Heterogeneity in lysosomal dynamics and metabolic functions along the kidney proximal tubule

Monika Kaminska; Imene B Sakhi; Nevena Jankovic; Marcello Polesel; Andrew M Hall

doi:10.1038/s41467-026-70306-5

. 2026 Mar 9;17:3677. doi: 10.1038/s41467-026-70306-5

Heterogeneity in lysosomal dynamics and metabolic functions along the kidney proximal tubule

Monika Kaminska ^1,^#, Imene B Sakhi ^1,^#, Nevena Jankovic ¹, Marcello Polesel ¹, Andrew M Hall ^1,^2,^3,^✉

PMCID: PMC13100021 PMID: 41803103

Abstract

The kidney proximal tubule is a highly specialized epithelium that transports metabolites and maintains body homeostasis. Cells lining this nephron segment are densely packed with lysosomes, but little is known about the dynamic activity of these organelles in situ. Here, using targeted sensors and live cell and intravital imaging we track acidified lysosomes along the mouse proximal tubule and uncover marked axial heterogeneity in their distribution, characteristics and organellar interactions. In the early part, cathepsin-rich lysosomes frequently contact with apical endosomes to receive and catabolize filtered plasma proteins. Conversely, in the later region, lipase-containing lysosomes traverse cells to mobilize and degrade mitochondria-associated lipid droplets and facilitate their extrusion into the tubular lumen. Acutely de-acidifying lysosomes dramatically alters their movement, causing major changes in tubular protein and lipid processing. Thus, lysosomes in proximal tubules are highly dynamic and adapted to perform distinct metabolic tasks within different specialized segments.

Subject terms: Nephrons, Kidney, Lysosomes

Cells lining the kidney proximal tubule are densely packed with lysosomes, but little is known about their dynamic activity in situ. Here, the authors use intravital imaging and targeted probes to study lysosomal behavior along the kidney proximal tubule, and discover substantial axial heterogeneity in lysosomal dynamics, organellar interactions and metabolic functions.

Introduction

The kidney proximal tubule (PT) is a critical metabolic hub of the body that transports and catabolizes circulating substances to preserve the inner milieu^1–3. PT cells are structurally optimized for this purpose, with an apical brush border containing many transporters, a basal zone packed with highly active mitochondria, and a sophisticated endo-lysosomal system (ELS) in between^4,5. Effective communication between different cellular regions and organelles is likely pivotal for performing integrated metabolic tasks in vivo.

Lysosomes are highly acidified organelles that play a key role in intracellular metabolite digestion, and lysosomal damage is increasingly recognized as an important cause of kidney diseases^6,7. The PT is divided into 3 segments (termed S1-3) based on gene expression profiles and ultrastructural appearances, and the morphology of lysosomes evolves from the beginning to the end, hinting at functional heterogeneity^4,8,9. A major task of the PT is to retrieve filtered plasma proteins by receptor mediated endocytosis (megalin and cubilin system) and degrade them to conserve important nutrients^5,10. Recent functional imaging studies have localized this process to the early part of the PT^11,12, but how the dynamic behavior of lysosomes is adapted to this or other metabolic pathways along the PT is not well understood.

Here, we have used pH-dependent probes, custom-designed delivery vectors and high-resolution intravital microscopy to label and track acidified lysosomes in mouse PTs in vivo, to derive their characteristics, movements and organellar interactions in different segments and investigate the metabolic consequences of acutely de-acidifying them.

Results

Generation of proteins labeled with pH sensitive fluorophores

As a first targeting vector to deliver fluorescent probes to the PT ELS, we used the small protein lysozyme, which we have previously shown is rapidly filtered from the blood and endocytosed in the mouse PT¹². Importantly, lysozyme is highly resistant to endo-lysosomal degradation¹³, and thus represents a stable carrier. We labeled lysozyme with either the pH sensitive fluorophore pHrodo red or the pH insensitive far red fluorophore Atto-647N (Fig. 1a). Crucially, both probes can be excited at a single wavelength (850 nm) when performing multiphoton imaging, allowing simultaneous collection of emitted signals. Experiments performed with the proteins in solution confirmed that the two signals provided a ratiometric readout of pH over the physiological range (4.5–7.5) (Fig. 1b, c).

Next, we incubated PT-derived cells (OK cells) with the proteins and observed that they were endocytosed and trafficked to lysosomes (Fig. 1d, e). We then clamped intracellular pH at different values using ionophores (Nigericin and Valinomycin) and a V-ATPase inhibitor (Bafilomycin A1), and confirmed that the protein signals yielded a ratiometric valuation of vesicular pH (Fig. 1f, g). Finally, to acutely alkalinize endo-lysosomes we treated cells with the lipophilic weak base hydroxychloroquine (HCQ, 20 µg/ml), which produced an abrupt rise in luminal pH (Fig. 1h, i).

Generation of a peptide for ratiometric pH imaging

As a second targeting approach, we generated a 20 amino acid peptide, which was custom-designed to allow selective tagging at each end, thus producing a dual labeled vector. This size was chosen because we have previously found that it is sufficient to trigger endocytic uptake in the PT¹¹ (small peptides are transported by other mechanisms¹⁴). Targeted dual labeling was achieved at one end by the reaction of the primary amine of the N-terminal with the N-hydroxysuccinimide ester of Atto 647 N, and at the other end by the reaction of the thiol SH of cysteine with the C-terminal of maleimide pHrodo red (Fig. 2a). The mass, purity and stability of the labeled peptide were confirmed by high pressure liquid chromatography (HPLC) and mass spectrometry (MS) (Supplementary Fig. 1). To verify labeling positions, the peptide was first cleaved using targeted trypsin digestion into two fragments, which were then analyzed separately by tandem mass spectrometry (MS/MS) (Supplementary Fig. 2).

Experiments performed with the labeled peptide in solution confirmed that it also provided a ratiometric readout of pH (Fig. 2b). Moreover, it was trafficked to lysosomes in OK cells (Fig. 2c, d), where it displayed expected signal changes across the physiological pH range (Fig. 2e, f), and in response to HCQ (20 µg/ml) (Fig. 2g).

Probe progression through the endo-lysosomal system

Before proceeding with intravital imaging experiments, we first wanted to determine the kinetics of protein progression through defined components of the PT ELS. We therefore stained tissue for early endosomes (EEs, Rab5), late endosomes (LEs, Rab7) and degrading lysosomes (cathepsin L), and determined their spatial locations relative to the actin brush border and their spatial profiles in a line scan analysis (Supplementary Fig. 3). Next, we injected mice with labeled lysozyme, fixed tissue at different time points assessed position and co-localization with the structural markers at different time points (Fig. 3a–c and Supplementary Fig. 4)^12,15. Of note, LEs in the PT have a characteristic enlarged structure and are also known as large apical vacuoles (LAVs)⁵.

Fig. 3 — a Immunolocalization of the brush border (Actin), early endosomes (Rab5), late endosomes/large vacuoles (Rab7) and lysosomes (Cathepsin L) in proximal tubules (PTs) in fixed mouse kidney tissue showed different morphology and localization of these endo-lysosomal system (ELS) compartments. Scale bars: 5 µm. b Tissue was fixed at different time points after Lysozyme-647 injection to show progression of the protein through ELS compartments. Arrows denote co-localization of signals. Scale bars: 5 µm (zoomed-out) and 2 µm (zoomed-in). This experiment was performed once for each time point with a clear progression of the protein with time. c Graphical representation of the data from A and B. Generated using Adobe Illustrator 2025 (version 29.7). d Labeled pH sensitive and insensitive probes were injected intravenously in mice and imaged with 2-photon microscopy. Created in BioRender. Kaminska, M. (2026) https://BioRender.com/n77wswc. e Lysozyme-pHrodo and Lysozyme-647N were filtered and reabsorbed in the PT and changes in pH were tracked through the ELS; arrows depict the brush border (BB) at 1 min, the early endosomes (EE) at 3 min, the late endosomes/large apical vacuole (LAV) at 6 min and lysosomes (Lyso) at 10 min. Scale bars: 10 µm (zoomed-out) and 5 µm (zoomed-in). Normalized fluorescence ratios are depicted for the different structures as mean ± SEM. Groups were compared using a one-way ANOVA followed by multiple comparison tests (n = 3 animals were considered as biological replicates with 20 to 21 regions of interest (ROIs) considered per replicate; ROIs from all replicates were merged to estimate a mean value for each compartment; p values as indicated on the chart). f Following intravenous injection, the dual-labeled peptide was filtered and endocytosed in proximal tubules; overlay of fluorescence was tracked in the BB (1 min), the EE (3 min), the LAV (6 min) and Lyso (10 min) as depicted by arrows. Scale bars: 10 µm (zoomed-out) and 5 µm (zoomed-in). Normalized fluorescence ratios are depicted for the different structures as mean ± SEM. Groups were compared using a one-way ANOVA followed by multiple comparison tests (n = 3 animals were considered as biological replicates with 17 to 20 regions of interest (ROIs) considered per replicate; ROIs from all replicates were merged to estimate a mean value for each compartment; p-values as indicated on the chart). g Graphical representation of the data from (e, f). Generated using Adobe Illustrator 2025 (version 29.7). Source data are provided as a source data file.

After 1 min, lysozyme was located at the base of the brush border, and at 3 min co-localized with Rab5 vesicles, denoting entry into EEs. After 6 min, lysozyme reached Rab7 positive LEs/LAVs lying beneath EEs. Finally, after 10 minutes, lysozyme entered cathepsin positive lysosomes, which were distributed in a narrow band just below the LEs/LAVs (Fig. 3a–c). These reference time points were subsequently used to map the location of lysozyme during intravital imaging experiments.

Intravital imaging of acidified endo-lysosomes

To target probes to the PT ELS in mice in vivo, we performed simultaneous intravenous injections of the two labeled lysozyme species and tracked their progression in real time with intravital multiphoton microscopy (Fig. 3d). As expected, the proteins were rapidly filtered from the bloodstream and endocytosed by early (S1) PT cells (Supplementary Fig. 3), which allowed us to assess the spatiotemporal evolution of pH changes as proteins sequentially traversed specific structures.

After 3 min, we observed a striking increase in the ratio of pHrodo to Atto-647N signals in small vesicles corresponding to the location of EEs (Fig. 3e). At 6 minutes, proteins were visible in larger vesicles beneath the EE layer, denoting trafficking into LEs/LAVs (Fig. 3e). Surprisingly, this was not associated with a marked change in the pHrodo/Atto-647N signal ratio, indicating only a minor pH difference between EEs and LEs. After 10 min, proteins were visible in small vesicles surrounding LEs/LAVs (Fig. 3e), corresponding to the morphology and location of cathepsin positive lysosomes. Importantly, these structures displayed a large, step increase in pHrodo/Atto-647N signal ratio, signifying that they were highly acidified (Fig. 3e).

To validate the results obtained with labeled lysozyme species, we injected mice intravenously with the custom-designed peptide, dual labeled with pHrodo and Atto-647N. Post injection, the peptide was rapidly filtered from the bloodstream and reabsorbed in early PTs. The progression of peptide through PT cells displayed a similar spatiotemporal profile to that of lysozyme, with clear increases in pHrodo/Atto-647N signals ratio occurring in vesicles corresponding to EEs and cathepsin positive lysosomes (Fig. 3f, g). Thus, by using two different vectors and antibody labeling specific structures, we were able to pinpoint where substantial pH changes occur within the PT ELS in vivo (Fig. 3g).

