Mechanoresilience of lysosomes conferred by TMEM63A

Lysosomal damage contributes to a wide variety of pathophysiological settings. These range from infection, where endocytosed pathogens grow or release effectors that perforate endomembranes (Gaillard et al., 1987), to neurodegenerative disease, which can be associated with the aberrant aggregation of luminal fibrils in lysosomes (Filimonenko et al., 2007). In turn, the damage to lysosomes leads to tissue-wide responses, notably inflammation (Duewell et al., 2010). If left unchecked, lysosomal damage causes cell death (Hornung et al., 2008). Accordingly, there is an emerging appreciation for mechanisms involved in sensing lysosomal stress or damage and repairing the injured membrane.

Many of these responses center on increases to cytosolic calcium, proposed to leach from compromised organelles. The local gain in cytosolic calcium can conceivably lead to lysosome repair in a number of ways. Cytosolic calcium activates scramblases on the lysosome, exposing sphingomyelin to neutral sphingomyelinase (Niekamp et al., 2022). The cleavage of sphingomyelin to generate conically shaped ceramide imbues the inward curvature required for intraluminal vesiculation, an effect that stands to sort damaged or damaging components inward. Calcium also recruits and activates ESCRT complexes to the lysosome that remodel the membrane and provide resilience to damage (Chen et al., 2024; Skowyra et al., 2018). Finally, calcium could support fusogenic complexes to stimulate the recruitment of additional membrane via fusion with other (endo)lysosomes (Westman et al., 2020). This latter effect is critical, as membrane bilayers are minimally distensible, stretching only 2–5% before tension exceeds what the lipid packing can tolerate before undergoing rupture (Evans et al., 1976; Koslov and Markin, 1984). An additional set of secondary response pathways involves the directional flow of lipids to tensed lysosomes from the ER across membrane contact sites. VPS13C, ATG2, and PDZD8 have been proposed to achieve these ends and can be important for the comparatively slow expansion of the delimiting lysosomal membrane upon stress (Cross et al., 2023; Tan and Finkel, 2022; Wang et al., 2025; Yang et al., 2025).

An important question emerging from the preceding observations is, what are the mechanisms that directly sense lysosomal stress to initiate the responses that prevent rupture? Even in the absence of infection or injury, lysosomes are subjected to threats to their integrity yet are not damaged. To date, the sensory machinery that provides these degradative organelles with resilience against acute rupture is unknown. The recent discovery that members of the OSCA/TMEM63 family of mechanosensitive cation channels localize to vacuoles/lysosomes in cells suggests the possibility that these channels could be direct sensors of lysosomal stress (Li et al., 2024; Murthy et al., 2018). By outwardly conducting cations upon mechanical gating, OSCA/TMEM63 could also support both osmotic and/or calcium-dependent aspects of the response.

By acutely increasing the hydrostatic pressure and membrane tension on lysosomes, we find that these organelles withstand damage by the activation of TMEM63A. TMEM63A relieves high membrane tension by rapidly alleviating osmotic pressure. This enables the secondary responses that deliver additional membrane surface to lysosomes either through homotypic fusion or the transport of lipids from the ER, leading to the lysosome enlargement that protects from rupture. Deletion of TMEM63A leaves cells highly vulnerable to lysosomal cell death. These results demonstrate that organelles are capable of regulated tension and volume decrease as a primary mechanism of resilience in response to stress.

Numerous experimental conditions have been established that lead to lysosome stress and rupture. We performed dose-response or time-course experiments in primary bone marrow–derived macrophages (BMDM) and RAW264.7 macrophages with the cell-permeant lysosomotropic agents L-leucyl-L-leucine methyl ester (LLOMe) or glycyl-L-phenylalanine 2-naphthylamide (GPN), as well as sucrose or alum, which slowly accumulate in lysosomes by fluid-phase endocytosis. If the limiting membrane of the lysosome undergoes rupture, it leads to the recruitment of galectin scaffolds that detect the exposed glycans that comprise the luminal lysosomal glycocalyx (Fig. 1 A), which is easily scored. Additionally, membrane rupture causes the release of its luminal contents including, small membrane-impermeant dyes loaded by endocytosis, e.g., sulforhodamine G (SRG), and alkalinization, which can be assessed with LysoTracker (Fig. 1 A). With these approaches, we determined the concentrations required for LLOMe or GPN to cause rupture of the lysosomes in macrophages (Fig. 1, B–I). In contrast, we did not see rupture of lysosomes in the cells at any time after their incubation with sucrose or alum (Fig. 1, J–L and Fig. S1, A–C).

