Abstract

   Differentiation is critical for cell fate decisions, but the signals
   involved remain unclear. The kidney proximal tubule (PT) cells reabsorb
   disulphide-rich proteins through endocytosis, generating cystine via
   lysosomal proteolysis. Here we report that defective cystine
   mobilization from lysosomes through cystinosin (CTNS), which is mutated
   in cystinosis, diverts PT cells towards growth and proliferation,
   disrupting their functions. Mechanistically, cystine storage stimulates
   Ragulator-Rag GTPase-dependent recruitment of mechanistic target of
   rapamycin complex 1 (mTORC1) and its constitutive activation.
   Re-introduction of CTNS restores nutrient-dependent regulation of
   mTORC1 in knockout cells, whereas cell-permeant analogues of L-cystine,
   accumulating within lysosomes, render wild-type cells resistant to
   nutrient withdrawal. Therapeutic mTORC1 inhibition corrects lysosome
   and differentiation downstream of cystine storage, and phenotypes in
   preclinical models of cystinosis. Thus, cystine serves as a lysosomal
   signal that tailors mTORC1 and metabolism to direct epithelial cell
   fate decisions. These results identify mechanisms and therapeutic
   targets for dysregulated homeostasis in cystinosis.

   Subject terms: Mechanisms of disease, TOR signalling, Lysosomes,
   Nutrient signalling
     __________________________________________________________________

   Cystinosis is a lysosomal storage disease that affects the kidney.
   Here, the authors use preclinical models and advanced profiling
   techniques to discover the mechanism by which defective cystine
   mobilization from lysosomes disrupts kidney cell function, offering
   insights into potential therapies.

Introduction

   The epithelial cells lining the kidney proximal tubule (PT) play a
   critical role in homeostasis by efficiently reabsorbing small proteins
   from the primary urine. The process is mediated by receptor-mediated
   endocytosis and lysosomal pathways, preventing the loss of precious
   nutrients in the urine^[46]1. The lysosomal degradation of
   disulphide-rich plasma proteins like albumin accounts for the supply of
   cystine that is exported to cytoplasm through the lysosomal cystine
   transporter cystinosin^[47]2. In the cytoplasm, cystine is reduced to
   cysteine and utilized for the maintenance of cellular redox
   homeostasis^[48]3. Additionally, the lysosome-mediated removal of
   dysfunctional mitochondria regulates the differentiating state of PT
   cells through junctional integrity and transcriptional programs,
   including the transcription factor Y box 3 (Ybx3)^[49]4. Recessive
   mutations in CTNS, the gene that encodes cystinosin (CTNS), cause
   cystinosis, a lysosomal storage disease characterized by loss of
   reabsorptive properties of PT cells, progressing towards kidney
   failure^[50]5. How, mechanistically, the increased levels of cystine
   within lysosomes, reflecting its defective export through CTNS, affect
   cell fate decisions remains unknown.

   A crucial step in nutrient sensing is the recruitment of the
   evolutionarily conserved mechanistic target of rapamycin complex 1
   (mTORC1) to the surface of the lysosome^[51]6. In the presence of
   nutrients, such as amino acids, glucose, and lipids, a multiprotein
   complex composed of heterodimeric Rag (Ras-related GTP binding) GTPases
   (consisting of RagA or RagB bound to RagC or RagD), Ragulator, and
   vacuolar H^+-ATPase (V-ATPase) tethers mTORC1 to the lysosomal surface,
   enabling its activation by the (growth factor-induced) GTP-binding
   protein Rheb (Ras homolog enriched in brain^[52]7,[53]8). The mTORC1
   signaling from lysosomes promotes growth and proliferation while
   repressing lysosome biogenesis and catabolic autophagy^[54]9. The
   proper regulation of mTORC1 sustains polarized transport properties in
   epithelial cell types^[55]10. This has been well illustrated in the
   kidney, where hypo and/or hyperactivation of mTORC1 pathway can drive
   pathological changes in podocytes and tubular cells, whereas blunting
   of hyperactive mTORC1 may prove beneficial in renal cell carcinoma and
   polycystic and/or diabetic kidney disease^[56]11.

   Like other lysosomal amino acid transporters, the CTNS protein
   physically interacts with the V-ATPase-Ragulator-Rag GTPase lysosomal
   scaffold that activates mTORC1 (ref. ^[57]12). During periods of
   starvation in Drosophila, mobilization of cystine from lysosomes fuels
   acetyl-CoA synthesis and limits mTORC1 reactivation, altogether
   balancing catabolic autophagy and growth in developing organs^[58]13.
   In the mammalian kidney, PT cells are particularly vulnerable to
   imbalances in metabolism, which can curb their differentiation and
   cause disease^[59]14. These findings raise the possibility that
   lysosomal cystine export may shape the response of mTORC1 and
   metabolism to fate decisions during the differentiation of the kidney
   tubule epithelium.

   Here, we combine genetic model organisms and physiologically relevant
   cell culture systems to test the causal role of an evolutionarily
   conserved pathway that tunes mTORC1 signaling and metabolism in
   response to levels of lysosomal cystine to instruct fate
   specialization. Together the results provide insights into the
   mechanisms that regulate homeostasis and the specialized functions in
   kidney tubular epithelial cells, and therapeutic targets for
   cystinosis.

Results

Cystine storage diverts differentiation towards proliferation and
dyshomeostasis

   The genetic deletion of Ctns in 24-week-old mouse kidneys induces
   cystine storage and lysosome abnormalities in PT segments (Fig. [60]1a,
   b), leading to defective receptor-mediated endocytosis of low molecular
   weight proteins (LMWPs; Fig. [61]1c) and their losses in the urine
   (Fig. [62]1d). Other markers of kidney tubular dysfunction, such as
   albuminuria and glucosuria, were significantly higher in Ctns knockout
   (KO) mice when compared with wild-type (WT) littermates (Supplementary
   Fig. [63]1a, b), confirming previous studies^[64]4,[65]5,[66]15. The
   expression of genes involved in the tubular reabsorption of albumin
   (Lrp2), glucose (Slc5a2), and LMWP/CC16 (Cubn) inversely correlated
   with solute loss in Ctns^KO compared with Ctns^WT animals
   (Supplementary Fig. [67]1c–e), reinforcing the link between loss of
   differentiation markers (i.e., transport systems) and development of
   kidney tubular dysfunction.

Fig. 1. Cystine storage diverts differentiation towards growth and
proliferation.

   [68]Fig. 1
   [69]Open in a new tab

   a Cystine levels in Ctns mouse kidneys; n = 6 animals in each
   experimental group. b Immunofluorescence and quantification of the
   Lamp1 mean fluorescence intensity (MFI) in the LTL^+ PT segments of
   Ctns mouse kidneys, n = 164 Ctns^WT and n = 137 Ctns^KO PT segments
   from 3 biologically independent animals. c Confocal microscopy and
   quantification of the Cy5-tagged β-lactoglobulin MFI in the LTL^+ PT
   segments of Ctns mouse kidneys, n = 356 Ctns^WT and n = 330 Ctns^KO PT
   segments from 3 biologically independent animals. d Quantification of
   LMW/Clara cell protein (CC16) levels in the urine samples of Ctns mice;
   n = 7 animals in each experimental group. e Volcano plot of genome-wide
   changes in PT segments of Ctns mouse kidneys. Red and green dots show
   the differentially expressed genes (DEGs; adjusted p value < 0.05;
   n = 3 animals in each experimental group). f GSEA of over- and
   under-represented biological pathways in the PT segments of Ctns^KO
   mouse kidneys. FDR q value was calculated from the enrichment pathway
   analysis in NetworkAnalyst. g Immunoblotting and quantification of the
   indicated proteins in the kidneys of Ctns mice; n = 3 animals in each
   experimental group. h Confocal microscopy of whole-kidney sections
   stained for Ccna2 (red), Lrp2 (blue), and PT marker (LTL, green);
   quantification of Ccna2^+ nuclei and of Lrp2 MFI in the PT segments of
   Ctns mouse kidneys. For Ccna2^+ nuclei: n = 31 Ctns^WT and n = 54
   Ctns^KO PT segments from  3 biologically independent animals. For Lrp2
   MFI: n = 60 Ctns^WT and n = 56 Ctns^KO PT segments from  3 biologically
   independent animals. i mRNA levels for the indicated genes in the
   kidneys of Ctns mice; n = 8 biologically independent animals. j
   Whole-kidney sections stained for PT apical marker (Aqp1, green) and PT
   basolateral marker (Laminin, red). Confocal microscopy and
   quantification of PT cell height in the kidneys of Ctns mouse; n = 90
   PT segments from 3 biologically independent animals. Plots represent
   mean ± SEM. Statistics calculated by unpaired two-tailed Student’s t
   test. Dotted white squares contain images at high magnification. Nuclei
   counterstained with DAPI (grey). Scale bars are 50 μm in b, c, and h
   and 20 μm in j. Source data are provided as a Source Data file.

   Analysis of the transcriptome of PT segments from 24-week-old
   littermates (Fig. [70]1e; Supplementary Fig. [71]1f) identified 61
   genes differentially expressed between Ctns^KO and Ctns^WT mice, of
   which 43 were upregulated and 18 downregulated, respectively. Gene set
   enrichment analysis revealed an over-representation of pathways
   involved in the cell cycle regulation, whereas membrane trafficking,
   vesicle-mediated/small molecule transport, and hemostasis were
   significantly under-represented (Fig. [72]1f). The dysregulation of
   these pathways was verified (Supplementary Fig. [73]1g), with increased
   expression of proteins regulating cell cycle (Ccna2, Cyclin A2, and
   Cdc20, Cell cycle division 20) and proliferation (Pcna, proliferating
   cell nuclear antigen), contrasting with decreased levels of Lrp2 (LDL
   receptor related protein 2)/megalin, the master coordinator of LMWP
   uptake^[74]16, in the PT segments of Ctns^KO kidneys compared with
   Ctns^WT littermates (Fig. [75]1g, h; Supplementary Fig. [76]1h). In
   line, decreased expression of PT cell differentiation related
   genes^[77]14 (Slc5a2 and Slc34a1) and increased
   dedifferentiation markers^[78]17,[79]18 (Foxm1, Sox9, and Vimentin)
   were observed in the PT segments of Ctns^KO mouse kidneys compared with
   Ctns^WT littermates (Fig. [80]1g, i; Supplementary Fig. [81]1i).
   Consistent with altered tubular differentiation programs, we noted that
   the PT cells from Ctns^KO mice changed from a cuboidal shape to a
   flattened and spread-out epithelium (Fig. [82]1j). Thus, the cystine
   storage secondary to the loss of CTNS triggers anabolic programs for
   growth and proliferation, disrupting the differentiating state of the
   PT cells and their reabsorptive properties.

Defective lysosomes lead to tubular cell fate changes

   In silico predictions of gene function and module associations
   (GeneBridge toolkit^[83]19; [84]www.systems-genetics.org) suggested
   that the lysosomally localized Ctns relates to lytic vacuole and
   organelle pathways beyond cystine transport (Supplementary
   Fig. [85]2a–c). This led us to investigate whether changes in the
   differentiation and tubular cell fate induced by CTNS loss and cystine
   storage reflect alterations in the lysosomal catabolic activities. We
   used primary cells derived from microdissected PT segments of Ctns
   mouse kidneys (henceforth mPTCs^[86]4) and transduced them with an
   adenovirus expressing the lysosomal transmembrane protein 192 (Tmem192)
   fused to 3×haemagglutinin (T192-3×HA^[87]20; Fig. [88]2a) to enable the
   immunoisolation and functional probing of pure and intact lysosomes
   (Supplementary Fig. [89]2d–h). The Ctns^KO mPTCs, which accumulate
   cystine within lysosomes (Fig. [90]2b), showed abnormal
   growth/proliferation rates (Fig. [91]2c), defective endocytosis of
   fluorescent ligands (Fig. [92]2d), and changes in the expression of PT
   cell dedifferentiation markers (Fig. [93]2e), closely mimicking the
   changes observed in vivo. Moreover, these cells showed
   abnormal/elongated primary cilium—an antenna-like sensory organelle
   whose signaling regulates differentiation and cell fate^[94]21—when
   compared with wild-type cells (Fig. [95]2f). These changes were
   paralleled by a marked decrease in the number of functionally
   degradative lysosomes in Ctns^KO mPTCs, both under full and
   nutrient-depleted conditions, when compared with Ctns^WT cells
   (Supplementary Fig. [96]3a–c). In agreement, the Ctns^KO cells showed
   higher numbers of Map1lc3b-positive autophagic puncta and
   autolysosomes, insensitive to treatment with Bafilomycin A1
   (Fig. [97]2g; Supplementary Fig. [98]3d). Comparative analysis of
   lysosome proteins from Ctns^KO versus Ctns^WT mPTCs revealed decreased
   proteolytic generation of the 32 kDa luminal mature cathepsin D (CtsD)
   and augmented levels of the lipidated, autophagosome-associated form
   Map1Lc3b-II (Fig. [99]2h). Similar alterations were also observed in
   Ctns^KO mouse kidneys (Supplementary Fig. [100]3e), as well as in the
   ctns-deficient zebrafish^[101]4 and Ctns^KO rat model^[102]22, both
   displaying significant cystine storage, metabolic defects, and PT cell
   dedifferentiation and tubular dysfunction (Supplementary
   Fig. [103]3f–i). The proteolytic defects were not linked to
   dysregulated lysosomal acidification as indicated by the measurements
   in Ctns mPTCs transiently expressing a bona fide biosensor
   (R2pH-LAMP1-3×HA^[104]23; Supplementary Fig. [105]4a–c) that readily
   measures the changes in luminal pH of the lysosome or stained with
   LysoTracker^[106]24 (Supplementary Fig. [107]4d, e). Thus, the tubular
   epithelial cells lacking CTNS/ accumulating cystine reprogramme their
   metabolism from catabolism to anabolism to fuel growth and
   proliferation, leading to dysregulated differentiation and
   dyshomeostasis.

Fig. 2. Defective lysosomes lead to loss of differentiation and dysregulated
homeostasis.

   [108]Fig. 2
   [109]Open in a new tab

   a Schematic of the lysosome purification using affinity-based capture
   from cells transiently expressing T192-3×HA. LC-MS/MS, liquid
   chromatography with tandem mass spectrometry. Images were drawn using
   pictures from Servier Medical Art, which is licensed under a Creative
   Commons Attribution 3.0 Unported License
   ([110]https://creativecommons.org/licenses/by/3.0/). b Quantification
   of cystine levels in the lysosomes of Ctns mPTCs; n = 2 biologically
   independent experiments. c Confocal microscopy and quantification of
   BrdU^+ cells (expressed as percentage of total cells); n = 60 Ctns^WT
   and n = 58 Ctns^KO randomly selected fields of views from
   3 biologically independent experiments. d Confocal microscopy and
   quantification of BSA^+ structures per cell. n = 209 Ctns^WT and
   n = 282 Ctns^KO cells from 3 biologically independent experiments. e
   mRNA levels for the indicated genes. For Slc7a13 and Slc5a2, n = 3
   Ctns^WT and n = 4 Ctns^KO; and for Slc3a1, Foxm1, Vimentin, and Sox9,
   n = 4 Ctns^WT and Ctns^KO biologically independent samples. f Maximum
   intensity projection of image stacks and quantification of the primary
   cilium length (Ac-tubulin, green) in Ctns mPTCs; n = 313 Ctns^WT and
   n = 285 Ctns^KO cells from 4 biologically independent animals. g
   Quantification of GFP/RFP^+ structures in Ctns mPTCs transiently
   expressing GFP-Map1lc3b and Lamp1-RFP (expressed as percentage of total
   lysosomes; n = 10 randomly selected fields of views from 3 biologically
   independent animals). h Immunoblotting and quantification of the
   indicated proteins in input samples and corresponding lysosome-enriched
   fractions (Lyso-IP) from Ctns mPTCs; n = 3 biologically independent
   experiments. Plots represent mean ± SEM. Statistics were calculated by
   unpaired two-tailed Student’s t test in c–f, h or by one-way ANOVA
   followed by Sidak’s multiple comparisons test in g. Dotted white
   squares contain images at high magnification. Nuclei counterstained
   with DAPI (grey). Scale bars are 10 μm. Source data are provided as a
   Source Data file.

Hyperactive mTORC1 signalosome drives the metabolic switch

   To identify the factors that drive the metabolic switch induced by CTNS
   deficiency, we conducted proteomics- and metabolomics-based profiling
   of Ctns mPTCs (Fig. [111]3a). PCA and hierarchical clustering revealed
   that proteome and metabolome of Ctns^KO mPTCs substantially differ from
   Ctns^WT control cells (Fig. [112]3b, c; Supplementary Fig. [113]5a, b).
   Protein-protein interaction (PPI) analysis of differentially expressed
   proteins revealed significant enrichment in metabolism, cell cycle and
   ribosome biogenesis, growth and proliferation, apoptosis regulation,
   signal transduction, and differentiation pathways in Ctns^KO versus
   Ctns^WT mPTCs (Supplementary Fig. [114]5c, d). Metabolite set
   enrichment analysis (MSEA) identified in Ctns^KO cells chemical
   categories predominantly containing amino acids, organic acids, sugars,
   and nucleotides (Supplementary Fig. [115]5e–g). Integrated biological
   network analysis crossing differentially produced proteins with
   metabolites (Fig. [116]3b, c) identified pathways linked to the
   regulation of mTORC1 signalosome, such as glucose and amino
   acids^[117]6, mTORC1 signaling itself, and downstream anabolic
   (protein, fatty acid, and nucleotide biosynthesis) or catabolic
   (autophagy) processes enriched in Ctns^KO mPTCs (Fig. [118]3d). An
   enrichment in pathways related to cancer cell metabolism and renal cell
   carcinoma, known to be driven by aberrant mTORC1 signaling^[119]25, was
   also observed in these mutant cells (Fig. [120]3d). These results were
   confirmed when applying an artificial intelligence (AI) engine
   (PandaOmics^[121]26) that uses machine learning (ML) tools and
   statistical validation to rank disease-target associations and
   prioritize actionable drug targets. Accordingly, this computational
   platform identified mTOR on the top of the list of predicted drug
   targets (Supplementary Fig. [122]6) potentially reversing
   dyshomeostasis and tubular dysfunction downstream of cystine storage.
   Inspired by the biological evidence and in silico predictions, we next
   examined the status of mTORC1 signaling in CTNS-defective PT cells. The
   CTNS loss rendered mTORC1 less sensitive to nutrient starvation and
   constitutively active, as indicated by persistent phosphorylation of
   the ribosomal protein S6 (pS6^Ser235/236) and eukaryotic translation
   initiation factor 4E (eIF4E)-binding protein 1 (p4EBP1^Ser65)
   (Fig. [123]3e, f). Increased phosphorylation rates of the canonical
   mTORC1 substrate (pS6^Ser235/236) were also observed in Ctns^KO mouse
   kidneys (Supplementary Fig. [124]7a, b). Prolonged starvation led to a
   partial and progressive reactivation of mTORC1 in wild-type cells,
   while this process was severely impaired in Ctns^KO mPTCs, as measured
   by phosphorylation of the mTORC1 substrate S6 (Supplementary
   Fig. [125]7c). Increasing the cytosolic cysteine levels through the
   supplementation of non-toxic concentrations of L-cysteine (1 mM for
   16 h; Supplementary Fig. [126]7d) or modified molecules such as
   N-acetylcysteine (NAC; 1 mM for 16 h; Supplementary Fig. [127]7e) did
   not blunt the hyperactive mTORC1 signaling in Ctns^KO PT cells
   (Supplementary Fig. [128]7f), suggesting that the recycling of
   lysosomal cystine through CTNS regulates mTORC1 signaling in the
   tubular epithelial cells. In agreement, CTNS deficiency augmented the
   lysosomal levels of components^[129]27 of the V-ATPase subunits
   Atp6v0d1 and Atp6v1b2, Ragulator complex subunit Lamtor2 and RagC
   GTPase that control mTORC1 activation (Fig. [130]3g). In Ctns^KO mPTCs,
   RagC GTPase was strongly clustered on lysosomal-associated membrane
   protein 1 (Lamp1)-positive structures compared with  Ctns^WT mPTCs
   (Fig. [131]3h), substantiating the regulatory effect of CTNS and
   cystine storage on mTORC1 activation. The impact of CTNS deficiency on
   mTORC1 activity was further examined in ctns-deficient zebrafish^[132]4
   and Ctns^KO rat model^[133]22, both displaying increased
   phosphorylation rates of pS6^Ser235/236 compared to controls
   (Fig. [134]3i–k), demonstrating the evolutionary conservation of this
   signaling connection.

Fig. 3. Hyperactive mTORC1 drives metabolic switch in CTNS-defective/cystine
accumulating PT cells.

   [135]Fig. 3
   [136]Open in a new tab

   a Schematic of the experimental workflow for multi-omics integration,
   and pathway/network enrichment analysis. b, c Volcano plot of (b)
   proteome and (c) metabolome changes induced by CTNS loss in the PT
   cells. Black and blue dots show the differentially produced proteins
   and metabolites, respectively. [Log[2] (fold change) <−0.5 and Log[2]
   (fold change) >1; and P value < 0.05]. (d, left) Kyoto Encyclopaedia of
   Genes and Genomes (KEGG) pathway analysis; n = 3 biologically
   independent experiments. Black dashed line shows the threshold of
   significant enrichment. Black circles show the size of enriched
   proteins and metabolites, respectively. P value was calculated from the
   enrichment analysis using Omicsnet2.0. d (right) Upstream regulators of
   mTORC1 signaling activity and downstream-regulated pathways. e, f
   Immunoblotting and quantification of the indicated proteins in Ctns
   mPTCs; n = 3 biologically independent animals. g Immunoblotting and
   quantification of the indicated proteins in input samples and
   corresponding Lyso-IP fractions from Ctns^WT and Ctns^KO mPTCs; n = 4
   biologically independent experiments. h Confocal microscopy and
   quantification of RagC/Lamp1^+ structures (expressed as percentage of
   total lysosomes; n = 15 randomly selected fields of views from
   3 biologically independent animals). Dotted white squares contain
   images at high magnification. i Confocal microscopy and quantification
   of ^pS235/236S6 MFI in Aqp1^+ (green) PT segments of Ctns rat kidneys;
   n = 227 Ctns^WT and n = 228 Ctns^KO PT segments pooled from 3
   biologically independent animals. j, k Immunoblotting and
   quantification of the indicated proteins in (j) 24-week-old Ctns rats
   or (k) 5 days post fertilization (dpf)-ctns zebrafish. For rat kidneys:
   n = 10 animals in each experimental group and for zebrafish: n = 7
   biologically independent samples (with each sample representing a pool
   of 10 zebrafish larvae). Plots represent mean ± SEM. Statistics
   calculated by unpaired two-tailed Student’s t test. Nuclei
   counterstained with DAPI (grey). Scale bars are 10 μm in h and 50 μm in
   i. Source data is provided as a Source Data file. Images in a and d
   were drawn using pictures from Servier Medical Art, which is licensed
   under a Creative Commons Attribution 3.0 Unported License
   ([137]https://creativecommons.org/licenses/by/3.0/).

Temporal mTORC1 activation by cystine storage directs kidney disease
development

   The possibility to monitor progressive PT dysfunction (Fig. [138]4a)
   and cystine storage (Fig. [139]4b) over time allowed us to test whether
   mTORC1 activation may drive the onset and progression of kidney disease
   in CTNS deficient mice (Fig. [140]4c). Analyses in mPTCs from Ctns^KO
   kidneys revealed nutrient-dependent mTORC1 activation, similar to
   control Ctns^WT mPTCs, at 6 weeks of age (Fig. [141]4d). Conversely, in
   mPTCs from pre-symptomatic 12-week-old Ctns^KO mice, mTORC1 activation
   was stronger than in control cells and largely resistant to nutrient
   starvation (Fig. [142]4e). Like wild-type cells, mTOR translocated to
   the surface of Lamp1-flagged lysosomes in presence of nutrients in
   mPTCs from 6-week-old Ctns^KO kidneys, but it readily dissociated from
   lysosomes following nutrient starvation (Fig. [143]4f), as expected in
   response to changes in nutrient availability. Conversely, in mPTCs from
   12-week-old Ctns^KO kidneys, mTOR remained clustered on Lamp1-flagged
   lysosomes following nutrient deprivation, contrasting with the
   dissociation observed in wild-type cells (Fig. [144]4g). These changes
   were accompanied by a reduced number of enlarged Lamp1^+ structures in
   mPTCs from kidneys of 6- to 12-week-old Ctns^KO mice (Supplementary
   Fig. [145]8a). Despite these metabolic and signaling alterations, the
   Ctns^KO mouse kidneys displayed mild tubular damage (Supplementary
   Fig. [146]8b, c), with no changes in kidney function (blood urea
   nitrogen and creatinine clearance; Supplementary Fig. [147]8d, e) and
   no alterations in kidney growth (as indicated by the kidney-to-body
   weight ratio; Supplementary Fig. [148]8f) in aged animals. The effect
   of CTNS deficiency on mTORC1 activity did not result from alterations
   in growth factor-directed or 5’-AMP activated protein (AMPK)-mediated
   signaling upstream of mTORC1, which were similar between wild-type and
   Ctns^KO mPTCs (Supplementary Fig. [149]8g). Thus, CTNS loss/cystine
   storage activates mTORC1 through mechanisms that are, at least in part,
   distinct from growth factor-induced signals and/or energy availability.

Fig. 4. Temporal mTORC1 activation by cystine buildup leads to kidney disease
development in CTNS-deficient mice.

   [150]Fig. 4
   [151]Open in a new tab

   a CC16 levels in the urine of Ctns mice at the indicated times; 6
   weeks: n = 7 Ctns^WT and n = 8 Ctns^KO animals; 12 weeks: n = 11
   Ctns^WT and n = 9 Ctns^KO animals; 16 weeks: n = 8 Ctns^WT and n = 10
   Ctns^KO animals; 24 weeks: n = 8 Ctns^WT and n = 10 Ctns^KO animals; 32
   weeks: n = 9 Ctns^WT and n = 7 Ctns^KO animals. b Cystine levels in
   Ctns mouse kidneys; 6 weeks: n = 9 animals in each experiment group; 12
   weeks: n = 8 Ctns^WT and n = 6 Ctns^KO animals; 24 weeks: n = 5 Ctns^WT
   and n = 4 Ctns^KO animals. c Schematic of the experimental workflow for
   monitoring mTORC1 activity in Ctns mice and in cultured PT cells. mPTCs
   from (d, f) asymptomatic (6 weeks of age), (e, g) pre-symptomatic (12
   weeks of age), and i symptomatic (24 weeks of age) Ctns^KO mice and
   their (age-matched) Ctns^WT littermates were cultured under fed and
   starved conditions. (d, e) Immunoblotting and quantification of the
   indicated proteins; n = 3 biologically independent animals. (f, g)
   Confocal microscopy and quantification of mTOR^+-Lamp1^+ signals;
   n = 10 randomly selected and non-overlapping fields of views from  3
   biologically independent animals. Nuclei counterstained with DAPI
   (blue). h Fed mPTCs were treated with cysteamine (100 nM) or Torin1
   (250 nM) for 16 h. Cysteamine enters the lysosomes through a distinct
   importer and reacts with cystine to exchange its disulfide bridge into
   a mixed (cysteamine-cysteine) disulfide, which exits via the cationic
   amino-acid exporter PQLC2, and generates free cysteine, which  exits
   from lysosomes through an unknown transporter. i Immunoblotting of the
   indicated proteins; n = 2 biologically independent experiments. j
   Cystine levels in Ctns mPTCs; n = 6 biologically independent samples.
   Plots represent mean ± SEM. Statistics were calculated by unpaired
   two-tailed Student’s t test. Scale bars, 10 μm. Source data are
   provided as a Source Data file. Images in c, h were drawn using
   pictures from Servier Medical Art, which is licensed under a Creative
   Commons Attribution 3.0 Unported License
   ([152]https://creativecommons.org/licenses/by/3.0/).

Lysosomal cystine regulates mTORC1 through Ragulator-Rag GTPase complex

   We next tested whether cystine storage per se might force mTORC1 to
   remain on the lysosomal surface, using pharmacologic and genetic tools.
   Emptying the cystine storage with the drug cysteamine^[153]28 blunted
   mTORC1 activation in mPTCs derived from 24-week-old Ctns^KO mice
   (Fig. [154]4h, i). Notably, the latter treatment corrected the mTORC1
   activation to the same extent than the selective mTOR inhibitor Torin1
   which, contrary to cysteamine, does not empty the cystine storage
   (Fig. [155]4i). To substantiate the causal link between CTNS loss,
   lysosomal cystine storage, and hyperactive mTORC1 signaling, we
   transduced Ctns^KO PT cells with an adenovirus expressing HA-tagged
   wild-type (HA-CTNS^WT) or Gly339Arg mutant (HA-CTNS^G339R) CTNS
   (Fig. [156]5a; Supplementary Fig. [157]9a). This disease-causing
   CTNS^G339R mutant^[158]29 preserves the delivery of mutant CTNS to the
   lysosomal surface while abrogating its transport function, thus causing
   cystine storage (Fig. [159]5b; Supplementary Fig. [160]9b). The
   re-introduction of CTNS^WT in Ctns^KO mPTCs rescued the cystine levels
   (Fig. [161]5b) and restored the nutrient-directed regulation of mTORC1
   compared to Ctns^KO mPTCs transduced with an empty vector
   (Fig. [162]5c; +HA-CTNS^WT). Conversely, the nutrient-dependent
   regulation of mTORC1 signaling activity was markedly abolished in
   Ctns^KO mPTCs expressing mutant CTNS^G339R/accumulating cystine
   (Fig. [163]5b), as evidenced by the elevated phosphorylation rate of
   mTORC1 substrate S6 (Fig. [164]5c; +HA-CTNS^G339R) in nutrient-depleted
   conditions. Immunoblotting analyses of anti-HA immunoprecipitates from
   Ctns^KO mPTCs transiently expressing diverse tagged CTNS forms revealed
   a decreased binding of CTNS^WT to endogenous members of Ragulator-Rag
   GTPases lysosomal scaffold (Lamtor2 and RagC GTPase) compared with
   mPTCs transduced with mutant CTNS^G339R (Fig. [165]5d). In line, the
   reconstitution with CTNS^WT, but not the mutant CTNS^G339R, rescued
   autophagy-lysosome degradation systems in Ctns^KO mPTCs (Supplementary
   Fig. [166]9c, d). To prove that lysosomal cystine is involved in this
   signaling loop, we bypassed the transport function of CTNS by culturing
   Ctns^WT mPTCs with cystine methyl ester derivates
   (CDME^[167]27,[168]30). These amino acid derivates rapidly enter the
   cells and are hydrolysed by esterases into native amino acids,
   ultimately accumulating within mPTCs and lysosomes (Fig. [169]5e–g).
   Exposure of nutrient-depleted Ctns^WT mPTCs with low doses of CDME
   stimulated mTORC1 activity in a time-dependent fashion (Fig. [170]5h;
   Supplementary Fig. [171]9e-g). This lag correlated with the time
   required for CDME to accumulate within lysosomes and could not be
   recapitulated by treating starved cells with unmodified L-cystine
   analogues that do not accumulate (Fig. [172]5f, g). Immunoprecipitation
   followed by immunoblotting analysis revealed that the heightened levels
   of lysosomal cystine increase the binding of HA-tagged CTNS to
   endogenous components of Ragulator-Rag GTPases scaffold complex
   (Lamtor2 and RagC) in CDME-treated cells compared with those exposed to
   unmodified L-cystine analogues (Fig. [173]5i). Through interactions
   with Ragulator-Rag GTPase lysosomal scaffold, CTNS regulates the
   activation of mTORC1 signaling landscape in response to variations in
   the levels of lysosomal cystine.

Fig. 5. Lysosomal cystine regulates mTORC1 activity through Ragulator-Rag
GTPase lysosomal scaffold.

   [174]Fig. 5
   [175]Open in a new tab

   (a–d) mPTCs from Ctns^KO mice were transduced with a Null or
   hemagglutinin-tagged CTNS^WT (CTNS^WT-HA) or HA-tagged-mutant CTNS
   (CTNS^G339R-HA) bearing adenoviral particles for 2 days. b Cystine
   levels in Ctns mPTCs; n = 6 biologically independent samples. (c, h)
   Cells were cultured under fed or starved conditions or restimulated by
   adding back amino acids and glucose (AA/G stim). c Immunoblotting of
   the indicated proteins. d Fed Ctns^KO mPTCs transiently expressing
   tagged CTNS^WT or CTNS^G339R were lysed, and the samples were subjected
   to HA immunoprecipitation and immunoblotting for the indicated
   proteins. (e–i) Starved Ctns^WT mPTCs were exposed to CDME (0.1 mM) or
   with unmodified l-cystine (0.1 mM) for the indicated times. f Cystine
   levels in the Ctns mPTCs; n = 7 (time point 0) and n = 5 (time points
   0.5 h and 6 h) biologically independent samples. g Cystine levels in
   whole cell lysates and purified lysosomes; n = 2 biologically
   independent samples. h Immunoblotting of the indicated proteins; n = 3
   biologically independent samples. Quantification of the indicated
   proteins is shown in Supplementary Fig. [176]9g. i The cells were
   lysed, and the samples were subjected to HA immunoprecipitation and
   immunoblotting for the indicated proteins. All data in c, d, i were
   obtained from 2 independent biological replicates. Plots represent
   mean ± SEM. Statistics calculated by one-way ANOVA followed by Tukey’s
   or Sidak’s multiple comparisons test in b, f. Source data are provided
   as a Source Data file. Images in a, e were drawn using pictures from
   Servier Medical Art, which is licensed under a Creative Commons
   Attribution 3.0 Unported License
   ([177]https://creativecommons.org/licenses/by/3.0/).

Blocking mTORC1 rescues lysosome proteolysis and differentiation downstream
of cystine storage

   We next tested whether mTORC1 inhibition restores lysosomal catabolic
   activities and the differentiating state of PT cells lacking
   CTNS/accumulating cystine. We treated Ctns^WT and Ctns^KO mPTCs with
   low, non-toxic doses of Torin1 (250 nM for 16 h; Supplementary
   Fig. [178]10a). Treatment with Torin1, which fails to correct cystine
   storage (Fig. [179]4j) while effectively blocking mTORC1 activity
   (Fig. [180]6a, Supplementary Fig. [181]10b), enhanced the nuclear
   translocation of the transcription factor E3 (Tfe3), the master
   modulator of lysosome biogenesis and autophagy^[182]9, in Ctns^KO mPTCs
   compared with those receiving the vehicle (Supplementary
   Fig. [183]10c). In turn, the increased nuclear import heightened
   Tfe3-targeted autophagy and lysosomal genes, re-activating lysosome
   function and catabolic autophagy (Fig. [184]6b–d; Supplementary
   Fig. [185]10d–f). In agreement, treatment with Torin1 ameliorated the
   mitochondrial bioenergetics and lowered the overproduction of
   mitochondria-derived reactive oxygen species (ROS), preventing the
   disruption of tight junctions manifested in Ctns^KO mPTCs
   (Fig. [186]6e, f; Supplementary Fig. [187]10g, h). This markedly
   repressed the nuclear translocation of Ybx3 and its driven activation
   of growth/proliferation (Fig. [188]6g–i; Supplementary Fig. [189]10i),
   while restoring PT cell differentiation markers and the endocytic
   activities of the tubular epithelial cells downstream of cystine
   storage (Fig. [190]6j–l). Similar functional rescues, including
   lysosomal catabolic activities, autophagy, proliferation, and
   differentiation were also observed in Ctns^KO cells treated with low
   concentrations of the allosteric mTOR inhibitor rapamycin (250 nM for
   16 h; Supplementary Fig. [191]11) or depleted of Raptor (Supplementary
   Fig. [192]12)—a gene encoding protein that docks the mTOR protein
   kinase on the surface of lysosomes^[193]6. The modulation of mTORC1
   signaling by either pharmacological or genetic means did not alter the
   tubular cell differentiation nor the Lrp2-mediated endocytosis in
   Ctns^WT mPTCs (Fig. [194]6g, j, k, l; Supplementary Fig. [195]11h;
   Supplementary Fig. [196]12f). In a similar vein, the treatment of Ctns
   rats with low doses of Rapamycin^[197]31 (1.5 mg/kg/day for 2 weeks;
   Fig. [198]7a) reduced/normalized mTORC1 activity and lysosomal
   abnormalities (Fig. [199]7b; Supplementary Fig. [200]13a, b), inhibited
   the growth and proliferation (Fig. [201]7c) while restoring the
   expression of tubular differentiation markers and cuboidal tubule cell
   morphology (Fig. [202]7d–f; Supplementary Fig. [203]13c), and the
   Lrp2-mediated endocytosis in the PT segments of Ctns^KO rat kidneys
   (Fig. [204]7g) compared with those receiving the vehicle. These
   functional rescues occurred in the absence of significant changes in
   the rat kidney cystine levels (Supplementary Fig. [205]13d). Thus,
   genetic and pharmacological suppression of mTORC1 signaling pathway
   corrects the catabolic activities of lysosomes and differentiating
   states of PT cells downstream of cystine storage.

Fig. 6. Suppressing mTORC1 rescues lysosome proteolysis and differentiation
downstream of CTNS loss and cystine storage.

   [206]Fig. 6
   [207]Open in a new tab

   (a–l) mPTCs from Ctns^KO and Ctns^WT mice were treated with vehicle
   (DMSO) or Torin1 (250 nM for 16 h). (a, b) Immunoblotting and
   quantification of the indicated proteins; n = 3 (a) and 4 (b)
   biologically independent experiments. Quantification of the indicated
   proteins in a shown in Supplementary Fig. [208]10b. c Confocal
   microscopy and d quantification of PepA^+ puncta per cell; n = 360
   (Ctns^WT + DMSO); n = 361 (Ctns^WT + Torin1); n = 247 (Ctns^KO + DMSO)
   and n = 340 (Ctns^KO + Torin1) cells from 3 biologically independent
   animals. e Oxygen consumption rates (OCRs) were measured under basal
   level and after the sequential addition of oligomycin (Oligo, 1 μM),
   FCCP (0.5 μM), and Rotenone (ROT; 1 μM) + Antimycin A (ANT; 1 μM);
   n = 3 biologically independent animals. f Confocal microscopy and
   quantification of MitoSOX fluorescence intensity per cell; n = 42
   (Ctns^WT + DMSO); n = 68 (Ctns^WT + Torin1); n = 89 (Ctns^KO + DMSO)
   and n = 122 (Ctns^KO + Torin1) cells from 3 biologically independent
   animals. Representative images are shown in Supplementary
   Fig. [209]10g. (g, j) mRNA levels for the indicated genes in mPTCs; g
   n = 9 or j n = 4 wells from 3 biologically independent animals. h
   Quantification of the number of Pcna^+ nuclei; n = 10 non-overlapping
   fields of views from 3 biologically independent animals. Representative
   images are shown in Supplementary Fig. [210]10j. i Growth curves in the
   Ctns mPTCs; n = 3 biologically independent experiments. k Maximum
   intensity projection of image stacks and quantification of the primary
   cilium length (Ac-tubulin, green) in Ctns mPTCs; n = 132
   (Ctns^WT + DMSO), n = 176 (Ctns^WT + Torin1), n = 177 (Ctns^KO + DMSO),
   and n = 306 (Ctns^KO + Torin1) cells from  3 biologically independent
   animals. l Confocal microscopy and quantification of BSA^+ structures
   per cell; n = 175 (Ctns^WT + DMSO), n = 139 (Ctns^WT + Torin1), n = 252
   (Ctns^KO + DMSO), and n = 274 (Ctns^KO + Torin1) cells from  3
   biologically independent animals. Plots represent mean ± SEM.
   Statistics calculated by one-way ANOVA followed by Tukey’s or Sidak’s
   multiple comparisons test in b, d, f, g, h, j, k, l; and by unpaired
   two-tailed Student’s t test in e, i. Nuclei counterstained with DAPI
   (blue or grey). Scale bars, 10 μm. Source data are provided as a Source
   Data file.

Fig. 7. Rapamycin rescues lysosome and differentiation defects in the kidney
PT of Ctns-deficient rats.

   [211]Fig. 7
   [212]Open in a new tab

   a Experimental design and drug treatment strategy in Ctns rats. Images
   were drawn using pictures from Servier Medical Art, which is licensed
   under a Creative Commons Attribution 3.0 Unported License
   ([213]https://creativecommons.org/licenses/by/3.0/). Ctns rats at 12
   weeks of age were subcutaneously implanted with a pellet providing a
   long-term release of rapamycin (Rapa, 1.5 mg/kg/day). After 2 weeks of
   treatment, the kidneys were harvested and analysed. b Immunoblotting
   and quantification of the indicated proteins; n = 7 biologically
   independent animals. c Confocal microscopy and quantification of the
   number of Pcna^+ nuclei (white) in Aqp1-positive PT segments of Ctns
   rat kidneys; n = 4 biologically independent animals. d mRNA levels for
   the indicated genes in the kidneys of Ctns rats; n = 7 (Ctns^WT +
   Vehicle) and (Ctns^KO + Vehicle); and n = 9 (Ctns^WT + Rapa) and
   (Ctns^KO + Rapa) biologically independent animals. e Confocal
   microscopy and quantification of Vimentin (red) MFI in Aqp1-positive PT
   segments of Ctns rat kidneys (green); n = 113 (Ctns^WT + Vehicle);
   n = 90 (Ctns^WT + Rapa); n = 117 (Ctns^KO + Vehicle); n = 107 (Ctns^KO
   + Rapa) PT segments from 4 biologically independent animals. f The
   whole-kidney sections stained for PT apical marker (Aqp1, green) and
   quantification of PT cell height in the Ctns rat kidneys; n = 60
   (Ctns^WT + Vehicle); n = 79 (Ctns^WT + Rapa); n = 77 (Ctns^KO +
   Vehicle); and n = 79 (Ctns^KO + Rapa) PT segments from 4 biologically
   independent animals. g Confocal microscopy and quantification of Lrp2
   (red) MFI in Aqp1 (green)-positive PT segments of Ctns rat kidneys;
   n = 164 (Ctns^WT + Vehicle); n = 202 (Ctns^WT + Rapa); n = 246 (Ctns^KO
   + Vehicle); and n = 272 (Ctns^KO + Rapa) PT segments pooled from 4
   biologically independent animals. Plots represent mean ± SEM.
   Statistics calculated by one-way ANOVA followed by Tukey’s or Sidak’s
   multiple comparisons test in b–g. Scale bars are 50 μm in c and 20 μm
   in e–g. Source data are provided as a Source Data file.

Therapeutic mTORC1 inhibition improves PT dysfunction in a zebrafish model of
cystinosis

   Finally, to test whether mTORC1 inhibition rescues the PT
   dysfunction/LMW proteinuria in vivo, we treated ctns zebrafish larvae
   expressing a bona fide LMW biosensor for PT function^[214]32 (lfabp::½
   vdbp-mCherry; Fig. [215]8a) with low, non-toxic doses of rapamycin
   (200 nM for 9 days). The short-term treatment with rapamycin, which
   efficiently blunts mTORC1 activity (Fig. [216]8b), restored the
   lysosome-directed processing of the ultrafiltered LMW proteins (as
   testified by decreases in mCherry fluorescence intensity and vesicle
   size; Fig. [217]8c) and improved the PT function in the ctns^KO
   zebrafish pronephros, as indicated by the substantial decreases in
   LMWP/mCherry levels in the urine of treated ctns^KO zebrafish compared
   with those receiving the vehicle (Fig. [218]8d). The phenotype rescues
   occurred without changes in cystine levels in ctns-deficient zebrafish
   (Supplementary Fig. [219]13e). Taken together, these proof-of-concept
   studies indicate that modulation of mTORC1 signaling may serve as an
   attractive therapeutic strategy for treating dysregulated homeostasis
   and PT dysfunction in cystinosis.

Fig. 8. Rapamycin rescues PT dysfunction in a zebrafish model of cystinosis.

   [220]Fig. 8
   [221]Open in a new tab

   a Schematic showing the generation of ctns transgenic zebrafish line
   expressing a bona fide biosensor for LMW proteinuria/PT function and
   drug treatment strategy. Fate of ½vdbp-mCherry under normal conditions:
   ½vdbp-mCherry (~50 kDa) is (1) produced in liver, (2) secreted into the
   bloodstream and filtered by the glomerulus, (3) reabsorbed by PT cells
   through receptor-mediated endocytosis, and (4) processed by lysosomes.
   EE, early endosome; LE, late endosome; Lys, lysosome. Images were drawn
   using pictures from Servier Medical Art, which is licensed under a
   Creative Commons Attribution 3.0 Unported License
   ([222]https://creativecommons.org/licenses/by/3.0/). b–d 5dpf-ctns
   zebrafish larvae stably expressing ½vdbp-mCherry were treated with
   vehicle or rapamycin (Rapa, 200 nM for 9 days). b Immunoblotting and
   quantification of the indicated proteins, with each lane representing a
   pool of 6 zebrafish; n = 4 biologically independent experiments. c
   Multiphoton microscopy and quantification of ½vdbp-mCherry fluorescence
   intensity and vesicle area in zebrafish pronephros; n = 14
   vehicle-treated ctns^WT zebrafish; n = 11 vehicle-treated ctns^KO
   zebrafish; and n = 15 Rapa-treated ctns^KO zebrafish. d After 9 days of
   treatment, urine samples were harvested and analysed. Quantification of
   urinary mCherry levels by ELISA; n = 11 vehicle-treated ctns^WT
   zebrafish; n = 26 vehicle-treated ctns^KO zebrafish; and n = 39
   Rapa-treated ctns^KO zebrafish. Plots represent mean ± SEM. Statistics
   calculated by one-way ANOVA followed by Sidak’s and Tukey’s multiple
   comparisons test in b and in c, d. Scale bars, 10 μm. Source data are
   provided as a Source Data file.

Discussion

   Here, we demonstrate that cystine mobilization from lysosomes is
   crucial for maintaining the differentiation and endocytic function of
   PT cells. Lysosomal cystine storage, typically caused by the functional
   loss of CTNS, diverts the differentiating trajectories of PT cells
   towards anabolic programs for growth and proliferation, disrupting
   homeostasis and their reabsorptive properties—with relevance across the
   spectrum of health and disease. Mechanistically, cystine storage
   stimulates Ragulator-Rag GTPase-dependent recruitment of mTORC1 and its
   constitutive activation. In cells and preclinical models of CTNS loss,
   the modulation of mTORC1 pathway restores lysosome function and
   catabolic autophagy, ultimately steering the proliferative and growing
   states of the cells towards differentiation and physiological
   homeostasis (Fig. [223]9).

Fig. 9. Role of CTNS-cystine-mTORC1 axis as a signaling node that instructs
fate specification in the epithelial cells of the kidney tubule.

   [224]Fig. 9
   [225]Open in a new tab

   The lysosomal cystine mobilization through CTNS serves as lysosomal
   signal that shapes the response of mTORC1 to direct metabolism and fate
   decisions in the kidney tubule epithelium. In the tubular cells lacking
   CTNS, the storage of cystine triggers the Ragulator-Rag-dependent
   translocation of mTORC1 and its constitutive activation at the surface
   of the lysosome. This diverts the catabolic trajectories and
   differentiating states of the tubular cells towards growth and
   proliferation, disrupting their endocytic functions and causing tubular
   dysfunction and kidney disease development in cystinosis. Therapeutic
   inhibition of hyperactive mTORC1 pathway rescues the lysosome
   proteolysis and rewires the proliferating trajectories of CTNS
   deficient/cystinosis-affected tubular epithelial cells towards
   differentiation and physiological homeostasis. EE early endosome, RE
   recycling endosome, Lys lysosome. Images were drawn using pictures from
   Servier Medical Art, which is licensed under a Creative Commons
   Attribution 3.0 Unported License
   ([226]https://creativecommons.org/licenses/by/3.0/).

   Our work indicates that lysosomes accumulating cystine due to CTNS loss
   display profound defects in their catabolic activities, with marked
   accumulation of intracellular constituents that are normally dismantled
   through autophagy^[227]33. The proteolytic impairment of CTNS-deficient
   lysosomes occurs through mechanisms that are independent of lysosomal
   acidification. In lysosomes, the maintenance of cysteine levels appears
   to promote the catabolic activities of the lysosomal proteases^[228]34.
   Thus, an abnormal storage of cystine (the oxidized dimer of cysteine)
   may fuel the oxidation of lysosomal thiols that preserve the folding of
   the catalytic sites of lysosomal proteases, impairing their degradative
   functions and organelle homeostasis. These catabolic anomalies are
   paralleled by the upregulation of anabolic programs for growth and
   proliferation while suppressing the differentiating states of the PT
   cells, hence leading to defective receptor-mediated endocytosis and
   tubular proteinuria, i.e., the earliest manifestation of the epithelial
   cell disease^[229]5. Thus, beyond its role in cystine transport, CTNS
   might act as a sensor of the lysosome function—a gatekeeper of
   metabolism and homeostasis in the epithelial cells lining the kidney
   tubule.

   How do CTNS-deficient/cystine accumulating lysosomes affect cell fate
   specialization and differentiation programs? Our integrative
   multi-omics analyses and AI/ML-target discovery pipeline, combined with
   cross-species validation, indicate that lysosomal cystine storage
   constitutively activates mTORC1 signaling in CTNS-deficient PT cells.
   These findings are in line with previous studies showing hyperactive
   mTORC1 pathway in other lysosomal storage diseases characterized by the
   abnormal accumulation of intracellular metabolites, such as cholesterol
   in Niemann Pick type C (NPC) and glycosaminoglycans in
   Mucopolysaccharidosis type VII^[230]9,[231]35–[232]37. Conversely,
   mTORC1 activity seems to be reduced/unchanged in transformed,
   proliferative tubular cells derived from cystinosis patients and in
   patient-derived iPSCs^[233]12,[234]38,[235]39. Such discrepancies may
   be related to immortalization protocols, which reduce per se the
   phosphorylation rates of canonical mTORC1 substrates^[236]40, or to use
   of dedifferentiated cellular systems that do not recapitulate the
   characteristics of native kidney tubular cells.

   The catabolic activities of the lysosome and the levels of mTORC1
   activity have a crucial role in the maintenance of autophagy. Our
   time-course studies in Ctns mice suggest that defects in lysosome and
   abnormal (mTORC1-driven) nutrient signaling pathways, which appear
   early in the course of the disease, may disrupt the autophagy-mediated
   turnover of dysfunctional and/or damaged mitochondria, in turn
   destroying metabolism and inactivating the actions of fate-determining
   transcriptional programs^[237]41. Because CTNS loss and the resulting
   cystine storage induce concomitant changes in the expression of genes
   associated with PT cell function and differentiation, as well as in the
   primary cilium, additional layers of (dys)regulation are likely
   involved.

   In addition to the regulation of glutathione synthesis^[238]42,
   exporting lysosomal cystine through CTNS may antagonize the
   reactivation of mTORC1 through the incorporation of carbons into the
   tricarboxylic acid cycle (TCA) cycle and limiting amino acid
   synthesis^[239]13. In line, our integrated multi-omics analyses have
   revealed an enrichment of intermediates of TCA in the tubular cells of
   Ctns^KO mouse kidneys, with accumulation of dysfunctional and/or
   damaged mitochondria and excessive oxidative stress. The tight
   integration of lysosome and mitochondrial quality control systems
   suggests that, in diseases driven by lysosome dysfunction, their
   aberrant crosstalk may synergize to disrupt (mTORC1-driven) regulation
   of metabolic signaling and the functions of specialized cell types.

   Despite the metabolic changes and signaling alterations, the Ctns^KO
   mice present a late onset and high variability in the extent of tubular
   dysfunction, mild structural changes, and no kidney
   failure^[240]4,[241]15,[242]43, in contrast with the Ctns^KO rat model
   that closely recapitulates the human situation in terms of timing,
   severity, and multisystem involvement^[243]22. The molecular basis of
   this inter-species difference may involve distinct levels of cystine
   storage in the lysosomes, reflecting specific transport or metabolic
   activities.

   The mTORC1 hyperactivation evidenced in CTNS-deficient mPTCs and in the
   three model organisms, together with time-course analysis, suggests
   that CTNS tailors the activation of mTORC1 signaling in response to
   lysosomal cystine levels. Treatment of CTNS-defective mPTCs with
   cysteamine, which facilitates the export of cystine from lysosomes
   through the cationic amino acid PQ-loop repeat-containing protein 2
   (PQLC2; ref. ^[244]5) or the exposure of wild-type mPTCs to CDME, which
   accumulates within lysosomes independently of CTNS, reduces and
   stimulates, respectively, mTORC1 activation at the lysosomal surface.
   Re-introduction of CTNS^WT, which reinstates the lysosomal cystine
   efflux compared to disease-causing mutant CTNS^G339R, successfully
   restored the sensitivity of mTORC1 to changes in nutrient availability
   in CTNS-defective PT cells. Conversely, CDME-induced cystine storage
   within lysosomes of wild-type PT cells, rendered mTORC1 insensitive to
   nutrient withdrawal. Our co-immunoprecipitation studies in Ctns^WT
   cells exposed to CDME or in Ctns^KO cells expressing CTNS^G339R mutant
   indicate that, in both contexts, increased levels of cystine enable the
   interaction of CTNS with components of V-ATPase and Ragulator-Rag
   GTPase scaffold complex that recruits mTORC1 at the surface of the
   lysosome. This suggests the possibility of conformational changes that
   allows CTNS to signal cystine sufficiency to mTORC1 via the
   Ragulator-Rag GTPase complex^[245]44. Thus, CTNS may function as a
   transceptor (e.g., transporter with a receptor-like function) involved
   not only in the mere transport of cystine but also in which the amino
   acid engagement is used for allosteric signal transduction, as recently
   described for another nutrient transporter systems^[246]45–[247]47,
   including the human solute transporter SLC38A9.

   Our study offers therapeutic targets to rewire physiological
   homeostasis in cystinosis, a life-threatening disease for which
   cysteamine has remained the standard of care despite major side
   effects^[248]5,[249]28. The inhibition of hyperactive mTORC1 signaling
   and disturbed homeostasis in diseased Ctns^KO kidneys may enhance
   Tfe3-mediated re-activation of cellular programs boosting lysosomes and
   catabolic autophagy, removing damaged (ROS-overproducing) mitochondria,
   stabilizing tight junctions, and repressing the Ybx3 signaling cascade
   that promotes dedifferentiation and growth/proliferation. However, the
   translatability of mTORC1 inhibitors is limited by poor availability,
   lack of specificity, and toxicity^[250]6. The rescue of cellular and
   functional defects observed with the use of rapamycin^[251]48 suggests
   therapeutic value for modulators of mTORC1 signaling in cystinosis.
   Given that dietary restriction of nutrients (e.g., low protein diet
   and/or reductions of essential and/or branched amino acids, including
   cystine) promote homeostasis by switching mTORC1 activity off^[252]49,
   combinations of nutritional interventions with existing and/or emerging
   drugs might have translatable potential in cystinosis and other
   currently intractable diseases.

   In summary, our work describes a fundamental role of
   CTNS-cystine-mTORC1 signaling node as a guardian of cellular
   homeostasis and identifies an evolutionarily conserved lysosome-based
   signal that instructs fate specification in the epithelial cells of the
   kidney tubule. Modulating this pathway may yield therapeutic strategies
   for cystinosis and other lysosomal disorders, while creating
   opportunities to regulate homeostasis in specialized cell types.

Methods

Ethical compliance

   All the animal experiments were approved by the Cantonal Veterinary
   Office Zurich (license no. ZH195/2020, ZH139/2022, and ZH230/2019) and
   performed in accordance with the guidelines of the University of
   Zurich.

Antibodies and reagents

   The following antibodies were used in this study: anti-LAMP1 (Santa
   Cruz Biotechnology, sc-19992, 1:1000); anti-LC3 (PM036, MBL, 1:200);
   anti-p62/SQSTM1 (PM045, MBL, 1:200); anti-phospho-S6 Ribosomal Protein
   (Ser235/236; Cell signaling technology, 4858, 1:500), anti-cathepsin D
   (Santa Cruz Biotechnology, sc-6486, 1:500), anti-S6 Ribosomal Protein
   (Cell Signaling Technology, 2217, 1:500), anti-phospho‒4E-BP1 (Ser65;
   Cell signaling technology, 9451, 1:500), anti-4E-BP1 (Cell Signaling
   Technology, 9644, 1:500), anti-phospho-p70 S6 kinase 1 (Thr389; Cell
   Signaling Technology, 9234, 1:500); anti-p70 S6 kinase 1 (Cell
   Signaling Technology, 2708, 1:500), anti-acetyl‒CoA carboxylase (Cell
   signaling, 3662, 1:500); anti-phospho‒acetyl‒CoA carboxylase (Ser79;
   Cell signaling, 3661, 1:500); anti-AMPKα (Cell signaling, 2793, 1:500);
   anti-phospho‒AMPKα (Thr172; Cell signaling, 2535, 1:500), anti-TFE3
   (Sigma, HPA023881, 1:500); anti-mTOR (Cell Signaling Technology, 7C10,
   1:400 for immunofluorescence studies), anti-Raptor (2280, Cell
   Signaling Technology, 2280, 1:400); anti−RagC (Cell Signaling
   Technology, 5466, 1:500); rabbit anti-phospho-AKT (Ser473; Cell
   Signaling Technology, 13038, 1:500), anti-AQP1(Aviva Systems Biology,
   OASA00210, 1:400), anti-PCNA (Dako, M0879, 1:500), anti-α-tubulin
   (Sigma-Aldrich, T5168; 1:1000), anti-Cyclin A2 (Abcam, [253]Ab181591;
   1:400), anti Cdc20 (Santa Cruz, sc-13162, 1:400), anti-Golgin-97 (Cell
   Signaling Technology, 13192; 1:400); anti-VDAC (Cell Signaling
   Technology; 4866, 1:400), anti-PMP70 (Sigma, SAB4200181, 1:500),
   anti-calreticulin (Cell Signaling Technology, 12238, 1:400),
   anti-β-actin (Sigma-Aldrich, A5441, 1:1000), anti-γ-tubulin
   (Sigma-Aldrich, clone GTU-88,1:10000), and anti-HA (Roche, 11867423001;
   1:500), anti-acetylated-tubulin (Sigma,T7451; 1:400), Anti-Laminin
   gamma 1 (clone A5; Thermo Fisher Scientific, [254]MA106100, 1:400),
   Atp6v0d1 (Thermo Fisher Scientific, PA5-103179, 1:200) ATP6V1B2 (D2F9R,
   Cell signaling,14617, 1:200), anti-vimentin (Abcam, Ab92547, 1:400),
   anti-Lipoclain2/NGAL (R&D systems, AF1857, 1:400), anti-cleaved caspase
   3 (Cell signaling), 9662, 1:400), anti-caspase 3 (Cell signaling, 9661,
   1:400), anti-megalin (1:1000) was kindly provided by P. Verroust and R.
   Kozyraki (INSERM, Paris, France), donkey-anti rabbit IgG,
   cross-absorbed secondary antibody, Alexa Fluor 488 (Thermo Fischer
   Scientific, A10040, 1:400), donkey-anti mouse IgG, cross-absorbed
   secondary antibody, Alexa Fluor 488 (Thermo Fischer Scientific, A21202,
   1:400), donkey-anti rat IgG, cross-absorbed secondary antibody, Alexa
   Fluor 488 (Thermo Fischer Scientific, A21202, 1:400), donkey-anti rat
   IgG, cross-absorbed secondary antibody, Alexa Fluor 647 (Thermo Fischer
   Scientific, [255]A31573, 1:400), donkey-anti goat IgG, cross-absorbed
   secondary antibody, Alexa Fluor 633 (Thermo Fischer Scientific, A21094,
   1:400), donkey-anti goat IgG, cross-absorbed secondary antibody, Alexa
   Fluor 546 (Thermo Fischer Scientific, A11056, 1:400), donkey-anti rat
   IgG, cross-absorbed secondary antibody, Alexa Fluor 546 (Thermo Fischer
   Scientific, A11081, 1:400). Compounds included Bafilomycin A1 (BfnA1;
   Enzo Life Sciences, ALX-380-030, 250 nM), ATP-competitive mTOR
   inhibitor Torin1 (TOCRIS Bioscience, CAS 1222998-36-8, 250 nM),
   allosteric mTOR inhibitor rapamycin (Sigma-Aldrich, R0395-1MG, 250 nM
   in cell treatment; and MedChemExpress HY-10219-1G), Cysteamine
   (Sigma-Aldrich, M6500-25G, 100μM), alcohol ester derivates of cystine
   (CDME, Sigma-Aldrich 857327-5 G, 0.05 mM), L-amino acid cystine
   (L-cystine, Sigma-Aldrich, C7602−25 G, 0.05 mM), L-cysteine (Sigma,
   C7352-25G, 1 mM), N-Acetyl-L-cysteine (NAC, Sigma, A9165-5GNAC, 1 mM),
   LysoTracker (Thermo Fisher Scientific, [256]L12492), and the
   biotinylated Lotus Tetragonolobus Lectin (LTL; Vector Laboratories,
   B−1325). The pCAG R2pH-LAMP1-3xFLAG plasmid was a gift from
   Massimiliano Stagi (Addgene plasmid # 157940;
   [257]http://n2t.net/addgene:157940; RRID: Addgene_157940). The cells
   used in this study were negatively tested for mycoplasma contamination
   using MycoAlert™ Mycoplasma Detection Kit (LT07-118, Lonza,
   Switzerland).

Generation and maintenance of ctns zebrafish

   The ctns-specific left and right TALENs (ctns-TALENs) were constructed
   in according to Golden Gate TALEN assembly protocol and using the
   Golden Gate TALEN and TAL Effector Kit 2.0 (Addgene, Kit #1000000024).
   CIscript-GoldyTALEN was a gift from Daniel Carlson & Stephen Ekker
   (Addgene, plasmid # 38142). TALENs were designed with the TAL Effector
   Nucleotide Targeter 2.0 software on the Website of Cornell University.
   The TALENs target the exon 3 of ctns zebrafish gene: left TALEN-F:
   TCTTTTAATCCTTTGTGTTCACA and right TALEN-R: CATCTGTAACGGTTTATTTCAAT. The
   spacer between two TALEN target sites is ~15 nucleotides and has an
   AciI restriction site in the middle, which is used for mutant
   screening. The TALEN expression plasmids were linearized with BamHI and
   then used for in vitro transcription (mMESSAGE mMACHINE T3 kit,
   Ambion). Approximately 1 nL of TALEN messenger ribonucleic acid (mRNAs,
   400 ng/µL) was injected into one-cell stage zebrafish (Danio rerio)
   embryos. After 24 h, genomic DNA was extracted from injected embryos
   with normal appearance. Targeted genomic loci were amplified by using
   primers designed to anneal approximately 240 base pairs, and mutant
   allele was detected by AciI digestion of PCR product. The TALEN
   injected embryos were raised to adulthood (F0) and outcrossed with
   wild-type zebrafish (AB background). The embryos were then raised to
   adulthood (F1) for screening of heterozygous carriers. We identified a
   heterozygous carrier harboring ctns^+/del8 mutation and (F1)
   generations were crossed for obtaining homozygous mutant carrying
   ctns^del8/del8. Zebrafish larvae (AB background) were kept at day/night
   cycle of 14/10 h at 28 °C and fed with a standard diet (Zebrafeed
   100-200, SPAROS, Portugal) starting at 5 d.p.f. Most of the experiments
   were performed on larvae at 5–14 d.p.f. Animal sex is not specified at
   this developmental stage. Stable zebrafish line expressing
   mCherry-tagged half vitamin D binding protein (½vdbp-mCherry) in the
   liver was established and outcrossed with ctns^+/del8 zebrafish to
   generate transgenic mutant line, which was crossed with ctns^+/del8
   zebrafish to produce homozygous larvae, and the measurement of
   ½vdbp-mCherry was assessed by ELISA. Where indicated, zebrafish larvae
   were treated at 5 dpf with zebrafish facility system water containing
   DMSO or rapamycin (200 nM) for 9 days.

Zebrafish urine collection and ELISA mCherry measurement

   Zebrafish larvae (AB background) were placed in a 48-well microplate
   with one larva/500 μL facility water/well and kept at 28 °C for
   16 hours, followed by urine collection for ELISA assay^[258]32. The
   levels of ½vdbp-mCherry were measured according to the manufacturer’s
   protocol. Briefly, 50 μL of fish pool water containing urine was
   distributed to a 96-well microplate pre-coated with anti-mCherry
   antibody. A 50 μL mixture of capture antibody and detector antibody was
   added to each well and incubated at room temperature for 1 hour. The
   wells were rinsed three times with washing buffer and incubated with
   100 μL of TMB development solution at room temperature for 10 minutes.
   The reaction was stopped by adding 100 μL of stop solution, followed by
   reading the absorbance of each well at 450 nm.

Rodent models

   Experiments were conducted on age- and gender-matched Ctns knockout
   mouse (C57BL/6 background^[259]4) or rat (Sprague-Dawley
   background^[260]22) lines, and their corresponding control littermates.
   Due to an effect of sex on cystine content (higher levels in female
   versus male kidneys mice^[261]5,[262]15,[263]43), only female mice at
   age 6, 12, and 24 weeks were used for primary cell cultures and
   mechanistic studies. Rats and mice were housed under specific pathogen
   free conditions and maintained under temperature (22–25 °C)- and
   humidity (50–60%)-controlled conditions with 12-h dark/light cycle with
   water and food provided ad libitum. Kidneys were collected for analyses
   at the time of sacrifice. Both male and female rats aged 12 weeks were
   subcutaneously implanted with a slow-release pellet (Innovative
   Research of America, FL, USA). The pellets were designed to allow a
   constant release of rapamycin (1.5 mg/kg B.W./day; R-5000, LC
   Laboratories, Woburn, MA). Control animals received a placebo pellet
   equivalent to their respective rapamycin-treated group. After 2 weeks
   of treatment, the rats were sacrificed, and the kidney tissues were
   harvested and collected for subsequent analyses.

Kidney function

   The mice were placed overnight in metabolic cages with ad libitum
   access to food and drinking water; urine was collected on ice, body
   weight, water intake, and diuresis were measured. Blood (from
   sublingual vein) was obtained after anesthesia with ketamine/xylazine
   or isoflurane. Urine and blood parameters were measured using UniCel
   DxC 800 pro Synchron (Beckman Coulter, Fullerton, CA, USA). Urinary
   levels of low-molecular weight Clara cell protein (CC16) were measured
   by using an enzyme-linked immunosorbent assay in according to the
   manufacturer’s instructions (BIOMATIK [264]EKU03200, Thermo Fischer
   Scientific, Waltham, MA).

Primary cultures of mouse proximal tubular cells

   The kidneys were harvested from Ctns knockout mice and from their
   corresponding control littermates: one kidney was transversally split,
   and one half was fixed and processed for immunostaining while the other
   half was flash-frozen, homogenized by Dounce homogenizer in 1 mL of
   RIPA buffer that contains protease and phosphatase inhibitors and
   processed for western blot analysis. The contralateral kidney was taken
   to generate primary cultures of mPTCs^[265]4. Freshly microdissected PT
   segments were seeded onto collagen-coated chamber slides (C7182,
   Sigma-Aldrich) and/or collagen coated 6- or 24-well plates (145380 or
   142475, Thermo Fisher Scientific), and cultured at 37 °C and 5% CO[2]
   in DMEM/F12 (21041-025, Thermo Fisher Scientific) with 0.5% dialyzed
   fetal bovine serum (FBS), 15 mM HEPES (H0887, Sigma-Aldrich), 0.55 mM
   sodium pyruvate (P2256, Sigma Aldrich), 0.1 ml L^−1 non-essential amino
   acids (M7145, Sigma Aldrich), hydrocortisone, human EGF, epinephrine,
   insulin, triiodothyronine, TF, and gentamicin/amphotericin (Single
   Quots® kit, CC‒4127, Lonza), pH 7.40, 325 mOsm kg^−1. The medium was
   replaced every 48 h. Confluent monolayers of mPTCs were expanded from
   the tubular fragments after 6–7 days, characterized by a high endocytic
   uptake capacity. These cells were negatively tested for mycoplasma
   contamination.

Cystine measurements

   Mouse and/or rat kidney tissue, and/or zebrafish larvae and/or cultured
   cells were homogenized and lysed with N-ethylmaleimide (NEM) solution
   containing 5.2 mmol l^−1 N-ethylmaleide in 10 mmol L^−1 potassium
   phosphate buffer adjusted to pH 7.4. The lysates were harvested,
   collected and precipitated with sulfosalicylic acid (12% w/v) and
   centrifuged at 10,000 × g for 10 min at 10 °C. The resulting
   supernatant was dissolved in citrate loading buffer (Biochrom Ltd,
   Cambridge, UK) and 50 μL of this solution was analysed by Biochrom 30
   Plus Amino Acid Analyzer (Biochrom Ltd). The protein pellet was
   dissolved in 0.1 mol L^−1 NaOH solution and the protein concentration
   was determined by Biuret method. The concentration of amino acids was
   measured by using a lithium high-performance physiological column
   followed by post-column derivatization with ninhydrin. The amino acids
   were identified according to the retention time and the ratio of the
   area between the two wavelengths (570 nm and 440 nm) and quantified by
   using EZChrom Elite software (Agilent Technologies Inc., Pleasanton,
   California, USA). Cystine concentration was normalized to the protein
   concentration and reported in nmol per mg protein.

Nutrient starvation protocols and cell treatments

   The withdrawal of serum, glucose, and amino acid was performed by
   washing mPTCs with Hank’s balanced salt solution (55021 C,
   Sigma-Aldrich) and placing them in normal growth and nutrient-deprived
   medium (amino acid-free RPMI, starvation medium) for the indicated
   times. For amino acid starvation and stimulation experiments, the
   cultured cells were rinsed twice with and incubated for 2 h starvation
   medium or incubated in starvation medium and then stimulated with
   acid-free RPMI supplemented for 60 minutes with a standard amino acid
   mixture composed of MEM non-essential amino acid solution, MEM
   essential amino acid solution, and L-glutamine (Invitrogen, Thermo
   Fisher Scientific) and with 0.5% fetal bovine serum that had been
   dialyzed against phosphate-buffered saline (PBS) in dialysis cassettes
   (Thermo Scientific) having an 3500 molecular weight cutoff. Where
   indicated, lysosomal proteolysis was inhibited by the addition of BfnA1
   (250 nM in cell culture medium for 2 h and/or 4 h). Where indicated,
   the cells were treated with either the allosteric mTORC1 inhibitor
   Rapamycin (250 nM in cell culture medium for 16 h) or the
   ATP-competitive mTORC1 inhibitor Torin1 (250 nM in cell culture medium
   for 16 h) or cysteamine (100 μM in cell culture medium for 16 h). Where
   indicated, mPTCs were exposed to alcohol ester derivates of cystine
   (CDME, 0.1 mM) and L-amino acid cystine (0.1 mM) for the indicated time
   points. Afterwards, the cells were processed and analysed as described
   below.

Adenovirus transduction

   For RNA interference studies, the adenovirus constructs include
   scrambled short hairpin (Scmb) or shRNAs targeting mouse Raptor. For
   expression studies, the adenoviral constructs include CMV (control
   vector, Ad-CMV-GFP, Vector Biolabs) or an individually carrying mouse
   hemagglutinin (HA) tagged-Ctns; or carrying mouse HA tagged-mutant Ctns
   (Gly339Arg; CTNS^G339R); or 3×HA-tagged TMEM192 or 3 × HA-tagged
   R2pH-LAMP1; or expressing mouse green fluorescent protein
   (GFP)-tagged-Map1lc3b or expressing mouse red fluorescent protein
   (RFP)-tagged-Lamp1. The adenoviral constructs were purchased from
   Vector Biolabs (University City Science Center, Philadelphia, USA). The
   cells were plated onto collagen-coated chamber slides and/or 6- and/or
   24-well tissue culture plates. 24 h after plating, adenovirus
   transduction was performed by incubating the cells (with approximately
   70-80% confluence) for 16 h at 37 °C. The cells were then challenged
   with fresh culture medium every 2 days and cultured for 2 days for the
   expression studies or 5 days for RNA interference studies, harvested
   and collected for subsequent analyses.

Microarray and bioinformatics analysis

   The Affymetrix (GeneChip mouse genome 430 A 2.0 array) hybridization
   experiments were performed in triplicate at the Coriell Genotyping and
   Microarray Center (Coriell Institute for Medical Research, Camden, New
   Jersey, USA) on total RNA extracted from microdissected proximal
   tubules of Ctns knockout mouse kidneys and controls. Normalized
   expression levels were generated from the raw intensities using the RMA
   method implemented in oligo Bioconductor package^[266]50.
   Differentially expressed probes between conditions (Ctns knockout
   versus controls) were identified using the Bioconductor package
   limma^[267]50. P values were corrected for multiple testing using the
   Benjamin-Hochberg method. The probes were annotated using the
   mouse430a2.db Bioconductor Annotation package. Pathway enrichment
   analysis was performed using the function enrichPathway from the
   Bioconductor package ReactomePA, run with the default parameters,
   selecting “mouse” as organism, and using all the genes on the
   microarray as universe. The analysis was performed separately for the
   upregulated and the down-regulated genes. The modules associated with
   mouse cystinosin gene were determined using GeneBridge toolkit^[268]19
   ([269]www.systems-genetics.org).

Bulk RNA sequencing

   Total RNA was isolated from kidney cortex samples from Ctns mouse
   (n = 4 animals in each experimental group) and rat (n = 8 animals in
   each experimental group) models using the Rneasy Plus Mini kit. RNA-seq
   libraries were prepared using the TruSeq Stranded mRNA-seq reagents
   (Illumina) using 100 ng of total RNA following the protocol provided by
   the supplier (Genetic Diagnostic and Sequencing Services, Tubingen,
   Germany). The quality of the total RNA and the RNA-seq libraries was
   assessed on Fragment Analyzer (Agilent). The libraries were sequenced
   on Illumina NovaSeq6000 using the 100-nucleotide paired-end-run
   configuration following the protocol provided by the supplier.

Proteomics

Sample preparation

   Cell pellets (~5 × 10^5 cells; n = 3 mice in each experimental group)
   were solubilized in 100 μL of lysis buffer, treated with High Intensity
   Focused Ultrasound (HIFU) for 1 minute at an ultrasonic amplitude of
   85%, and boiled at 95 °C for 10 minutes. Proteins were extracted using
   a tissue homogenizer (Tissue Lyser II, QUIAGEN) for 2 cycles of
   2 minutes at 30 Hz, boiled at 95 °C for another 10 minutes, and
   subjected to a last cycle of 1 min of HIFU. The samples were
   centrifuged at 20,000 × g for 10 min, and the protein concentration was
   determined using a Lunatic (Unchained Labs) instrument. For each
   sample, 50 µg of protein were taken. The samples were digested by
   adding 10 μL of the ‘Digest’ solution. After 60 min of incubation at
   37 °C the digestion was stopped with 100 μL of Stop solution. The
   solution was transferred to the cartridge and were removed by
   centrifugation at 3800 × g, while the peptides were retained by the
   iST-filter. Finally, the peptides were washed, eluted, dried and
   re-solubilized in 40 μL of 3% acetonitrile, 0.1% FA. 1 µl of iRT
   peptides (Biognosys) at 1:100 dilution was added to each sample.

Liquid chromatography-mass spectrometry analysis

   Mass spectrometry analysis was performed on an Orbitrap Fusion Lumos
   (Thermo Scientific) equipped with a Digital PicoView source (New
   Objective) and coupled to a M-Class UPLC (Waters). Solvent composition
   at the two channels was 0.1% formic acid for channel A and 0.1% formic
   acid, 99.9% acetonitrile for channel B. For each sample 1 μL of
   peptides were loaded on a commercial MZ Symmetry C18 Trap Column
   (100 Å, 5 µm, 180 µm × 20 mm, Waters) followed by nanoEase MZ C18 HSS
   T3 Column (100 Å, 1.8 µm, 75 µm × 250 mm, Waters). The peptides were
   eluted at a flow rate of 300 nL min^−1 by a gradient from 5 to 22% B in
   80 min and 32% B in 10 min after an initial hold at 5% B for 3 min. The
   column was washed with 95% B for 10 min and afterwards the column was
   re-equilibrated to starting conditions for additional 10 min. Samples
   were acquired in a randomized order. The mass spectrometer was operated
   in data-dependent mode (DDA) acquiring a full-scan MS spectra
   (300–1500 m/z) at a resolution of 120,000 at 200 m/z after accumulation
   to a target value of 500,000. Data-dependent MS/MS were recorded in the
   linear ion trap using quadrupole isolation with a window of 0.8 Da and
   HCD fragmentation with 35% fragmentation energy. The ion trap was
   operated in rapid scan mode with a target value of 10,000 and a maximum
   injection time of 50 ms. Only precursors with intensity above 5000 were
   selected for MS/MS and the maximum cycle time was set to 3 s. Charge
   state screening was enabled. Singly, unassigned, and charge states
   higher than seven were rejected. Precursor masses previously selected
   for MS/MS measurement were excluded from further selection for 20 s,
   and the exclusion window was set at 10 ppm. The samples were acquired
   using internal lock mass calibration on m/z 371.1012 and 445.1200. The
   mass spectrometry proteomics data were handled using the local
   laboratory information management system (LIMS).

Protein identification and label-free protein quantification

   The acquired raw MS data were processed by MaxQuant (version 1.6.2.3),
   followed by protein identification using the integrated Andromeda
   search engine. Spectra were searched against a Swissprot canonical
   mouse proteome (version from 2019-07-09), concatenated to its reversed
   decoyed fasta database and common protein contaminants.
   Carbamidomethylation of cysteine was set as fixed modification, while
   methionine oxidation and N-terminal protein acetylation were set as
   variable. Enzyme specificity was set to trypsin/P allowing a minimal
   peptide length of seven amino acids and a maximum of two missed
   cleavages. MaxQuant Orbitrap default search settings were used. The
   maximum false discovery rate (FDR) was set to 0.01 for peptides and
   0.05 for proteins. Label-free quantification was enabled and a 2-minute
   window for match between runs was applied. In the MaxQuant experimental
   design template, each file is kept separate in the experimental design
   to obtain individual quantitative values. Protein fold changes were
   computed based on Intensity values reported in the proteinGroups.txt
   file. A set of functions implemented in the R package SRM Service was
   used to filter for proteins with 2 or more peptides allowing for a
   maximum of three missing values, and to normalize the data with a
   modified robust z-score transformation and to compute p values using
   the t test with pooled variance. If all measurements of a protein are
   missing in one of the conditions, a pseudo fold change was computed
   replacing the missing group average by the mean of 10% smallest protein
   intensities in that condition. Pathway analysis using NetworkAnalyst
   ([270]http://www.networkanalyst.ca) with the default setting was
   performed to the proteins significantly changed in the mPTCs derived
   from Ctns mouse kidneys.

Metabolomics

Sample preparation

   Cell pellets (~5 × 10^5 cells; n = 3 mice in each experimental group)
   were lysed and extracted with 500 μL lysis solution (methanol: water
   4:1, v/v) by vortexing for 30 min at 4 °C. Precipitated proteins were
   pelleted by centrifugation (16,000 × g, for 15 min at 4 °C) and 50uL of
   the supernatants (50 μL) were transferred to a clean Eppendorf vial.
   The transferred aliquot of the supernatants was kept at 35 °C and dried
   down under a gentle flow of nitrogen. The dried extracts were
   reconstituted in 20 μL water and 80 μL injection buffer. 50 μL of the
   reconstituted extract was transferred to a glass vial with narrowed
   bottom (Total Recovery Vials, Waters) for LC-MS injection. In addition,
   method blanks, QC standards, and pooled samples were prepared in the
   same way to serve as quality controls for the measurement. Injection
   buffer was composed of 90 parts of acetonitrile, 9 parts of methanol,
   and 1 part of 5 M ammonium acetate.

Liquid chromatography/mass spectrometry analysis

   Metabolites were separated on a nanoAcquity UPLC (Waters) equipped with
   a BEH Amide capillary column (150 μm × 50 mm, 1.7 μm particle size,
   Waters), applying a gradient of 5 mM ammonium acetate in water (A) and
   5 mM ammonium acetate in acetonitrile (B) from 5% A to 50% A for
   12 min. The injection volume was 1 μL. The flow rate was adjusted over
   the gradient from 3 to 2 μl/min. The UPLC was coupled to Synapt G2Si
   mass spectrometer (Waters) by a nanoESI source. MS1 (molecular ion) and
   MS2 (fragment) data were acquired using negative polarization and MS^E
   over a mass range of 50 to 1200 m/z at MS1 and MS2 resolution of
   >20,000.

Untargeted metabolomics data analysis

   Metabolomics data sets were evaluated in an untargeted fashion with
   Progenesis QI software (Nonlinear Dynamics, Waters), which aligns the
   ion intensity maps based on a reference data set, followed by a peak
   picking on an aggregated ion intensity map. Detected ions were
   identified based on accurate mass, detected adduct patterns and isotope
   patterns by comparing with entries in the Human Metabolome Data Base
   (HMDB). A mass accuracy tolerance of 5mDa was set for the searches.
   Fragmentation patterns were considered for the identifications of
   metabolites.

   All biological samples were analysed in triplicate and quality controls
   were run on pooled samples and reference compound mixtures to determine
   technical accuracy and stability. Pathway analysis using MetaboAnalyst
   ([271]http://www.metaboanalyst.ca) with the default setting was
   performed for the metabolites significantly changed in mPTCs derived
   from wild-type and Ctns^KO mouse kidneys. The ranked list of
   significantly changed proteins and metabolites had been crossed each
   other, and the enrichment pathway analysis was performed using the open
   source OmicsNet 2.0 software^[272]51 ([273]http://www.omicsnet.ca).

PandaOmics TargetID platform for target identification

   In silico-based PandaOmics target discovery/scoring approach^[274]52
   was applied to identify novel molecular targets for cystinosis. This
   approach is based on the combination of multiple scores derived from
   text and omics data. Text-based scores are derived from various sources
   including scientific publications, grants, patents, clinical trials,
   and the key opinion leaders, and thus represent how strongly a
   particular target is associated with a disease. Specifically,
   text-based scores contain Attention, Trend, Attention Spike, Evidence,
   Grant funding, Funding per Publication, Grant Size, Average Hirsch,
   Impact Factor, and Credibility attention index scores. In contrast,
   omics scores are based on the differential expression, GWAS studies,
   somatic and germline mutations, interactome topology, signaling pathway
   perturbation analysis algorithms, knockout/overexpression experiments,
   and omics-data sources, and thus represent the target-diseases
   association according to molecular connections between proposed target
   and disease of interest. Omics scores include thirteen models
   (Heterogeneous Graph Walk, Matrix Factorization, Interactome Community,
   Causal inference, Overexpression/Knockout, Mutated/Disease Submodules,
   Mutations, Pathway, Network Neighbors, Relevance, and Expression) that
   can be subdivided into classic bioinformatics approaches and complex
   AI-based models. For example, the Expression score relies on the
   combination of each gene’s fold change difference in disease versus
   control samples, the statistical significance of this change, and basal
   expression in the disease-relevant tissue. On the other hand, AI-based
   omics model called Heterogeneous Graph Walk (HeroWalk) is a guided
   random walk-based approach that is applied to a heterogeneous graph.
   The model learns node representations and then identifies gene nodes,
   which are close to the reference disease node. The “walks” are sampled
   with a predefined metapath, i.e., fixed sequence of node types in a
   walk, e.g., “gene”-“disease”-“gene”. The node degree controls the
   probability of transition between the nodes while sampling, following
   by the SkipGram model that learns the representation of each node based
   on the resulting corpus of walks. The cosine similarity between the
   specific disease and all genes produces a ranked list of genes. The top
   genes from this list are predicted to be promising target hypotheses.
   Combination of described scores results in a ranked list of targets
   proposed for a given disease and can be filtered out based on their
   novelty, small molecules synthesis availability, availability of PDB
   structure, and other useful filters. Description of all mentioned
   scores and filters is available in the User manual section of
   PandaOmics ([275]https://insilico.com/pandaomics/help). Targets that
   are not druggable by the small molecules and do have red flags in terms
   of safety were excluded from the analysis. After filtering, a ranked
   list of proposed targets was extracted from PandaOmics and presented as
   a heatmap.

Reverse transcription-quantitative PCR

   Total RNA was extracted from mouse tissues using Aurum^TM Total RNA
   Fatty and Fibrous Tissue Kit (Bio-Rad, Hercules, CA). DNAse I treatment
   was performed to cut genomic DNA contamination. Total RNA was extracted
   from cell cultures with RNAqueous^R kit (Applied Biosystems, Life
   Technologies). One μg of RNA was used to perform the reverse
   transcriptase reaction with iScript^TM cDNA Synthesis Kit (Bio-Rad).
   The changes in mRNA levels of the target genes were determined by
   relative RT-qPCR with a CFX96^TM Real-Time PCR Detection System
   (Bio-Rad) using iQ^TM SYBR Green Supermix (Bio-Rad). The analyses were
   performed in duplicate with 100 nM of both sense and anti-sense primers
   in a final volume of 20 µL using iQ^TM SYBR Green Supermix (Bio-Rad).
   Specific primers were designed using Primer3 (Supplementary
   Tables [276]1, [277]2). PCR conditions were 95 °C for 3 min followed by
   40 cycles of 15 sec at 95 °C, 30 sec at 60 °C. The PCR products were
   sequenced with the BigDye terminator kit (Perkin Elmer Applied
   Biosystems) using ABI3100 capillary sequencer (Perkin Elmer Applied
   Biosystems). The efficiency of each set of primers was determined by
   dilution curves. The program geNorm version 3.4 was applied to
   characterize the expression stability of the candidate reference genes
   in kidneys and six reference genes were selected to calculate the
   normalization factor. The relative changes in targeted genes over Gapdh
   mRNAs were calculated using the 2^−ΔΔCt formula.

Endocytosis assay in mouse kidneys and derived PT cells

Mouse kidneys

   Receptor-mediated endocytosis was monitored in the PT segments of mouse
   kidneys by measuring the uptake of β-lactoglobulin (L3908, Sigma).
   Briefly, β-lactoglobulin was tagged with Cy5 using TM2 Ab labelling kit
   (Amersham) following the manufacturer’s instructions. Fifteen minutes
   after tail-vein injection of Cy5-β-lactoglobulin (1 mg/kg B.W.) mice
   were anesthetized and their kidneys were harvested and processed by
   confocal microscopy.

Primary cultured cells

   The endocytic uptake was monitored in mPTCs cells following incubation
   for 60 min at 4 °C with 50 μg × mL^−1 bovine serum albumin (BSA)-Alexa
   Fluor-647 ([278]A34785, Thermo Fisher Scientific) in complete
   HEPES-buffered Dulbecco’s modified Eagle’s medium. The cells were given
   an acid wash and warmed to 37 °C in growth cell medium for 15 min
   before being fixed and processed for immunofluorescence analyses.

Lysosomal degradation

   The detection of lysosomal activity was performed in mPTCs by using
   Bodipy-FL-PepstatinA ([279]P12271, Thermo Fischer Scientific) or
   MagicRed-(RR)[2] substrate (MR-CtsB; 938, Immunochemistry Technologies)
   according to the manufacturer’s specifications. The cells were pulsed
   with 1 μM Bodipy-FL-Pepstatin A or with 1 μM MagicRed-(RR)[2] in Live
   Cell Imaging medium for 1 h at 37 °C, fixed, and subsequently analyzed
   by confocal microscopy. The number of PepstatinA or MagicRed- positive
   structures per cell were quantified by using the open-source cell image
   analysis software CellProfiler^TM as described below.

Luminal lysosomal pH

   Cells expressing an empty or 3×HA-tagged R2pH-LAMP1 were treated in the
   presence and in the absence of non-saturating concentrations of
   Bafilomycin A1 and analysed by confocal microscopy in a chamber heated
   to 37 °C at 5% CO[2].

   The mCherry and pHluorin fluorophores were excited at 561 nm and
   488 nm, respectively and both channels acquired simultaneously to
   minimize misalignment between channels. Where indicated, the Ctns mPTCs
   were also stained with LysoTracker dye (1 μM for 4 h). After washing,
   the cells were subsequently analysed by confocal microscopy in a
   chamber heated to 37 °C at 5% CO[2]. Imaging settings were maintained
   with the same parameters for comparison between different experimental
   conditions. Images were acquired using a Leica SP8 confocal laser
   scanning microscope (Center for Microscopy and Image Analysis,
   University of Zurich) and the fluorescence intensity was quantified by
   the open-source image processing software Fiji (ImageJ, NIH) as
   described below.

Fluorescence microscopy

Mouse and rat tissue slides

   Fresh mouse and rat kidneys were fixed by perfusion with 50–60 mL of 4%
   paraformaldehyde in PBS (158127, Sigma-Aldrich), dehydrated, and
   embedded in paraffin at 58 °C. Paraffin blocks were sectioned into
   consecutive five μm-thick slices with a Leica RM2255 rotary microtome
   (Thermo-Fisher Scientific) on Superfrost Plus glass slides
   (Thermo-Fisher Scientific). Before staining, slides were deparaffinized
   in Xylenes (534056, Sigma-Aldrich) and rehydrated. Antigen retrieval
   was carried out by heating the slides at 95 °C for 10 min in 10 mM
   sodium citrate buffer (pH 6.0). The slides were quenched with 50 mM
   NH[4]Cl, blocked with 3% BSA in PBS Ca/Mg (D1283, Sigma-Aldrich) for
   30 min, and stained with primary antibodies diluted in blocking buffer
   overnight at 4 °C. After two washes in 0.1% Tween 20 (v/v in PBS), the
   slides were incubated with the corresponding fluorophore-conjugated
   Alexa secondary antibodies (Invitrogen) diluted in blocking buffer at
   room temperature for 1 h and counterstained with 1μg Biotinylated Lotus
   Tetragonolobus Lectin (LTL; B-1325 Vector Laboratories) and 1 µM
   4’,6-Diamino-2-phenylindole dihydrochloride (DAPI; D1306, Thermo
   Fischer Scientific). The slides were mounted in Prolong Gold Anti-fade
   reagent ([280]P36930, Thermo Fisher Scientific) and analyzed by
   confocal microscopy. The images were acquired using Leica SP8 confocal
   laser scanning microscope (Center for Microscopy and Image Analysis,
   University of Zurich) equipped with a Leica APO ×63 NA 1.4 oil
   immersion objective at a definition of 1024 × 1024 pixels (average of
   eight or sixteen scans), adjusting the pinhole diameter to 1 Airy unit
   for each emission channel to have all the intensity values between 1
   and 254 (linear range). The micrographs were processed with Adobe
   Photoshop (version CS5, Adobe System Inc., San Jose, USA) software.
   Quantitative image analysis was performed by selecting randomly ~5–10
   visual fields per each slide that included at least 3–5 PTs
   (LTL-positive or AQP1-positive where indicated), using the same setting
   parameters (pinhole, laser power, and offset gain and detector
   amplification below pixel saturation). For Pcna staining on rat kidney
   tissue following mTOR inhibition, the slides were acquired using Leica
   Dmi8 wide-field fluorescence microscope equipped with a Leica HC PL APO
   ×20/0.80 objective, and a Leica DFC9000 GTC camera. The full kidney
   slides were scanned and Pcna fluorescence signal in the cortex of each
   kidney was quantified using the LASX software.

Cultured mPTCs

   The cells were fixed for 10 min with 4% PFA in PBS, quenched with 50 mM
   NH[4]Cl, and permeabilized for 20 min in blocking buffer solution
   containing 0.1% Triton X-100 and 0.5% BSA dissolved in PBS.
   Subsequently, cells were incubated overnight with the appropriate
   primary antibodies at 4 °C. After repeated washing with PBS, the slides
   were incubated for 45 min with the suitable fluorophore-conjugated
   Alexa secondary antibodies (Invitrogen), counterstained with 1 µM DAPI
   for 5 min, mounted with the Prolong Gold Anti-fade reagent and analyzed
   by a Leica SP8 confocal laser scanning microscope (Center for
   Microscopy and Image Analysis, University of Zurich) using the settings
   described above. Quantitative image analysis was performed by selecting
   randomly 5 visual fields pooled from biological triplicates, with each
   field including at least 10–15 cells, using the same setting parameters
   (e.g., pinhole, laser power, and offset gain and detector amplification
   below pixel saturation). The quantitative cell image analyses of
   subcellular structures were determined by using the open-source cell
   image analysis software CellProfiler^[281]53. The specific module
   “Measure-Object-Intensity-Distribution” was used to score the number of
   Map1Lc3b, or PepA or MR-positive structures. The pipeline
   “Cell/particle counting and scoring the percentage of stained objects”
   was used to score the fractions of Lamp1-positive structures that were
   also positive for Map1Lc3b. The “Cytoplasm-Nucleus Translocation Assay”
   was used to score the numbers of Ybx3 and Pcna-positive nuclei. For the
   quantification of colocalization, non-overlapping images were acquired
   from each coverslip. Raw, unprocessed images were imported into FIJI
   v.2.0.0-rc-69/1.52i and converted to 8-bit images, and images of
   individual channels were thresholded independently to eliminate
   background and non-specific staining noise and converted to binary
   masks. Co-localization between Lamp1 and mTOR or RagC and Lamp1 was
   determined using the “AND” function of the image calculator. An
   open-source image processing software Fiji v.2.0.0 (ImageJ, NIH) was
   used for to measure the cilia length and the tubular height. The number
   of cells and/or fields of views used per each condition and the number
   of biologically independent experiments are indicated in the figure
   legends.

Imaging of zebrafish larvae

   Quantitative analysis of fluorescent signals was performed with one
   whole pronephric tubule for each larva using the same setting
   parameters. A multi-photon fluorescence microscope (Leica SP8 MP DIVE
   Falcon; Leica Microsystems, Heerbrugg, Switzerland) was used to acquire
   high-resolution images to analyse the fluorescent vesicles in the
   proximal tubules by using ×25/1.0 NA water immersion objective (HC
   IRAPO L, Leica). The excitation wavelength was 1040 nm tunable laser
   for imaging of mCherry. The acquired data were processed by Huygens
   software (Scientific Volume Imaging, Hilversum, The Netherlands) for
   deconvolution, followed by segmentation using Ilastik software (EMBL,
   Heidelberg, Germany) and quantification of fluorescence intensity,
   vesicle number, and vesicle area for each stack using an open-source
   image processing software Fiji (ImageJ, NIH).

Immunoprecipitation

   The cells (~5 × 10^6 cells) were lysed in a buffer containing 25 mM
   Tris/HCl adjusted to pH 7.4, 150 mM NaCl, 1% NP-40, and 1 mM EDTA, 5%
   glycerol, protease, and phosphatase inhibitors for 10 min at 4 °C.
   Lysates were centrifuged at 13,800 × g for 10 min and the supernatants
   were incubated with 50 µL anti-HA magnetic beads (Thermo Fisher
   Scientific, 88836) with end-over-end rotation for 16 h at 4 °C. The
   protein-bound beads were washed four times with fractionation buffer
   and proteins immunoprecipitated were reduced by the addition of 40 μl
   of Laemmli sample buffer, heated at 95 °C for 5 min, and resolved on
   12% SDS polyacrylamide gel electrophoresis (PAGE) gel, and analyzed by
   immunoblotting.

Lysosome immunoprecipitation

   Cells transiently expressing T192-3×HA were lysed, and intact lysosomes
   were immunoprecipitated using anti-HA-conjugated Dynabeads (Thermo
   Scientific, 88837). Briefly, the cells were seeded in a six-well plate
   at a density appropriate for them to reach confluency after 24 h. The
   medium was removed, the cell monolayers were scraped into 10 ml KPBS
   (136 mM KCl and 10 mM KH2PO4, pH 7.25, containing protease (Roche,
   1836153001) and phosphatase inhibitors (PhosSTOP Sigma, 04906845001)
   and collected by centrifugation at 200 × g. for 5 min. The pelleted
   cells were resuspended in a total volume of 1 ml KPBS and ~4 × 10^6
   cells were mechanically broken by spraying 8–10 times through a 23 G
   needle attached to a 1 ml syringe, and then spun down at 400 g for
   10 min at 4 °C, yielding a post nuclear supernatant (PNS). The
   post-nuclear supernatant was harvested, quantified using Bradford assay
   (Thermo-Fisher, 23246). Equal quantities of proteins from each
   condition were incubated with 50 µL anti-HA magnetic beads with
   end-over-end rotation for 1 h at 4 °C. The lysosome-bound beads were
   washed four times with KPBS. For immunoblotting analysis, anti-HA
   immunoprecipitates were reduced by the addition of 40 μL of Laemmli
   sample buffer, heated at 95 °C for 5 min, and resolved on 12% SDS
   polyacrylamide gel electrophoresis (PAGE) gel, and analyzed by western
   blotting.

   For cystine measurements on immunopurified lysosomes, 50 μL of 80% MeOH
   was added to the beads bound to organelles and incubated for 5 mins at
   room temperature. The free amino acids were further derivatized before
   MS analysis. For whole cell lysate samples, 30 μL of supernatant were
   taken and mixed with 3 μL of 5 N NaOH in borate buffer to ensure the pH
   was below pH 9. For Lyso-IP, the samples were dissolved in 40 μL borate
   buffer from Waters. A mixture of 30 μL samples (both WCL and LysoIP),
   5 μL borate buffer, 5 μL MSK-CAA-A2 internal standard mixture (1:5
   dilution), and 10 μL AQC solution was added, giving a total volume of
   50 μL and mixed thoroughly immediately after addition. After 10 mins
   incubation at 55 °C, samples were loaded onto a Waters UPLC H-Class
   Plus system with an Acquity UPLC Qardupole Solvent Manager and Sample
   Manager. Derivatized amino acids were detected on a Waters QDa single
   quadrupole mass detector in positive mode. The Qda was operated by one
   MS, followed by selected ion monitoring (SIR) for each individual amino
   acid with a defined retention window, shown in the table below. The
   column temperature was maintained at 55 °C. The injection volume was
   1 μL. Gradient elution was performed using 0.1% formic acid in water as
   eluent A and 0.1% formic acid in acetonitrile as eluent B. The flow
   rate was kept constant at 0.5 mL × min^−1 with the following gradient
   (expressed as solvent B): Initial conditions: 1.0% B, 0.0–1 min: 1% B,
   1–4 min: 13.0% B, 4–8.5 min: 15.0% B, 8.5−9.5 min: 95.0% B,
   9.5–11.5 min: 95% B, 12–15 min: 1% B. The data acquisition and data
   analysis were done by Masslynx 4.2 (Waters).

Immunoblotting

   Proteins were extracted from animal tissues or cultured cells, lysed
   using a buffer containing protease (Roche) and phosphatase inhibitors
   (PhosSTOP Sigma), followed by sonication and centrifugation at
   16,000 × g for 10 min at 4 °C. The samples were thawed on ice,
   normalized for protein (20 μg/lane), dissolved in Laemmli sample
   buffer, and separated by SDS-PAGE under reducing conditions. After
   blotting onto PVDF and blocking with 5% non-fat milk (1706404, Bio-Rad
   Laboratories), the membranes were incubated overnight at 4 °C with
   primary antibody, washed, incubated with peroxidase-labeled secondary
   antibody, and visualized with enhanced chemiluminescence (WBKLS0050,
   Millipore, Life Technologies) and acquired using ChemiDoc MP imaging
   system (Bio-Rad Laboratories). Signal intensity was assessed by
   measuring the relative density of each band normalized to β-actin,
   GAPDH, γ-tubulin, or α-tubulin with ImageJ software.

Cell viability assay

   The viability of mPTCs after drug treatments were assessed via MTT
   assay (ab211091, Abcam) in accordance with the manufacturer’s
   instructions. Briefly, the cells were washed three times with PBS and
   then incubated with 0.5 mg/mL of MTT diluted in cell media. After 4 h
   of incubation at 37 °C and until appearance of intracellular purple
   formazan crystals, the remaining crystals were dissolved with dimethyl
   sulfoxide (276855, Sigma-Aldrich) or SDS, and the absorbance was
   measured at 570 nm.

Cell growth and proliferation

   To measure proliferation and growth, the cells were seeded in 24-well
   plates at a density of 2.0 × 10^4 cells per well. The cells were
   cultured for 4 days, and the cell medium was renewed daily. Where
   indicated, the cells were treated with Torin1 (16 h at the indicated
   concentrations), then trypsinized every 24 h and quantified using the
   countess automated cell counter TC10 automated cell counter (BIO-RAD).
   To assess the rate of proliferation in mPTCs, we used the Click-iT Plus
   EdU Alexa Fluor 488 Imaging Kit ([282]C10637 Life Technologies). Cells
   were incubated with EdU solution for 16 h at 37 °C and processed in
   accordance with the manufacturer’s instructions. DAPI
   (4′,6-diamino-2-phenylindole dihydrochloride) was used to counterstain
   nuclei. The slides were analysed by a Leica SP8 confocal laser scanning
   microscope and the labeling index was evaluated as the percentage of
   EdU-labeled cells in relation to the DAPI-flagged nuclei by ImageJ
   software.

Statistics and reproducibility

   The plotted data were presented as mean ± standard error of the mean
   (SEM). Statistical comparisons between experimental groups were
   determined by using one-way analysis of variance (ANOVA) followed by
   Tukey’s or Sidak’s or Dunnett’s multiple comparison test, when
   appropriate. When only two groups were compared, two-tailed unpaired or
   paired Student’s t tests were used as appropriate. The normality
   criteria (calculated by D’Agostino and Pearson omnibus normally test)
   were met. Non-parametric data were analysed using a Kruskal-Wallis test
   with Dunn’s multiple comparison correction. The levels of statistical
   significance are indicated by symbols, and the P values are indicated
   in the figure legends along with the statistical tests. All experiments
   reported here were performed at least two to three times independently
   (unless otherwise indicated in the figure legends) and all attempts at
   replication were successful. The investigators were not blinded to
   allocation during the experiments and outcome assessment. Graph Pad
   Prism software v. 9.4.1 (GraphPad Software) was used for generating all
   statistical analyses. Parts of Figs. [283]2a, [284]3a, d, [285]4c, h,
   [286]5a, e, [287]7a, [288]8a, [289]9 and supplementary figs. [290]2d,
   [291]3a, f, i, [292]4a, [293]6a, [294]7a, c were drawn by using
   pictures from Servier Medical Art. Servier Medical Art by Servier is
   licensed under a Creative Commons Attribution 3.0 Unported License
   ([295]https://creativecommons.org/licenses/by/3.0/).

Reporting summary

   Further information on research design is available in the [296]Nature
   Portfolio Reporting Summary linked to this article.

Supplementary information

   [297]Supplementary Information^ (4MB, pdf)
   [298]Reporting Summary^ (5.8MB, pdf)

Acknowledgements