We used C57Bl/6 Wild-type mice (P2X7^+/+, WT) and P2X7 knockout mice (P2X7^-/- from Jackson Laboratories – JAX stock #005576 Mice and Services, Bar Harbor, ME, USA). P2X7 knockout mice that are homozygous for the targeted allele (P2X7R ^Δ506–532) are viable, fertile, normal in size, and do not display any gross physical or behavioral abnormalities (25). The procedures for the care and use of animals were according to the Brazilian College of Animal Experimentation guidelines. The Commission for the Ethical Use of Research Animals from the Federal University of Rio de Janeiro (UFRJ) approved all experiments (protocol number: IBCCF166/18). Roswell Park Memorial Institute Medium 1640 (RPMI), Lucifer Yellow (LY), Lysotracker Green (LG), mefloquine (MFQ), adenosine-5’-triphosphate disodium salt hydrate (ATP), ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA) and P2X7 antagonist A740003 were obtained from Sigma-Aldrich (St. Louis, USA). Fetal bovine serum (FBS) and penicillin-streptomycin were purchased from Gibco-BRL (São Paulo, SP, Brazil). BAPTA-AM and probenecid were from Invitrogen Molecular Probes (Eugene, OR, USA). Magic Red Cathepsin B Assay was from ImmunoChemistry (Bloomington, USA). Phosphate buffer solution (PBS) (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 2 mM KH₂PO_4, pH = 7.4), Z-Phe-Arg-AMC (Z-FR-AMC) (C9521), and Z-Arg-Arg-AMC (Z-RR-AMC) (C5429) were from Sigma-Aldrich (St. Louis, USA). Hoechst 33342 was from Life Technologies (Eugene, OR. USA). Goat anti-mouse pannexin-1 antibody (K-20) (Santa Cruz Biotechnology, CA, USA), anti-vinculin monoclonal antibody (Sigma Aldrich, Missouri, USA), mouse anti-goat IgG (H+L)-Alexa Fluor^® 546 and mouse anti-goat IgG-HRP secondary antibodies (ThermoFisher Scientific, MA, USA), fura-2 AM (Thermo Fisher Scientific, MA, USA), ECL Plus Western Blotting Detection System (GE Healthcare Life Sciences, NY, USA), SuperSignal™ ELISA Femto Substrate (Thermo Fisher Scientific, MA, USA).

Non-stimulated peritoneal macrophages were obtained from C57Bl/6 WT or P2X7^-/- male mice, 8–10 weeks old, after peritoneal lavage. Cells were centrifuged at 300 x g for 10 minutes and plated at 2 x 10⁵ cells/mL in RPMI-1640 medium for 1 h at 37°C and 5% CO₂. Then these cells were washed three times with PBS pH 7.4 to remove non-adherent cells, and adherent cells were cultured in RPMI-1640 medium supplemented with 10% FBS and 1% penicillin-streptomycin. The peritoneal macrophages were maintained in culture at 37°C and 5% CO₂ in culture plates with a diameter of 35 mm and a glass slide of 10 mm in diameter at the bottom divided into four compartments for microscopic observation (Greiner Bio-One, North Carolina, USA). For the enzymatic activity assays, the macrophages were plated on 12-well cell culture plates (ThermoFisher Scientific, MA, USA). Phenotypic analyses were performed by staining the cells with CD11b and F480 macrophage markers, indicating a purity of more than 90%.

For the endocytosis assay, the peritoneal macrophages were incubated for 45 minutes in RPMI-1640 medium with L-glutamine, without fetal bovine serum at 37°C, 5% CO₂ in the presence of the following fluorescent probes in independent experiments: 1 mM Lucifer Yellow (495/519) nm, 50 nM Lysotracker Green (504/511) nm or 10 µM Yo-Pro 1 (491/509 nm) (26, 27). In the assays using inhibitors, cells were pretreated (30 minutes) with the following: 100 nM A7400043, 100 µM mefloquine, 50 µM carbenoxolone, 20 µM 10Panx, 150 µM probenecid, and 50 µM 2-APB or 50 µM verapamil. After the endocytosis assay, the cells were washed three times with physiological extracellular solution (145 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 10 mM Hepes, 1 mM CaCl₂, pH 7.4) to remove the probe that was not endocytosed. Finally, the peritoneal macrophages were stained with the nuclear dye Hoechst 33342 (350/461 nm) (Life Technologies, OR, USA) or with CellMask™ Deep Red Plasma Membrane Stain (659/676 nm) (Life Technologies, OR, USA). Live cells were kept in physiological extracellular solution (145 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 10 mM Hepes, 1 mM CaCl₂, pH 7.4) and observed before and after ATP treatment at various concentrations (50 µM, 500 µM, and 5 mM) or UTP (100 µM) for 10 minutes at 37°C. For the experiments with calcium chelators 2 mM de EGTA e 10 µM de BAPTA-AM were added to physiological extracellular solution. The cells were observed in a confocal microscope - rotating disk Zeiss Cell Observer Yokogawa (Cell Observer SD, Carl Zeiss, located on the Gustavo De Oliveira Castro light optical microscopy platform, from the Carlos Chagas Filho Biophysics Institute - UFRJ). The images were obtained in three dimensions (Z-stack), and in all tests, 14 cellular fields were analyzed per well at 100x magnification.

The images of live cells were processed using Zen Lite Blue software (Zen Digital Imaging for Light Microscopy), and the Open-Source software ImageJ focused on biological-image analysis (Fiji software). Initially, images were processed to reduce background noise, and measurements of mean fluorescence intensity from a region of interest were performed to assess the fluorescence level only at that point. Fluorescent puncta counts (fluorescent-labelled vesicles) were performed on all sections of the Z-stack image using the Fiji plugin ‘cell counter’. For image composition, orthogonal projections were made, consisting of the union of all sections obtained from the 3D images created for the living cells.

Macrophage supernatants were collected before and after each treatment. For the negative control, cells were pre-incubated for 10 min with 100 µM cysteine proteases inhibitor E-64 before adding extracellular ATP. Cells were incubated with cathepsin assay buffer and the various fluorogenic cathepsin substrates [50 µM Z-Phe-Arg-AMC (Z-FR-AMC) (Sigma C9521), 50 µM Z-Arg-Arg-AMC (Z-RR -AMC)] (Sigma C5429). The reading of enzymatic kinetics was performed for 90 min at 37°C excitation/emission of 380/460 nm on a SpectraMax Gemini XPS (Molecular Devices, LLC. First Street San Jose, CA, USA).

The Magic Red™-MR-RR2 kit (Immunochemistry Technologies, USA) was used to quantify and monitor cathepsin B activity in cultured peritoneal macrophages. Magic Red 1:10 has been added directly into the culture (RPMI - not supplemented) containing 2 x 10⁵ cells/mL. When cathepsin is active, this substrate (MR-RR2) is cleaved and the cresyl violet fluorophore becomes fluorescent by excitation 592/628 nm (28). Cells were also incubated with 50 nM Lysotracker green for 30 minutes at 37°C and then incubated for 10 minutes with Hoechst 33342 to detect nuclear morphology. Subsequently, live cells were observed by spinning disk confocal microscopy before and after 5 mM ATP treatment.

Peritoneal macrophages plated in 25 mm round coverslips were loaded with loading solution containing 4 µM fura-2 AM (Thermo Fisher Scientific, MA, USA), Pluronic 0,04% at 37°C for 45 minutes. The coverslips were mounted in a 200 µl open imaging chamber in a PDMI-2 perfusion incubator (Warner Instruments) maintained at 37°C. Solutions were fed to the incubator with a peristaltic pump: physiological extracellular solution (145 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 10 mM HEPES, 1 mM CaCl₂, pH 7.4; control solution); physiological extracellular solution + 5 mM ATP in the absence or presence of 1 mM CaCl₂ and physiological extracellular solution + 2 mM EGTA, (0-Ca²⁺). Ratiometric imaging was performed through a 20×, 0.45 n.a. objective (Nikon Ti-U) and fura-2 filter set (Semrock) with an Evolve 512 EMCCD camera (Photometrics) and Lambda DG-4 illuminator (Sutter), under control of MetaFluor software (Molecular Devices). Fluorescence image pairs (F340 and F384) were obtained at 0.5 Hz and analyzed with Fiji-ImageJ. The image sequences were corrected for small movement artefacts (field translation), elliptical regions of interest (ROI) were drawn on each cell in the field, and the F340/F384 average intensity ratios (R) within each ROI was calculated without background subtraction. The R traces were further analyzed with Clampfit software (Molecular Devices). The peak changes of intracellular Ca²⁺ (ΔR) were estimated as the difference between the maximum R value (three-point moving average) during the ATP pulse and the 30-second average value preceding the pulse. For the ATP pulse in Ca²⁺-free medium, during which R was still slowly decaying due to recovery from the first ATP pulse, a decaying linear trend was subtracted from each trace before measuring the response. The decay rates were estimated by linear regression in 200-s segments of the R traces.

After pharmacological treatments described above, samples (2 x 10 ⁵cells/mL) were fixed with 4% paraformaldehyde and 4% sucrose for 15 min at room temperature and blocked with 10% horse serum (Thermo Fisher Scientific, MA, USA and 1% BSA in PBS for 30 min at room temperature. Samples were then incubated (overnight at 4°C) with the following primary antibodies (in 0.1% BSA in PBS): goat pannexin-1 antibody (K-20) (Santa Cruz Biotechnology, CA, USA) diluted 1:500. Cells were then washed and incubated at room temperature for 1 h with the following secondary antibody (diluted 1:300, in 0.1% BSA in PBS): anti-goat IgG (H+L)-Alexa Fluor^® 546 (Thermo Fisher Scientific, MA, USA). Finally, samples were mounted with Prolong Gold Antifade Mountant with DAPI (Invitrogen) and examined in a fluorescence microscope Zeiss AxioVert 200M and the three-dimensional images (z-stack) in a Spinning Disk Confocal Microscope ZEISS Cell Observer SD (Peabody, MA, USA).

Heart fragments were macerated in HEPES buffer (50 mM Hepes, 10 mM EDTA, 1 mM MgCl₂, 1% Triton X-100 and protease/phosphatase inhibitor cocktail - pH 7.4) using the Ultra-Turax T25 basic homogenizer (IKA, Germany) at 9500 L/min. This macerate was incubated on ice for 30 minutes and centrifuged at 12000 g for 15 minutes at 4°C. Non-stimulated peritoneal macrophages and SAOS-2 cells were scrubbed in 200 µl of RIPA buffer (20 mM HEPES pH 7.9, 20% glycerol, 200 mM KCI, 0.5 mM EDTA, 0.5% NP-40, 0.5 mM DTT, 1% protease/phosphatase inhibitor cocktail) using a cell scraper, then transferred to a dounce homogenizer. The samples were homogenized on ice for 15 min and left incubating for another 15 min, and then the cell lysate was centrifuged at 15000 g for 10 minutes at 4°C. The supernatant was used for protein dosage by the Bradford method, and 35 μg of protein were resolved in 12% SDS-PAGE at 150 V and transferred to a nitrocellulose membrane (Bio-Rad, CA, USA) for 65 minutes at 350 mA. After transfer, the membrane was blocked with 5% nonfat dry milk in 1X TBS-T [10 mM Tris, pH 7,4, 150 mM NaCl, 0.1% Tween20] for 2 hours and incubated overnight at 4°C with primary antibodies: goat anti-mouse pannexin-1 (K-20) (Santa Cruz Biotechnology, CA, USA) and anti-vinculin monoclonal antibody (Sigma Aldrich, Missouri, USA). The next day the membrane was washed with TBS-T and incubated mouse anti-goat IgG-HRP secondary antibody (ThermoFisher Scientific, MA, USA) for 1 hour at room temperature. After incubation, the washing procedure was repeated and the membrane developed using the chemiluminescent ECL Plus Western Blotting Detection System (GE Healthcare Life Sciences, NY, USA) or SuperSignal™ ELISA Femto Substrate (Thermo Fisher Scientific, MA, USA).

The data were analyzed using GraphPad Prism version 8.2, and values represent the mean value with standard deviation. Differences between means were assessed using the Student’s t-test, one-way analysis of variance followed or not by the Bonferroni test, all considering p < 0.05.

First, we investigated the effect of extracellular ATP (50 μM, 500 μM, and 5 mM) on non-stimulated peritoneal macrophage late endosome and endolysosome stability. These intracellular organelles were formed after endocytosis of 1 mM Lucifer Yellow (LY) fluorescent dye (26, 27). Spinning Disk Confocal Microscopy was used to observe continuously the same cell areas before and after eATP treatment. As depicted in Figure 1A , 50 μM ATP (concentration that does not activate the P2X7 receptor) did not induce late endosome or endolysosome stability changes. By contrast, 500 μM (p < 0.001) and 5 mM ATP (p < 0.0001) treatments significantly decreased the number of late endosomes/endolysosomes marked with LY ( Figure 1A , representative images; Figure 1C , quantification), suggesting that P2X7 receptor activation mediates lysosomal leakage. In parallel, a time dependent increase in the fluorescent content of macrophages’ cytoplasm and cell supernatant after the addition of eATP was observed (data not shown). This phenomenon was not detected in macrophages from P2X7 knockout mice. To verify the participation of P2 receptors such as P2Y₂, macrophages labeled with LY were treated with 100 μM UTP. However, this nucleotide did not diminish the number of LY-stained endolysosomes ( Supplementary Figures 1A, B ).

Next, the fluorescent probe Lysotracker Green was used in endocytosis assays to investigate the ATP effects on lysosomal stability. We observed a significant decrease in the number of lysosomes in peritoneal macrophages treated with 500 μM (p < 0.05) and 5 mM ATP (p < 0.01) but not with 50 μM ATP ( Figures 1B, C , representative images, Figure 1D , quantification).

To confirm that ATP-induced lysosomal destabilization leads to the release of organelle content to the macrophage cytoplasm, we performed an endocytosis assay with the YO-PRO^®-1 DNA-staining dye impermeable to the cellular membrane (29). After being exposed to YO-PRO^®-1 marker for 45 min, non-stimulated macrophage cells were washed three times with PBS to discard the free dye. In the absence of ATP ( Figure 1E - left), no fluorescence signal related to the YO-PRO^®-1 dye was detected. However, treatment with 5 mM ATP for 10 min induced YO-PRO^®-1 escape from endocytic vesicles staining the nucleus (p < 0.001) ( Figure 1E -right). The mean fluorescence intensity quantification of the cell nucleus confirmed that the fluorescence signal was significantly greater after 5 mM ATP treatment when compared in the absence of ATP ( Figure 1F ). The Cell Mask Deep Red Plasma Membrane Stain shows preserved cell membrane integrity.

The role of the P2X7 receptor in lysosomal leakage was confirmed by performing endocytosis assays with macrophages obtained from P2X7 knockout mice (P2X7^-/-) and macrophages from wild-type (WT) mice in the presence of 100 nM A740003, a specific P2X7 antagonist.

Peritoneal macrophages from P2X7^-/- animals treated with 5 mM ATP did not show a significant decrease in the number of endolysosomes stained with LY, suggesting an absence of leakage ( Figure 2B , representative images; Figure 2D quantification) when compared to macrophages from WT animals (p < 0.0001) ( Figures 2A, D ). The same result was obtained in WT macrophages pre-incubated with 100 nM A740003 for 10 min before 5 mM ATP treatment ( Figures 2C, D ). These results suggest that the effect triggered by 5 mM extracellular ATP is mediated by P2X7 receptor activation.

To determine the effect of P2X7 receptor activation on cathepsin B release from endolysosomes lumen to the macrophage cytoplasm, we used a fluorogenic substrate for cathepsin B (Magic Red MR-RR2). This substrate allows the detection of cathepsin B location and activation. Magic Red MR-RR2 substrate cleavage by cathepsin B in living cells releases the red fluorescent compound cresyl violet as a product (28). Macrophages from WT mice were also previously incubated with Lysotracker Green, allowing the visualization of the colocalization of endolysosomes and cathepsin B product (yellow dots resulting from the merge of the green and red labeling) before 5 mM ATP treatment ( Figure 2E - left). After ATP treatment, there was a significant decrease in the yellow signal inside lysosomes and a presence of an intense fluorescent red signal dispersed in the cell cytoplasm emitted by the MR-RR2 cleavage product, suggesting the extravasation of cathepsin from lysosomes to the cytoplasm ( Figure 2E - right).

Cathepsin B enzyme activity assays were also performed on cell supernatants before and after treatment with 5 mM ATP for 10 minutes. Extracellular ATP triggers the exit of active cathepsin B from WT macrophages ( Figure 2F ). This effect was significantly inhibited (p < 0.05) in WT macrophages treated with P2X7 receptor antagonist (100 nM A740003) or in macrophages from P2X7^-/- mice (p < 0.0001) ( Figure 2F ). These results reinforce the notion that P2X7 receptor activation leads to endolysosomal leakage, releasing cathepsin B and endocytosed fluorescent compounds from the lumen of these organelles.

Next, we investigated whether pannexin-1 pore participates in lysosomal rupture triggered by P2X7 receptor activation using the pannexin-1 inhibitors mefloquine and 10Panx. We found that peritoneal macrophages pretreated with 100 µM mefloquine ( Figure 3B ) or with 20 µM of the mimetic peptide pannexin-1 ¹⁰Panx ( Figure 3C ) did not show evidence of lysosomal rupture/destabilization after ATP treatment, as opposed to the control (no ATP) ( Figure 3A ). There was no significant decrease in the number of lysosomes after ATP treatment when cells were pretreated with pannexin-1 pore blockers ¹⁰Panx or mefloquine (p < 0.001) ( Figure 3D ). Pannexin-1 expression in non-stimulated peritoneal macrophages was confirmed by western blot and immunofluorescence analysis ( Supplementary Figures 2A, B ). Cardiac tissue lysates and osteoblasts were used as positive and negative controls, respectively.

Cathepsin B enzymatic assays were also performed in WT macrophages pretreated with 100 µM mefloquine (p < 0.001) and 150 µM probenecid as pannexin-1 blockers before ATP treatment ( Figure 3E ). Pannexin-1 inhibition diminished the effects of extracellular ATP on cathepsin B release. These results suggest that pannexin-1 participates in the lysosomal leakage and cell permeabilization triggered by P2X7 receptor activation.

P2X7 receptor activation triggers an increase in intracellular calcium concentration (30). This calcium mobilization induces the opening of the pannexin-1 pore, contributing to the effect of extracellular ATP in endolysosomes and cell permeabilization. Therefore, we tested the effect of calcium signals on ATP-induced endolysosomal leakage. Peritoneal macrophages previously incubated with LY were maintained in an external solution containing 2 mM EGTA, a Ca²⁺ chelator, and observed by confocal microscopy before and during 5 mM ATP treatment for 10 minutes ( Figures 4A, B ). EGTA significantly blocked the effect of extracellular ATP on endolysosome stability ( Figures 4B, D ). The same results were obtained using 10 μM BAPTA-AM, another calcium chelator ( Figures 4C, D ). These findings suggest the relevance of calcium signals for the effect of extracellular ATP on endolysosomal destabilization (p < 0.0001).

We have also ascertained that eATP opens Ca²⁺-permeable channels in WT non-stimulated macrophages using fura-2 ratiometric fluorescence imaging ( Supplementary Figures 3A, B ). ATP induced a fast rise in intracellular Ca²⁺ concentration in nearly all cells examined and the Ca²⁺ level decayed very slowly after removal of ATP for several minutes. After changing to a Ca²⁺-free extracellular medium, ATP could not evoke a Ca²⁺ response, demonstrating a requirement of the cation influx. This change was largely reversible since another response could be evoked after washing with the medium containing Ca²⁺, showing the integrity of macrophage cells during the assay ( Supplementary Figures 3A, B ).

The ¹⁰Panx is described as a pannexin-1 inhibitor that does not interfere directly with ATP-induced P2X7 responses (17). Considering that prolonged rise in intracellular Ca²⁺ induced by ATP could be partly sustained by influx through “large”, non-selective pannexin-1 channels we evaluated the effects of ¹⁰Panx in ATP-mediated Ca²⁺ responses. The addition of 50 µM ¹⁰Panx accelerated the decay of the ATP-induced Ca²⁺ response ( Supplementary Figures 4A-C ). Interestingly, in all 15 (of 70) cells which showed a slight initial fast rise followed by a further slow increase in R (positive slope), ¹⁰Panx caused a decay, reversing the trend. This result supports the interpretation that the P2X7-gated Ca²⁺ influx was followed by a secondary Ca²⁺ influx mediated by pannexin-1.

We also hypothesized that calcium-dependent/permeable channels present in the endolysosome membrane, including TRPM2 and TPC channels, would also contribute to eATP-induced organelle leakage. We used the TRPM2 inhibitor (50 μM 2-APB) to test this hypothesis. The 2-APB inhibited ATP-induced lysosome destabilization ( Figures 5A, B ). We also used a TPC channel inhibitor (verapamil 50 μM), and the same result was observed ( Figures 5C, D ). Finally, we determined whether the release of cathepsin induced by ATP treatment would be inhibited by 50 μM verapamil. The TPC inhibitor significantly reduced cathepsin release caused by 5 mM ATP treatment ( Figure 5E , p < 0.0001). These findings suggest that these two calcium-permeable channels might be involved in the mechanism associated with lysosomal leakage triggered by extracellular ATP acting via the P2X7 receptor. However, we do not exclude the possibility that P2X7 receptor activation also induces lysosomal alkalinization or exocytosis in these settings, as previously suggested (31, 32).