Effects of de-acidifying endo-lysosomes

To assess the consequences of acutely alkalinizing endo-lysosomes on protein reabsorption in S1 PTs, we injected mice intravenously with HCQ (20 mg/kg) 30 minutes before injection of lysozyme. Treatment with HCQ induced a severe defect in protein uptake and a corresponding increase in urinary protein excretion (Fig. 4a, b). To investigate the potential cause of this, we performed antibody staining for the multi-ligand receptor megalin, which binds luminal proteins to facilitate their internalization. We observed a significant shift in the location of megalin away from the apical membrane and an increased co-localization with the LE/LAV marker Rab7 (Fig. 4c–f), suggesting that HCQ alters its recycling, which likely contributes to the striking impact on protein reabsorption (Fig. 4g).

Fig. 4 — a Lysozyme-647N emitted fluorescence in proximal tubules (PTs) in control and hydroxychloroquine (HCQ, 20 mg/kg, i.v.) treated mice indicated a reduced protein endocytosis in the latter. Scale bars: 20 µm. This experiment was repeated on 2 animals with similar results. b Coomassie staining of a gel containing urines from mice injected with HCQ (20 mg/kg) for 3 h indicating increased proteinuria when total protein signal was normalized by urinary creatinine. Groups were compared using a two-tailed t student test. Data depicted are mean ± SEM (n = 5 animals were considered as biological replicates; p = 0.0337). c Immunolocalization of the brush border (actin) and megalin in proximal tubules (PTs) from control and HCQ (20 mg/kg) treated mice; Scale bars: 10 µm (zoomed-out) and 2 µm (zoomed-in). This experiment was repeated on 3 individual animals. (d Object-based analysis was used to segment actin and megalin structures and showed decreased colocalization and increased separation between these in HCQ-treated mice. Groups were compared using a two-tailed student t test. Data depicted are mean ± SEM (n = 3 animals were considered as biological replicates; objects were identified on 4 images per replicate and all data extracted from this analysis merged to estimate a mean value per group; p < 0.0001). Scale bars: 5 µm. e Immunolocalization of megalin and the late endosome/large apical vacuole marker Rab7 in 3D view in control and HCQ treated mice on a 3.5 µm thick z-stack with a z-step of 180 nm. Scale bars 2 µm. f Objected-based analysis from 2D images revealed increased co-localizing areas between Rab7 and megalin in HCQ-treated mice. Groups were compared using a two-tailed student t test. Data depicted are mean ± SEM (n = 3 animals were considered as biological replicates; objects were identified on 3 images per replicate and all data extracted from this analysis merged to estimate a mean value per group; p < 0.0001). g Graphical representation of the data from a-e. Generated using Adobe Illustrator 2025 (version 29.7). Source data are provided as a source data file.

Lysosomes in S1 proximal tubules degrade filtered proteins

To visualize lysosomal trafficking in S1 PTs in vivo, we injected mice with lysozyme labeled with Atto-565, a bright and photostable dye that is optimized for high-resolution time-lapse imaging¹¹. From this, we observed small cargo containing vesicles detaching from larger vacuoles in the main band of ELS structures before returning and re-attaching a few minutes later (Fig. 5a). Next, to pinpoint where catabolism of filtered proteins occurs, we injected mice with lactoglobulin highly labeled with Atto-532, where the emitted fluorescence signal remains quenched until the protein undergoes degradation¹¹. When co-injected with lysozyme (labeled with Atto-647N), we noted that unquenching of lactoglobulin occurred in small vesicles surrounding LEs/LAVs, matching the morphology and location of the previously identified, highly acidified vesicles (Fig. 5b).

Fig. 5 — a Intravital imaging of mouse kidneys following intravenous injection of Lysozyme-565. Temporal color coding denotes trafficking of cargo containing vesicles in the proximal tubule over time. In the example depicted, a discrete vesicle can be observed detaching from the main region of endo-lysosomes, before returning again and re-attaching. Scale bars: 20 µm (zoomed-out, left image) and 5 µm (zoomed-in, right image). This experiment was repeated on 3 individual animals. b Highly labeled lactoglobulin-532 was co-injected with Lysozyme-647 in mice. After 6 min, only Lysozyme-647 was visible in proximal tubules. After 15 minutes, fluorescence from lactoglobulin-532 appeared in small vesicles surrounding larger structures. Scale bars: 10 µm (zoomed-out) and 2 µm (zoomed-in). This experiment was performed on one mouse co-injected with the 2 probes. c Fluorescence signal ratios of Lysozyme-pHrodo to Lysozyme-647N in early endosomes (EE) and large apical vacuoles (LAV) after treatment with hydroxychloroquine (HCQ, 20 mg/kg, i.v.). Discrete lysosomes surrounding LAVs could not be visualized. Scale bar: 10 µm. Groups were compared using a one-way ANOVA followed by multiple comparison tests. Data depicted are mean ± SEM (n = 3 animals were considered as biological replicates with 15 to 20 regions of interest (ROIs) considered per replicate; ROIs from all replicates were merged to estimate a mean value for each compartment; p(BB to EE) = 0.0002, p(EE to LAV) = 0.1049, p(BB to LAV) < 0.0001). d Following intravenous injection, Lysozyme-565 was filtered and endocytosed in early proximal tubules. Temporal color coding denotes trafficking of cargo-containing vesicles over time (arrows). HCQ was injected (20 mg/kg) and after 25 min vesicular trafficking into the basal region was reduced (n = 3 animals were considered as biological replicates). Scale bars: 10 µm (zoomed-out) and 5 µm (zoomed-in). e Immunolocalization of the late endosome/large apical vacuole (LAV) marker Rab7 and the lysosomal protease Cathepsin L in control and HCQ treated mice, showing increased contact between these structures in the latter. The left image depicts a single plane view, and the right 3D view reconstructed from a z-stack of 1.4 µm with a z-step of 150 nm (n = 3 animals were considered as biological replicates). Scale bar: 5 µm. f Cathepsin L signal in the basal region of proximal tubules was lower in mice treated with HCQ (20 mg/kg); the total area of Cathepsin L positive pixels was quantified in the basal half of the cells defined using an apical to basal line as depicted. Groups were compared using a two-tailed student t test. Data depicted are mean ± SEM (n = 3 animals were considered as biological replicates; objects were identified on 2 to 3 images per replicate and all data extracted from this analysis merged to estimate a mean value per group; p = 0.0039). Source data are provided as a source data file.

Deacidification of lysosomes in S1 proximal tubules

Next, we investigated the effects of alkalinization on lysosomal dynamics in S1. To overcome the protein uptake defect caused by HCQ, we injected an increased amount of labeled lysozyme and observed that it was trafficked through EEs and into the LE/LAV layer in PTs (Fig. 5c). Although early trafficking kinetics appeared similar to controls, we cannot exclude that these may have been slightly altered by HCQ exposure. However, as expected, endosomal pHrodo/Atto-647N signal ratios were much smaller compared to untreated animals, denoting an increase in pH (Fig. 5c). Moreover, post HCQ we could no longer discern discrete, acidified vesicles surrounding LEs/LAVs (Fig. 5c). Time lapse imaging with Lysozyme-Atto-565 confirmed loss of trafficking vesicles moving in and out of the main ELS region (Fig. 5d). Meanwhile, antibody staining in fixed tissue revealed extensive contact between Rab7 and cathepsin positive vesicles, resulting in a corresponding disappearance of the latter outside the apical zone (Fig. 5e–g).

To investigate dynamic interactions between LEs and lysosomes in more detail, we performed live confocal imaging in human PT-derived (HK-2) cells transfected with a Rab7-GFP encoded construct to label the former. By co-imaging with pH- (Lysotracker) or cathepsin- (SiR Lysosome) dependent probes, we could identify highly mobile acidified vesicles with protein degrading activity, which frequently connected with Rab7 positive structures before detaching again (Fig. 6a, b). 3D reconstruction of signals in fixed cells confirmed evidence of extensive membrane contact between these structures, with apparent partial invagination of lysosomes into LEs (Fig. 6c, d). Next, we treated cells with HCQ (20 µM) to determine the effects of deacidification. After 30 minutes, Rab7 and SiR Lysosome positive vesicles were unable to divide (Fig. 6e, h, i), leading to dramatic and widespread contact between structures by 60 minutes (Fig. 6e–g).

Fig. 6 — a Late endosomes (LEs) were labeled in HK-2 cells by transfection with Rab7-GFP, and acidified lysosomes with Lysotracker. The image sequence depicts a single LE connecting with a lysosome, before detaching and subsequently reconnecting. Scale bar: 1 µm. This experiment was performed 3 times leading to the same observation. b Lysosomes containing active cathepsins were labeled with SiR-Lysosome. Interactions between these structures and LEs were also detected. This experiment was performed 3 times leading to the same observation. Scale bar: 1 µm. c, d 3D reconstruction of Rab7-GFP and SiR-Lysosome signals in fixed cells revealed a large surface area of contact between connecting LEs and lysosomes. Scale bar: 1 µm. d Three different views are provided of a single lysosome and LE in contact. Volume rendering shows regions of overlap between signals, with colocalized voxels highlighted in blue (Pearson’s r = 0.74). In this example, volume-based analysis confirmed that 69.52% of the SiR-positive volume overlapped with Rab7, and 51.02% of the Rab7-positive volume overlapped with SiR. Scale bar: 0.2 µm. e, f Treatment of cells with hydroxychloroquine (HCQ, 20 µM) prevent detachment of lysosomes from LEs. This experiment was performed 3 times leading to the same observation. Scale bar: 0.8 µm. g 3D view after 60 min of HCQ exposure, showing extensive contact between LEs and lysosomes. Scale bar: 1 µm. h Time-lapse imaging showing dynamic contacts between Rab7-GFP late endosomes and SiR-labeled lysosomes. Representative frames illustrate lysosomes either maintaining contact with endosomes (magenta) or moving without contact (violet). Grey outline denotes the lysosomal track. Scale bar: 0.5 µm. i Raster plots showing the frequency and duration of contacts between lysosomes (SiR) and LEs (Rab7-GFP) in control cells (top) and after treatment with HCQ (30 min; bottom). Color coding indicates contact duration (s). Source data are provided as a source data file.

In summary, these observations imply the existence of extensive, transient contacts between LEs and lysosomes in S1 PT cells, which might facilitate efficient transfer and subsequent degradation of reclaimed plasma proteins, and are dependent on the existence of substantial pH gradients between the organelles.