We went on to determine if tension is increased before rupture in the LLOMe or GPN experiments, which we have previously characterized for sucrose-laden lysosomes (Cai et al., 2024). This was done using the recently described Lyso-Flipper probe (Colom et al., 2018). Briefly, cells were treated with concentrations of LLOMe or GPN for time periods that did not cause lysis of the lysosomes. Under these settings, we noted increased lifetime of the Lyso-Flipper probe, suggesting the limiting membrane incurs increases in its lateral tension upon such perturbations (Fig. 2 A). An osmotic gain by the membrane-impermeant products (glycyl-L-phenylalanine and 2-naphthylamine) of GPN produced upon its cleavage by cathepsins is expected to generate proportional increases in membrane tension. Like GPN, LLOMe is cleaved by lysosomal cathepsins, but this results in a hydrophobic (Leu–Leu)n-OMe polymer (Thiele and Lipsky, 1990); the increased tension upon LLOMe may either result from pressure produced by the polymers directly or their membrane-disruptive properties.

Under such conditions, mechanosensitive channels could be opened. To determine which channels would be positioned to respond to increases in lysosome membrane tension, we first analyzed the expression of TMEM63A, B, and C in macrophages. TMEM63A and B were both well expressed in macrophages; TMEM63A was expressed at threefold higher levels in macrophages compared with murine fibroblasts (Fig. 2 B). We found that TMEM63A showed pronounced localization to lysosomes and phagosomes as indicated by coincidence with LAMP1-GFP (Fig. 2, C and D) (Li et al., 2024), suggesting that TMEM63A may play a role in the lysosomes of highly endocytic cell types.

To test for the role of TMEM63A in lysosomes of macrophages, we generated TMEM63A KO cells and TMEM63A-floxed LysM-Cre conditional KO mice (Fig. S1, D–F). We initially investigated the morphology of the tubular lysosome network in the macrophages derived from the WT versus conditional KO mice. The morphology of lysosomes in the TMEM63A KO cells appeared to be normal, and these retained 10-kDa dextran and SRG (Fig. 2, E and F). We also noted no differences in lysosomal pH in the TMEM63A KO cells (Fig. 2 G) or differences in LAMP1 or cathepsin C expression (Fig. 2 H). DQ-BSA was also taken up and degraded normally in the TMEM63A KO cells (Fig. 2 I), and the cells were capable of phagocytosis, even when challenged with a large number of phagocytic targets (Fig. 2, J and K). Taken together, these data suggested that, under steady-state conditions, there were no obvious defects in the lysosomes in the TMEM63A KO macrophages. We therefore proceeded to determine the function of TMEM63A under conditions in which it would presumably become activated, i.e., under high tension.

Tension on the limiting membrane of lysosomes increases upon their slow accumulation of solutes, including sucrose, which cannot be digested nor outwardly transported. The same occurs with the vaccine adjuvant alum. We found that neither insult caused the overt rupture of lysosomes in WT macrophages despite causing their growth, ostensibly by fusion (Fig. 3, A–E). We also did not find that the lysosomes in TMEM63A KO cells underwent any signs of rupture upon accumulating sucrose or alum (Fig. 3, A–E; and Fig. S2, C and D) or upon phagocytosis of latex microspheres (not shown), suggesting that the channel is not required to preserve (phago)lysosomal integrity when solutes/particles slowly accumulate or during phagocytosis itself. These data also suggest that TMEM63A is not required for maturation in the endocytic pathway or the fusion of lysosomes.

In stark contrast to the stress accumulated on lysosomes when they cannot digest particles or transport solutes, sudden shifts in hydrostatic pressure can cause swift increases in lysosome tension. At high enough concentrations, agents like GPN and LLOMe cause swelling and rupture of lysosomes in minutes. We therefore challenged the TMEM63A KO macrophages with GPN and LLOMe across a range of concentrations established previously (see Fig. 1).

Under these conditions, we determined lysosome rupture by the loss of luminal contents (SRG) and the accumulation of galectin-3, indicative of membrane lysis. Remarkably, the lysosomes in TMEM63A KO macrophages underwent rupture at concentrations of lysosomotropic agents far below those required to rupture lysosomes in cells with the channel expressed (Fig. 3, F–I and L–M; and Fig. S2, A and B). While the re-repression of the WT version of human TMEM63A rescued the TMEM63A KO cells from lysosomal damage, as measured by galectin-3 recruitment, expression of disease-causing variants of TMEM63A, i.e., those causing hypomyelinating leukodystrophies (Yan et al., 2019), did not (Fig. S3 E).

