Role of the phagosomal redox-sensitive TRP channel TRPM2 in regulating bactericidal activity of macrophages

Macrophages, upon ingesting invading micro-organisms, have the primary function of killing bacteria through maturation of phagosomes as defined by their acidification (Russell et al., 2009; Vieira et al., 2002). The mature phagosome is also rich in proteolytic enzymes, ensuring effective degradation of internalized microbes (Kinchen and Ravichandran, 2008). During phagosome maturation, phagosomes recruit vacuolar proton (H⁺)-ATPases (vATPases) that transfer H⁺ into phagosomes (Flannagan et al., 2009). To optimize H⁺ accumulation, efflux of cations and influx of anions such as chloride (Hara-Chikuma et al., 2005) are needed to dissipate the development of a restrictive electrical gradient induced by the H⁺ influx (Steinberg et al., 2007a). However, the identity of these channels remains unclear (Steinberg et al., 2010).

Transient receptor potential melastatin 2 (TRPM2), a redox-sensitive non-selective cation channel localized in the plasma membrane (Hara et al., 2002; Knowles et al., 2013; Perraud et al., 2001, 2005; Syed Mortadza et al., 2015; Toth and Csanady, 2010), plays an important role in bacterial clearance due in part to its ability to activate cytokine generation (Knowles et al., 2011; Qian et al., 2014). Impaired production of cytokines in Trpm2^−/− mice led to high mortality following Listeria monocytogenes (LM) challenge (Knowles et al., 2011). TRPM2 may also control the killing of internalized Escherichia coli through its role as a plasma membrane Ca²⁺-influx channel and induce phagosome–lysosome fusion (Zhang et al., 2017). Because of the importance of macrophages in killing bacteria, here, we addressed the question of whether TRPM2 is expressed in phagosomal membrane of macrophages and whether it functions as an efflux of cation channel to regulate phagosomal pH and thereby bacterial killing.

Studies using TRPM2-knockout (Trpm2^−/−) mice have previously shown that the protective role of TRPM2 in polymicrobial sepsis (Qian et al., 2014) and Listeria monocytogenes (LM) infection (Knowles et al., 2011) is mediated by TRPM2-dependent bacterial clearance. However, the question remains as to whether TRPM2 is expressed in phagosomes of macrophages and whether its phagosomal expression contributes to bacterial killing. Here, we studied the role of TRPM2 in phagosomes of macrophages for killing of the gram-negative bacteria Pseudomonas aerugenosa (PA), a common pathogen mediating respiratory infection (Sadikot et al., 2005). We first examined the role of TRPM2 in mice challenged intra-tracheally with PA. We observed markedly augmented lung PA retention in Trpm2^−/− compared to in Trpm2^+/+ mice (Fig. 1A). Mortality was also greater in Trpm2^−/− relative to Trpm2^+/+ mice in response to intra-tracheal administration of PA as well as to cecal ligation puncture (CLP), a model of polymicrobial sepsis (Fig. 1B,C). We next determined whether TRPM2 in isolated macrophages contributed to killing of PA. Upon infecting mouse bone marrow-derived macrophages (BMDMs) with either gram-negative PA bacteria or gram-positive Staphylococcus aureus (SA) bacteria, we observed significantly greater bacterial survival in BMDMs from Trpm2^−/− mice exposed to PA or SA than BMDMs from Trpm2^+/+ mice (Fig. 1D,E).

Although TRPM2 is known to be expressed in the plasma membrane of macrophages (Di et al., 2011), its distribution in phagosomes is unclear. We identified marked TRPM2 immunostaining in human alveolar macrophages (Fig. 2A–C, green). We also demonstrated, by feeding human alveolar macrophages latex beads, that TRPM2 was specifically distributed in the phagosome membrane (Fig. 2D). Furthermore, TRPM2 staining was seen in phagosomes of BMDMs from Trpm2^+/+ but not in Trpm2^−/− mice (Fig. 2E–I). Since lysosomes contribute to phagosomal maturation (Flannagan et al., 2009; Vieira et al., 2002), we next determined whether TRPM2 is present in lysosomes. Macrophages were immunostained with anti-TRPM2 antibody and antibody against the lysosome marker lysosomal-associated membrane protein 1 (LAMP1) (Meikle et al., 1997). However, we observed that TRPM2 staining was absent in lysosomes (Fig. S1).