Lysosomal function in S2 proximal tubules

Because our custom-designed protein/peptide probes were reabsorbed mainly in the early PT (S1), we deployed a different strategy to visualize lysosomes in downstream segments. The commercially available pH-dependent dye Lysotracker contains a fluorescein group, which we reasoned should facilitate uptake into PT cells via basal organic ion transporters. In agreement with this notion, following intravenous injection in mice, Lysotracker entered PTs from the basal side (i.e. directly from the bloodstream) and rapidly accumulated in discrete vesicles (Fig. 7a). By co-injecting with labeled lysozyme, we found that lysotracker signal in S1 cells localized to small vesicles surrounding LEs/LAVs (Fig. 7a), thus corroborating readouts from our custom-designed probes. In contrast, lysotracker positive vesicles in S2 appeared larger and more distributed across cells (Fig. 7b). Moreover, time lapse imaging revealed them to be highly dynamic, trafficking deep into the basal region of S2 cells (Supplementary Movie 1).

Fig. 7 — a Lysozyme-565 was injected into mice to label S1 proximal tubular segments and lysotracker injected to label lysosomes. Overlay of signals identified small acidified lysosomes in S1 (arrows), lying just below the main band of protein-containing endosomes. Scale bar: 20 µm. This experiment was repeated on 3 individual animals. b Lysotracker positive lysosomes in S2 appeared larger than in S1 and were more dispersed across cells (n = 4 animals were considered as biological replicates). Scale bar: 20 µm. c Mouse kidney tissue stained with BODIPY 493/503 revealed basal lipid containing structures (arrows). Scale bars: 10 µm. This experiment was performed on 4 individual kidney slices stained for BODIPY. d Transmission electron microscopy image of an S2 mouse tubule showing a lipid droplet (arrow) in the mitochondrial region of the cell. Scale bars: 1 µm. This experiment was performed on 2 individual control mice. e Immunolocalization of the lysosomal hydrolases cathepsin L (CatL) and lysosomal lipase A (LAL) along proximal tubules. Scale bar: 200 µm. Zoomed in view showing that segments with high LAL signal were positive for the S2 marker OAT1. Scale bar: 20 µm. A 3D view of LAL signal in S2 reconstructed from a z-stack of 900 nm with a z-step of 130 nm showing distribution of signal across the cell. Scale bar: 5 µm. The histogram depicts mean ± SEM in regions of proximal tubule, ordered from highest cathepsin signal to lowest (n = 3 animals were considered as biological replicates; 20 tubules were identified on 1 to 2 images per replicate). f Incubation of mouse kidney tissue for 1 h with the lysosomal lipase inhibitor Lalistat-2 (1 mM) increased BODIPY 493/503 signal in vesicles (arrows) in the sub-apical region of proximal tubules, denoting an increase in lipids. Scale bars: 20 µm. Groups were compared using a two-tailed student t test. Data depicted are mean ± SEM (n = 3 slices extracted from 2 animals were considered as biological replicates as they were treated in individual experiments; 14 to 15 regions of interest were drawn on 1 to 2 image per replicate and all data extracted from this analysis merged to estimate a mean value per group; p < 0.0001). g Treatment of HK-2 cells with Lalistat-2 (100 µM) for 24 h caused an accumulation of lipids (BODIPY 493/503) in the region of lysosomes (labeled with SiR-Lysosome). Groups were compared using a two-sided unpaired t-test, t⁵⁸ = 27.66, P < 0.0001. Data are mean ± SEM (n = 3 biological replicates per condition, 10 ROIs analyzed per image). Scale bars: 10 µm (zoomed-out) and 5 µm (zoomed-in). h 3D reconstruction of lipid droplets (BODIPY 493/503), lysosomes (SiR-Lysosome), and mitochondria (MitoTracker Orange CM-H2TMRos) in HK-2 cells. Scale bars: 2 µm. i, j Time-lapse sequence and trajectory tracing depicting a lysosome trafficking to a cluster of lipid droplets and making contact, before detaching again. Scale bars: 1 µm. k 3D reconstruction of an HK-2 cell showing a LAL positive lysosome wrapping around a lipid droplet (BODIPY 493/503). Scale bars: 2 µm. l Transmission electron microscopy images of mouse PTs showing lysosomes apparently engulfing lipid droplets and a progressive digestion of the latter to multi-lamellar bodies (MLBs), as depicted by arrows (images 1-4). Secretion of MLBs was denoted by their appearance in tubular lumens (arrows in images 5 and 6). Scale bars: 1 µm. These examples are extracted from 2 individual control animals. Source data are provided as a source data file.

These observations prompted us to consider that lysosomes in S2 might be adapted to metabolic functions other than protein processing, which could involve interactions with organelles spatially displaced from the apical compartment. Of note, lipids are thought to be a major fuel for PT cells^16,17, but unlike proteins their spatial transport profile has not yet been well defined. However, recent single cell sequencing studies have uncovered a high expression of genes related to lipid transport and metabolism in S2 cells⁸. To evaluate the distribution of lipid droplets in PT cells, we incubated kidney tissue with BODIPY 493 and identified that they were predominantly located in the basal region (Fig. 7c). Co-staining with Sudan Black and for the organic anion transporter OAT1 - which is both an S2 marker and lipid transporter¹⁸ – confirmed the presence of lipid droplets in S2s (Supplementary Fig. 6). Furthermore, electron microscopy revealed a close association to mitochondria (Fig. 7d), suggesting a possible association with mitochondrial metabolism (however, spatial proximity alone does not prove a direct functional connection).

Next, we performed antibody staining for lysosomal acid lipase – which degrades lipids at acidic pH¹⁹ - and observed a striking inverse relationship in abundance to cathepsin L along the PT. Whereas the latter showed the highest abundance in S1, the former was much more prominent in OAT1 positive S2 cells (Fig. 7e). Furthermore, incubating mouse kidney slices with the lysosomal lipase inhibitor Lalistat-2 produced an increase in lipid vesicles in PTs (Fig. 7f), indicating that the enzyme is active. Meanwhile, co-imaging with a lysosomal marker in HK-2 cells revealed that Lalistat-2 induced lipid accumulation occurred specifically in the vicinity of lysosomes (Fig. 7g).

To investigate the possibility of dynamic interplay between lysosomes and lipid droplets in PT-derived cells, we incubated HK-2 cells with oleic acid to increase lipid uptake and usage, which triggered the formation of lipid droplets in mitochondrial regions (Fig. 7h). Using Structured Illumination Microscopy (SIM), a form of super resolution imaging with a high acquisition rate, we tracked lysosomes and – as in vivo - found them to be highly motile. Crucially, we observed that they traffick specifically towards lipid droplets, before making contact with them (Fig. 7i, j).

We subsequently detected several different types of physical interactions between lysosomes and lipid droplets. For example, following the initiation of contact, lysosomes were observed to crawl around the surface of lipid droplets, and fuse with other lysosomes to encircle them (Supplementary Movie 2). 3D imaging of fixed cells provided further evidence of acid lipase containing lysosomes wrapping around lipid droplets (Fig. 7k). Moreover, during live imaging lysosomes appeared to both push and pull lipid droplets through the cytosol, thus mobilizing these stationary structures (Supplementary Movies 2 and 3).

Finally, analysis of EM images from healthy mice identified examples of lysosomes apparently surrounding and engulfing lipid droplets in S2 cells, providing evidence that these interactions occur in vivo. Moreover, we detected various stages of lipid droplet degradation to whirl-like multilamellar bodies (MLBs), which were also observed in tubular lumens (Fig. 7l), implying that they are eventually secreted.

Taken together, these observations suggest that lysosomal trafficking across S2 cells provides a dynamic mechanism to target mitochondria-related lipid droplets in the basal region, facilitate their breakdown and extrude their contents into the urine.

De-acidification of lysosomes in S2 proximal tubules

Next, we wanted to investigate the consequences of acutely de-acidifying lysosomes in S2 cells. We therefore injected mice with HCQ, which inhibited the movement of lysotracker positive vesicles into the basal region of PT cells (Fig. 8a). HCQ also completely abolished trafficking of lysosomes to lipid droplets in HK-2 cells (Supplementary Fig. 7). Conversely, HCQ triggered an acute release of lysotracker positive vesicles into tubular lumens (Fig. 8b and supplementary Movie 4), consistent with the known capacity of HCQ to stimulate lysosomal exocytosis²⁰. Moreover, we stained fixed kidney tissue with Toluidine blue to label acidic structures²¹, and observed that HCQ stimulated their apical release from PT cells (Fig. 8c). Meanwhile, electron microscopy revealed enlargement and secretion of MLBs in response to HCQ (Fig. 8d), and staining with BODIPY 493/503 confirmed a dramatic accumulation of lipid in tubular lumens (Fig. 8e).

Fig. 8 — a Lysotracker was injected intravenously to label lysosomes in S2 cells. Temporal color-coding depicts movement of lysosomes over time (arrow). Basal trafficking of lysosomes was reduced after 25 minutes post injection of HCQ (20 mg/kg). Scale bars: 10 µm (zoomed-out) and 5 µm (zoomed-in) (n = 3 animals were considered as biological replicates). b HCQ treatment triggered secretion of lysotracker positive lysosomes into the tubular lumen; images depict pre and 30 min post HCQ injection. Scale bars: 30 µm. Quantification of luminal lysotracker signals was performed at these 2 time points. Groups were compared using a two-tailed paired student t test. Data depicted are mean ± SEM (n = 3 animals were considered as biological replicates with 34 to 38 regions of interest (ROIs) considered per replicate; ROIs from all replicates were merged to estimate a mean value for each group; p < 0.0001). c Confocal imaging of Toluidine blue fluorescence signal revealed apparent secretion of acidified vesicles (arrows) into the tubular lumen in HCQ treated mice. Scale bars: 5 µm. d Transmission electron microscopy revealed an enlargement of multi-lamellar bodies (MLBs) in S2 cells following treatment with HCQ (20 mg/kg for 3 h). Groups were compared using a two-tailed student t test. Data depicted are mean ± SEM (n = 2 animals in control group and n = 3 animals in HCQ group were considered as biological replicates; 36 to 196 MLBs per animal were considered for the quantification; data from all replicates were merged to estimate a mean value for each group; p < 0.0001). e Mouse kidney tissue was incubated for 1 h with hydroxychloroquine (HCQ, 0.7 mg/ml) and stained with BODIPY 493/503 to label lipids. Lipid containing vesicles (arrows) were observed in the sub-apical region of proximal tubules. Luminal accumulation of lipid also occurred in response to HCQ. Scale bars: 20 µm (zoomed-out) and 5 µm (zoomed-in). Quantification of luminal BODIPY signal intensity was performed. Groups were compared using a two-tailed student t test. Data depicted are mean ± SEM (n = 3 slices extracted from 2 animals were considered as biological replicates as they were treated in individual experiments; 6 to 13 regions of interest were drawn on 1 to 2 image per replicate and all data extracted from this analysis merged to estimate a mean value per group; p < 0.0001). Source data are provided as a source data file.

Finally, we performed lipidomic analysis on urine specimens and observed that HCQ treatment increased excretion of multiple lipid species, including membrane lipids originating from organelles (Supplementary Fig. 8). In contrast, triglyceride excretion was not elevated, arguing against a direct secretion of lipid droplets.

In summary, these findings suggest that alkalinization drastically alters lysosomal dynamics in S2 cells, re-routing them from basal trafficking to apical extrusion, and thus causing release of intracellular lipids into the lumen.

