine macrophages. Infect. Immun. 67:3199. natural resistance to infection with intracellular parasites is abrogated by disrup- 40. Brown, F. D., N. Thompson, K. M. Saqib, J. M. Clark, D. Powner, tion of the Nramp1 gene. J. Exp. Med. 182:655. N. T. Thompson, R. Solari, and M. J. Wakelam. 1998. Phospholipase D1 localizes 14. Gomes, M. S., and R. Appelberg. 1998. Evidence for a link between iron me- to secretory granules and lysosomes and is plasma-membrane translocated on tabolism and Nramp1 gene function in innate resistance against Mycobacterium cellular stimulation. Curr. Biol. 8:835. avium. Immunology 95:165. 41. Jones, A. T., and M. J. Clague. 1997. Regulation of early endosome fusion by 15. Zwilling, B. S., D. E. Kuhn, L. Wikoff, D. Brown, and W. Lafuse. 1999. Role of phospholipase D activity. Biochem. Biophys. Res. Commun. 236:285. iron in Nramp1-mediated inhibition of mycobacterial growth. Infect. Immun. 67: 42. Milella, M., A. Gismondi, P. Roncaioli, G. Palmieri, S. Morrone, M. Piccoli, 1386. ␤ L. Frati, M. G. Cifone, and A. Santoni. 1999. 1 integrin cross-linking inhibits 16. Hackam, D. J., O. D. Rotstein, W. Zhang, S. Gruenheid, P. Gros, and CD16-induced phospholipase D and secretory phospholipase A2 activity and S. Grinstein. 1998. Host resistance to intracellular infection: mutation of natural granule exocytosis in human NK cells: role of phospholipase D in CD16-trig- resistance-associated macrophage protein 1 (Nramp1) impairs phagosomal acid- gered degranulation. J. Immunol. 162:2064. ification. J. Exp. Med. 188:351. 43. Kinkaid, A. R., R. Othman, J. Voysey, and D. C. Wilton. 1998. Phospholipase D 17. Sato, K., T. Akaki, and H. Tomioka. 1998. Differential potentiation of anti- and phosphatidic acid enhance the hydrolysis of phospholipids in vesicles and in mycobacterial activity and reactive nitrogen intermediate-producing ability of cell membranes by human secreted phospholipase A2. Biochim. Biophys. Acta murine peritoneal macrophages activated by interferon-␥ (IFN-␥) and tumor ne- 1390:173. crosis factor-␣ (TNF-␣). Clin. Exp. Immunol. 112:63. 44. Bauldry, S. A., and R. E. Wooten. 1997. Induction of cytosolic phospholipase A2 18. Rook, G. A., J. Taverne, C. Leveton, and J. Steele. 1987. The role of ␥-interferon, activity by phosphatidic acid and diglycerides in permeabilized human neutro- vitamin D3 metabolites and tumor necrosis factor in the pathogenesis of tuber- phils: interrelationship between phospholipases D and A2. Biochem. J. 322:353. culosis. Immunology 62:229. 45. Sato, T., T. Ishimoto, S. Akiba, and T. Fujii. 1993. Enhancement of phospho- 19. Lammas, D. A., C. Stober, C. J. Harvey, N. Kendrick, S. Panchalingam, and lipase A2 activation by phosphatidic acid endogenously formed through phos- D. S. Kumararatne. 1997. ATP-induced killing of mycobacteria by human mac- pholipase D action in rat peritoneal mast cell. FEBS Lett. 323:23. rophages is mediated by purinergic P2Z (P2X7) receptors. Immunity 7:433. 46. Balboa, M. A., J. Balsinde, C. A. Johnson, and E. A. Dennis. 1999. Regulation 20. Kusner, D. J., and J. Adams. 2000. ATP-induced killing of virulent Mycobacte- of arachidonic acid mobilization in lipopolysaccharide-activated P388D(1) mac- rium tuberculosis within human macrophages requires phospholipase D. J. Im- rophages by adenosine triphosphate. J. Biol. Chem. 274:36764. munol. 164:379. 47. Mayorga, L. S., M. I. Colombo, M. Lennartz, E. J. Brown, K. H. Rahman, 21. Molloy, A., P. Laochumroonvorapong, and G. Kaplan. 1994. Apoptosis, but not R. Weiss, P. J. Lennon, and P. D. Stahl. 1993. Inhibition of endosome fusion by necrosis, of infected monocytes is coupled with killing of intracellular bacillus phospholipase A2 (PLA2) inhibitors points to a role for PLA2 in endocytosis. Calmette-Gue´rin. J. Exp. Med. 180:1499. Proc. Natl. Acad. Sci. USA 90:10255.