Upon investigating the recruitment of ESCRT, we found that the ESCRT-associated protein ALIX was recruited to an equal extent in TMEM63A KO cells upon lower doses of LLOMe as in WT cells (Fig. 3, J and K), indicating that a failure to recruit ESCRT was not ostensibly the cause of increased rupture. Indeed, a majority of the ALIX recruitment to stressed lysosomes in WT cells occurred without their rupture, as judged by galectin-3 (Fig. S2 F), which is consistent with its role in protection. The recruitment of ALIX in TMEM63A KO cells could occur only after damage (Fig. S2 F), perhaps not excluding the possibility that TMEM63A activation recruits ESCRT.

Lysosome-mediated cell death is a consequence of permeabilization of the limiting lysosomal membrane. Given the preceding observations, we tested if the TMEM63A KO macrophages were more susceptible to death upon treatment with titrated doses of LLOMe. Based on PI positivity, we found that the TMEM63A KO macrophages underwent cell death at much lower concentrations of LLOMe (Fig. 3 N), consistent with the findings of the increased lysosome damage sensitivity.

Pathogens, including the opportunistic fungal pathogen Candida albicans, can grow within phagolysosomes by shifting to a hyphal growth state (Fig. 3 O). We challenged primary macrophages with two different strains of C. albicans for 2.5 h and investigated the permeability of the phagolysosome, judged by the leaching of SRG preloaded to the compartment into the cytosol. Both WT and TMEM63A KO macrophages maintained the integrity of the phagolysosomes containing the yeast-locked C. albicans. However, when phagosomes contained Δece1 C. albicans, a strain that does not produce candidalysin but can undergo hyphal growth, we found that the loss of TMEM63A caused a significant fraction of the phagosomes to undergo rupture (Fig. 3, P and Q). This result is consistent with the notion that TMEM63A confers protection of the lysosomal membrane only upon mechanical challenge.

There are several putative mechanisms by which TMEM63A, a nonselective cation channel known to transport calcium, sodium, and potassium (Li et al., 2024), could protect lysosomes from damage. First, TMEM63A could outwardly transport Ca²⁺ to exert protective effects, for example, by recruiting ALG-2 (Chen et al., 2024). This was tested by removing Ca²⁺ from the lysosomes (and the cells) entirely: in the presence of Ca²⁺-free medium, cells were treated with thapsigargin to inhibit the SERCA pump and release Ca²⁺-stores from the ER (Fig. 4 A), reportedly required to supply lysosomes with Ca²⁺ (Calvo et al., 2025; Garrity et al., 2016). We tested if Ca²⁺ was depleted from the lysosomes in this condition by treating the cells with high concentrations of LLOMe to rupture the lysosomes. Here, we did not observe increases in GCaMP6s (Fig. 4 B), a Ca²⁺ biosensor expressed in the cytosol, indicating that the lysosomes were indeed depleted of Ca²⁺ under these conditions. We then treated such cells with concentrations of GPN and LLOMe that caused lysosomal rupture only in the absence of TMEM63A. In lysosomes devoid of Ca²⁺, we did not observe any loss in the protective effects of TMEM63A (Fig. 4, C–E). This would suggest that Ca²⁺ is not a relevant cation to confer TMEM63A lysosome protection.

In addition to Ca²⁺ flux, nonselective cation channels like TMEM63A could also result in an H⁺-leak from lysosomes, which could in turn activate lysosome repair pathways. To determine if TMEM63A could transport H⁺, we treated cells with 30 mM of sucrose for 1 h, which caused swelling and increased tension as judged by the Lyso-Flipper probe (Fig. S3 A). We noted that the lysosomes in TMEM63A KO cells in this condition had a modestly more acidic pH (Fig. S3 B). To monitor the leak of H⁺ in the tensed but intact lysosomes, we inhibited the activity of the V-ATPase with concanamycin A. The rate of lysosomal alkalinization, indicative of the ongoing H⁺ leak, was subtly decreased in the TMEM63A KO cells, suggesting that some of the H⁺ leak was attributed to TMEM63A (Fig. S3, B and C). We therefore investigated the main pathway activated in response to the stabilization of the V-ATPase and increases in its activity, i.e., the recruitment of LC3 to lysosomes via CASM (Durgan and Florey, 2022). We found that LC3 was clearly recruited to lysosomes upon LLOMe treatment but that this occurred in both WT and TMEM63A KO cells equally (Fig. S3 D). Alternative insults that selectively ruptured TMEM63A KO cells, on the other hand, did not lead to LC3 recruitment to lysosomes. Thus, it would seem unlikely that TMEM63A protects from lysosome damage by mechanisms related to the H⁺-leak. Instead, we noted that complete inhibition of the V-ATPase with concanamycin A rendered the lysosomes more susceptible to damage (Fig. S3 E). These results implicate secondary ion gradients, established by the H⁺ gradient, in the mechanoresilience of lysosomes.