Discussion

In this study, we deployed targeted fluorescent sensors and high-resolution microscopy to perform a detailed assessment of the landscape of lysosomal dynamics and organellar interactions in the kidney PT in vivo. We have uncovered evidence of striking axial differences in lysosomal characteristics and behavior, and that lysosomes play integral roles in both protein and lipid processing, in a segment specific manner. These findings indicate the existence of considerable metabolic heterogeneity along the PT, and shed new light on how lysosomal function is adapted to specific tasks within this highly specialized epithelium.

Starting from the development of custom-designed delivery vectors, we mapped pH changes in the PT ELS in the protein-reabsorbing region (S1), using a combination of morphological, kinetic and molecular criteria to assign identities to structures. Due to the proximity of ELS vesicles and their extensive interactions, it is possible that there may have been some mixing of populations but based on detailed assessments of the dynamics of protein progression we expect to have a majority of designated structures at each time point used for recordings. We were able to identify significant pH drops in EEs and lysosomes; surprisingly, we did not observe a substantial decrease in LAVs, which should prompt future exploration of other ion gradients that might be more relevant for their specialized sorting/trafficking roles. Acutely raising vesicular pH with HCQ severely disrupted megalin recycling and protein uptake, which might reflect the pH dependence of conformational changes required for receptor-ligand dissociation²².

We identified that lysosomes in S1 traffic in and out of the endosomal layer and frequently connect with LEs/LAVs, providing a potential mechanism for efficient cargo transfer in specialized protein-processing cells. Acidification of lysosomes optimizes cathepsin activity²³; accordingly, we observed onset of protein degradation in these organelles. Fusion of endosomes and lysosomes temporarily equalizes pH, so we speculate that detachment of the latter then allows re-acidification and catabolism of cargo. Alternatively, protein degradation might occur in fused endo-lysosomes, but in this scenario we would expect to see onset in larger vesicles, which was not the case. Importantly, de-acidification triggered widespread contact between LEs and lysosomes and loss of trafficking, suggesting that maintenance of normal pH gradients is integral for their dynamic interplay. These observations provide a rationale for the high expression of V-ATPase subunits and other proteins involved in vesicular acidification in PTs, and may help explain their localization to specific ELS structures^24–27.

In contrast to S1, in the downstream (S2) part of the PT we observed that lysosomes are characterized by a high lipase content and are extremely motile, trafficking across cells to interact with lipid droplets associated with mitochondria. However, it is important to stress that lipid droplets can arise from different structures and proximity to mitochondria alone does not prove their origin. Moreover, lipid droplet production might be a protective response in the setting of mitochondrial dysfunction²⁸. Because lipid droplets appeared static, lysosomal trafficking is essential to initiate this contact. Moreover, lysosomes were seen to push and pull lipid droplets, thus mobilizing them within the cytosol. Using super-resolution microscopy, we identified that lysosomes patrol along the membrane of lipid droplets and can fuse together to engulf these larger structures. Finally, S2 cells contained numerous MLBs, which have been shown in previous studies to consist of lipids, lysosomal proteins and autophagic markers^29,30, providing further evidence that they represent remnants of degraded lipid droplets, which can be extruded from cells.

These findings imply that directed movement of lysosomes provides a metabolic bridge between apical and basal compartments of polarized PT cells, and a dynamic pathway for excreting lipids. It therefore follows that impairment of lysosomal trafficking and/or function might contribute to intracellular lipid accumulation in disease states¹⁷. In line with this concept, it has been reported that poisoning with the microtubule inhibitor colchicine produces a massive basal accumulation of lipid droplets in human PTs³¹. Moreover, enlargement of renal parenchymal lipid droplets and decreased urinary lipid excretion was recently described in mice with a genetic defect in the PT ELS^32,33, while we have shown that metabolic acidosis induces severe endocytic defects and lipid accumulation in PTs³⁰.

Importantly, we observed that acute de-acidification triggered the luminal secretion of lysosomes, MLBs and lipid into tubular lumens, resulting in increased urinary lipid excretion. Interestingly, recent studies have suggested that fatty acid loading alkalinizes lysosomes³⁴, and that urinary excretion of MLBs increases in response to high-fat feeding³⁵. Based on this, we speculate that changes in lysosomal pH might mediate apical extrusion of lipids from PTs to limit intracellular accumulation and subsequent lipotoxicity. Intriguingly, several clinical studies have reported lower blood lipid levels in patients treated with HCQ³⁶.

Taken together, our study suggests that lysosomal activity is adapted to specific tasks within different specialized segments of the PT (Fig. 9), which are metabolically much more diverse than previously appreciated. Further work will be required to decipher the factors that shape axial patterns of lysosomal behavior, but gradients in nutrient microenvironment (e.g., luminal protein concentration) and metabolite uptake (e.g., lipid transport) are probably important variables. It will therefore be interesting to ascertain to what extent PT lysosomes display plasticity in the context of homeostatic disturbances; for example, in the setting of glomerular proteinuria or dyslipidemias.

Our study has several potential limitations. First, 2D tracking of vesicles may be confounded by movements in and out of the focal plane. Second, endosomal identities were assigned by cross-reference to fixed tissue, rather than direct labeling. Third, findings in mice might not translate to other species. Fourth, it was not possible to calibrate the fluorescent signals in vivo, meaning that we could only derive qualitative data regarding pH changes. Fifth, due to their deep location is not possible to visualize S3 PT segments with intravital microscopy. Finally, we only focused on the acute effects of disrupting lysosomal function – the longer-term consequences remain to be explored.

In conclusion, we show that lysosomes within the PT are highly dynamic and interact with other organelles in a segment-specific manner. These findings shed new light on how lysosomal function is integrated with tubular metabolism and how lysosomal defects might potentially lead to disease phenotypes.

Methods

The research in this study complies with all relevant ethical regulations, and the protocols for animal experiments were approved by The Zurich Cantonal Veterinary Office (ZH076/2021). All images are presented as single plane or 3D reconstruction of single planes.

Labelling of proteins

Recombinant human lysozyme (#L1667; Sigma-Aldrich) was labelled with Atto-647N NHS ester (#AD 647N-31; Atto-Tec) or pHrodo iFL Red STP ester (# P36010 Invitrogen) to obtain two populations of dye-protein probes. Labelled proteins were purified using benchtop PD midiTrap G-25 Sephadex chromatography columns (#28–9180–08; GE Healthcare) and concentrated with 3-kD Amicon Ultra-4 centrifugal filters (#Z740186; Sigma-Aldrich). They were then analyzed and if necessary re-purified by LCMS-2020 Shimadzu (Waters XBridge column C18 3.5 μm 4.6 × 250 mm 130 Å, flow 1.0 ml/min, temperature 65 C, solvent A: water 0.1% v/v TFA, solvent B: acetonitrile 0.1% v/v TFA, mobile phase gradients: 5% 10 min, 5–30% 12 min, 30-70% 48 min, 70-95% 50 min, 95% 55 min, 95-5% 56 min, 5% 65 min) and freeze-dried. Concentration of dissolved probes was measured by UV/Vis absorption spectroscopy (Libra S70; Biochrom) and labelling was confirmed by a nano ESI-MS mass spectrometer (Synapt G2 Si, Waters). Lysozyme-565 and highly labeled β-lactoglobulin-532 were generated as previously reported¹¹.

Labelling of peptide

The custom peptide was synthetized and characterized by Biomatik Corporation. Exact peptide mass and purity ( > 90%) were confirmed by LCMS (Inertsil ODS-SP column 4.6 × 250 mm, flow 1,0 ml/min, mobile phase A: acetonitrile 0.1% v/v TFA, B water: 0.1% v/v TFA). The peptide was dissolved in PBS adjusted to pH 7.0 and labelled with pHrodo Red Maleimide (#P35371; Invitrogen). Labelling conditions: 400 µg of the peptide in 250 µl of the buffer, 10 ul of 10 μg/ul pHrodo Red Maleimide, 1 h in the ultrasonic bath. In the next step the second dye Atto-647N NHS ester (#AD 647N-31; Atto-Tec) was added (15 μl of 10 μg/μl). Before adding the second dye, the pH of the buffer was checked and adapted to 7.0 to label the primary amine of the N-terminal.

The dual labelled peptide was then purified by high pressure liquid chromatography and immediately characterized by in-line mass spectrometry in flowing conditions: LCMS-2020 Shimadzu, Waters XBridge column C18 3.5 μm 4.6x 250 mm 130 Å, flow 1.0 ml/min, temperature 65 C and mobile phase gradients were adapted: solvent A: water 0.1% v/v TFA, solvent B: acetonitrile 0.1% v/v TFA, mobile phase gradients: 5% 5 min, 5–45% 10 min, 45–55% 55 min, 55–95% 65 min, 95% 70 min, 95-5% 71 min, 5% 100 min.

Exact peptide mass and purity ( > 95%) were confirmed by LCMS (Waters XBridge column C18 3.5μm 4.6x 250 mm 130 Å, flow 1,0 ml/min, mobile phase A: water 0.1% v/v TFA, B: acetonitrile 0.1% v/v TFA, gradient: 5% 0–2 min, 5–95% 2–17 min, 95% 17–20 min, 5% 20–22 min, 60 C). Detections were performed using a photodiode array detector (SPD‑M40; polarity +). PDA spectra were acquired using D2 & W lamps over 190–800 nm (spectrum resolution 512; slit width 8 nm; cell temperature 40 °C) with an acquisition sampling rate of 3.125 Hz (320 ms; time constant 0.640 s). MS data were acquired in positive‑ion ESI mode (m/z 200–2000; scan speed 1500 u s⁻¹; event time 0.15 s; tuning voltage 4.5 kV) and raw ESI-MS spectra were deconvoluted using MagTran. Labelled peptide concentrations were measured using UV/Vis absorption spectroscopy (Libra S70; Biochrom).

The labelling position was confirmed by ESI-MS/MS analysis. Samples were first desalted using a C18 Zip Tip (Millipore, USA) and analyzed in MeOH:2-PrOH:0.2% FA (30:20:50). The solution was infused through a fused silica capillary (ID75 um) at a flow rate of 1 uLmin⁻¹ and sprayed through a Pico Tips (ID30um), obtained from New Objective (Woburn, MA). Nano ESI-MS analysis of the samples was performed on a Synapt G2_Si mass spectrometer and the data were recorded with the MassLynx 4.2 Software (both Waters, UK). Mass spectra were acquired in the positive-ion mode by scanning an m/z range from 50 to 3000 da with a scan duration of 1 s and an interscan delay of 0.1 s. The spray voltage was set to 3 kV, the cone voltage was set to 35 V and source temperature 100 °C. The species of interest were detected as multiply charged ions and subjected for structural elucidation by MS/MS. The peptide was also digested with trypsin (30 min) and both N-terminal and C terminal fragments subjected for structural elucidation by MS/MS.

Characterization of labelled proteins and peptide in solution

Fluorescence intensities of dye-protein conjugates and labelled peptides were measured on a Cytation 5 Microplate reader (Biotec) and using 96-well microplates (Greiner Bio-One UV-Star™ 96-Well Microplates). The measurements were performed using an excitation of 560/10, emission: 585/20 and excitation: 645/10, emission: 670/10.