We next used patch clamping to monitor the current in phagosomes isolated from BMDMs (Samie et al., 2013). ADP-ribose (ADPR; 200 µM) was added to the patch pipette solution to activate TRPM2 (Perraud et al., 2001). ADPR activated a non-selective phagosomal cation current with a linear current–voltage plot, and isosmotic replacement of external Na⁺ with the impermeant organic cation N-methyl-D-glucamine (NMDG) abolished the inward current (Fig. 3A–C). Phagosomes from Trpm2^–/– BMDMs failed to respond to ADPR (Fig. 3C). ADPR added to the bath solution also failed to activate a current (Fig. 3A–C). These findings demonstrated the intra-luminal localization of the TRPM2 C- and N-terminals. Since TRPM2 is a redox-sensitive channel (Hara et al., 2002; Perraud et al., 2005), it can also be activated by oxidants (Wehage et al., 2002). We showed that H₂O₂ (100 µM) added into phagosomes via the micropipette elicited a non-selective cation current in Trpm2^+/+ BMDMs and isosmotic replacement of external Na⁺ by NMDG significantly reduced the inward current, but not the outward current (Fig. 3D–F). Furthermore, the H₂O₂-induced current was significantly reduced in Trpm2^–/– BMDMs.

Since there is uncertainty about whether oxidants and oxidant-mediated production of ADPR activate TRPM2 (Knowles et al., 2013), we next examined whether H₂O₂-induced activation of TRPM2 occurred secondary to production of ADPR. We observed that 8-Br-ADPR (an antagonistic analog of ADPR; Partida-Sanchez et al., 2007) reduced H₂O₂-induced current in Trpm2^+/+ BMDMs to the level seen in Trpm2^−/− BMDMs (Fig. 3D–F). However, 8-Br-ADPR did not modify the H₂O₂-induced current in Trpm2^−/− BMDMs (Fig. 3D–F), consistent with the essential role of generation of H₂O₂ in activating phagosomal TRPM2 secondary to ADPR generation.

We next addressed the postulated mechanism by which reactive oxygen species (ROS) induced production of ADPR in phagosomes. The mechanism involves H₂O₂-induced generation of ADPR through the oxidant sensitive-enzyme NADase (also known as NAD glycohydrolase), which catalyzes hydrolysis of NAD into ADPR (Albeniz et al., 2004), and poly(ADP-ribose) polymerase (PARP), which binds single-stranded and double-stranded DNA breaks and catalyzes breakdown of NAD into poly-ADPR (Fonfria et al., 2004; Hecquet and Malik, 2009). We observed in patch clamp experiments that included nicotinamide, an inhibitor of NADase (Alano et al., 2010), along with H₂O₂ in the intra-phagosomal pipette solution did not show a H₂O₂-activated current (Fig. 3E,F). Thus, oxidant-induced ADPR generation in phagosomes involved oxidant-sensitive transition of NAD into ADPR mediated by NADase.

Since phagosomal TRPM2 faces the inner aspect of the phagosome, as evident by the electrophysiological measurements described above (Fig. 3), we next determined the relationship between TRPM2 recruitment to the phagosome and phagosomal acidification during phagosome maturation. In this study, we simultaneously monitored both phagosomal TRPM2 activation and acidification. Intra-phagosomal Ca²⁺ was determined using fluo3 and phagosomal acidification was determined using the pH-sensitive fluorescent dye pHrodoRed (hereafter pHrodo) conjugated to SA (Fig. 4). We observed greater intra-phagosomal Ca²⁺ accumulation (a green fluorescence increase) in Trpm2^−/− mouse (m)BMDMs compared to Trpm2^+/+ mBMDMs during phagosomal acidification (Fig. 4A,C; Movies 1,2). Further, we observed weaker increase in intensity of fluorescence from pHrodo in Trpm2^−/− BMDMs as compared to Trpm2^+/+ BMDMs (Fig. 4A,D; Movies 1,2), indicating decreased phagosomal acidification in Trpm2^−/− macrophages. These findings thus demonstrate that TRPM2 activation, as reflected by efflux of Ca²⁺ in Trpm2^+/+ BMDMs, occurred in parallel with phagosomal acidification.

To further evaluate the role of TRPM2 in regulating phagosomal pH, we also measured phagosomal pH using dual fluorescence flow cytometry (Sokolovska et al., 2012). SA was conjugated to ratiometric fluorescent dyes containing both the pH-sensitive component Oregon Green 488 (OG), having a low pK_a (4.8) (Han and Burgess, 2010), and the acidification-insensitive component carboxy-tetramethyl-rhodamine (TAMRA) (Sokolovska et al., 2012) (Fig. 5A). The intensity ratios of OG and TAMRA versus pH were obtained by flow cytometry. Calibration of fluorescence was interpolated on a standard curve for the intensity ratios of OG and TAMRA versus pH determined in vitro using a series of buffers (Fig. 5A,B). BMDMs containing SA from Trpm2^−/− mice showed greater pH values (6.53±0.28, mean±s.e.m.) in phagosomes containing SA than the Trpm2^+/+ BMDMs (5.32±0.01) measured after phagocytosis (Fig. 5C). Trpm2^+/+ BMDMs treated with diphenyleneiodonium (DPI; an inhibitor of NOX) (Cross and Jones, 1986), however, showed increased phagosomal pH (6.71±0.46) similar to the pH value (6.87±0.12) of phagosomes from Trpm2^−/− BMDMs treated with DPI (Fig. 5D). These findings together show the central role of oxidant-mediated TRPM2 activation in the mechanism of phagosomal pH drop.

To address whether defective phagosomal acidification was responsible for the PA-killing defect seen in Trpm2^−/− macrophages (Fig. 1), we next determined PA killing following treatment of BMDMs with bafilomycin, which inhibits H⁺ influx (Lukacs et al., 1990). Bafilomycin-treated macrophages from Trpm2^+/+ mice showed significantly greater PA survival than control untreated cells, and moreover, the level of bacterial killing was similar to Trpm2^−/− BMDM (Fig. 5E).

We next examined phagosomal acidification and bacterial killing in the presence of 8-Br-ADPR, an analog of ADPR (Partida-Sanchez et al., 2007) that does not activate TRPM2. Phagosomal pH and bacterial survival increased in Trpm2^+/+ BMDMs treated with 8-Br-ADPR to the same extent as seen in Trpm2^−/− BMDMs (Fig. 5F,G). These observations showed that oxidant-mediated ADPR production was essential for the TRPM2 activation in phagosomes and thereby for bacterial killing by macrophages (Fig. 5H).

Since the fusion of lysosomes with phagosomes may regulate phagosomal pH (Vieira et al., 2002), we next queried whether phagosomal TRPM2 was involved in the fusion process. Phagosomes were traced by pHrodo conjugated to SA and lysosomes were traced with the fluorescent dye Alexa Fluor 488 conjugated to dextran as described previously (Steinberg et al., 2010). We first examined phagosomes positive for dextran–Alexa-Fluor-488 staining as the percent of all phagosomes detected by staining with pHrodo. We then examined the strength of lysosome fluorescence (Alexa Fluor 488) of each phagosome containing pHrodo fluorescence (Fig. S2). However, we did not observe any difference in the fluorescence intensity distribution of the lysosomal dye (Alexa Fluor 488) for each phagosome containing pHrodo, or a difference in proportion the phagosomes positive for dextran–Alexa-Fluor-488 staining as a percentage of all phagosomes detected between Trpm2^+/+ and Trpm2^−/− mBMDMs (Fig. S2G,H). Phagosomes containing SA from Trpm2^+/+ and Trpm2^−/− mBMDMs showed similar loading of dextran (Fig. S2G,H). These studies demonstrated no fusion defect in either group.

Macrophages serve as an essential host-defense function in killing invading bacteria secondary to phagosomal maturation and decrease in pH (Vieira et al., 2002). We previously showed that TRPM2 functioning as a non-selective cation channel at the plasma membrane of macrophages increased survival of endotoxemic mice and decreased inflammatory lung injury secondary to activation of influx of cations and depolarization of the plasma membrane (Di et al., 2011). However in the present study, in profiling TRPM2 expression in macrophages, we also found that TRPM2 was present in the phagosomal membrane. Furthermore, we showed that the phagosomally distributed TRPM2 was required for killing gram-negative bacteria PA and gram-positive bacteria SA secondary to promoting acidification of phagosomes. Specifically, macrophage phagosomes obtained from Trpm2^−/− mice failed to acidify, and treatment of Trpm2^+/+ macrophages with the vATPase inhibitor bafilomycin showed the same PA-killing defect as seen in Trpm2^−/− macrophages. Furthermore, macrophages treated with the TRPM2 antagonist 8-Br-ADPR (Partida-Sanchez et al., 2007), which, unlike ADPR, fails to bind TRPM2 in the C-terminal NUDT9-H (ADPRase) domain of the channel (Perraud et al., 2001), displayed the same defects in phagosomal acidification and bacterial killing as seen in Trpm2^−/− macrophages. These results collectively demonstrated an essential role of phagosomal TRPM2 in killing bacteria secondary to phagosomal acidification.

Zhan et al. have also observed defective E. coli killing in stimulated peritoneal Trpm2^−/− macrophages, which they ascribed to defective phagosome–lysosome fusion resulting from impaired Ca²⁺ influx via the plasma membrane-localized TRPM2 (Zhang et al., 2017). In the present study however, on carefully examining the distribution and function of TPRM2, we observed localization of TRPM2 in phagosomes but not in lysosomes of macrophages derived from bone marrow. We demonstrated that the impaired phagosomal acidification and bacteria killing of Trpm2^−/− macrophages were due solely to the lack of counter conductance for proton influx. Furthermore, phagosome–lysosome fusion in both Trpm2^+/+ and Trpm2^−/− BMDMs was similar, indicating that the defective phagosome–lysosome fusion in the present study cannot explain the impaired killing defect of Trpm2^−/− macrophages. The reasons for the discrepancy in the findings are not clear. One possibility is that the stimulated peritoneal macrophages used by Zhang et al. (2017) may have been activated prior to their use, and thus this favored fusion of phagosomes and lysosomes.

In a previous study addressing the role of TRPM2-mediated inhibition of ROS generation (Di et al., 2011), augmented E. coli. killing was observed in lungs and blood of Trpm2^−/− compared to Trpm2^+/+ mice after a 4 h period of infection, whereas there was no difference between Trpm2^+/+ and Trpm2^−/− mice seen at 24 h after infection. In the study using mice infected with LM, the liver and spleen bacterial killing in Trpm2^−/− was less than in Trpm2^+/+ mice (Knowles et al., 2011). The Trpm2^−/− peritoneal macrophages, however, showed normal killing of LM (Knowles et al., 2011). In the present study, we clearly showed an essential role for TRPM2 localized in phagosomes for killing of both gram-positive bacteria SA and gram-negative bacteria PA. It appears from these disparate observations that the role of TRPM2 in mediating bacterial killing is complex. A possible explanation is that TRPM2-mediated bactericidal mechanisms vary depending on bacteria type, the state of activation of macrophages, the type of macrophages used for the study and the distinct functions of TRPM2 present in the plasmalemmal versus in the phagosome.

Our patch clamp results in isolated phagosomes showed the luminal orientation of the TRPM2 termini in phagosomes. That the N- and C-terminal regions of TRPM2 did not face the outer side of phagosomes suggests that phagosomal TRPM2 did not originate from the plasma membrane after phagocytosis. Phagosomal maturation requires spatial and temporal orchestration of proteins (for a review, see Levin et al., 2016). During the formation of phagosomes, proteins such as transferrin receptor are shuttled back to the plasma membrane whereas proteins such as the GTPase Rab5 are recruited into phagosomes (Levin et al., 2016). Other phagosomal proteins undergo retrograde traffic to the trans-Golgi network via tubulovesicular structures (Levin et al., 2016). The mechanism of TRPM2 positioning in phagosomes is not known. Our results on the timecourse of phagosome maturation showed that phagosomal TRPM2 activation and acidification occurred together, thus indicating that the recruitment of TRPM2 was directly coupled to phagosome maturation.

We also addressed the proximal basis of TRPM2 activation in phagosomes. Since TRPM2 is a redox-sensitive channel (Hara et al., 2002; Perraud et al., 2005), a possible mechanism of phagosomal TRPM2 activation may involve generation of oxidants and the subsequent production of ADPR (Hara et al., 2002; Knowles et al., 2013; Perraud et al., 2005; Syed Mortadza et al., 2015; Toth and Csanady, 2010; Wehage et al., 2002). We observed that the H₂O₂-activated current in Trpm2^+/+ BMDMs treated with an antagonistic analog of ADPR was reduced to the level seen in Trpm2^−/− BMDMs, suggesting that H₂O₂ activated phagosomal TRPM2 through ADPR generation. The production of ADPR from mitochondria and nuclei (DNA) is considered to be primarily responsible for oxidative stress-induced TRPM2 activation (Hecquet and Malik, 2009). In phagosomes containing bacteria, DNA from nuclei may be the source for ROS-induced and NADase-mediated generation of ADPR (Bricker et al., 2002; Huang et al., 2009). Since NADase is ROS sensitive (Krapp et al., 1997; Ziegler et al., 1997), a tenable presumption is that H₂O₂ induces ADPR generation through activating NADase (Albeniz et al., 2004). This was confirmed by the observation that nicotinamide (an inhibitor of oxidant-sensitive NADase) (Alano et al., 2010) prevented H₂O₂-induced current, suggesting that ADPR generation in phagosomes involved the oxidant-sensitive transition of NAD into ADPR. Thus, our results support the model that ROS generation activated during phagocytosis mediated ADPR generation, and thereby signaled TRPM2 activation in phagosomes.

Phagosomes acquire an antimicrobial arsenal during the course of phagosomal maturation (Flannagan et al., 2009). This includes recruiting the vATPase that pumps H⁺ into phagosomes, and assembly and activation of phagosomal membrane NADPH oxidases (NOXs) that produce ROS (Russell et al., 2009). In light of the role of ROS in bacteria killing (Pollock et al., 1995; Segal, 1996; Segal et al., 2000), the previous observation that the NOX-inhibiting function of TRPM2 promoted bacterial killing during phagocytosis at the plasma membrane (Di et al., 2011; Zhang et al., 2017) led us to consider the possibility that TRPM2 could also mediate bacterial killing through a phagosomal acidification mechanism. Phagosomal acidification in macrophages is essential for activation of lysosomal proteases, microbe killing and protein degradation (Aronson and De Duve, 1968; Coffey and De Duve, 1968; Pillay et al., 2002; Savina et al., 2006). The role of the ROS and NOX pathway in intra-phagosomal bacterial killing has been proposed in part based on findings in chronic granulomatous disease (CGD) (Pollock et al., 1995; Segal, 1996; Segal et al., 2000). CDG is characterized by severe, protracted and fatal infection resulting from the failure of the NOX system to produce superoxide (O₂ ⁻) and other ROS in neutrophils (Pollock et al., 1995; Segal, 1996; Segal et al., 2000); however, it was later discovered that defective bacterial killing in CGD neutrophils was secondary to dysregulated phagosomal acidification mechanism (Dri et al., 2002; Geiszt et al., 2001; Henriet et al., 2013; Segal et al., 1981; Thrasher and Segal, 2011). We previously showed that TRPM2 inhibited NOX-mediated ROS generation, specifically that there was overproduction of ROS in Trpm2^−/− macrophages infected with E. coli (Di et al., 2011). In the present study, however, DPI treatment of Trpm2^−/− mBMDMs failed to alter significantly phagosomal pH, indicating that the increased pH seen in Trpm2^−/− mBMDMs was caused by TRPM2 deletion rather than increased ROS production secondary to TRPM2 deletion (Di et al., 2011). Thus, the dysregulated phagosomal pH in Trpm2^−/− BMDMs, as opposed to the overproduction of ROS, was essential for the phagosomal killing of bacteria as was shown to be the case for CDG neutrophils (Dri et al., 2002; Geiszt et al., 2001; Henriet et al., 2013; Segal et al., 1981; Thrasher and Segal, 2011).

In summary, we demonstrated that TRPM2 localized in phagosomes is required for mediating bacterial killing through facilitating acidification of phagosomes in macrophages. This finding suggests that strategies aimed at activating TRPM2-mediated cation efflux, such as through ADPR generation, might enhance acidification and thereby the bactericidal function of macrophage.

Age (4–6 weeks)- and sex (both male and female)-matched mice were used in these studies. TRPM2-knockout (Trpm2^−/−) mice used in these studies were generously provided by Dr. Yasuo Mori (Department of Technology and Ecology, Kyoto University, Japan; Yamamoto et al., 2008); these mice were generated by deleting exons corresponding to transmembrane (TM) spans 5 and 6, including the pore loop. C57BL6 mice were obtained from Charles River Laboratory. All mice were housed in the University of Illinois Animal Care Facility in accordance with institutional and NIH guidelines. All experiments were approved by the University of Illinois Animal Resources Center.

Mouse bone marrow-derived macrophages (mBMDMs) were induced and cultured as described previously (Zhang et al., 2008). Pseudomonas aeruginosa (PA) strain 103, was a gift from Dr Yuru Liu (University of Illinois, Chicago, IL) and Staphylococcus aureus strain USA 300 (SA) was a gift from Dr Nancy Freitag (University of Illinois, Chicago, IL). Bacteria were suspended in medium containing 20% glycerol, and frozen in aliquots at −70°C. Before each experiment, an aliquot was thawed and cultured in Luria-Bertani (LB) broth at 37°C with continuous shaking, and the exact number of bacteria was determined by plating serial dilutions onto LB agar plates for overnight culture. The original culture was kept at 4°C and was used for infection once the concentration of live bacteria was determined overnight (∼12 h). Unlabeled SA (Wood strain without protein A) bio-particles (SA, S-2859), Oregon Green 488 carboxylic acid succinimidyl ester (OG, O6149), 6-carboxy-tetramethyl-rhodamine (TAMRA, C6122), Fluo3-AM (F-14218), Fluo-3 pentapotassium salt (F3715) and pHrodo (A10010) were purchased from Life Technologies (NY). Rabbit anti-human TRPM2 antibody (ab11168) was purchased from Abcam (MA, USA) and used at 1:100. 8-Br-ADPR (B 051-05) was purchased from Biolog/Axxora, LLC (Enzo Life Sciences, Inc.). Bafilomycin (B1793), latex beads (3.0 µm, LB30), Nicotinamide (N3376) and other chemicals were purchased from Sigma (MO).

Use of human subjects for broncho-alveolar lavage was approved by the Institutional Review Board of University of Illinois Chicago Hospitals (Chicago, IL). Informed consent was obtained for all tissue donors and all clinical investigation have been conducted according to the principles expressed in the Declaration of Helsinki.

Mouse polymicrobial sepsis was induced by cecal ligation puncture (CLP) as described previously (Cuenca et al., 2010) using a 16-gauge needle. For survival studies, mice were monitored twice daily for 6 days.

Lung infection with PA was induced by intra-tracheal administration of PA (10⁵ in 30 µl). For survival studies, mice were monitored twice daily for 6 days after PA administration. Bactericidal activity in lungs was assessed as described previously (Sadikot et al., 2006). Mice were killed and lungs were removed 48 h after PA administration. The lung was removed aseptically and placed in 1 ml of sterile HBSS. Lungs were homogenized in a tissue homogenizer in a vented hood under sterile conditions. Serial dilutions of the homogenates were made and 10 μl of each dilution were plated in agar plates. Plates were incubated for 18 h at 37°C, and the number of PA colonies was counted.

Bactericidal activity of macrophage was analyzed by performing a colony-forming unit (CFU) assay as described previously (Hamrick et al., 2000). Pseudomonas aeruginosa strain 103 (PA) and Staphylococcus aureus strain USA 300 (SA) were grown in tryptic soy broth for 14 h at 37°C, and bacteria concentrations were determined by serial dilutions and counting CFUs. In each sample, 2×10⁵ macrophages on culture dishes were infected with bacteria at a multiplicity of infection (MOI) of 20 for PA or an MOI of 10 for SA for 30 min at 37°C. Cells were gently scraped off the bottom of the culture dish into the medium and centrifuged at 500 g for 10 min at 4°C. Then, supernatant was aspirated and pellets were re-suspended in DMEM with 10% fetal bovine serum (FBS). Non-internalized bacteria were killed by adding 100 µg/ml gentamicin to the medium and incubating for 1 h at 37°C. It was tested that a 1 h pulse with gentamicin is sufficient to kill these bacteria. Samples were centrifuged (500 g for 10 min), and the pellets re-suspended in medium with 10 µg/ml gentamicin. Macrophage bactericidal activity was monitored at 37°C over several time points (0, 2 or 3 h). At each time point, cells were centrifuged at 500 g for 5 min and the cell pellet was re-suspended in 1% Triton X-100 in PBS. Cells were lysed for 2 min at room temperature (RT) and 50, 100 and 500 µl aliquots of lysate were spread onto agar plates. Surviving bacteria were grown for 16 h and the number of colony-forming units was tabulated.

Bone marrow-derived macrophage (BMDMs) from both Trpm2^−/− and Trpm2^+/+ mice plated on 25 mm coverslips were treated with 3 µm latex beads at 37°C for 30 min. Phagocytosis was stopped by transferring coverslips to ice-cold PBS, and washing five times with PBS to remove any remaining external beads, and fixed for 20 min in 3% paraformaldehyde in HBBS. Macrophages were stained with anti-TRPM2 antibody for 1 h and localization of TRPM2 and phagocytosed beads were imaged with a Zeiss LSM 710 confocal microscope using 488 and 561 nm lasers and a 63× objective lens (NA 1.3). Images were acquired with LCS software and images were processed with Fiji software.

Whole phagosomal current was measured as described previously (Samie et al., 2013). Red blood cells (RBCs) opsonized with IgG were incubated with BMDMs at 37°C and 5% CO₂ for 10 min to lead to the formation of phagosomes. To release phagosomes from cells, a patch pipette was used to open the cell by slicing the cell membrane. Currents from the isolated phagosome were elicited using a series of test pulses ranging from −110 to +110 mV (test pulses were 200 ms in duration and delivered at 2 s intervals). The holding potential was 0 mV. The pipette solution contained (in mM): 135 CsSO₃CH₃, 8 NaCl, 2 MgCl₂, 0.5 CaCl₂, 1 EGTA and 10 HEPES pH 7.4. ADPR (200 µM) or H₂O₂ (100 µM) was included in the pipette solution to induce TRPM2 current. The bath solution contained (in mM): 145 NaCl, 2 CaCl₂, 1 MgC₂ and 10 HEPES pH 7.4. Phagosomal current was analyzed using IGOR software (WaveMetrics, Lake Oswego, OR).

Acidification of phagosomes containing SA was measured as described previously (Sokolovska et al., 2012). SA was labeled with both OG and TAMRA as above. Macrophages were cultured on a 96-well plate (Nunclon Delta surface, Thermo Scientific) at a concentration of 5×10⁵ cells/well, and were incubated with labeled SA particles at a MOI of 5 for 10 min at 37°C for flow cytometry analysis. The ratio of the mean fluorescence from OG over mean fluorescence of TAMRA was determined by flow cytometry (LSR Fortessa, BD, NJ) with a 488 nm and a 561 nm laser. The intra-phagosome pH values were interpolated on a standard curve for SA fluorescence versus pH determined in vitro using a series of buffers as above. For phagosomal pH calibration, phagosomal pH was equalized by incubating cells for 15–20 min at 37°C in calibration buffer (120 mM KCl, 20 mM NaCl, 1 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES with the pH adjusted from 4 to 8) containing the ionophores carbonyl cyanide m-chlorophenylhydrazone (CCCP, 10 µM), nigericin (10 µM) and the H⁺ vATPase pump inhibitor bafilomycin (0.1 µM). Changes in pH were measured by assessing the intensity ratio of OG to TAMRA in each phagosome. OG:TAMRA ratios were plotted against pH values and curves were fitted with the sigmoidal Boltzman equation. In each independent experiment, results from three samples (each containing about 5×10⁵ cells) of the same group were averaged, and averaged results from six independent experiments (n=6) were presented.

Cellular and intra-phagosomal Ca²⁺ was monitored with fluo3 as the Ca²⁺ indicator (Minta et al., 1989), and phagosomal acidification was monitored by using the pH-sensitive fluorescent dye pHrodo conjugated to SA. BMDMs from both Trpm2^+/+ and Trpm2^−/− were first loaded with 5 µΜ fluo3-AM (membrane permeant) for 30 min at 37°C and cells were then infected with SA conjugated to the pH-sensitive fluorescent dye pHrodo (Han and Burgess, 2010) with a MOI of 10. The pHrodo dye has the property that its fluorescence increases as pH decreases from neutral to acidic (Inc., 2014); however, importantly it does not fluorescence outside the cell (Miksa et al., 2009) and therefore is useful in identifying particles internalized in phagosomes. Dye emission was determined with a Zeiss LSM 710 confocal microscope by using a 488 nm and a 561 nm laser, a 63× objective lens (NA 1.3) and an emission bandwidth of 500–535 nm and 600–680 mm. Images were collected every 10 s with LCS software and were processed and analyzed with Fiji software.

Images were analyzed using Fiji software (NIH). Summary data are expressed as mean±s.e.m. with the number of experiments listed in parentheses or indicated in the text. Significance between groups was determined using the Student’s t-test (two-tailed) and asterisks indicate values as shown.

We thank Dr Sergio Grinstein of Hospital for Sick Children and Nic Demaraux of University of Geneva for insights into the work and Dr Balaji Ganesh (University of Illinois at Chicago) for flow cytometry experiment design and data analysis.