If not Ca²⁺ or H⁺ release, TMEM63A could conceivably protect from lysosomal damage by forcing the exit of water from lysosomes to relieve high membrane tension. This could be achieved by the outward transport of alkali cations, major osmoticants of biological fluids, including for lysosomes, which contain high luminal Na⁺ (Wang et al., 2012). Volume regulation of lysosomes has indeed been shown to be regulated by VRAC and its transport of Cl^-, notably under conditions of osmotic stress (Li et al., 2020), which would require the parallel transport of cations to maintain electroneutrality. This seemed like a reasonable possibility since increases in hydrostatic pressure alone caused elevations in endomembrane tension (Fig. 5 A) and led to the loss of luminal Na⁺ as judged by Natrium Green, a pH-insensitive dye of the same size as SRG, precluding the possibility that it would be lost from the lumen (Fig. 5 B). The same treatment with hypotonic solutions caused a rapid loss of lysosome tubulation, which was restored over time in control cells (Fig. 5 E). TMEM63A was in fact required for lysosomes to withstand high osmotic pressure: the deletion of TMEM63A caused a pronounced loss of lysosome integrity and rupture upon challenging the cells with hypotonic solutions (Fig. 5, C–E). As was also shown for responses to LLOMe, the re-expression of human TMEM63A in TMEM63A KO cells rescued the lysosomal rupture upon hypotonic shock, but loss-of-function variants did not (Fig. S3 G).

There is a very small preference for Na⁺ over K⁺ transport by TMEM63A (Li et al., 2024). Outward transport of Na⁺ would therefore be balanced by the inward transport of K⁺ without any clear net loss of osmolytes if the channel were working in isolation. Instead, the parallel transport of Cl^- together with the inside positive membrane potential of the lysosome (Saminathan et al., 2021) would, in theory, result in the directional and electroneutral exit of cations and anions and oblige the exit of water. We therefore opted to inhibit lysosome-resident VRAC using the membrane-permeable inhibitor DCPIB; as a control, we used the membrane-impermeable VRAC inhibitor, DIDS, which prevented regulatory volume decrease (RVD) as expected (Fig. 5 F). Convincingly, we found that the inhibition of lyso-VRAC with DCPIB but not surface VRAC with DIDS rendered lysosomes highly susceptible to damage upon hypotonic shock (Fig. 5, G and H). By comparison, when the lysosome-to-cytosol gradient of Na⁺ was blunted by inhibiting the Na⁺/K⁺ pump with ouabain for extended periods, we did not observe increased susceptibility to damage to the same extent, suggesting that either Na⁺ or K⁺ transport suffices for the protective effects of TMEM63A (Fig. 5, G and H). Interestingly, ALIX was only recruited to damaged lysosomes in TMEM63 KO cells and not in WT lysosomes exposed to hypotonic solutions, indicating that lysosome repair pathways involving ESCRT were not invoked in this response unless TMEM63A was absent (Fig. 5, I and J).

We then sought to forcibly regulate the volume and resultant tension of lysosomes in the TMEM63A KO macrophages. This was achieved by shifting the osmolarity of the medium upward in order to drive the exit of water from the cells and from the lysosomes at the time of the addition of lysosomotropic agents. These results were revealing, as the hypertonic solutions strongly reverted the rupture of lysosomes in the TMEM63A KO macrophages as judged by decreased galectin-3 puncta and decreased release of lysosome contents into the cytosol (Fig. 5, K and L). Taken together, these results implicate a model for TMEM63A in sensing and relieving high tension on lysosomes (Fig. 5 M).

Threats to lysosomal integrity can increase tension on the membrane that delimits the organelle, an effect that causes its rupture if left unregulated. To offset sudden increases in tension, we propose that TMEM63A decreases organelle volume, reduces hydrostatic pressure, and effectively relieves high membrane tension. This could provide membrane slack and time for the organelle to obtain additional membrane via fusion with other lysosomes or lipid transport from the ER. Supporting this model, lysosomes experiencing sudden osmotic stress undergo a period of growth/expansion before undergoing secondary volume loss. These stages are not observed in the absence of TMEM63A because the stressed organelles undergo rapid lysis. The slow accumulation of solutes, on the other hand, proceeds without damage and without the need for TMEM63A. In such cases, fusion and/or the addition of ER lipids clearly suffice to maintain organelle integrity and occur independently of TMEM63A.