Characterization of labelled proteins and peptide in cells

Proximal tubule derived OK cells (gift from Prof Devuyst, University of Zurich, available from atcc.org: CRL-1840) were transfected with 5 μL of BacMam 2.0 reagent (#C10596, CellLight Lysosomes-GFP, BacMam 2.0, Invitrogen) per 100 μl buffer. Cells were then incubated with labeled lysozyme (10 µg/ml) or peptide (0,2 nmol/100 μl) for 30 min. Imaging was performed with a Leica DMI6000B, Model SP8 inverted confocal microscope. The following wavelengths were used: GFP (ex. 488 nm, em. 500–540 nm), pHrodo Red (ex. 560 nm, em. 570–620 nm); Atto 647 N (ex. 647 nm, em 660–710 nm). Images were acquired with a pixel size of 0.04 µm × 0.04 µm (XY).

To calibrate pH within cells, the Calibration Buffer Kit (#P35379 Invitrogen™) was used to clamp intracellular pH with extracellular buffers at pH 4.5, 5.5, 6.5 and 7.5 containing Valinomycin 10 μM, Nigericin 10 μM and Bafilomycin A1 200 nM (#BML-CM110-0100 Enzo). HCQ (20 mg/ml, Sigma-Aldrich, H0915) was incubated on OK cells to disrupt intravesicular pH 20 min post-probe mix. Fluorescence ratio from pHrodo and Atto647 was then followed at different time points. Each condition was measured in a different well. Ilastik v1.3.3post3 was trained to detect vesicles using the pixel classification protocol on a minimum of 2 images per condition in an experiment. Objects smaller than 40 pixels² on 41.7 µm x 41.7 µm images at 1024 pixels x 1024 pixels (showing one to three cells) were considered as artifacts and excluded. Subsequent fluorescence intensity given by each dye was quantified as mean grey value in these objects using Fiji. Outliers were then removed based on ROUT 1% on GraphPad prism and fluorescence ratios converted into pH values using the calibration curve equation. When necessary, aberrant values (imaging artifacts) with a pH <3.5 or pH > 7.8 were manually removed.

All live cell experiments were conducted at 37 °C, 5% CO₂.

Imaging of late endosomes and lysosomes in cells

HK-2 cells (CRL-2190, ATCC) were transfected with 1.5 µg of mScarlet-Rab7 plasmid (Addgene, #169068) using Lipofectamine 3000 (Thermo Fisher Scientific) according to the manufacturer’s instructions. For lysosome labeling, cells were incubated with SiR-Lysosome dye (SpiroChrome, #SC012) at a final concentration of 100 nM (overnight) or 1 µM (2 h) in complete medium at 37 °C. For acidic vesicle detection, cells were incubated with LysoTracker™ Deep Red (Thermo Fisher Scientific, #L12492) at 50 nM for 30 min at 37 °C prior to imaging. Imaging was performed using a Zeiss LSM 980 confocal microscope equipped with Airyscan detection. Live time-lapse imaging was acquired with a pixel size of 0.049 µm × 0.049 µm (XY). Fixed-cell 3D imaging of control samples was performed over a total Z range of 12.35 µm (Z step ≈ 0.13 µm; voxel size = 0.035 µm × 0.035 µm × 0.130 µm). For HCQ-treated samples, 3D stacks were acquired over a total Z range of 7.48 µm (Z step ≈ 0.17 µm; voxel size = 0.049 µm × 0.049 µm × 0.170 µm). Image processing and visualization were carried out using ZEN 3.1 (Zeiss) and Imaris 10.2.0 (Oxford Instruments).

Three-dimensional visualization and colocalization of Rab7-positive compartments and SiR-labeled lysosomes were performed using Zen 3.1 (Zeiss) and Imaris (Bitplane, Oxford Instruments). Volume renderings were generated in Zen with the Surface tool, and colocalization analyses were performed in Imaris using the Coloc module. Pearson’s correlation coefficients were calculated from colocalization channels, and the proportion of overlapping volumes was quantified relative to the total volume of each compartment.

Tracking analysis was performed using Imaris (Bitplane, Oxford Instruments). SiR-labeled vesicles were identified as “spots” and tracked over time using manual tracking. Rab7-positive vesicles were segmented using threshold-based selection. Contacts between lysosomes and Rab7-positive vesicles were defined when the centroid of a SiR-labeled vesicle was within 350 nm of a Rab7 object.

Imaging of lipid droplets, mitochondria, and lysosomes in cells

HK-2 cells were incubated with 150 µM oleic acid (OA; Merck, #O1383) for 24 hours. For organelle labeling, lysosomes were stained by incubatng cells with SiR-Lysosome dye (1 μM) for 2 h. Mitochondria were labeled using MitoTracker Orange CM-H2TMRos (ThermoFisher, #M7511) at a concentration of 100 nM for 30 minutes. After staining, cells were fixed with 4% paraformaldehyde for 15 min at room temperature. Lipid droplets were stained with BODIPY 493/503 (ThermoFisher, #D3922) at a concentration of 0.5 µg/mL for 15 min. To label lysosomal acid lipase cells were incubated with a primary antibody (mouse anti-LAL sc-58374 Santa Cruz Biotech) and then a fluorophore-conjugated secondary antibody (Goat-anti-mouse Alexa Fluor 568, 1:500, ThermoFisher, A11004). 3D imaging was performed using a Zeiss LSM 980 microscope with Airyscan with a voxel size of 0.043 µm × 0.043 µm × 0.150 µm (XYZ).

Live Imaging of Lipid Droplet–Lysosome Interactions in HK-2 Cells

HK-2 cells were incubated with 150 µM oleic acid for 24 h. For lysosomal labeling, cells were incubated with SiR-Lysosome dye (1:1000 dilution) for 2 hours. Lipid droplets were stained with LipidSpot 488 (#70065 Biotium) (1:1000 dilution) for 1 hour. Live-cell imaging was performed using a ZEISS Elyra 7 Lattice SIM² microscope, equipped with a Plan-Apochromat 63×/1.4 oil immersion objective (DIC M27). High-speed super-resolution structured illumination microscopy (Lattice SIM mode) was used to capture dynamic interactions. Images were acquired using a camera model: pco.edge 4.2 sCMOS. Images were acquired with a pixel size of 0.03 µm × 0.03 µm (XY). Image processing was performed using ZEISS ZEN 3.0 SR software, applying SIM² Processing (2D, Leap mode), ZEISS ZEN 3.10 deconvolution. Lysosome movement was analyzed using the manual tracking tool and time-color coding in Imaris software.

For Supplementary Movies 2 and 3, imaging was conducted using 2D time-lapse acquisition with an exposure time of 50 ms. Lipid droplets were stained with BODIPY 558/568 and lysosomes were labeled with SiR-Lysosome dye. Time-lapse were acquired with a pixel size of 0.03 x µm × 0.03x µm (XY). Image processing was performed using ZEISS ZEN 3.0 SR software, applying SIM² Processing (2D + , Burst Mode, SIM² low contrast, detrend).

Imaging of cells treated with Lalistat-2

HK-2 cells were treated with 100 µM Lalistat-2 (#053-5MG, Sigma) for 24 h. Low-density lipoprotein (LDL, 100 µg/mL; Sigma-Aldrich #SAE0053) was then added while maintaining Lalistat-2 treatment, resulting in a total incubation time of 48 hours. Control cells were incubated with LDL but without Lalistat-2. For lysosomal labeling, cells were incubated with SiR-Lysosome dye (1:1000 dilution) for 2 h at 37 °C. Lipids were stained with BODIPY 493/503 (0.5 µg/mL) for 15 min before imaging. Images were acquired with a pixel size of 0.035 µm × 0.035 µm (XY).

Animal experiments

All in vivo experiments were performed on 8–15-week-old male C57Bl/6JRj mice (Janvier, Le Genest, France). Mice were housed in the University of Zurich animal facility, at ambient temperature on a 12 h life cycle, with free access to food (Kliba Nafag formula 3436) and water. They were anaesthetized with 2.5% Isoflurane (Attane, Provet AG, Switzerland) in 100 ml/min Oxygen. At the end of experiments animals were euthanized by exsanguination under anaesthesia.

A mixture of the lab designed fluorescent reporters Lysozyme-647N and Lysozyme-pHrodo Red was intravenously injected following a 1:6 ratio. In control experiments, 5 µg of Lysozyme-647N and 30 µg of Lysozyme-pHrodo Red were co-injected; in HCQ experiments, 7 µg of Lysozyme-647N and 42 µg of Lysozyme-pHrodo Red were used. For peptide experiments, 3 nmol of the dual labelled custom-made peptide described above was intravenously injected.

HCQ (Sigma-Aldrich, H0915) 20 mg/kg was injected intravenously to disrupt pH gradients in vivo, 30 min prior to injections of fluorescent proteins or tissue perfusion fixation. When evaluating lysosome trafficking, 25 µg of Lysozyme-565 and 18 nmol of Lysotracker Deep red (Thermo Fisher Scientific, L12492) were injected intravenously. Temporal-color coded images were generated using Fiji on an image sequence (with a frame rate of 10 s) over a period of 6.5 min. To visualize protein degradation, 20 µg of highly labeled β-lactoglobulin-532¹¹ was co-injected with 25 µg Lysozyme-647.

Urine samples were analyzed for their protein content by loading 10 µl of urine on a 12% acrylamide gel and staining with Imperial protein stain (Thermo Scientific, 24615). Total urinary proteins were quantified using Fiji and data normalized by urinary creatinine evaluated using the creatinine colorimetric assay (Cayman chemical, 500701). Statistical analyses were performed using a t student test on n = 5 animals.

Intravital imaging

The left kidney was externalized for imaging as described in previous protocols³⁷. Briefly, the internal jugular vein was cannulated to allow intravenous injections of molecules. Imaging was performed on a custom-built multiphoton microscope operating in an inverted mode, using ScanImage software. A broadband tunable laser (InSight DS Dual, Spectraphysics, Santa Clara, USA) was used as excitation source. A × 25 water immersion objective with a numerical aperture of 1.05 and a working distance of 2 mm was used for intravital imaging (Olympus, Tokyo, Japan). In most experiments, 850 nm was used as the excitation wavelength and emissions acquired simultaneously in two different channels at 555–655 nm and 625–775 nm. For co-imaging β-lactoglobulin-532 and Lysozyme-647, 820 nm was used as the excitation wavelength.

Image processing was performed using Fiji. For estimation of pH gradients, the quantification of fluorescence intensities (in mean grey values) from each fluorescent reporter was performed on single frames at specific time points, in structures identified based on appearance, localization along the apical-basal axis and time. A minimum of hand-drawn 20 regions of interest were selected for each animal and statistical analyses performed using a One-Way ANOVA test on n = 3 animals.

To determine whether ELS de-acidification has an effect on lysosomal secretion in vivo, luminal regions of interest were drawn on S2 PTs bearing lysotracker before and 30 min after HCQ injection. The mean gray value from lysotracker was measured using Fiji and data analyzed with a paired student t test on n = 3.

Images from intravital microscopy were acquired with a pixel size of 228.45 nm or 114.225 nm.

Immunostaining in fixed kidney tissue

Kidney tissue was perfusion fixed with 3% paraformaldehyde in a phosphate buffered solution. When tracking protein endocytosis rate on fixed tissues, 10 µg of Lysozyme-647 were injected intravenously and kidneys fixed 1, 3, 6 or 10 min post-injection. Stainings were performed on 5 µm thick cryosections incubated overnight with the following antibodies: mouse anti-Megalin (H10) (1:100, Santa cruz Biotech, sc-515772), rabbit anti-Rab5 (1:50, Cell signaling, #3547), rabbit anti-Rab7 (1:50, Abcam, ab137029), goat anti-Cathepsin L (1:100, R&D Systems, AF1515), rabbit anti-OAT1 (1:100, Alpha diagnostic international, OAT11-A). LAL staining was performed in paraffin embedded tissues, following a citrate pH 6 buffer antigen retrieval step at high temperature and pressure, using mouse anti-LAL (1:100, Santa cruz Biotech, sc-58374). The following secondary conjugated antibodies were used (all at 1:500): Alexa Fluor 488 donkey anti-rabbit (Jackson ImmunoResearch, 711-546-152), Alexa Fluor 568 goat anti-mouse (ThermoFisher Scientific, A-11004), Alexa Fluor 647 donkey anti-mouse (ThermoFisher Scientific, A-31571), Alexa Fluor 647 donkey anti-goat (Jackson ImmunoResearch, 705-606-147), Alexa Fluor 555 donkey anti-goat (Abcam, ab150130). Brush-border actin filaments were stained with ActinRed 555 ReadyProbes reagent (Invitrogen, R37112) according to manufacturer’s instructions.

Images were acquired using a Leica SP8 inverse STED 3X confocal microscope with a pixel size of 75 nm and the widefield slidescanner Zeiss Axio Scan.Z1 with a pixel size of 325 nm. The brightness and contrast of cathepsin/LAL signals was adjusted equally in all images for clarity; raw images are depicted in Supplementary Fig. 9. Super-resolution 3D images were acquired using the Zeiss LSM 980 Airyscan microscope with a pixel size of 35 nm and processed using ZEN software.

In order to determine ELS markers distribution from apical to basal poles in mouse PT cells, Rab5/Rab7/Cathepsin L were co-stained with Actin. The distance between the arbitrary determined center of each compartment and the brush border was measured using ZEN software (in µm). Data were compared using a One-Way ANOVA test on n = 3 animals.

To show Lysozyme-647 trafficking across the ELS at the above-mentioned time points, DiAna plugin on Fiji was used³⁸. Briefly, fluorescent signals from Lysozyme-647 and Rab5/Rab7/Cat L were segmented and identified as objects. The percentage of colocalizing volume between objects was then measured to highlight the progression of the protein through the different ELS structures. A similar approach was used to determine LRP2/megalin localization relative to Actin and Rab7 as a response to HCQ treatment in mice.

To identify lipid containing structures in kidney tissue, 240 µm thick slices were cut a Microm HM 650 V vibratome (Thermo Scientific, Waltham, MA, USA) and fixed with PFA 4%. Fixed slices were then permeabilized with Triton X100 0.1%, blocked with BSA 1% and donkey serum 10% and incubated with BODIPY 493/503 (1:1000, Cayman Chemicals, 121207-31-6) overnight at 4 °C. Samples were then mounted between slide and coverslip using two 120 µm thick spacers (Sigma-Aldrich, GBL654004) and Glycergel (Agilent, C056330-2). To investigate the effects of disrupting lysosomal function, slices were kept at 4 °C in a HEPES-buffered solution (pH 7.4) gassed with carbogen (95% O2/5% CO₂), containing (in mM): 118 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.8 CaCl₂, 1.44 MgSO₄, 5 glucose, 10 NaHCO₃, 10 HEPES, 5 pyruvate, 2.5 sodium butyrate, and 2.5 sodium lactate, and incubated for 1 h with 0.7 mg/ml HCQ or 1 mM Lalistat-2 or vehicle. They were then fixed in PFA 4% and incubated with BODIPY 493/503 (1:1000, Cayman Chemicals, 121207-31-6) overnight at 4 °C. For both experiments, signals were acquired using a Leica SP8 MP DIVE FALCON microscope and the HC IRAPO L 25x/1.0 W motCORR objective. 850 and 950 nm were used as exciting wavelength and images were processed and analysed using LASX software.

The determination of lipid content in S2 PTs was performed on 5 µm-thick immersion fixed frozen kidney sections stained with Sudan Black B (Merck, 199664) following OAT1 immunostaining. Stainings were imaged using the widefield Leica Thunder imager and LASX software.

Transmission electron microscopy and toluidine blue staining experiments

Kidney tissue was perfusion fixed with 3% paraformaldehyde in a phosphate buffered solution (0.1 M). After kidney extraction, 3 mm slices were immersed in 2.5% glutaraldehyde in phosphate buffer then half of them were frozen in propane gas while the others were transferred into 0.1 M cacodylate buffer containing 2.5% glutaraldehyde for several days. The latter were fixed with 1% OsO₄ for 1 h in 0.1 M cacodylate buffer at 0 °C, and 1% aqueous uranyl acetate for 1 h at 4 °C. Samples then were dehydrated in an ethanol series followed by 2 steps with 100% Propylene oxide. Samples were embedded in Epon/Araldite (Sigma-Aldrich) 66% in propylene oxide overnight, 100% for 1 h at RT and polymerized at 60 °C for 20 h. Ultrathin (70 nm) sections were post-stained with lead citrate and examined with a Talos 120 transmission electron microscope at an acceleration voltage of 120 KV using a bottom mounted Ceta camera and the MAPS software for automatic image acquisition (Thermo Fisher Scientific, Eindhoven, The Netherlands). The diameter of MLBs was measured on MAPS software. Toluidine blue staining was performed on 5 µm cryosections then deep red fluorescence acquired on single planes using a Leica SP8 inverse STED 3X confocal microscope.

Lipidomics

Mouse urine collection and lipid extraction

Urine was collected from control and HCQ-treated (20 mg/kg, 3 h) mice (n = 6) and stored at –80 °C until analysis. Samples were thawed on ice prior to lipid extraction, which was performed as described previously³⁹ with modifications. Creatinine concentrations were determined for each sample and used to normalize lipid abundances. An internal standard mixture (EquiSPLASH, Avanti Polar Lipids) was spiked into each sample before extraction. Methanol, methyl tert-butyl ether (MTBE), and water were added sequentially to induce phase separation. Samples were vortexed thoroughly, incubated, and centrifuged at 14,000 g for 10 min at 4 °C. The upper organic phase was collected, evaporated to dryness under nitrogen, and reconstituted in LC starting conditions for lipidomics.

The lower aqueous phase was used for creatinine determination. Cold methanol was added, followed by shaking (1000 rpm, 4 °C). Samples were centrifuged at 16,000 g for 10 min, and the supernatant was dried. Dried extracts were resuspended in 5 µL water, vortexed, and placed on ice for 10 min, followed by addition of 45 µL acetonitrile (final solution: 90% ACN). The mixture was vortexed, incubated on ice for 10 min, shaken on a thermomixer (1000 rpm, 4 °C, 10 min), and centrifuged at 14,000 rpm ( ≈ 16,000 g) for 10 min at 4 °C. The supernatant was transferred to glass vials for LC–MS analysis. Aliquots of all samples were pooled to generate a quality-control (QC) mixture, which was injected at regular intervals during LC–MS/MS acquisition to monitor retention time and signal stability.

LC–MS/MS acquisition

Lipidomics was performed as described previously^40–42 with modifications. Analyses were conducted on a Thermo Vanquish Horizon Binary Pump UHPLC system coupled to a Thermo Orbitrap Exploris 240 mass spectrometer with a heated electrospray ionization (HESI) source. Chromatographic separation was achieved on a Waters ACQUITY Premier BEH C18 column (1.7 µm, 2.1 × 50 mm) maintained at 60 °C. Mobile phase A was acetonitrile:water (60:40, v/v) with 5 mM ammonium acetate, and mobile phase B was acetonitrile:isopropanol (10:90, v/v) with 5 mM ammonium acetate. The following gradient was applied at 0.6 mL min⁻¹: 0 min, 15% B; 1.87 min, 30% B; 2.08 min, 48% B; 9.17 min, 82% B; 9.59 min, 99% B; 10.08–12.6 min, 99% B; followed by re-equilibration to 15% B. The injection volume was 10 µL. The Orbitrap Exploris 240 was operated in positive ionization mode with the following source settings: sheath gas = 30, auxiliary gas = 13, spray voltage = 3,800 V, vaporizer temperature = 350 °C, ion transfer tube temperature = 300 °C. Internal mass calibration was performed at the start of each run, and mild trapping was enabled to minimize in-source fragmentation. MS1 settings: resolution 60,000; scan range 150–2000 m/z; AGC target 1e6; maximum injection time 20 ms. MS2 settings: isolation window 1 m/z; stepped normalized collision energy 10, 30, and 40%; resolution 30,000; AGC target 1e5; maximum injection time 10 ms; RF lens 50%.

Data processing and lipid identification

Data analysis was conducted using the Thermo Compound Discoverer software 3.3.3.200. Lipid identification was achieved by matching to the lipid blast library, followed by visual inspection of the mirror plots (library vs. data) to confirm the match and the presence of diagnostic fragments according to the Lipid Standards Initiative guidelines, confirming the fragmentation rules. Only lipids that matched the expected fragmentation spectra, mass accuracy threshold, and RT expected pattern were included in the analysis. Areas under the peaks were normalized using respective creatinine concentration of the samples and statistical analysis was performed comparing the two different treatment groups. The differential analysis (volcano plot) was generated in Python. A lipid ontology enrichment analysis was performed using LION/web application^43,44 with a two-tailed t-test and a lipid classification based on cellular compartments.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary Information^{(4.6MB, pdf)}

41467_2026_70306_MOESM2_ESM.pdf^{(237.1KB, pdf)}

Description of Additional Supplementary Files

Supplementary movie 1^{(5MB, mov)}

Supplementary movie 2^{(1.4MB, mov)}

Supplementary movie 3^{(3.4MB, mov)}

Supplementary movie 4^{(1.4MB, mov)}

Reporting Summary^{(104.8KB, pdf)}

Transparent Peer Review file^{(322.8KB, pdf)}

Source data

Source data^{(1.6MB, xlsx)}

Acknowledgements

The authors acknowledge support from The Center for Microscopy and Image Analysis, University of Zurich, for imaging experiments and from the Functional Genomics Center Zurich, University of Zurich and ETH Zurich, for lipidomic experiments. In particular, we are grateful to Dr. Alaa Othman and Dr. Annika Jagels for performing the lipidomic analysis. Funding was from The Swiss National Centre for Competence in Research (NCCR) Kidney Control of Homeostasis (183774), The Swiss National Science Foundation (310030_184688) and The Hartmann Muller foundation to A.M.H. Figures were partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license, Adobe Illustrator, and BioRender.

Author contributions

Conceptualization: A.M.H. Experimentation: M.K., I.B.S., N.J., and M.P. Data analysis: M.K. and I.B.S. Funding acquisition: A.M.H. Project supervision: A.M.H. Writing—original draft: A.M.H., M.K., and I.B.S. Writing—review and editing: all authors.