The OSCA/TMEM63 family of mechanosensitive channels is highly conserved through evolution; members can be found in plants, including on their vacuoles. Osmotic stresses in plants—incurred upon injuries to the cell wall—lead to the mechanical gating of these channels, presumably enabling the free passage of cations (Cao et al., 2020; Miao et al., 2022; Yuan et al., 2014). The expression of OSCA channels ultimately preserves vacuolar integrity upon cell wall damage and supports the viability of plants, which is akin to our model for TMEM63A on lysosomes.

While calcium would stand to be released through TMEM63A, and this could increase membrane fusion, we did not find there to be a role in whole-cell calcium for preserving lysosome integrity in very acute responses to lysosomotropic agents. Rather than using BAPTA-AM, which is cleaved by cytosolic esterases and accumulates in cells together with its toxic byproducts, formaldehyde and acetic acid, we instead depleted cells of their Ca²⁺ stores entirely, including those in lysosomes. Calcium is a critical mediator of plasma membrane and endomembrane repair, notably by its activation of the ESCRT machinery (Vietri et al., 2020). We found that the recruitment of ALIX, a protein that recruits the ESCRT-III to endomembranes, was unaffected in TMEM63A KO cells, suggesting that Ca²⁺ release or ESCRT-III recruitment are not responsible for the immediate protective effects of TMEM63A. A steady leak of cations or the decrease in membrane tension facilitated by TMEM63A and its role in ESCRT recruitment and activity before rupture is nevertheless possible.

A critical, outstanding question is how lysosomes gain surface to respond to insults. One dramatic example occurs in contexts in which phagosomes need to expand to accommodate rapidly growing pathogens harbored in their lumen. In such instances, lysosomes undergo substantive fusion with the phagosome, which begins to consume the entire population of LAMP1-positive compartments (Westman et al., 2020). Such homotypic fusion is also initiated when PIKfyve is inhibited, leading to vacuolation (Choy et al., 2018). An alternative or perhaps additional mechanism comes from lipid delivery from the ER via lipid-binding proteins, e.g., ATG2, VPS13C, and PDZD8. Lipid transport would need to also be coupled to scrambling. In that regard, it is interesting to note that TMEM63B can function as a lipid scramblase (Miyata et al., 2025), which also targets to lysosomes (Li et al., 2024). It may be interesting to investigate the potential coupling between TMEM63A, B, and the bridge-like lipid transfer proteins.

The role of TMEM63A in protecting lysosomes from damage to their membrane has immediate implications in inflammation and antigen cross-presentation (Canton et al., 2021). Since the channel protects from lysosomal damage, inactivation of the channel may increase inflammatory processes and the release of tumor-associated antigens. Other mechanisms of lysosomal resilience are surely also in place, including biophysical mechanisms that protect the bilayer like the lysosomal glycocalyx, membrane stabilization provided by ESCRT machinery, and potentially yet-to-be-determined ion transport pathways (Scott et al., 2025). This study establishes a framework whereby lysosomal resilience is considered greater than first appreciated, which would need to be taken into account when understanding the lost protection in pathology associated with lysosomal damage.

Heterozygous Tmem63a (flox/wt) mice were from obtained from Cyagen (S-CKO-05381) and crossed with LysMCre (+/−) mice in our facility. Tmem63a^flox/flox-LysMCre^+/− from these crosses were established as breeders. Offspring were genotyped following the manufacture’s protocol.

2−5 littermates per cage were housed in ventilated cages under standard housing conditions (40% humidity, 22°C, 12-h reversed light/dark cycle) with free access to a chow diet. All procedures were approved by the Animal Care Committee at the Hospital for Sick Children and were performed in accordance with regulations from the Animals for Research Act of Ontario and the Canadian Council on Animal Care. Mice were kept at the Animal Facility in the Peter Gilgan Center for Research and Learning. Primary murine macrophages were derived from the marrow of femoral bones from 6- to 8-wk-old C57BL/6 WT, Tmem63a^flox/flox-LysMCre^+/−, and littermate Tmem63a^flox/flox-LysMCre^−/− (WT) control mice.

BMDMs were harvested by centrifugation and grown in DMEM with L-glutamine–containing 10% heat-inactivated fetal calf serum (FBS) and 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin. BMDMs were plated at 1 × 10⁶ cells ml⁻¹ and differentiated using 10% L929-conditioned medium for 5 days at 37°C with 5% CO₂ on Petri dishes. Cells were lifted using 5 mM EDTA in PBS, pH 7.4, and seeded onto coverslips.

RAW264.7 were purchased from ATCC (TIB-71) and cultured in DMEM + 10% FBS at 37°C with 5% CO₂. RAW264.7 TMEM63A KO cells were generated by CRISPR-Cas9 as previously described (Cai et al., 2024). Transfection of RAW264.7 cells using FuGene HD was done according to the manufacturer’s protocol.

Cells seeded onto coverslips were pulsed with 5 mg/ml sulforhodamine G (S-6957; Molecular probes) up to overnight and chased for >1 h before the addition of LLOMe, GPN, alum, sucrose, pathogens, or hypotonic solutions. Cells were imaged live in HBSS (311-512 CL; Wisent). The mean fluorescence of the cytosolic SRG signal of each cell was determined using three regions of interest, subtracting background signal, and normalizing to the highest value (i.e., lysosomal) for each experiment.

A 63×1.35 numerical aperture oil immersion objective using an Olympus IX81 Quorum spinning disc confocal microscope was used to image. The microscope was equipped with a CSU-X1-A Yokogawa spinning-disc unit and a Hamamatsu C9100-13 EM-CCD camera controlled by Volocity 6.1.2 software (PerkinElmer). Fluorescence images shown as volume-render projections belong to Z-stacks acquired at 0.4 μm.

RAW264.7 cells and BMDMs were seeded on 18-mm coverslips in 12-well tissue culture plates 16 h prior to experimentation. Cells were fixed and permeabilized with ice-cold 100% MeOH for 5 min at −20°C and blocked in 1% BSA for 1 h. Subsequently, the cells were incubated with the indicated primary and secondary antibodies in 1% BSA for 1.5 h at room temperature. All primary antibodies and Alexa488-, Cy3-, or Cy5-conjugated secondary antibodies (Jackson ImmunoResearch) were diluted to 1:1,000 in blocking solution. Three washes with 1X PBS were performed after each incubation. Slides were mounted using mounting media containing DAPI (P36931; Invitrogen).

RAW264.7 cells were seeded on 18-mm coverslips in 12-well tissue culture plates 16 h prior to experimentation. Cells were incubated with 2 μm Lyso Flipper TR (CY-SC022; Cytoskeleton) at 37°C for 10 min in serum-free DMEM before washing three times with PBS and imaging in HBSS. Fluorescence-lifetime imaging microscopy (FLIM) was acquired using the Leica Stellaris 8 FALCON w/STED. The Leica LAS X software (Stellaris) was used for analysis.

RAW264.7, BMDM, and mouse embryonic fibroblasts were plated on 6-well tissue culture plates and harvested 16 h after. Cellular RNA was extracted using the GeneJET RNA purification kit (Thermo Fisher Scientific). Equal RNA from each condition was used to generate cDNA with the Superscript VILO cDNA Synthesis Kit (Thermo Fisher Scientific). Quantitative real-time PCR was then performed with TaqMan probes specific to each assayed gene on a QuantStudio 3 Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific) in 96-well plates. TaqMan probes targeting mouse genes Tmem63a (Mm00522653_m1; no. 4331182), Tmem63b (Mm01269695_m1; no. 4351372), and Tmem63c (Mm00553595_m1; no. 4331182) were from Thermo Fisher Scientific (no. 4448892). Readouts for target and reference (Abt1) genes were performed in duplicate. Relative gene expression was then calculated based on the 2^−ΔΔCt method by comparing the relative quantification of the control gene with each target gene.

Cells seeded on coverslips were pulsed with 25 µg/ml of 10-kDa Oregon green dextran (D7170; Thermo Fisher Scientific) overnight and chased for >1 h before measurement. The cells were excited at 490 and 440 nm wavelengths in HBSS buffer, followed by incubation of pH standard solutions with pH values at 4.5, 5.5, 6.5, and 7.5 to obtain a standard pH curve. The solutions were made with 143 mM KCl, 5 mM glucose, 1 mM MgCl₂, and 1 mM CaCl₂ buffered with 20 mM acetic acid (pH 4.5), 20 mM MES (pH 5.5 and 6.5), or 20 mM HEPES (pH 7.5). A concentration of 10 μM nigericin (11437; Cayman chemical) and 5 μM monensin (475895; Sigma-Aldrich) was added to each pH buffer before use. The curve was graphed from the background-subtracted values for the ratiometric fluorescence (490/440 nm). The corresponding baseline pH was calculated from the standard curve.

Imaging was performed on a microscopy system that consists of a microscope (Axio Observer, Zeiss), a cooled CCD camera (Evolve 512; Photometrics), and a fluorescent light source (HXP 120V, Zeiss). Ratiometric imaging was performed with two filter cubes: a 470/40 nm and a 436/20 nm. The microscope is operated using ZEN 2 blue edition software (Zeiss), and images were acquired with a 63×/1.4 NA oil objective (Zeiss).

BMDMs were seeded on 18-mm coverslips in 12-well tissue culture plates 16 h prior to imaging. Cells were loaded with 10 µg/ml DQ-BSA Red (D12051; Invitrogen) for 1 h and imaged in HBSS using confocal microscopy.

BMDMs were seeded on 18-mm coverslips in 12-well tissue culture plates 16 h prior to imaging. Cells were then incubated with human IgG-opsonized 4.98-μm silica beads (SS05003; Bangs Laboratories Inc.) for 30 min. Cultures were then fixed with 3% PFA for 20 min and incubated with fluorescent antihuman secondary antibodies for 1 h at room temperature to visualize beads that were not ingested by the macrophages. Cells were then permeabilized with 0.1% Triton X-100 for 5 min and incubated with a different fluorescent antihuman secondary antibody to visualize internalized beads for 1 h. F-actin was stained with phalloidin to delineate individual macrophages.

RAW264.7 cells were grown to 80% confluency in 6-well plates and challenged with the indicated concentrations of LLOMe for 30 min. After 25 min, 0.3 µg/ml of PI (P1304MP; Invitrogen) was added for 5 min. Three images of different fields of view of each well were taken in both RFP and bright-field settings of an EVOS microscope (Thermo Fisher Scientific). Quantification of PI positivity was done in ImageJ.

C. albicans cultures were grown at 30°C overnight in liquid yeast peptone dextrose (YPD) medium (Bioshop). Overnight C. albicans cultures were diluted to OD_600nm ∼0.1 in YPD media and grown at 30°C until OD_600nm reached 1.0. Prior to infection, C. albicans were opsonized with human serum for 60 min at room temperature with rotating. The subculture was diluted into DMEM containing 10% FBS. Using a 270-gauge needle, C. albicans aggregates were dispersed.

RAW264.7WT and TMEM63A KO cells were seeded on 18-mm coverslips in 12-well plates, and were pulsed with SRG overnight. For infections, the medium was removed from the wells and replaced with 1 ml of C. albicans containing DMEM. Plates were centrifuged at 1,500 rpm for 1 min, then incubated at 37°C and 5% CO₂ for 2.5 h. Cells were then imaged live in HBSS.

RAW264.7 cells expressing GCaMP6s were washed with PBS three times, followed by incubation in Ca²⁺ free Tyrode’s buffer (10 mM HEPES, 10 mM d-glucose, 5 mM KCl, 140 mM NaCl, 1 mM MgCl₂, and 1 mM EGTA) for 5 min. Cells were recorded every 15 s in Ca²⁺-free Tyrode’s buffer and were treated with either 2 µM thapsigargin or 3 mM of LLOMe. Samples were imaged by spinning disc confocal microscopy using a 25× objective. The change in GCaMP6s fluorescence was quantified by normalizing to the highest value for each cell.

WT and TMEM63A KO RAW264.7 cells seeded on 18-mm coverslips in 12-well tissue culture plates were incubated with 10-kDa Oregon Green dextran overnight and chased for >1 h before measurement. The pH of the lysosomes was measured ratiometrically over 18 min, and cells were acutely treated with 500 nM concanamycin A (11050; Cayman Chemical) at the 4 min time point.

BMDMs were seeded on 18-mm coverslips in 12-well tissue culture plates and were loaded with 25 µg/ml of 10-kDa Alexa Fluor-647 dextran (D22914; Thermo Fisher Scientific) and 5 mg/ml ING-2 TMA+ salt (ION Biosciences) overnight and chased for >1 h before measurement. Cells were imaged in HBSS using live confocal microscopy.

RAW264.7 cells were grown to 80% confluency in 6-well plates. After, they were lifted mechanically. 0.5 ml of collected cell suspension of each condition was diluted with 3 ml Isoton II Diluent 531 (Beckman Coulter) and 6 ml distilled water immediately before analysis with a Coulter counter (Beckman Coulter Life Sciences, Multisizer 4). The median cell diameter for >10,000 cells ranging from 5 to 25 µm was measured at 0, 3, 10, and 15 min. The diameter was converted to volume by assuming that the suspended cells were spherical. Values were normalized to the 0 min time point.

Confocal microscopy was performed on a spinning-disk system (Quorum Technologies Inc.), consisting of a microscope (Axiovert 200 M, Zeiss), cooled CCD camera (ORCA-Fusion BT, Hamamatsu), five-line laser module (Spectral Applied Research) with 405-, 443-, 491-, 561-, and 655-nm lines, and a filter wheel (MAC5000; Ludl). The microscope is operated using Volocity v6.3 software (Perkin Elmer). Images were acquired with a 63×/1.4 NA oil objective (Zeiss).

TMEM63A-OFP was purchased from Sino Biological (MG51287-ACR). G168E and G567S were generated with site-directed mutagenesis using Q5 High-Fidelity DNA Polymerase (New England BioLabs, M0494). LAMP1-GFP was a gift from Ron Vale, Whitehead Institute, Boston, MA, USA (# 16290; Addgene plasmid). EEA1-GFP was a gift from Silvia Corvera, U Mass Chan Medical School, Worchester, MA, USA (# 42307; Addgene plasmid). GCaMP6s was a gift from Douglas Kim, NIH, Bethesda, MD, USA and GENIE Project (# 40753; Addgene plasmid).

L-Leucyl-L-Leucine methyl ester (L7393; Sigma-Aldrich), Gly-phe-β-naphthylamide (14634; Cayman Chemical), alum hydroxide (21645-51-2; Invivogen), and sucrose (57-501; Millipore Sigma) were used at indicated concentrations. Primary antibodies against galectin-3 (ab76245; Abcam) and LAMP1 (1D4B; DSHB) were used at 1:1,000 (vol/vol) for immunofluorescence. Anti-ALIX (634502; BioLegend) and LC3B (53566; Cell Signaling) were used at 1:500 (vol/vol) for immunofluorescence. LysoTracker Red DND-99 (L7528; Thermo Fisher Scientific) was used at 1 µM. Primary antibodies against LAMP1 (1D4B; DSHB), Cathepsin C (sc-74590; Santa Cruz) and GAPDH (365062; Santa Cruz) were used at 1:1,000 (vol/vol) for western blotting. Thapsigargin (586005; Calbiochem), ouabain (O3152; Sigma-Aldrich), DCPIB (1540; Tocris), and DIDS (D3514; Sigma-Aldrich) were used at 2 µM, 2 mM, 100 µM, and 250 µM respectively.

All experiments and data were conducted with at least three biological replicates or cells with different passage numbers. Number of experiments and cells quantified are indicated in the figure legends. Error bars represent mean ± SD unless stated otherwise. P < 0.05 was considered statistically significant. All statistical analyses were performed in Prism software (GraphPad). An unpaired two-tailed Student’s t test was used to compare the two groups. For multiple comparisons, one-way ANOVA or two-way ANOVA with Tukey’s multiple comparisons test was used.

Fig. S1 shows that lysosomes are not damaged in macrophages that accumulate of sucrose over time. Additionally, the strategy and validation of TMEM63A KO macrophages derived from TMEM63A-floxed;LysM-Cre cKO mice is shown. Fig. S2 shows lysosomal integrity/damage in WT versus TMEM63A KO or variant-expressing RAW264.7 cells. Fig. S3 assesses the H⁺ leak in WT or TMEM63A KO macrophages. Additionally, the CASM pathway and lysosomal integrity/damage in WT versus TMEM63A KO or variant-expressing RAW264.7 cells upon hypotonic shock is determined.

Original data are available from the corresponding authors upon reasonable request.

S.A. Freeman is the recipient of a Canada Research Chair and supported by grants PJT-169180 from the Canadian Institutes of Health Research and RGPIN-2022-04485 from the Natural Sciences and Engineering Research Council of Canada.

Author contributions: Angelina S. Kim: formal analysis, investigation, methodology, validation, visualization, and writing—original draft, review, and editing. Jing Ze Wu: conceptualization, methodology, and supervision. Ruiqi Cai: conceptualization, investigation, methodology, project administration, resources, supervision, validation, and writing—review and editing. Spencer A. Freeman: conceptualization, funding acquisition, supervision, and writing—original draft, review, and editing.