Peer review

Peer review information

Nature Communications thanks Maria Antonietta De Matteis who co-reviewed with Leopoldo Staiano, and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The data generated in this study are provided in the Source data file. The metabolomics data have been deposited in the Center for Computational Mass Spectrometry database, University of California, San Diego under dataset ID MassIVE MSV000100733 [https://massive.ucsd.edu/Proteo<div class="pi">show [QJ]SAFe/dataset.jsp?task=4f0db3d4a29b423d8ff892e7a7b42a72]. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Monika Kaminska, Imene B. Sakhi.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70306-5.

References

1.Chrysopoulou, M. & Rinschen, M. M. Metabolic Rewiring and Communication: An Integrative View of Kidney Proximal Tubule Function. Annu. Rev. Physiol.86, 405–427 (2024). [DOI] [PubMed]
2.Hoenig, M. P., Brooks, C. R., Hoorn, E. J. & Hall, A. M. Biology of the proximal tubule in body homeostasis and kidney disease. Nephrol. Dialysis Transplant.0, 1–10 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Jang, C. et al. Metabolite Exchange between Mammalian Organs Quantified in Pigs. Cell Metab.30, 594–606.e3 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Christensen, E. I., Wagner, C. A. & Kaissling, B. Uriniferous tubule: Structural and functional organization. Compr. Physiol.2, 805–861 (2012). [DOI] [PubMed] [Google Scholar]
5.Eshbach, M. L. & Weisz, O. A. Receptor-Mediated Endocytosis in the Proximal Tubule. Annu. Rev. Physiol.79, 425–448 (2017). [DOI] [PMC free article] [PubMed]
6.Gros, F. & Muller, S. The role of lysosomes in metabolic and autoimmune diseases. Nat Rev. Nephrol.19, 366–383 (2023). [DOI] [PubMed]
7.Hall, A. M. Protein handling in kidney tubules. Nat. Rev. Nephrol. 2025, 1–12 10.1038/s41581-024-00914-1.(2025). [DOI] [PubMed]
8.Faivre, A. et al. Spatiotemporal Landscape of Kidney Tubular Responses to Glomerular Proteinuria. J Am. Soc. Nephrol.35, 854–869 (2024). [DOI] [PMC free article] [PubMed]
9.Maunsbach, A. B. Observations on the segmentation of the proximal tubule in the rat kidney. Comparison results phase contrast, fluorescence electron Microsc. J. Ultrasructure Res.16, 239–258 (1966). [DOI] [PubMed] [Google Scholar]
10.Christensen, E. I., Birn, H., Storm, T., Weyer, K. & Nielsen, R. Endocytic receptors in the renal proximal tubule. Physiology27, 223–236 (2012). [DOI] [PubMed] [Google Scholar]
11.Polesel, M. et al. Spatiotemporal organisation of protein processing in the kidney. Nat. Commun. 202213, 1–13 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Schuh, C. D. et al. Combined structural and functional imaging of the kidney reveals major axial differences in proximal tubule endocytosis. J. Am. Soc. Nephrol.29, 2696–2712 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Maack, T., Johnson, V., Kau, S. T., Figueiredo, J. & Sigulem, D. Renal filtration, transport, and metabolism of low-molecular-weight proteins: A review. Kidney Int16, 251–270 (1979). [DOI] [PubMed] [Google Scholar]
14.Daniel, H. & Rubio-Aliaga, I. An update on renal peptide transporters. Am. J. Physiol.-Ren. Physiol.284, F885–F892 (2003). [DOI] [PubMed] [Google Scholar]
15.Shipman, K. E. et al. An adaptable physiological model of endocytic megalin trafficking in OK cells and mouse kidney proximal tubule. Function (Oxf). 3, zqac046 (2022). [DOI] [PMC free article] [PubMed]
16.Mitrofanova, A., Merscher, S. & Fornoni, A. Kidney lipid dysmetabolism and lipid droplet accumulation in chronic kidney disease. Nat. Rev. Nephrol. 2023 1919, 629–645 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Lee, L.E., Doke, T., Mukhi, D. & Susztak, K. The key role of altered tubule cell lipid metabolism in kidney disease development. Kidney Int. 106, 24–34 (2024). [DOI] [PMC free article] [PubMed]
18.Granados, J. C., Nigam, A.K., Bush, K. T., Jamshidi, N. & Nigam, S.K. A key role for the transporter OAT1 in systemic lipid metabolism. J. Biol. Chem.296, 100603 (2021). [DOI] [PMC free article] [PubMed]
19.Zhang, H. Lysosomal acid lipase and lipid metabolism: new mechanisms, new questions, and new therapies. Curr. Opin. Lipido.29, 218–223 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Chen, X. & Geiger, J. D. Janus sword actions of chloroquine and hydroxychloroquine against COVID-19. Cell Signal73, 109706 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Sridharan, G. & Shankar, A. A. Toluidine blue: A review of its chemistry and clinical utility. J. Oral. Maxillofac. Pathol.16, 251 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Beenken, A. et al. Structures of LRP2 reveal a molecular machine for endocytosis. Cell186, 821–836.e13 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Turk, V., Turk, B. & Turk, D. Lysosomal cysteine proteases: facts and opportunities. EMBO J.20, 4629–4633 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Hurtado-Lorenzo, A. et al. V-ATPase interacts with ARNO and Arf6 in early endosomes and regulates the protein degradative pathway. Nat. Cell Biol.8, 124–136 (2006). [DOI] [PubMed] [Google Scholar]
25.Wartosch, L., Fuhrmann, J. C., Schweizer, M., Stauber, T. & Jentsch, T. J. Lysosomal degradation of endocytosed proteins depends on the chloride transport protein ClC-7. FASEB J.23, 4056–4068 (2009). [DOI] [PubMed] [Google Scholar]
26.Hennings, J. C. et al. A mouse model for distal renal tubular acidosis reveals a previously unrecognized role of the V-ATPase a4 subunit in the proximal tubule. EMBO Mol. Med4, 1057–1071 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Eaton, A. F., Merkulova, M. & Brown, D. The H+-ATPase (V-ATPase): from proton pump to signaling complex in health and disease. Am. J. Physiol. Cell Physiol.320, C392–C414 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Ge, M., Fontanesi, F., Merscher, S. & Fornoni, A. The Vicious Cycle of Renal Lipotoxicity and Mitochondrial Dysfunction. Front Physiol.11, 559904 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Rampanelli, E. et al. Excessive dietary lipid intake provokes an acquired form of lysosomal lipid storage disease in the kidney. J. Pathol.246, 470–484 (2018). [DOI] [PubMed] [Google Scholar]
30.Bugarski, M., Ghazi, S., Polesel, M., Martins, J. R. & Hall, A. M. Changes in NAD and Lipid Metabolism Drive Acidosis-Induced Acute Kidney Injury. J Am. Soc. Nephrol. 32, 342–356 (2021). [DOI] [PMC free article] [PubMed]
31.Ichimata, S., Hata, Y., Hirota, K. & Nishida, N. Histopathology of acute colchicine intoxication: novel findings and their association with clinical manifestations. J. Toxicol. Pathol.35, 255–262 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Sakhi, I. B. et al. A novel transgenic mouse model highlights molecular disruptions involved in the pathogenesis of Dent disease 1. Gene928, 148766 (2024). [DOI] [PubMed] [Google Scholar]
33.de Combiens, E., Sakhi, I. B. & Lourdel, S. A Focus on the Proximal Tubule Dysfunction in Dent Disease Type 1. Genes (Basel)15, 1175 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Pierre, L. et al. AMPK protects proximal tubular epithelial cells from lysosomal dysfunction and dedifferentiation induced by lipotoxicity. Autophagy21, 860–880 (2025). [DOI] [PMC free article] [PubMed]
35.Nakamura, J. et al. TFEB-mediated lysosomal exocytosis alleviates high-fat diet–induced lipotoxicity in the kidney. JCI Insight8, 17 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Wahlin, B., Braune, A., Jönsson, E., Wållberg-Jonsson, S. & Bengtsson, C. Beneficial effects of hydroxychloroquine on blood lipids and glycated haemoglobin: A randomised interventional study in patients with rheumatoid arthritis and systemic lupus erythematosus. PLoS One19, e0312546 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Schuh, C. D. et al. Long wavelength multiphoton excitation is advantageous for intravital kidney imaging. Kidney Int89, 712–719 (2016). [DOI] [PubMed] [Google Scholar]
38.Gilles, J. F., Dos Santos, M., Boudier, T., Bolte, S. & Heck, N. DiAna, an ImageJ tool for object-based 3D co-localization and distance analysis. Methods115, 55–64 (2016). [DOI] [PubMed] [Google Scholar]
39.Alshehry, Z. H. et al. An Efficient Single Phase Method for the Extraction of Plasma Lipids. Metabolites 20155, 389–403 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Cajka, T. & Fiehn, O. Increasing lipidomic coverage by selecting optimal mobile-phase modifiers in LC–MS of blood plasma. Metabolomics12, 1–11 (2016). [Google Scholar]
41.Ganguin, A. A., Skorup, I., Streb, S., Othman, A. & Luciani, P. Formation and Investigation of Cell-Derived Nanovesicles as Potential Therapeutics against Chronic Liver Disease. Adv. Health. Mater.12, 2300811 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Zareba, J. et al. NPC1 links cholesterol trafficking to microglial morphology via the gastrosome. Nat. Commun.15, 1–15 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Molenaar, M. R. et al. LION/web: a web-based ontology enrichment tool for lipidomic data analysis. Gigascience8, 1–10 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Molenaar, M. R., Haaker, M. W., Vaandrager, A. B., Houweling, M. & Helms, J. B. Lipidomic profiling of rat hepatic stellate cells during activation reveals a two-stage process accompanied by increased levels of lysosomal lipids. J. Biol. Chem.299, 103042 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information^{(4.6MB, pdf)}

41467_2026_70306_MOESM2_ESM.pdf^{(237.1KB, pdf)}

Description of Additional Supplementary Files

Supplementary movie 1^{(5MB, mov)}

Supplementary movie 2^{(1.4MB, mov)}

Supplementary movie 3^{(3.4MB, mov)}

Supplementary movie 4^{(1.4MB, mov)}

Reporting Summary^{(104.8KB, pdf)}

Transparent Peer Review file^{(322.8KB, pdf)}

Source data^{(1.6MB, xlsx)}

Data Availability Statement

[CR1] 1.Chrysopoulou, M. & Rinschen, M. M. Metabolic Rewiring and Communication: An Integrative View of Kidney Proximal Tubule Function. Annu. Rev. Physiol.86, 405–427 (2024). [DOI] [PubMed]

[CR2] 2.Hoenig, M. P., Brooks, C. R., Hoorn, E. J. & Hall, A. M. Biology of the proximal tubule in body homeostasis and kidney disease. Nephrol. Dialysis Transplant.0, 1–10 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Jang, C. et al. Metabolite Exchange between Mammalian Organs Quantified in Pigs. Cell Metab.30, 594–606.e3 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Christensen, E. I., Wagner, C. A. & Kaissling, B. Uriniferous tubule: Structural and functional organization. Compr. Physiol.2, 805–861 (2012). [DOI] [PubMed] [Google Scholar]

[CR5] 5.Eshbach, M. L. & Weisz, O. A. Receptor-Mediated Endocytosis in the Proximal Tubule. Annu. Rev. Physiol.79, 425–448 (2017). [DOI] [PMC free article] [PubMed]

[CR6] 6.Gros, F. & Muller, S. The role of lysosomes in metabolic and autoimmune diseases. Nat Rev. Nephrol.19, 366–383 (2023). [DOI] [PubMed]

[CR7] 7.Hall, A. M. Protein handling in kidney tubules. Nat. Rev. Nephrol. 2025, 1–12 10.1038/s41581-024-00914-1.(2025). [DOI] [PubMed]

[CR8] 8.Faivre, A. et al. Spatiotemporal Landscape of Kidney Tubular Responses to Glomerular Proteinuria. J Am. Soc. Nephrol.35, 854–869 (2024). [DOI] [PMC free article] [PubMed]

[CR9] 9.Maunsbach, A. B. Observations on the segmentation of the proximal tubule in the rat kidney. Comparison results phase contrast, fluorescence electron Microsc. J. Ultrasructure Res.16, 239–258 (1966). [DOI] [PubMed] [Google Scholar]

[CR10] 10.Christensen, E. I., Birn, H., Storm, T., Weyer, K. & Nielsen, R. Endocytic receptors in the renal proximal tubule. Physiology27, 223–236 (2012). [DOI] [PubMed] [Google Scholar]

[CR11] 11.Polesel, M. et al. Spatiotemporal organisation of protein processing in the kidney. Nat. Commun. 202213, 1–13 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Schuh, C. D. et al. Combined structural and functional imaging of the kidney reveals major axial differences in proximal tubule endocytosis. J. Am. Soc. Nephrol.29, 2696–2712 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Maack, T., Johnson, V., Kau, S. T., Figueiredo, J. & Sigulem, D. Renal filtration, transport, and metabolism of low-molecular-weight proteins: A review. Kidney Int16, 251–270 (1979). [DOI] [PubMed] [Google Scholar]

[CR14] 14.Daniel, H. & Rubio-Aliaga, I. An update on renal peptide transporters. Am. J. Physiol.-Ren. Physiol.284, F885–F892 (2003). [DOI] [PubMed] [Google Scholar]

[CR15] 15.Shipman, K. E. et al. An adaptable physiological model of endocytic megalin trafficking in OK cells and mouse kidney proximal tubule. Function (Oxf). 3, zqac046 (2022). [DOI] [PMC free article] [PubMed]

[CR16] 16.Mitrofanova, A., Merscher, S. & Fornoni, A. Kidney lipid dysmetabolism and lipid droplet accumulation in chronic kidney disease. Nat. Rev. Nephrol. 2023 1919, 629–645 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Lee, L.E., Doke, T., Mukhi, D. & Susztak, K. The key role of altered tubule cell lipid metabolism in kidney disease development. Kidney Int. 106, 24–34 (2024). [DOI] [PMC free article] [PubMed]

[CR18] 18.Granados, J. C., Nigam, A.K., Bush, K. T., Jamshidi, N. & Nigam, S.K. A key role for the transporter OAT1 in systemic lipid metabolism. J. Biol. Chem.296, 100603 (2021). [DOI] [PMC free article] [PubMed]

[CR19] 19.Zhang, H. Lysosomal acid lipase and lipid metabolism: new mechanisms, new questions, and new therapies. Curr. Opin. Lipido.29, 218–223 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Chen, X. & Geiger, J. D. Janus sword actions of chloroquine and hydroxychloroquine against COVID-19. Cell Signal73, 109706 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Sridharan, G. & Shankar, A. A. Toluidine blue: A review of its chemistry and clinical utility. J. Oral. Maxillofac. Pathol.16, 251 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Beenken, A. et al. Structures of LRP2 reveal a molecular machine for endocytosis. Cell186, 821–836.e13 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Turk, V., Turk, B. & Turk, D. Lysosomal cysteine proteases: facts and opportunities. EMBO J.20, 4629–4633 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Hurtado-Lorenzo, A. et al. V-ATPase interacts with ARNO and Arf6 in early endosomes and regulates the protein degradative pathway. Nat. Cell Biol.8, 124–136 (2006). [DOI] [PubMed] [Google Scholar]

[CR25] 25.Wartosch, L., Fuhrmann, J. C., Schweizer, M., Stauber, T. & Jentsch, T. J. Lysosomal degradation of endocytosed proteins depends on the chloride transport protein ClC-7. FASEB J.23, 4056–4068 (2009). [DOI] [PubMed] [Google Scholar]

[CR26] 26.Hennings, J. C. et al. A mouse model for distal renal tubular acidosis reveals a previously unrecognized role of the V-ATPase a4 subunit in the proximal tubule. EMBO Mol. Med4, 1057–1071 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Eaton, A. F., Merkulova, M. & Brown, D. The H+-ATPase (V-ATPase): from proton pump to signaling complex in health and disease. Am. J. Physiol. Cell Physiol.320, C392–C414 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Ge, M., Fontanesi, F., Merscher, S. & Fornoni, A. The Vicious Cycle of Renal Lipotoxicity and Mitochondrial Dysfunction. Front Physiol.11, 559904 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Rampanelli, E. et al. Excessive dietary lipid intake provokes an acquired form of lysosomal lipid storage disease in the kidney. J. Pathol.246, 470–484 (2018). [DOI] [PubMed] [Google Scholar]

[CR30] 30.Bugarski, M., Ghazi, S., Polesel, M., Martins, J. R. & Hall, A. M. Changes in NAD and Lipid Metabolism Drive Acidosis-Induced Acute Kidney Injury. J Am. Soc. Nephrol. 32, 342–356 (2021). [DOI] [PMC free article] [PubMed]

[CR31] 31.Ichimata, S., Hata, Y., Hirota, K. & Nishida, N. Histopathology of acute colchicine intoxication: novel findings and their association with clinical manifestations. J. Toxicol. Pathol.35, 255–262 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Sakhi, I. B. et al. A novel transgenic mouse model highlights molecular disruptions involved in the pathogenesis of Dent disease 1. Gene928, 148766 (2024). [DOI] [PubMed] [Google Scholar]

[CR33] 33.de Combiens, E., Sakhi, I. B. & Lourdel, S. A Focus on the Proximal Tubule Dysfunction in Dent Disease Type 1. Genes (Basel)15, 1175 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Pierre, L. et al. AMPK protects proximal tubular epithelial cells from lysosomal dysfunction and dedifferentiation induced by lipotoxicity. Autophagy21, 860–880 (2025). [DOI] [PMC free article] [PubMed]

[CR35] 35.Nakamura, J. et al. TFEB-mediated lysosomal exocytosis alleviates high-fat diet–induced lipotoxicity in the kidney. JCI Insight8, 17 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Wahlin, B., Braune, A., Jönsson, E., Wållberg-Jonsson, S. & Bengtsson, C. Beneficial effects of hydroxychloroquine on blood lipids and glycated haemoglobin: A randomised interventional study in patients with rheumatoid arthritis and systemic lupus erythematosus. PLoS One19, e0312546 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Schuh, C. D. et al. Long wavelength multiphoton excitation is advantageous for intravital kidney imaging. Kidney Int89, 712–719 (2016). [DOI] [PubMed] [Google Scholar]

[CR38] 38.Gilles, J. F., Dos Santos, M., Boudier, T., Bolte, S. & Heck, N. DiAna, an ImageJ tool for object-based 3D co-localization and distance analysis. Methods115, 55–64 (2016). [DOI] [PubMed] [Google Scholar]

[CR39] 39.Alshehry, Z. H. et al. An Efficient Single Phase Method for the Extraction of Plasma Lipids. Metabolites 20155, 389–403 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Cajka, T. & Fiehn, O. Increasing lipidomic coverage by selecting optimal mobile-phase modifiers in LC–MS of blood plasma. Metabolomics12, 1–11 (2016). [Google Scholar]

[CR41] 41.Ganguin, A. A., Skorup, I., Streb, S., Othman, A. & Luciani, P. Formation and Investigation of Cell-Derived Nanovesicles as Potential Therapeutics against Chronic Liver Disease. Adv. Health. Mater.12, 2300811 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Zareba, J. et al. NPC1 links cholesterol trafficking to microglial morphology via the gastrosome. Nat. Commun.15, 1–15 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Molenaar, M. R. et al. LION/web: a web-based ontology enrichment tool for lipidomic data analysis. Gigascience8, 1–10 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Molenaar, M. R., Haaker, M. W., Vaandrager, A. B., Houweling, M. & Helms, J. B. Lipidomic profiling of rat hepatic stellate cells during activation reveals a two-stage process accompanied by increased levels of lysosomal lipids. J. Biol. Chem.299, 103042 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Heterogeneity in lysosomal dynamics and metabolic functions along the kidney proximal tubule

Monika Kaminska

Imene B Sakhi

Nevena Jankovic

Marcello Polesel

Andrew M Hall

Abstract

Introduction

Results

Generation of proteins labeled with pH sensitive fluorophores

Fig. 1. Characterization of fluorescently labelled proteins as a ratiometric readout of endo-lysosomal pH measurement in vitro.

Generation of a peptide for ratiometric pH imaging

Fig. 2. Characterization of a custom-designed dual-labelled peptide as a ratiometric readout of endo-lysosomal pH measurement in vitro.

Probe progression through the endo-lysosomal system

Fig. 3. Intravital imaging of endo-lysosomal pH evolution in the kidney proximal tubule.

Intravital imaging of acidified endo-lysosomes

Effects of de-acidifying endo-lysosomes

Fig. 4. Hydroxychloroquine injection inhibits megalin receptor recycling and protein uptake in the kidney proximal tubule.

Lysosomes in S1 proximal tubules degrade filtered proteins

Fig. 5. De-acidification of the endo-lysosomal system alters lysosomal dynamics in the early proximal tubule.

Deacidification of lysosomes in S1 proximal tubules

Fig. 6. Dynamic interactions between lysosomes and late endosomes in proximal tubule-derived cells.

Lysosomal function in S2 proximal tubules

Fig. 7. Axial heterogeneity of lysosomal function along the proximal tubule.

De-acidification of lysosomes in S2 proximal tubules

Fig. 8. Lysosomal de-acidification impacts on lipid metabolism in the S2 proximal tubule.

Discussion

Fig. 9. Summary diagram.

Methods

Labelling of proteins

Labelling of peptide

Characterization of labelled proteins and peptide in solution

Characterization of labelled proteins and peptide in cells

Imaging of late endosomes and lysosomes in cells

Imaging of lipid droplets, mitochondria, and lysosomes in cells

Live Imaging of Lipid Droplet–Lysosome Interactions in HK-2 Cells

Imaging of cells treated with Lalistat-2

Animal experiments

Intravital imaging

Immunostaining in fixed kidney tissue

Transmission electron microscopy and toluidine blue staining experiments

Lipidomics

Mouse urine collection and lipid extraction

LC–MS/MS acquisition

Data processing and lipid identification

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions

Peer review

Peer review information

Data availability

Competing interests

Footnotes

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases