Cellular and Molecular Life Sciences (2020) 77:2839–2857 https://doi.org/10.1007/s00018-019-03330-y Cellular and Molecular Life Sciences ORIGINAL ARTICLE Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in NPC1 cells Elsa Meneses‑Salas1,2 · Ana García‑Melero1 · Kristiina Kanerva3,4 · Patricia Blanco‑Muñoz1,2 · Frederic Morales‑Paytuvi1,2 · Júlia Bonjoch1,2 · Josefina Casas5 · Antonia Egert6 · Syed S. Beevi6 · Jaimy Jose6 · Vicenta Llorente‑Cortés7,8,9 · Kerry‑Anne Rye10 · Joerg Heeren11 · Albert Lu12 · Albert Pol1,2,13 · Francesc Tebar1,2 · Elina Ikonen3,4 · Thomas Grewal6 · Carlos Enrich1,2 · Carles Rentero1,2 Received: 16 July 2019 / Revised: 30 September 2019 / Accepted: 1 October 2019 / Published online: 29 October 2019 © The Author(s) 2019 Abstract Cholesterol accumulation in late endosomes is a prevailing phenotype of Niemann-Pick type C1 (NPC1) mutant cells. Like- wise, annexin A6 (AnxA6) overexpression induces a phenotype reminiscent of NPC1 mutant cells. Here, we demonstrate that this cellular cholesterol imbalance is due to AnxA6 promoting Rab7 inactivation via TBC1D15, a Rab7-GAP. In NPC1 mutant cells, AnxA6 depletion and eventual Rab7 activation was associated with peripheral distribution and increased mobil- ity of late endosomes. This was accompanied by an enhanced lipid accumulation in lipid droplets in an acyl-CoA:cholesterol acyltransferase (ACAT)-dependent manner. Moreover, in AnxA6-deficient NPC1 mutant cells, Rab7-mediated rescue of late endosome-cholesterol export required the StAR-related lipid transfer domain-3 (StARD3) protein. Electron microscopy revealed a significant increase of membrane contact sites (MCS) between late endosomes and ER in NPC1 mutant cells lack- ing AnxA6, suggesting late endosome-cholesterol transfer to the ER via Rab7 and StARD3-dependent MCS formation. This study identifies AnxA6 as a novel gatekeeper that controls cellular distribution of late endosome-cholesterol via regulation of a Rab7-GAP and MCS formation. Keywords Cholesterol · Late endosomes · Rab7 · NPC1 · Annexin A6 · Membrane contact sites Abbreviations ER Endoplasmic reticulum A431 Human epidermoid carcinoma cells FYCO1 FYVE and coiled-coil domain containing 1 ACAT​ Acyl-CoA:cholesterol acyltransferase GST Glutathione S-transferase AnxA6 Annexin A6 LE/Lys Late endosome/lysosome (endolysosomes) CHO Chinese hamster ovary LPDS Lipoprotein-deficient serum CHO M12 NPC1 mutant CHO cell line MCS Membrane contact sites CMA Chaperone-mediated autophagy MEF Mouse embryonic fibroblasts MOSPD2 Motile sperm domain containing 2 NPC1 Niemann-Pick type C1 Elsa Meneses-Salas and Ana García-Melero contributed equally to ORP1L Oxysterol-related protein 1L this work. OSBP Oxysterol-binding protein Electronic supplementary material The online version of this PFO Perfringolysin O article (https​://doi.org/10.1007/s0001​8-019-03330​-y) contains RILP Rab interacting lysosomal protein supplementary material, which is available to authorized users. SREBP Sterol regulatory element binding protein StARD3 StAR-related lipid transfer domain-3 * Thomas Grewal thomas.grewal@sydney.edu.au TBC1D15 TBC1 domain family member 15 WT Wild type * Carlos Enrich enrich@ub.edu VAP-A Vesicle-associated membrane protein-associ- ated protein A * Carles Rentero carles.rentero@ub.edu Vps13 Vacuolar protein sorting-associated protein 13 Extended author information available on the last page of the article 13 Vol.:(0123456789) 2840 E. Meneses‑Salas et al. Introduction accumulation [36, 38]. Given the complexity of these observations, we reasoned that yet unknown tethers or The transmembrane NPC1 protein is essential for the effi- scaffolding proteins could control the dynamics of late cient export of cholesterol from endolysosomes. Several endosome MCS, raising the possibility that yet unidenti- other cytoplasmic players also contribute, via vesicular fied player(s) or “gatekeepers” may fine-tune alternative and/or non-vesicular pathways, to the exit of cholesterol late endosome-cholesterol transport routes in concert with from this compartment [1–3], including members of the NPC1. In fact, a recent publication identifies Gramd1b, oxysterol-binding protein (OSBP) family, such as oxys- an ER-sterol transport protein, interacting with NPC1 and terol-related protein 1L (ORP1L), the small GTPases transferring cholesterol from LE to the ER [39]. Rab7, Rab8 and Rab9, as well as the late endosome, AnxA6, the largest member of the annexin family, has membrane-anchored StARD3 and StARD3 N-terminal been implicated in the regulation of endo- and exocytic path- like (StARD3NL) proteins [3–6]. Despite the ability of ways, cholesterol homeostasis and the formation of multifac- these multiple players to contribute to cholesterol homeo- torial signaling complexes [40–42]. Like other annexins, the stasis, loss-of-function mutations in the NPC1 protein are majority of AnxA6 binds to membranes in a ­Ca2+-dependent dominant. Some of the cytoplasmic proteins that bind and manner, yet cholesterol loading of late endosomes, using transport cholesterol are also engaged in the formation the NPC1 inhibitor U18666A or low-density lipoproteins and functions of membrane contact sites (MCS) that are (LDL), led to the recruitment of significant amounts of emerging as important non-vesicular transfer mediators AnxA6 to the surfaces of late endosomes [43, 44]. AnxA6 for lipids, cholesterol or calcium (­ Ca2+) between compart- was also enriched in late endosomes lacking functional ments [7–20]. NPC1 [45]. Moreover, AnxA6 overexpression led to the StARD3 and Rab7 are critical for the regulation of accumulation of cholesterol in late endosomes. Although MCS formation as well as cholesterol transfer between late these studies link AnxA6 with cholesterol export from endosomes and the endoplasmic reticulum (ER) [21–23]. endolysosomes [46], the underlying molecular mechanisms StARD3 is ubiquitously expressed and anchored to the remain unclear. Here, we show that AnxA6 depletion allevi- membrane of a late endosome subpopulation [24, 25], ates the NPC1 mutant phenotype through two critical mech- where it binds to the ER-resident vesicle-associated mem- anisms: it triggers endogenous Rab7 activation by sequester- brane protein-associated protein A (VAP-A) protein [26, ing the Rab7-GTPase activating protein, TBC1D15; it also 27]. Like other members of the START family, StARD3 enables StARD3 to facilitate the function of MCS between could facilitate transport of cholesterol between late late endosomes and the ER, aiding cholesterol export from endosomes and other compartments such as the ER, mito- endolysosomes. Our findings implicate AnxA6 inhibition as chondria or plasma membrane [28–30]. However, despite a novel strategy to rescue late endosome-cholesterol accu- its participation in cholesterol transfer between compart- mulation and identify the AnxA6/TBC1D15 complex as a ments via MCS, StARD3 overexpression did not increase potential therapeutic target for NPC disease. cholesterol esterification via acyl-CoA:cholesterol acyl- transferase (ACAT) in the ER [31, 32] and was unable to rescue late endosome-cholesterol accumulation in NPC1 Materials and methods mutant cells [23, 24]. At the same interface that connects late endosomes Materials and ER compartments, the GTPase Rab7 regulates mem- brane trafficking, cholesterol homeostasis and contributes, For primary and secondary antibodies, recombinant DNA, together with protrudin and FYVE and coiled-coil domain siRNA, chemicals and commercial assays, see Supple- containing 1 (FYCO1), in MCS dynamics [9–12]. In addi- mentary Table 1. Low-density lipoproteins (LDL, density tion, Rab7 is responsible for endocytic transport between 1.025–1.05 g/ml) were isolated from the plasma of nor- early endosomes, late endosomes, lysosomes, phago- and molipidemic volunteers by two sequential density gradient autolysosomes [3, 33, 34]. Within these compartments, ultracentrifugation in KBr gradients [47]. Lipoprotein-defi- Rab7 contributes to late endosome motility [35], choles- cient fetal calf serum (LPDS) was prepared by ultracen- terol egress [36, 37], as well as early endosome matu- trifugation as described [48]. Before experiments, LDL and ration. While Rab7-GTP levels appear downregulated in LPDS were dialyzed extensively against PBS and stored at cholesterol overloaded endosomes of NPC1 mutant cells, 4 °C until use. ectopic expression of wild type, constitutively active Rab7, LDL protein concentration was determined by the or the adenoviral protein RIDα, can bypass (at least in bicinchoninic acid (BCA) method (Bio-Rad). Glutathione part), the NPC1 defect to reduce late endosome-cholesterol S-transferase (GST) and GST-fusion proteins (GST–AnxA6, Rab interacting lysosomal protein (RILP)-C33–GST, 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2841 GST–perfringolysin O (PFO)) were produced in E. coli 10 µg/ml aprotinin). Lysates were boiled in 1 × sample BL21 cells and purified using glutathione Sepharose 4B buffer, resolved on SDS-PAGE and transferred to nitrocel- beads (GE Healthcare) as reported previously [49]. lulose (Bio-Rad) or Immobilon-P (Millipore) membranes. Membranes were blocked in 5% non-fat milk, incubated Cell culture and transfections overnight in primary antibodies, washed in TBST, incu- bated with HRP-conjugated secondary antibodies (Bio-Rad Chinese hamster ovary wild type (CHO-WT), CHO-AnxA6 or Abcam, see Supplementary Table 1) and developed using [43], CHO M12 and CHO 2-2 cells were grown in F12 enhanced chemiluminescence EZ-ECL (Biological Indus- (HAM) supplemented with 10% fetal bovine serum (FBS, tries) and Fuji Medical X-ray films (Fujifilm). ImageJ soft- Biological Industries), 2 mM l-glutamine (Sigma Aldrich), ware was used for quantitative analysis of WB bands [53]. 100 units/ml penicillin (Biological Industries) and 100 μg/ ml streptomycin (Biological Industries) at 37 °C, 5% C­ O 2. RNA extraction and quantitative real‑time PCR CHO M12 and CHO 2-2 were kindly provided by Dr. L. Lis- cum (Tufts University School of Medicine, USA) and Dr. D. Total RNA was extracted using RNeasy Mini Kit (Qiagen) Ory (Washington University, USA), respectively. A431-WT, in accordance with the manufacturer’s protocol. 1 μg RNA A431-A6 [50], mouse embryonic fibroblasts from wild was reverse-transcribed using High Capacity cDNA Reverse type (MEF-WT) and AnxA6 KO-mice (MEF-A6ko) [51], Transcription Kit (Applied Bioscience). In a final volume COS-1 cells were cultured in DMEM supplemented with of 20 μl real-time PCR Brilliant SYBRGreen QPCR Mas- 10% (A431, COS-1) or 5% FBS (MEF), 2 mM l-glutamine ter Mix (Agilent Technologies, Stratagene), 10 μl of 1:20 (Sigma Aldrich), 100 units/ml penicillin (Biological Indus- diluted cDNA was used as a template for PCR analysis using tries) and 100 μg/ml streptomycin (Biological Industries) at the LightCycler system (Roche Diagnostics), specific prim- 37 °C, 5% ­CO2. ers (see Supplementary Table 1) and standard PCR amplifi- For transient transfections with fluorescently labeled cation protocol (10 min at 95 °C; 45 cycles of 30 s at 95 °C, AnxA6, TBC1D15 and Rab7 proteins, cells were incubated 15 s at 60 °C and 30 s at 72 °C; and 10 s at 95 °C and 60 s with GenJet Plus Reagent (SigmaGen Laboratories) fol- at 65 °C) according to manufacturer’s instructions. Values lowing manufacturer’s instructions. For siRNA-mediated were normalized to Rpl13 gene in each sample. knockdown of AnxA6, TBC1D15 and StARD3, cells were transfected with 100 μM siRNA targeting mouse AnxA6, TBC1D15 and StARD3 (Santa Cruz) using Lipofectamine Preparation of liver homogenates RNAiMax (Invitrogen) according to the manufacturer’s instructions. Studies were conducted 24 h (siTBC1D15) or Mouse liver tissues were placed in Lysing Matrix tubes (MP 72 h (siAnxA6, siStARD3) after transfection. Scrambled Biomedicals) with homogenization buffer (10 mM Tris, siRNA served as negative control (Dharmacon). 150 mM NaCl, 5 mM EDTA, pH 7.5) supplemented with protease/phosphatase inhibitors cocktail (see above). Sam- Generation of CHO M12‑A6ko cells using ples were then homogenized in a FastPrep120 homogenizer the CRISPR/Cas9 system (MP Biomedicals) and stored at − 20 °C. For immunopre- cipitations (see below), liver homogenates were pre-cleaned For AnxA6 gene depletion in CHO M12 cells using with Protein A-agarose beads for 90 min at 4 °C before anti- CRISPR/Cas9 editing technology, guide RNAs targeting body incubation. hamster AnxA6 were designed as described [52], and CHO M12 cells were transfected with pSpCas9(BB)-2A-Puro v2 Immunoprecipitation (Addgene) carrying gRNAs against hamster AnxA6. 24 h after transfection, cells were selected for 48 h in puromy- Cells were grown on 10-mm dishes, washed with PBS and cin (50 µg/ml). Clones were isolated by dilution and single solubilized in lysis buffer (50 mM Tris, 150 mM NaCl, 1% clones were screened for AnxA6 gene knockout by western Triton X-100, 0.1 mM ­CaCl2, pH 7.4), supplemented with blotting and sequencing. protease/phosphatase inhibitors cocktail (see above). After centrifugation at 12,000g for 6 min at 4 °C, proteins from Immunoblotting supernatants (200–400 µg) were incubated with 2 µg of rab- bit polyclonal anti-AnxA6, rabbit polyclonal anti-TBC1D15 Cells were lysed in lysis buffer (50 mM Tris–HCl, 150 mM (Abcam) or rabbit IgG for 2 h at 4 °C, followed by 60 min NaCl, 1% Triton X-100, 0.1 mM ­CaCl2, pH 7.4) supple- with Protein A-agarose beads (Thermo Scientific). Immu- mented with protease/phosphatase inhibitors cocktail (1 mM noprecipitates were washed three times with lysis buffer and ­Na3VO4, 10 mM NaF, 1 mM PMSF, 10 µg/ml leupeptin, analyzed by western blotting. 13 2842 E. Meneses‑Salas et al. Pull‑down assays TCS SP5 laser scanning confocal microscope equipped with a DMI6000 inverted microscope, blue diode (405 nm), Cells were solubilized in pull-down buffer (50 mM Tris, Argon (458/476/488/496/514 nm), diode pumped solid 150 mM NaCl, 1% Triton X-100, 0.1 mM C ­ aCl2, pH 7.3) state (561 nm), HeNe (594/633 nm) lasers and APO 63x supplemented with protease/phosphatase inhibitors cock- oil immersion objective lens or a Leica DMI6000B epifluo- tail (see above). Samples were centrifuged at 12,000g for rescence inverted microscope equipped with an HCX PLA 10 min at 4 °C. Proteins from post-nuclear supernatants APO 63× oil immersion objective lens. (400–700 µg) were incubated with glutathione Sepharose 4B beads (GE Healthcare) coated with purified recombinant LDL‑cholesterol transport studies AnxA6–GST or RILP-C33–GST (40–70 µg) fusion protein for 2 h at 4 °C. GST was used as a negative control. Samples To analyze the cellular fate of LDL-cholesterol, cells were were washed three times, collected in 30 µL of 1× loading plated on coverslips and grown in F-12 (HAM) supple- buffer and analyzed by western blotting. mented with 5% LPDS for 48 h. Cells were then loaded with 50 µg/µl of LDL ± 10 µg/ml ACAT inhibitor (Sandow Subcellular fractionation 58-035) for 24 h, fixed with 4% PFA for 1 h. Free cholesterol was stained with 0.05 mg/ml of filipin (Sigma Aldrich) and Late endosomes were isolated using sucrose gradients as neutral lipids were stained with 1 µg/ml of BODIPY 493/503 described previously [43, 54]. Briefly, 25 × 106 CHO-WT (Molecular Probes) for 20 min at RT. Coverslips were and CHO-A6 cells were used for each gradient. Cells were mounted in Mowiol (Calbiochem, Merck). Alternatively, washed twice with cold PBS and collected. Cells were pel- cellular cholesterol was stained with recombinant GST–PFO leted and resuspended in homogenization buffer (250 mM as follows: cells were fixed with 4% PFA for 15 min, perme- sucrose, 3 mM imidazole, pH 7.4) supplemented with pro- abilized with 0.1% Triton X-100 (Sigma Aldrich) for 5 min tease/phosphatase inhibitors cocktail (see above). Next, cells and blocked with 3% fat free BSA (Sigma Aldrich) PBS for were homogenized by 15–20 passages through a 22 G nee- 30 min at RT. Cells were incubated with 10 µg/ml of puri- dle at 4 °C. Complete homogenization was confirmed under fied recombinant GST–PFO in blocking buffer for 1 h at RT. the phase microscope. The homogenate was centrifuged for Immunostaining with anti-GST (Abcam) and fluorescently 15 min at 1000g at 4 °C. The post-nuclear supernatant was labeled antibody was performed as above. collected and quantified by Bradford and 3 mg of PNS were brought to a final 40.2% sucrose (w/v) concentration by add- Live‑cell LDL‑BODIPY‑cholesteryl linoleate transport ing 2.5 M sucrose and loaded at the bottom of a 13.2-ml assay tube (Beckman UltraClear). Then 3 ml of 35% sucrose, 3 ml of 25% sucrose and 2.5 ml of homogenization buffer were NPC1-deficient CHO M12 cells seeded onto glass-bottom overlaid stepwise on top. The gradient was centrifuged for dishes (Nunc LabTek 4-well chambered coverglass) were 90 min at 150,000g, 4 °C in a Beckman SW 41 Ti rotor. transfected with non-targeting control and AnxA6 siRNAs After centrifugation, 1.5-ml fractions were collected from in DMEM/F-12 supplemented with 5% LPDS. The transfec- top to bottom and protein was precipitated using trichloro- tions were carried with Lipofectamine RNAiMax (Thermo acetic acid/acetone to determine the TBC1D15 and Rab7 Scientific). 6 h later, 50 μg/ml Alexa Fluor 568-dextran distribution by western blotting. (10,000 MW; Thermo Scientific) was added to the cells to label late endosomal organelles. 22 h after transfection, the Immunofluorescence cells were pulse-labeled for 2 h with 50 μg/ml BODIPY- cholesteryl linoleate-labeled LDL in serum-free medium, Cells grown on coverslips were fixed with 4% paraform- washed and chased in serum-free C ­ O2-independent medium aldehyde (PFA, Electron Microscopy Sciences) for 20 min (Gibco) for the indicated times. Synthesis of BODIPY-cho- at room temperature (RT), washed with PBS, permeabi- lesteryl linoleate was carried by Dr. Young Ah Kim (Queens lized with 0.1% saponin for 10 min and blocked with 1% College, New York) and labeling of human LDL was per- bovine serum albumin (BSA) for 5 min. Coverslips were formed as previously described [55]. incubated with primary antibody diluted in 0.02% sapo- The chase was followed by confocal live-cell imaging on nin, 0.1% BSA in PBS for 1 h at RT, washed intensively a Leica TCS SP8 attached to a motorized DMI 6000 inverted and then incubated with the adequate secondary antibody microscope with 63× HC PL APO CS2 water objective labeled with Alexa Fluor-555 (Invitrogen) for 45 min at RT. (1.20 NA). Experiments were performed at 37 °C in a fully After staining, coverslips were mounted in Mowiol (Cal- enclosed temperature-controlled environmental chamber. biochem, Merck). Samples were visualized using a Leica Data were acquired with Leica LAS X (Leica Microsystems) 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2843 and the efflux of late endosome BODIPY-cholesterol was number of cells scored for each condition was 25 and the quantified from background subtracted images with ImageJ average number of sections (fields) 40. by analyzing mean intensity of BODIPY-cholesterol fluo- rescence per cell. Statistical analysis For late endosome mobility analysis, time-lapse series were obtained with image acquisition frame rate of 370 ms. Unless mentioned in the figure legend, group data are pre- From the resulting live-cell videos, late endosome mobil- sented as mean ± SD. Comparison between 2 groups were ity was assessed by measuring the Pearson colocalization analyzed by Student’s t test; comparison between more than between subsequent frames/cell, and decreased degree of 2 groups were analyzed by one-way ANOVA with a Bon- colocalization was considered indicative of increased late ferroni post hoc test, and comparison between groups and endosome mobility. Ten initial frames from time-lapse condition were analyzed by Bonferroni post-tested two-way acquisitions were included in the analysis for each cell. All ANOVA for condition and group differences using Graph- the data are expressed as mean ± SEM. Pad Prism software. Differences were considered statistically significant at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001. Image analysis Results Image analysis was performed with NIH ImageJ software [53]. When comparing different treatments, images were Interaction of AnxA6 with the Rab7‑GAP TBC1D15 captured and systematically screened using identical micro- scope settings. We previously demonstrated that AnxA6 overexpression For number, size, fluorescence intensity and cellular dis- led to late endosome-cholesterol accumulation, a phenotype tribution of late endosome and lipid droplets, a semi-auto- reminiscent of the NPC1 mutant phenotype [46, 62]. This mated ImageJ macro was designed and used. Specifically, was accompanied by an increased recruitment of AnxA6 fluorescence microscopy images were locally thresholded to cholesterol-laden late endosomes upon pharmacologi- and vesicles were selected through the ImageJ particle analy- cal NPC1 inhibition, using U18666A, or loading with LDL sis function. Number, size and fluorescence intensity from [43–45]. In order to unravel the underlying mechanism, we raw images were then calculated. Cellular distribution was reasoned that cells lacking NPC1-dependent cholesterol analyzed using the 3D ImageJ Suite [61]. export pathways would be the most promising model to address how AnxA6 could affect late endosome-cholesterol levels. Electron microscopy Most strikingly, in the NPC1 mutant cell line CHO M12, depleting AnxA6 using a small interfering RNA (siRNA) or For conventional electron microscopy, pellets from frac- CRISPR/Cas9 technology (CHO M12-A6ko) led to a signifi- tions enriched with late endosomes from discontinuous cant reduction of late endosome-cholesterol accumulation sucrose gradients or cells in culture were washed in PBS in the perinuclear region, determined using both filipin and and fixed overnight (gradient fractions) or for 1 h (cells) PFO labeling to visualize unesterified (free) cholesterol [63] in 2.5% glutaraldehyde in 0.1 M phosphate buffer (PB) at (Fig. 1a, b; quantified in c–e). In addition, loss of late endo- RT. Next, samples were slowly and gently scraped and pel- some-cholesterol accumulation in AnxA6-depleted CHO leted in 1.5 ml tubes. Pellets were washed in PB and incu- M12 cells was associated with increased late endosome bated with 1% ­OsO4 for 90 min at 4 °C. Then samples were positioning towards the cell periphery (Fig. 1f), a feature dehydrated, embedded in Spurr and sectioned using Leica common to late endosomes with normal cholesterol content. ultramicrotome (Leica Microsystems). Ultrathin sections We performed a yeast two-hybrid screen (Hybrigenics (50–70 nm) were stained with 2% uranyl acetate for 10 min, Services, France) using the N-terminal region of human a lead-staining solution for 5 min and observed using a trans- AnxA6 (aa1–273) as bait against a cDNA library from mission electron microscope, JEOL JEM-1010 fitted with a human liver to identify AnxA6 interaction partners that Gatan Orius SC1000 (model 832) digital camera. could explain the rescue of the NPC1 mutant phenotype Perimeter and areas of contact between late endosome/ upon AnxA6 depletion. These studies identified TBC1D15, lysosomes (LE/Lys) and ER–LE/Lys were identified by mor- a Rab7-GAP [64, 65–69], as a possible AnxA6-binding phology was measured with ImageJ [53]. At least 20–50 protein. cells were analyzed per experiment and data were analyzed Previous reports suggested that ectopic expression of from duplicate or triplicate separate experiments. At least Rab7 could restore cholesterol re-esterification and neu- two grids were used for each condition. The minimum tral lipid deposition in NPC1 mutants [36]. Therefore, 13 2844 E. Meneses‑Salas et al. Fig. 1  AnxA6 depletion rescues late endosomal cholesterol accumu- ing, the outline and shape of cells is indicated. Representative cells lation in NPC1 mutant cells. a CHO M12 cells expressing control and AnxA6 protein levels in lysates from CHO M12 and CHO M12- siRNA (siRNA Ctrl) or siRNA targeting AnxA6 (siRNA AnxA6) A6ko cells are shown. Scale bar, 10 μm. c–f Dot-plot of PFO-stained were fixed and stained with filipin. For better comparison of fil- vesicle number (c), size (d), PFO relative staining intensity (e) and ipin staining, the outline and shape of transfected cells is indicated. vesicle distance to the nucleus (f) of a representative experiment from AnxA6 protein levels in lysates from CHO M12 cells ± siRNA CHO M12 and CHO M12-A6ko cells (n > 60 cells, 3 experiments). AnxA6 is shown. Scale bar, 10 μm. b Stable AnxA6 gene deletion For quantification details see “Materials and methods”. ***p < 0.001 in CHO M12 cells (CHO M12-A6ko) using CRISPR/Cas9 genome by two-tailed Student’s t test (c–f). All data are presented as editing technology. Cholesterol was visualized using GST–perfrin- mean ± SD in red golysin O (PFO) staining. For better comparison of GST–PFO stain- the potential interaction of AnxA6 with a Rab7-GAP was To provide additional evidence for AnxA6-regulated pursued further. Reciprocal co-immunoprecipitations with association of TBC1D15 with late endosomes, subcellular antibodies against AnxA6 and TBC1D15 confirmed the fractionation was performed to compare the cellular distribu- interaction between AnxA6 and TBC1D15 in mouse liver tion of TBC1D15 in CHO-WT and CHO-A6 cells (Fig. S1b). homogenates (Fig. 2a), cell lysates from CHO-WT, AnxA6 Significant enrichment of TBC1D15 in the late endosome overexpressing CHO (CHO-A6) [43], and NPC1 mutant cell fraction (F2) of CHO-A6 cells was observed (Fig. S1b). lines CHO M12 and CHO 2-2 (Fig. 2b). Notably, AnxA6 Electron microscopy of the late endosome fraction from protein levels were elevated in NPC1 mutant cells (CHO CHO cells (Fig. S1c) detected vesicular structures, some- M12, CHO 2-2), compared with CHO-WT cells (Fig. 2c) times with internal membrane fragments and/or multilamel- (see “Discussion”). Pull-down assays with AnxA6–GST lar prototypical endolysosomes. Importantly, mitochondria, fusion protein further established the ability of AnxA6 to peroxisomes or microsomes (ER) were completely absent directly interact with TBC1D15 (Fig. 2d) and use of trun- from this fraction. These data suggested that AnxA6 targets cated TBC1D15 mutants mapped the interaction of AnxA6 TBC1D15 to Rab7-positive endosomes, and correlated with to the N-terminal (1–200aa) region of TBC1D15 (Fig. 2e). reduced filipin staining in AnxA6-depleted CHO M12 cells We next addressed a possible involvement of AnxA6/ (Fig. 1). Thus, high levels of AnxA6 appear to induce com- TBC1D15 interaction in Rab7-mediated late endosome- plex formation with TBC1D15 and its recruitment to Rab7- cholesterol egress. AnxA6 depletion in CHO M12 cells GTP-positive late endosomes and lysosomes, and removal significantly reduced the association and colocalization of of the AnxA6/TBC1D15 complex to restore late endosome- YFP–TBC1D15 with late endosome and lysosome struc- cholesterol export in NPC1 mutants. tures expressing the constitutively active Rab7 mutant GFP–Rab7-Q67L (Fig. S1a). In CHO M12 cells, quantifi- AnxA6 interferes with Rab7 activity to impair late cation showed that 68 ± 18% of YFP-labeled ring structures endosome‑cholesterol egress colocalized with GFP-positive vesicles (see also line profile in Fig. S1a). In contrast, CHO M12-A6ko cells contained To determine whether increased AnxA6 levels could pro- significantly fewer GFP-positive vesicles that colocalized mote Rab7-GAP activity, we examined Rab7-GTP levels with YFP-labeled ring formations (36 ± 23%). using RILP-C33–GST pull-down assays (Fig. 2f–h) in cell 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2845 lysates containing moderate (CHO-WT) or elevated (CHO- requirement for active Rab7 protein for the rescue of the A6, M12, 2-2) AnxA6 protein levels (see Fig. 2c). In addi- NPC1 mutant phenotype (Fig. S2d). Collectively, these tion, we also compared Rab7-GTP amounts in A431wt cells observations clearly indicate that AnxA6 modulates late that lack endogenous AnxA6, with those from a well-char- endosome-cholesterol levels by regulation of Rab7 activity. acterized A431 line stably expressing AnxA6 (A431-A6) Importantly, in line with elevated Rab7 activity (Fig. 2f) (Fig. S2a) [50, 70], as well as in MEFs from wild type and increased late endosome positioning towards the cell (MEF-WT) and AnxA6-KO (MEF-A6ko) mice (Fig. S2b) periphery (Fig. 1f) in AnxA6-depleted M12 cells, we [40, 51]. Both A431-A6 and MEF-WT express AnxA6 lev- observed the re-establishment of late endosome motility as els commonly found in other cell lines and tissues [50, 70]. judged by live-cell microscopy of M12 cells labeled with Indeed, cells with elevated AnxA6 protein expression dis- LDL-derived BODIPY-cholesterol upon AnxA6 depletion played a substantial and significant reduction in Rab7-GTP (Fig. 2j; quantified in k and Movies S1 and S2). One out- levels (Fig. 2f). On the other hand, AnxA6 knockdown in come of this stimulated late endosome-trafficking could the NPC1 mutant CHO M12 or CHO 2-2 cells was associ- be the redistribution of late endosome-cholesterol to other ated with effectively increased Rab7-GTP levels (Fig. 2g). destinations. Indeed, analysis of BODIPY-cholesterol efflux Furthermore, TBC1D15 depletion strongly increased Rab7- from late endosomes in the presence of extracellular choles- GTP amounts in NPC1 mutant M12 and 2-2 cells (Fig. 2h). terol acceptors revealed a faster removal of BODIPY-cho- Thus, elevated AnxA6 levels create an environment that lesterol from AnxA6-depleted CHO M12 (siRNA-AnxA6) favors TBC1D15-mediated Rab7 inactivation. These stud- cells compared with control cells (Fig. 2l). Previous studies ies reveal for the first time that AnxA6 can regulate Rab7 implicated ectopic and non-physiological elevation of Rab7 GTPase via direct binding and recruitment of a member of levels as a prerequisite to restore neutral lipid storage and the TBC/Rab7GAP-family, TBC1D15 to late endosomes, cholesterol esterification in NPC1 mutants [36]. However, thereby inhibiting Rab7 activity. our studies strongly suggest that elevation of endogenous To further confirm that AnxA6- and/or TBC1D15- Rab7-GTP levels, through the depletion of AnxA6 (or induced downregulation of Rab7-GTP levels would inter- TBC1D15), is sufficient to re-establish the ability of NPC1 fere with the ability of overexpressed Rab7 to rescue the mutant cells to export late endosome-cholesterol to the cell NPC1 mutant phenotype [36], RFP-Rab7 was ectopically surface (Fig. 2l), or store as neutral lipid in lipid droplets co-expressed with GFP, AnxA6–GFP or YFP–TBC1D15 (see below). in NPC1 mutant CHO M12 cells. To visualize late endo- some-cholesterol accumulation, cells were fixed and stained AnxA6 depletion restores cholesterol trafficking with filipin (Fig. 2i). In agreement with previous data [36], in NPC1 mutant cells transient overexpression of Rab7 drastically reduced late endosome-cholesterol accumulation in CHO M12 cells, In wild type cells, endocytosed, esterified LDL-cholesterol yet co-expression of Rab7 with AnxA6 or TBC1D15 not is hydrolyzed into free cholesterol in late endosomes and only blocked Rab7-mediated rescue of late endosome-cho- lysosomes, to be delivered to other cellular sites, includ- lesterol export (YFP–TBC1D15), but also increased late ing the ER for re-esterification and subsequent storage as endosome-cholesterol accumulation (AnxA6–GFP). Strik- cholesteryl esters in lipid droplets [3, 5]. In NPC1 mutant ingly, TBC1D15 depletion in CHO M12 cells restored late cells, late endosome-cholesterol accumulation is accompa- endosome-cholesterol export even upon ectopic co-expres- nied by reduced cholesterol re-esterification and neutral lipid sion of AnxA6–GFP (Fig. 2i). These findings show une- deposition [36]. To address the trafficking of late endosome- quivocally that AnxA6, acting through TBC1D15, reduces cholesterol in AnxA6-depleted CHO M12 cells, we set up Rab7-GTP and blocks late endosome-cholesterol egress in experimental conditions to monitor the trafficking of LDL- NPC1 mutant cells. In support of this, expression of the derived cholesterol out of late endosomes and lysosomes. constitutively active Rab7 mutant Rab7-Q67L was sufficient CHO-WT and NPC1 mutant cells were grown in LPDS- to release accumulated late endosome-cholesterol in CHO containing media for 48 h before loading with LDL for addi- M12 cells (Fig. S2c). Moreover, ectopic expression of the tional 24 h (Fig. S3a) [38]. After 48 h, a significant reduction YFP–TBC1D15(1–200) deletion mutant, which still interacts of filipin and neutral lipid stain (BODIPY 493/503-positive with AnxA6 (see Fig. 2e), yet lacks the GAP domain to structures, green) in both CHO-WT and NPC1 mutant cell inactivate Rab7 (and therefore acts as a dominant-negative lines (CHO M12) was observed, pointing to strongly reduced mutant), also showed a significant reduction of late endo- late endosome-cholesterol and neutral lipid levels in lipid some-cholesterol in CHO M12 cells (Fig. S2d). In contrast, droplets. As expected, subsequent LDL loading caused late ectopic expression of the dominant-negative GFP–Rab7- endosome-cholesterol accumulation only in CHO M12 cells, T22N mutant inhibited late endosome-cholesterol egress while CHO-WT cells showed cholesterol redistribution and in AnxA6-depleted CHO M12 cells, demonstrating the robust BODIPY staining (Fig. S3b). 13 2846 E. Meneses‑Salas et al. Given that AnxA6 depletion reduced perinuclear late neutral lipid storage in lipid droplets in these cells. Indeed, endosome-cholesterol accumulation in CHO M12 cells compared with a control siRNA (siCtrl), AnxA6 deple- (Fig. 1), we next investigated if the loss of AnxA6 could tion not only reduced late endosome-cholesterol accumu- also rescue the delivery of late endosome-cholesterol for lation, as judged by reduced filipin staining (quantified in 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2847 ◂Fig. 2  AnxA6 interacts with TBC1D15. a Mouse liver homogenates were Fig. S4a), but significantly increased neutral lipid stain immunoprecipitated with antibodies against AnxA6, TBC1D15 or control (BODIPY-positive structures, green) in CHO M12 cells antibody (normal rabbit serum, NRS) and analyzed by western blotting for co-immunoprecipitation. Representative images of AnxA6 and TBC1D15 (Fig. 3a; quantified in c and d). A slightly increased num- levels (5% of total input) and immunoprecipitations are shown (n = 2). b ber of lipid droplets was observed under normal growth Cell lysates from CHO-WT, CHO-A6, CHO M12 and CHO 2-2 cells were conditions in AnxA6-depleted CHO M12 cells (Fig. S4a). immunoprecipitated with antibodies against AnxA6, TBC1D15 or control Colocalization with anti-adipophilin, a well-established antibody (NRS) as indicated and analyzed by western blotting for co-immu- noprecipitation. Representative images of TBC1D15, AnxA6, Rab7 and lipid droplet marker [71], confirmed the identity of these actin levels in the cell lysates (10% of total input, left panel) and immuno- structures in AnxA6-depleted CHO M12 cells (Fig. S4b). precipitations are shown (n = 3). c Quantification of AnxA6 levels in CHO- Consistent with our previous results (Fig. 1), decreased WT, CHO-A6, CHO M12 and CHO 2-2 cell lysates (n = 3). d Coomassie filipin staining was evident in CHO M12-A6ko cells blue staining of purified glutathione S-transferase (GST) and GST–AnxA6 used in the pull-down assays shown in D–E. Cell lysates (2% of total input) (quantified in Fig. S4a). Furthermore, and in strong sup- from CHO-WT, CHO-A6, CHO M12 and CHO 2-2 cells were incubated port of these findings, conventional electron microscopy with GST or GST–AnxA6 fusion protein in pull-down assays and analyzed showed abundant lipid droplets after LDL loading not only by western blotting with anti-TBC1D15 as indicated (n = 3). e Scheme of in CHO-WT, but also in CHO M12-A6ko cells; in contrast, YFP-tagged TBC1D15 wild type (1–674aa) and deletion mutants (1–200, 201–333, 334–557, 558–674) used to map the interaction of TBC1D15 as expected [36], very few lipid droplets were observed in with AnxA6. The TBC domain in TBC1D15 (334–557) is indicated. Cell LDL-loaded CHO M12 cells (Fig. S4c; see quantification). lysates from COS-1 cells ectopically expressing YFP–TBC1D15 wild type These results indicate that AnxA6 depletion overcomes and mutants were incubated ± GST–AnxA6 fusion protein in pull-down defective transport of LDL-cholesterol to the ER in NPC1 assays and analyzed by western blotting for interaction with anti-GFP as indicated. Expression levels of YFP-tagged TBC1D15 deletion mutants mutants. Increased cholesterol levels in the ER suppress pro- (5% of total input) in COS-1 cell lysates are shown (n = 2). f Cell lysates cessing of the major transcriptional regulator of cholesterol from CHO-WT, CHO-A6, CHO M12 and CHO 2-2 cells were subjected homeostasis, sterol regulatory element binding protein 2 to Rab interacting lysosomal protein (RILP)-C33–GST pull-down assays to (SREBP2), enabling mature SREBP (mSREBP2) to enhance determine active Rab7 (Rab7-GTP) levels. g Rab7-GTP levels determined as above with lysates from CHO M12 cells expressing non-targeting con- the transcription of its target genes [72]. In line with pub- trol siRNA (siCtrl) or siRNA targeting AnxA6 (siA6). Total levels of Rab7, lished data [73], LDL-cholesterol failed to suppress SREBP2 AnxA6, and actin in cell lysates and the quantification of relative Rab7 maturation in NPC1 mutants. However, AnxA6 depletion activity are shown (5% of total input, n = 3). h Representative western blot in CHO M12 was associated with decreased amounts of showing Rab7-GTP levels determined as above with lysates from CHO M12 and CHO 2-2 cells expressing non-targeting control siRNA (siCtrl) or mSREBP2 upon LDL loading (Fig. S5). To further verify siRNA targeting TBC1D15 as indicated. Total levels of Rab7, TBC1D15 that late endosome-cholesterol delivery to the ER, followed and actin in cell lysates (5% of total input) and the quantification of relative by ACAT-mediated cholesterol esterification, would drive Rab7 activity are shown (n = 3). i CHO M12 cells were co-transfected with transfer of neutral lipids into newly formed lipid droplets in RFP–Rab7 (red) and empty vector (GFP), AnxA6–GFP, YFP–TBC1D15 or AnxA6–GFP together with siRNA targeting TBC1D15 (siTBC1D15) AnxA6-depleted CHO M12 cells, a pharmacological ACAT (green) as indicated. Cells were fixed and stained with filipin (blue). For inhibitor (Sandoz 58-035) was employed [74]. ACAT inhi- better comparison of filipin staining, the outline and shape of transfected bition completely abrogated neutral lipid (cholesteryl ester) cells is indicated. Merged images are shown. Scale bar, 10 μm. The mean accumulation and lipid droplet formation in AnxA6-depleted relative filipin intensity of at least 20 transfected cells from 3 independent experiments was quantified (n = 3). j Confocal images of CHO M12 cells M12 cells, while a concomitant increase of filipin staining expressing control siRNA (siRNA Ctrl) or siRNA targeting AnxA6 (siRNA was observed (Fig. 3b; quantified in c and d). AnxA6). Cells were grown in lipoprotein-protein deficient serum (LPDS), Finally, employing the same experimental conditions and then pulse-labeled with LDL-BODIPY-cholesteryl linoleate (see described above (Fig. 3a), we addressed if TBC1D15 deple- “Materials and methods” for details). Vesicles labeled with LDL-derived BODIPY-cholesterol (BC) were imaged in live cells over time. Repre- tion would also impact the transfer of neutral lipids into sentative images tracing pseudocoloured individual BC-stained vesicles for lipid droplets in NPC1 mutant CHO M12 cells. Indeed, con- CHO M12 ± AnxA6 are shown (0–8 s) after 24 h chase. For representative sistent with TBC1D15 silencing leading to elevated Rab7- frames from live-cell videos see Movies S1 and S2. k Quantitation of BC- GTP levels (Fig. 2h) and restoration of late endosome-cho- stained vesicle mobility in CHO M12 ± AnxA6 (siCtr, siA6) after pulse- labeling with LDL-BODIPY-cholesteryl linoleate, followed by 24 h chase. lesterol efflux (Fig. 2i), depletion of TBC1D15 in CHO M12 Bars: 1 − [Pearson’s colocalization coefficient between subsequent frames/ cells was accompanied with increased BODIPY-positive cell] ± SEM (n = 24–26 cells, two experiments). l Analysis of late endoso- lipid droplet numbers (Fig. S2e, f; quantified in g and h). mal (LE) BODIPY-cholesterol removal. CHO M12 cells were transfected Thus, both TBC1D15 and AnxA6 depletion enabled neu- with control siRNA (siCtrl) or siRNA targeting AnxA6 (siAnxA6). 22 h after transfection, cells were pulse-labeled for 2 h with 50 μg/ml LDL- tral lipid storage in LDL-loaded CHO M12, confirming the BODIPY-cholesteryl linoleate. Cells were washed and chased in medium hypothesis that both proteins contribute to a role for Rab7 with 5% lipoprotein-deficient serum for 0–48 h. The efflux of LDL-derived in cholesterol homeostasis. Taken together, these data indi- BODIPY-cholesterol was quantified by analyzing mean fluorescence cate that AnxA6 as well as TBC1D15 deficiency in NPC1 intensity per cell (n = 31–33 cells, two experiments). *p < 0.05; **p < 0.01; ***p < 0.001 by one-way ANOVA with Bonferroni post hoc test (c, f, i), mutant cells promotes transport of free cholesterol from late two-tailed Student’s t test (g–i, k) or two-way ANOVA with Bonferroni post endosomes to lipid droplets in an ACAT-dependent manner. hoc test (l). All data are shown as mean ± SEM 13 2848 E. Meneses‑Salas et al. Fig. 3  AnxA6 knockdown elicits late endosome-cholesterol release t = 48 (LPDS) and t = 72 h (LDL) (merged and split channels) are and neutral lipid accumulation in NPC1 mutant cells. a CHO M12 shown. Enlarged regions of interest are shown. For better comparison cells expressing control siRNA (siCtrl) or siRNA targeting AnxA6 of filipin and BODIPY staining, the outline and shape of cells is indi- (siAnxA6) were grown in 10% fetal calf serum (0 h, control), then cated. Scale bar, 10 μm. A representative western blot showing siRNA starved in 5% lipoprotein-deficient serum (LPDS) for 48 h before AnxA6 depletion in CHO M12 cells is provided. Actin served as load- loading with 50 µg/ml LDL (± 10 μg/ml ACAT inhibitor Sandoz ing control. c–d Dot-plot of number and area of filipin-stained (LE) 58–035 in b) for 24 h (see scheme in Fig. S3a, and “Materials and and BODIPY-stained (LD) vesicles per cell of a representative experi- methods” for details). At each time point (0, 48 and 72 h), cells were ment (n > 60, 3 experiments). For quantification details see “Materials fixed, stained with filipin (cholesterol, red) and BODIPY 493/503 and methods”.*p < 0.05; ***p < 0.001 by one-way ANOVA with Bon- (neutral lipids, green). Representative fields of cells at t = 0 (control), ferroni post hoc test (d, e). All data are presented as mean ± SD in red 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2849 StARD3 is required to rescue late HeLa cells [23]. Consequently, AnxA6 depletion rescued endosome‑cholesterol export in NPC1 mutant cells late endosome-cholesterol accumulation in NPC1 mutant lacking AnxA6 cells via StARD3-dependent cholesterol transport routes. Further supporting a role for StARD3 downstream of the It was previously reported that enlarged and cholesterol- AnxA6–TBC1D15–Rab7 axis, simultaneous depletion laden late endosomes have impaired vesicular trafficking of TBC1D15 and StARD3 in CHO M12 cells was asso- [35]. In addition, active Rab7 is required to promote MCS ciated with late endosome-cholesterol accumulation in formation between late endosomes and lysosomes and the more scattered late endosomes (Fig. 5e). Hence, despite ER [12], providing protein–protein interactions within MCS upregulated Rab7 activity due to TBC1D15 depletion, for the bidirectional transfer of cholesterol and other lipids StARD3 depletion interfered with late endosome-choles- between late endosomes and ER [75, 76]. Given elevated terol egress and consequently, these cells did not display Rab7-GTP levels and increased late endosome motility in lipid droplets (Fig. 5e; quantified in g and h). In strong AnxA6-depleted NPC1 mutant cells (Fig. 2), we reasoned support of these findings, and in line with published data that increased MCS formation could aid cholesterol transfer [36], conventional electron microscopy showed abundant in these cells. However, CHO M12 lack NPC1 and do not lipid droplets in LDL-loaded CHO-WT cells, but not CHO express ORP1L (Fig. S6), excluding the ORP1L/VAP-A- M12 cells. Lipid droplet formation upon LDL loading was dependent cholesterol transfer route [4, 6, 15, 16, 77–79]. restored in AnxA6-depleted CHO M12 cells, yet depletion We examined the impact of AnxA6 depletion on MCS of StARD3 in these cells resulted in a lack of lipid drop- between late endosomes/lysosomes and the ER in CHO lets, further validating the requirement of StARD3 to store M12 cells. Since the size of MCS is in the nanometer LDL-derived neutral lipids as cholesteryl esters (Fig. S4c range (~ 5–30 nm), electron microscopy is currently the and quantification). Altogether, AnxA6 deficiency in CHO best approach to identify and quantify these membrane M12 cells correlated with a significant increase of sur- domains. Therefore, CHO-WT, CHO M12 cells with and face contacts between late endosomes and the ER (MCS), without AnxA6 were prepared for conventional elec- possibly facilitating the transfer of cholesterol out of late tron microscopy and ultrathin sections were analyzed endosomes. and images quantified using ImageJ. For quantifications, Although it was suggested that non-functional NPC1 may we selected vesicle sections harboring prototypical late impair MCS formation between ER and late endosomes [13], endosomes and lysosomes with electron-dense mem- this has not been proven experimentally. Here we identified branes of variable size containing bits of electron-dense decreased MCS numbers in LDL-loaded NPC1 mutant CHO material [80], similar to those structures observed by oth- M12 cells compared with CHO-WT cells (Fig. S4d) (CHO- ers after treatment with U18666A. Figure 4a shows rep- WT cells contain less AnxA6 than CHO M12; see Fig. 2c). resentative late endosome images from CHO-WT, CHO However, those MCS numbers were significantly expanded M12 and AnxA6-depleted CHO M12 cells, where the sur- by the depletion of AnxA6 in NPC1 mutant cells, creating face contacts between late endosomes and the ER were an increased surface contact between late endosomes and quantified by stereology. These data sets show reduced the ER. More remarkably, in these settings no changes in MCS formation or stability of NPC1-mutant CHO cells the percentage of MCS were observed when StARD3 was and clearly demonstrate that AnxA6 depletion increased depleted, despite the transfer of cholesterol to lipid drop- MCS formation in CHO M12 cells. In fact, this is the first lets being completely blocked under these conditions. This quantitative study showing diminution of MCS in NPC1 strongly indicates that tethering and cholesterol transport at mutant cells (Fig. 4b). the MCS interface are two independent functions. Indeed, Finally, we investigated whether the StARD3/VAP-A StARD3-depleted CHO M12-A6ko cells, showed no differ- protein complex that can facilitate cholesterol transfer ences in MCS numbers (compared with CHO M12-A6ko), between late endosomes and the ER [75, 76] could con- though no lipid droplets were observed (Fig. 5a, S4c). tribute to the rescue of late endosome-cholesterol export in CHO M12-A6ko cells. The ability of AnxA6 depletion to reduce late endosome-cholesterol accumulation in CHO Discussion M12 cells was lost upon StARD3 depletion and concomi- tantly, strongly interfered with neutral lipid accumulation We have shown here that the loss of NPC1 function and in lipid droplets of CHO M12-A6ko cells (Fig. 5a, quan- the concomitant accumulation of cholesterol in the late tified in c and d; also, electron microscopy in Fig. S4c). endocytic compartment can be rescued by AnxA6 deple- This observation was accompanied by a wide distribution tion. We found that AnxA6 interacts with TBC1D15 of LDL-containing and filipin-positive late endosomes (Rab7-GAP), enabling TBC1D15 to inactivate Rab7. Con- (Fig. 5a), as shown previously for StARD3-deficient sequently, AnxA6 as well as TBC1D15 depletion lead to 13 2850 E. Meneses‑Salas et al. Fig. 4  Increased membrane contact sites in CHO M12 cells after ple) between ER and LE/Lys structures is shown. b Quantitative ste- AnxA6 depletion. a Representative transmission electron micros- reology of ER–LE/Lys contacts in TEM sections is given: perimeter copy (TEM) images of late endosomes/lysosomes (LE/Lys) from of LE/Lys contacts (in microns) and percentage of endosome surface CHO-WT, CHO M12 expressing control siRNA (siRNA Ctrl), in contact with the ER per cell (n > 30). LE late endosomes, ER endo- siRNA targeting AnxA6 (siRNA AnxA6) and CHO M12-A6ko cells. plasmic reticulum, Mit mitochondria, PM plasma membrane. Scale Arrowheads indicate endoplasmic reticulum (ER). A schematic repre- bar, 200 nm. *p < 0.05; ***p < 0.001 by one-way ANOVA with Bon- sentation of these images with highlighted membrane contacts (pur- ferroni post hoc test. All data are presented as mean ± SD in red elevated Rab7-GTP levels, which facilitates late endosome- cell lines shown here, human NPC1 mutant skin fibroblasts cholesterol egress, increased late endosome motility and a (GM03123, data not shown), and U18666A-treated CHO- concurrent increase of neutral lipid accumulation in lipid WT cells [44] can possibly be explained by a KFERQ-motif droplets, in an ACAT-dependent manner. Mechanistically, in AnxA6, which targets AnxA6 via chaperone-mediated we showed that StARD3 is instrumental for the transfer of autophagy (CMA) or endosomal microautophagy [81–83]. late endosome-cholesterol to the ER in NPC1 mutants lack- However, CMA is markedly inhibited upon late endosome- ing AnxA6. cholesterol accumulation [83–88], which might explain ham- Therefore, AnxA6 is a novel component of the cellu- pered CMA-mediated AnxA6 degradation. lar machinery regulating cellular cholesterol homeostasis On the other hand, it is now well established that cho- (Fig. 6). Late endosome-cholesterol accumulation in NPC1 lesterol levels in endolysosomes can regulate the position- mutant cells is associated with elevated AnxA6 protein lev- ing of this organelle. A sophisticated ensemble of tethers, els, which is detrimental for the trafficking and dynamics OSBPs, motor proteins and components of the cytoskeleton of the late endocytic compartment because it blocks Rab7 accomplishes the spatio-temporal re-organization of late activation. Increased AnxA6 levels in NPC1 mutant CHO endosomes and lysosomes between the perinuclear area to 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2851 Fig. 5  StARD3 contributes to late endosome-cholesterol egress were starved in 5% LPDS for 48 h and loaded with 50 µg/ml LDL for and neutral lipid accumulation in AnxA6-deficient NPC1 mutant 24 h. Then cells were fixed, stained with filipin (cholesterol, red) and cells. a CHO M12-A6ko cells expressing control siRNA (siCtrl) or BODIPY 493/503 (neutral lipids, green), and representative fields siRNA targeting StARD3 (siStARD3) were starved in 5% LPDS for (merged and split channels) are shown. Enlarged regions of interest 48 h before loading with 50 µg/ml LDL for 24 h. Cells were fixed are shown. For better comparison of filipin and BODIPY staining, and stained with filipin (cholesterol, red) and BODIPY 493/503 (neu- the outline and shape of cells is indicated. Scale bar, 10 μm. f West- tral lipids, green). Enlarged regions of interest are shown. For better ern blot analysis of TBC1D15, StARD3 and actin ± siRNA-mediated comparison of filipin and BODIPY staining, the outline and shape of TBC1D15 and StARD3 knockdown from lysates of CHO M12 cells cells is indicated. Scale bar, 10 μm. b Western blot analysis (StARD3, as indicated. g–h Dot-plot of number, area and relative intensity of tubulin) and RNA quantification determined by qPCR (n = 3) of filipin-stained (LE) and BODIPY-stained (LD) vesicles per cell of a StARD3 knockdown in CHO M12-A6ko cells is shown. c–d Dot- representative experiment in CHO M12 cells ± siTBC1D15 and siSt- plot of number, area and relative intensity of filipin-stained (LEs) and ARD3 as indicated (n > 60, 3 experiments). For quantification details BODIPY-stained (LDs) vesicles per cell of a representative experi- see “Materials and methods”.**p < 0.01; ***p < 0.001 by two-tailed ment (n > 60, 3 experiments). For quantification details see “Materials Student’s t test (b–d, g, h). Data are shown as mean ± SEM (b) or as and methods”. e CHO M12 cells expressing control siRNA (siCtrl) mean ± SD in red (c, d, g, h) or siRNA targeting TBC1D15 and StARD3 (siTBC1D15 siStARD3) 13 2852 E. Meneses‑Salas et al. Fig. 6  Model. Scheme of the proposed mechanism for AnxA6 in cholesterol esterification in an ACAT-dependent manner and choles- Rab7 inactivation and membrane contact site (MCS) functioning. teryl-ester storage (neutral lipids) in lipid droplets (LDs). StARD3/ Increased amounts of AnxA6 in cholesterol-laden late endosomes VAP-A seems to be instrumental for the cholesterol transfer from (LE) of NPC1 mutant cells enables the recruitment of the Rab7-GAP, late endosomes to ER through MCS. This concomitantly reduces TBC1D15, which inactivates Rab7. Lowering AnxA6 levels in late late endosome-cholesterol accumulation in NPC1 mutant cells. Most endosomes of NPC1 mutant cells leads to elevated amounts of Rab7- likely other tethers also operate at this interface, since StARD3 deple- GTP and upregulation of StARD3. This facilitates the formation of tion interferes with cholesterol transfer from LE to LD yet MCS for- MCS to establish LDL-cholesterol transfer to the ER, followed by mation was not affected (see text for details) the cell periphery [15, 16, 77]. The currently best understood droplets of CHO M12-A6ko cells. These findings are con- mechanism involves cholesterol driving ORP1L interaction sistent with StARD3 upregulation in NPC1 mutant CHO with Rab7 and phosphoinositides to facilitate minus-end cells and livers of NPC1 KO-mice [93] and StARD3 overex- transport, leading to enlarged, cholesterol-rich late endo- pression inducing MCS formation between late endosomes some/lysosome vesicles in the perinuclear region. In con- and the ER [26]. trast, when cholesterol levels are low, ORP1L undergoes Although a number of MCS constituents between late a conformational change that allows interaction with VAP endosomes and the ER have been identified (for a recent proteins in the ER and MCS formation in the cell periph- review see: [20]), the understanding of their regulation and ery, permitting cholesterol transfer between these two com- dynamics remains elusive, and additional tethering/scaffold- partments [27, 78, 89–91]. Hence, in NPC1 mutant cells ing proteins, such as MOSPD2 [75], AnxA1 [90], VPS13 late endosome-cholesterol accumulation is responsible for proteins [94], Gramd1b [39] or lipids (PtdIns) have also been late endosome/lysosome clustering and transport collapse implicated [10]. With regard to cholesterol homeostasis, at the minus-end [27]. Although ORP1L and Rab7 are the unsolved mechanisms include the kinetics and directional- main drivers in this process, NPC1 activity is required for ity of cholesterol transfer. For example, in HeLa cells the ORP1L function [78]. These findings are based on studies StARD3/VAP-A complex mediates cholesterol transport in MelJuSo [27], HeLa [78, 90, 91] and A549 [92] cells, from the ER to late endosomes independently of ORP1L yet NPC1 mutant CHO cell lines examined in the current [75, 76]; however, using the same protein machinery, trans- study do not express significant amounts of ORP1L. Hence port of cholesterol from late endosomes to the ER was also other proteins and mechanisms that allow for MCS forma- demonstrated [92]. Based on our findings, one can envis- tion and cholesterol transfer, including StARD3, need to be age that late endosome-cholesterol accumulation caused by considered. loss of NPC1 would trigger an elevation of AnxA6 levels, The involvement of MCS in Rab7-dependent late endo- possibly due to inhibition of CMA [83], and a concomitant some functionality was strongly supported by the increased increased recruitment of AnxA6 to late endosomes [43–45]. contact surface between the ER and late endosomes. In CHO M12 cells, this enlarged pool of AnxA6 proteins StARD3 depletion blocking late endosome-cholesterol could further potentiate late endosome-cholesterol accumu- export was in line with neutral lipid deposition in lipid lation and possibly impair their ability to fuse with other 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2853 structures, for example autophagosomes or phagosomes. In thankful to the staff from Centres Científics i Tecnològics (CCiTUB), line with this model, active Rab7-GTP was not detected on Universitat de Barcelona, Campus Casanova, Unitat de Microscòpia Òptica Avançada and Unitat de Microscòpia Electrònica. We are grate- phagosomes from cholesterol-laden cells [95]. This would ful to Maria Molinos (University of Barcelona) and Jennifer Reichardt not only result in an increased ability of AnxA6/TBC1D15 (School of Pharmacy, University of Sydney) for technical assistance to inactivate Rab7, but would also interfere with StARD3- and Mikko Liljeström and Biomedicum Imaging Unit (HiLIFE) for dependent MCS formation/functioning, thereby inhibiting help with live-cell microscopy. alternative late endosome-cholesterol export routes in NPC1 Author contributions E.M-S., A.G-M., K.K., P.B-M., J.C., A.E., mutant cells. S.S.B., J.J., F.M-P., V.Ll-C., K-A.R., J.H., performed all experiments Active GTP-bound lysosomal Rab7 is also involved in and prepared reagents. A.L. provided expertise, PFO constructs and MCS formation via its direct interaction with protrudin valuable suggestions and help in latest experiments for the revision. [9–12, 18, 19]. This points to dual mechanistic tasks for C.E. conducted EM studies. C.R. designed automated software for image quantification. A.P. and F.T. and all authors participated in data activated Rab7 after AnxA6 depletion in NPC1 mutant cells: analysis and discussion. T.G., C.E., E.I. and C.R. conceived the study (i) to confer increased late endosome motility and choles- and designed experiments. C.E. and T.G. wrote the manuscript with terol transport and (ii) to stabilize MCS through Rab7/pro- input from all other authors. trudin/VAP-A complex formation. This implicates that cho- lesterol transport, which is inhibited by StARD3 depletion Compliance with ethical standards in AnxA6-depleted M12 cells, and MCS formation, which is not inhibited in these settings, are regulated separately. Conflict of interest The authors declare no competing interests. Depletion of AnxA6 could stabilize StARD3/VAP-A and Open Access This article is distributed under the terms of the Crea- Rab7/protrudin/VAP-A complexes, ensuring re-establish- tive Commons Attribution 4.0 International License (http://creat​iveco​ ment of MCS between the ER and late endosomes, and even- mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribu- tually the transfer of late endosome-cholesterol to the ER in tion, and reproduction in any medium, provided you give appropriate cells with non-functional NPC1. Although several annexins credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. contribute to endosomal membrane dynamics [62, 96, 97], only the AnxA1/S100A11 protein complex has yet been associated with MCS formation [90]. While AnxA2 and AnxA6 are also well known to bind S100 proteins, to our References knowledge there is no data linking AnxA2 with MCS [90, 1. Urano Y, Watanabe H, Murphy SR, Shibuya Y, Geng Y, Peden 98]. On the other hand, in the present study, the presence of AA, Chang CC, Chang TY (2008) Transport of LDL-derived AnxA6 seems to confer untethering of MCS between LE/ cholesterol from the NPC1 compartment to the ER involves the Lys and ER in NPC1 mutant cells. Indeed, recent findings trans-Golgi network and the SNARE protein complex. Proc Natl support that untethering of mitochondria–lysosome con- Acad Sci USA 105(43):16513–16518. https​://doi.org/10.1073/ pnas.08074​50105​ tacts is mediated by the recruitment of TBC1D15 to elicit 2. Mesmin B, Maxfield FR (2009) Intracellular sterol dynamics. Bio- Rab7-GTP hydrolysis and thereby release contacts [18, 19]. chim Biophys Acta 1791(7):636–645. https​://doi.org/10.1016/j. Alternatively, AnxA6 may interact with Rab7-GTP similar bbali​p.2009.03.002 to ORP1L, which binds Rab7 via its ANK domain, exclud- 3. Ikonen E (2008) Cellular cholesterol trafficking and compart- mentalization. Nat Rev Mol Cell Biol 9(2):125–138. https​://doi. ing a direct effect on Rab7 GTPase activity [99]. org/10.1038/nrm23​36 We propose a model (Fig. 6) in which lowering AnxA6 4. Ikonen E (2018) Mechanisms of cellular cholesterol compartmen- levels on late endosomes of NPC1 mutant cells, character- talization: recent insights. Curr Opin Cell Biol 53:77–83. https​:// ized by upregulated StARD3 expression [93], leads to elevated doi.org/10.1016/j.ceb.2018.06.002 5. Gomez NM, Lu W, Lim JC, Kiselyov K, Campagno KE, Grish- Rab7-GTP levels. This enables the formation of MCS to estab- chuk Y, Slaugenhaupt SA, Pfeffer BA, Fliesler SJ, Mitchell CH lish LDL-cholesterol transfer to the ER, followed by increased (2018) Robust lysosomal calcium signaling through channel cholesterol delivery to lipid droplets, ultimately decreasing late TRPML1 is impaired by lysosomal lipid accumulation. FASEB J endosome-cholesterol accumulation in NPC1 mutant cells. 32(2):782–794. https​://doi.org/10.1096/fj.20170​0220R​R 6. Luo J, Jiang L, Yang H, Song BL (2017) Routes and mechanisms of post-endosomal cholesterol trafficking: a story that never ends. Acknowledgements This study was supported by grants BFU2015- Traffic 18(4):209–217. https​://doi.org/10.1111/tra.12471​ 66785-P, Consolider-Ingenio (CSD2009-00016 and BFU2016-81912- 7. Levine T (2004) Short-range intracellular trafficking of small mol- REDC) from the Ministerio de Economía y Competitividad (Spain) ecules across endoplasmic reticulum junctions. Trends Cell Biol to CE. TG is supported by the University of Sydney (U7113, RY253, 14(9):483–490. https​://doi.org/10.1016/j.tcb.2004.07.017 U3367), Sydney, Australia. EI is supported by the Academy of Finland 8. Phillips MJ, Voeltz GK (2016) Structure and function of ER mem- (grants 282192, 307415, 312491) and KK is thankful to grants from the brane contact sites with other organelles. Nat Rev Mol Cell Biol Finnish Cultural Foundation, Häme Regional Fund and University of 17(2):69–82. https​://doi.org/10.1038/nrm.2015.8 Helsinki. CR is supported by the Serra Húnter Programme (Generalitat 9. Raiborg C, Wenzel EM, Pedersen NM, Olsvik H, Schink KO, de Catalunya). We thank Dr A.L. Edinger, Dr M. Matsuda, Dr R.J. Schultz SW, Vietri M, Nisi V, Bucci C, Brech A, Johansen T, Youle and Dr C. Bucci for kindly providing cDNA constructs. We are 13 2854 E. Meneses‑Salas et al. Stenmark H (2015) Repeated ER-endosome contacts pro- novel molecular tether between late endosomes and the ER. J Cell mote endosome translocation and neurite outgrowth. Nature Sci 126(Pt 23):5500–5512. https​://doi.org/10.1242/jcs.13929​5 520(7546):234–238. https​://doi.org/10.1038/natur​e1435​9 27. Rocha N, Kuijl C, van der Kant R, Janssen L, Houben D, Janssen 10. Raiborg C, Wenzel EM, Pedersen NM, Stenmark H (2016) Phos- H, Zwart W, Neefjes J (2009) Cholesterol sensor ORP1L contacts phoinositides in membrane contact sites. Biochem Soc Trans the ER protein VAP to control Rab7-RILP-p150 Glued and late 44(2):425–430. https​://doi.org/10.1042/BST20​15019​0 endosome positioning. J Cell Biol 185(7):1209–1225. https​://doi. 11. Raiborg C, Wenzel EM, Stenmark H (2015) ER-endosome org/10.1083/jcb.20081​1005 contact sites: molecular compositions and functions. EMBO J 28. Charman M, Kennedy BE, Osborne N, Karten B (2010) MLN64 34(14):1848–1858. https​://doi.org/10.15252​/embj.20159​1481 mediates egress of cholesterol from endosomes to mitochondria in 12. Wijdeven RH, Jongsma ML, Neefjes J, Berlin I (2015) ER con- the absence of functional Niemann–Pick Type C1 protein. J Lipid tact sites direct late endosome transport. BioEssays 37(12):1298– Res 51(5):1023–1034. https​://doi.org/10.1194/jlr.M0023​45 1302. https​://doi.org/10.1002/bies.20150​0095 29. Elustondo P, Martin LA, Karten B (2017) Mitochondrial choles- 13. van der Kant R, Neefjes J (2014) Small regulators, major conse- terol import. Biochim Biophys Acta 1862(1):90–101. https​://doi. quences - Ca(2)(+) and cholesterol at the endosome-ER interface. org/10.1016/j.bbali​p.2016.08.012 J Cell Sci 127(Pt 5):929–938. https​://doi.org/10.1242/jcs.13753​9 30. Vassilev B, Sihto H, Li S, Holtta-Vuori M, Ilola J, Lundin J, 14. Pfisterer SG, Peranen J, Ikonen E (2016) LDL-cholesterol trans- Isola J, Kellokumpu-Lehtinen PL, Joensuu H, Ikonen E (2015) port to the endoplasmic reticulum: current concepts. Curr Opin Elevated levels of StAR-related lipid transfer protein 3 alter cho- Lipidol 27(3):282–287. https:​ //doi.org/10.1097/MOL.000000​ 0000​ lesterol balance and adhesiveness of breast cancer cells: potential 00029​2 mechanisms contributing to progression of HER2-positive breast 15. Luo J, Jiang LY, Yang H, Song BL (2018) Intracellular cholesterol cancers. Am J Pathol 185(4):987–1000. https​://doi.org/10.1016/j. transport by sterol transfer proteins at membrane contact sites. ajpat​h.2014.12.018 Trends Biochem Sci. https​://doi.org/10.1016/j.tibs.2018.10.001 31. Borthwick F, Allen AM, Taylor JM, Graham A (2010) Overex- 16. Wu H, Carvalho P, Voeltz GK (2018) Here, there, and every- pression of STARD3 in human monocyte/macrophages induces where: the importance of ER membrane contact sites. Science. an anti-atherogenic lipid phenotype. Clin Sci (Lond) 119(7):265– https​://doi.org/10.1126/scien​ce.aan58​35 272. https​://doi.org/10.1042/CS201​00266​ 17. Ridgway ND, Zhao K (2018) Cholesterol transfer at endosomal- 32. Liapis A, Chen FW, Davies JP, Wang R, Ioannou YA (2012) organelle membrane contact sites. Curr Opin Lipidol 29(3):212– MLN64 transport to the late endosome is regulated by binding to 217. https​://doi.org/10.1097/MOL.00000​00000​00050​6 14-3-3 via a non-canonical binding site. PLoS One 7(4):e34424. 18. Wong LH, Eden ER, Futter CE (2018) Roles for ER:endosome https​://doi.org/10.1371/journ​al.pone.00344​24 membrane contact sites in ligand-stimulated intraluminal vesi- 33. Bucci C, Thomsen P, Nicoziani P, McCarthy J, van Deurs B cle formation. Biochem Soc Trans 46(5):1055–1062. https​://doi. (2000) Rab7: a key to lysosome biogenesis. Mol Biol Cell org/10.1042/BST20​17043​2 11(2):467–480 19. Wong YC, Ysselstein D, Krainc D (2018) Mitochondria-lysosome 34. Ganley IG, Wong PM, Gammoh N, Jiang X (2011) Distinct contacts regulate mitochondrial fission via RAB7 GTP hydrolysis. autophagosomal-lysosomal fusion mechanism revealed by thap- Nature 554(7692):382–386. https​://doi.org/10.1038/natur​e2548​6 sigargin-induced autophagy arrest. Mol Cell 42(6):731–743. https​ 20. Scorrano L, De Matteis MA, Emr S, Giordano F, Hajnoczky G, ://doi.org/10.1016/j.molce​l.2011.04.024 Kornmann B, Lackner LL, Levine TP, Pellegrini L, Reinisch K, 35. Lebrand C, Corti M, Goodson H, Cosson P, Cavalli V, Mayran Rizzuto R, Simmen T, Stenmark H, Ungermann C, Schuldiner M N, Faure J, Gruenberg J (2002) Late endosome motility depends (2019) Coming together to define membrane contact sites. Nat on lipids via the small GTPase Rab7. EMBO J 21(6):1289–1300. Commun 10(1):1287. https​://doi.org/10.1038/s4146​7-019-09253​ https​://doi.org/10.1093/emboj​/21.6.1289 -3 36. Choudhury A, Dominguez M, Puri V, Sharma DK, Narita K, 21. Hulce JJ, Cognetta AB, Niphakis MJ, Tully SE, Cravatt BF Wheatley CL, Marks DL, Pagano RE (2002) Rab proteins medi- (2013) Proteome-wide mapping of cholesterol-interacting pro- ate Golgi transport of caveola-internalized glycosphingolipids and teins in mammalian cells. Nat Methods 10(3):259–264. https​:// correct lipid trafficking in Niemann–Pick C cells. J Clin Invest doi.org/10.1038/nmeth​.2368 109(12):1541–1550. https​://doi.org/10.1172/JCI15​420 22. Alpy F, Latchumanan VK, Kedinger V, Janoshazi A, Thiele C, 37. Linder MD, Uronen RL, Holtta-Vuori M, van der Sluijs P, Peranen Wendling C, Rio MC, Tomasetto C (2005) Functional charac- J, Ikonen E (2007) Rab8-dependent recycling promotes endoso- terization of the MENTAL domain. J Biol Chem 280(18):17945– mal cholesterol removal in normal and sphingolipidosis cells. Mol 17952. https​://doi.org/10.1074/jbc.M5007​23200​ Biol Cell 18(1):47–56. https​://doi.org/10.1091/mbc.E06-07-0575 23. Holtta-Vuori M, Alpy F, Tanhuanpaa K, Jokitalo E, Mutka AL, 38. Cianciola NL, Greene DJ, Morton RE, Carlin CR (2013) Ade- Ikonen E (2005) MLN64 is involved in actin-mediated dynam- novirus RIDalpha uncovers a novel pathway requiring ORP1L ics of late endocytic organelles. Mol Biol Cell 16(8):3873–3886. for lipid droplet formation independent of NPC1. Mol Biol Cell https​://doi.org/10.1091/mbc.E04-12-1105 24(21):3309–3325. https​://doi.org/10.1091/mbc.E12-10-0760 24. Alpy F, Stoeckel ME, Dierich A, Escola JM, Wendling C, Chenard 39. Hoglinger D, Burgoyne T, Sanchez-Heras E, Hartwig P, Colaco A, MP, Vanier MT, Gruenberg J, Tomasetto C, Rio MC (2001) The Newton J, Futter CE, Spiegel S, Platt FM, Eden ER (2019) NPC1 steroidogenic acute regulatory protein homolog MLN64, a late regulates ER contacts with endocytic organelles to mediate cho- endosomal cholesterol-binding protein. J Biol Chem 276(6):4261– lesterol egress. Nat Commun 10(1):4276. https:​ //doi.org/10.1038/ 4269. https​://doi.org/10.1074/jbc.M0062​79200​ s4146​7-019-12152​-2 25. van der Kant R, Zondervan I, Janssen L, Neefjes J (2013) Cho- 40. Garcia-Melero A, Reverter M, Hoque M, Meneses-Salas E, Koese lesterol-binding molecules MLN64 and ORP1L mark distinct M, Conway JR, Johnsen CH, Alvarez-Guaita A, Morales-Paytuvi late endosomes with transporters ABCA3 and NPC1. J Lipid Res F, Elmaghrabi YA, Pol A, Tebar F, Murray RZ, Timpson P, Enrich 54(8):2153–2165. https​://doi.org/10.1194/jlr.M0373​25 C, Grewal T, Rentero C (2016) Annexin A6 and late endosomal 26. Alpy F, Rousseau A, Schwab Y, Legueux F, Stoll I, Wendling cholesterol modulate integrin recycling and cell migration. J Biol C, Spiegelhalter C, Kessler P, Mathelin C, Rio MC, Levine TP, Chem 291(3):1320–1335. https​://doi.org/10.1074/jbc.M115.68355​7 Tomasetto C (2013) STARD3 or STARD3NL and VAP form a 41. Enrich C, Rentero C, de Muga SV, Reverter M, Mulay V, Wood P, Koese M, Grewal T (2011) Annexin A6-linking Ca(2+) signaling 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2855 with cholesterol transport. Biochim Biophys Acta 1813(5):935– Acad Sci USA 107(45):19338–19343. https​://doi.org/10.1073/ 947. https​://doi.org/10.1016/j.bbamc​r.2010.09.015 pnas.10105​54107​ 42. Gerke V, Creutz CE, Moss SE (2005) Annexins: linking Ca2+ sig- 57. Itoh RE, Kurokawa K, Fujioka A, Sharma A, Mayer BJ, Mat- nalling to membrane dynamics. Nat Rev Mol Cell Biol 6(6):449– suda M (2005) A FRET-based probe for epidermal growth factor 461. https​://doi.org/10.1038/nrm16​61 receptor bound non-covalently to a pair of synthetic amphipathic 43. Grewal T, Heeren J, Mewawala D, Schnitgerhans T, Wendt D, helixes. Exp Cell Res 307(1):142–152. https​://doi.org/10.1016/j. Salomon G, Enrich C, Beisiegel U, Jackle S (2000) Annexin VI yexcr​.2005.02.026 stimulates endocytosis and is involved in the trafficking of low 58. Yamano K, Fogel AI, Wang C, van der Bliek AM, Youle RJ density lipoprotein to the prelysosomal compartment. J Biol Chem (2014) Mitochondrial Rab GAPs govern autophagosome biogen- 275(43):33806–33813. https​://doi.org/10.1074/jbc.M0026​62200​ esis during mitophagy. Elife 3:e01612. https​://doi.org/10.7554/ 44. de Diego I, Schwartz F, Siegfried H, Dauterstedt P, Heeren J, eLife​.01612​ Beisiegel U, Enrich C, Grewal T (2002) Cholesterol modulates 59. Cantalupo G, Alifano P, Roberti V, Bruni CB, Bucci C (2001) the membrane binding and intracellular distribution of annexin Rab-interacting lysosomal protein (RILP): the Rab7 effector 6. J Biol Chem 277(35):32187–32194. https​://doi.org/10.1074/ required for transport to lysosomes. EMBO J 20(4):683–693. https​ jbc.M2054​99200​ ://doi.org/10.1093/emboj​/20.4.683 45. te Vruchte D, Lloyd-Evans E, Veldman RJ, Neville DC, Dwek RA, 60. Das A, Goldstein JL, Anderson DD, Brown MS, Radhakrishnan Platt FM, van Blitterswijk WJ, Sillence DJ (2004) Accumulation A (2013) Use of mutant 125I-perfringolysin O to probe trans- of glycosphingolipids in Niemann-Pick C disease disrupts endo- port and organization of cholesterol in membranes of animal somal transport. J Biol Chem 279(25):26167–26175. https​://doi. cells. Proc Natl Acad Sci USA 110(26):10580–10585. https​:// org/10.1074/jbc.M3115​91200​ doi.org/10.1073/pnas.13092​73110​ 46. Cubells L, Vila de Muga S, Tebar F, Wood P, Evans R, Ingelmo- 61. Ollion J, Cochennec J, Loll F, Escude C, Boudier T (2013) Torres M, Calvo M, Gaus K, Pol A, Grewal T, Enrich C (2007) TANGO: a generic tool for high-throughput 3D image analysis Annexin A6-induced alterations in cholesterol transport and cave- for studying nuclear organization. Bioinformatics 29(14):1840– olin export from the Golgi complex. Traffic 8(11):1568–1589. 1841. https​://doi.org/10.1093/bioin​forma​tics/btt27​6 https​://doi.org/10.1111/j.1600-0854.2007.00640​.x 62. Rentero C, Blanco-Munoz P, Meneses-Salas E, Grewal T, Enrich 47. Havel RJ, Eder HA, Bragdon JH (1955) The distribution and C (2018) Annexins-coordinators of cholesterol homeostasis in chemical composition of ultracentrifugally separated lipopro- endocytic pathways. Int J Mol Sci. https​://doi.org/10.3390/ijms1​ teins in human serum. J Clin Invest 34(9):1345–1353. https​://doi. 90514​44 org/10.1172/JCI10​3182 63. Kwiatkowska K, Marszalek-Sadowska E, Traczyk G, Koprowski 48. Goldstein JL, Basu SK, Brown MS (1983) Receptor-mediated P, Musielak M, Lugowska A, Kulma M, Grzelczyk A, Sobota A endocytosis of low-density lipoprotein in cultured cells. Methods (2014) Visualization of cholesterol deposits in lysosomes of Nie- Enzymol 98:241–260 mann-Pick type C fibroblasts using recombinant perfringolysin O. 49. Cubells L, Vila de Muga S, Tebar F, Bonventre JV, Balsinde J, Orphanet J Rare Dis 9:64. https:​ //doi.org/10.1186/1750-1172-9-64 Pol A, Grewal T, Enrich C (2008) Annexin A6-induced inhibition 64. Peralta ER, Martin BC, Edinger AL (2010) Differential effects of of cytoplasmic phospholipase A2 is linked to caveolin-1 export TBC1D15 and mammalian Vps39 on Rab7 activation state, lyso- from the Golgi. J Biol Chem 283(15):10174–10183. https​://doi. somal morphology, and growth factor dependence. J Biol Chem org/10.1074/jbc.M7066​18200​ 285(22):16814–16821. https​://doi.org/10.1074/jbc.M110.11163​3 50. Grewal T, Evans R, Rentero C, Tebar F, Cubells L, de Diego 65. Chen YN, Gu X, Zhou XE, Wang W, Cheng D, Ge Y, Ye F, Xu I, Kirchhoff MF, Hughes WE, Heeren J, Rye KA, Rinninger F, HE, Lv Z (2017) Crystal structure of TBC1D15 GTPase-acti- Daly RJ, Pol A, Enrich C (2005) Annexin A6 stimulates the vating protein (GAP) domain and its activity on Rab GTPases. membrane recruitment of p120GAP to modulate Ras and Raf-1 Protein Sci 26(4):834–846. https​://doi.org/10.1002/pro.3132 activity. Oncogene 24(38):5809–5820. https​://doi.org/10.1038/ 66. Frasa MA, Koessmeier KT, Ahmadian MR, Braga VM (2012) sj.onc.12087​43 Illuminating the functional and structural repertoire of human 51. Alvarez-Guaita A, Vila de Muga S, Owen DM, Williamson D, TBC/RABGAPs. Nat Rev Mol Cell Biol 13(2):67–73. https​:// Magenau A, Garcia-Melero A, Reverter M, Hoque M, Cairns doi.org/10.1038/nrm32​67 R, Cornely R, Tebar F, Grewal T, Gaus K, Ayala-Sanmartin J, 67. Fukuda M (2011) TBC proteins: Gaps for mammalian small Enrich C, Rentero C (2015) Evidence for annexin A6-dependent GTPase Rab? Biosci Rep 31(3):159–168. https:​ //doi.org/10.1042/ plasma membrane remodelling of lipid domains. Br J Pharmacol BSR20​10011​2 172(7):1677–1690. https​://doi.org/10.1111/bph.13022​ 68. Nottingham RM, Pfeffer SR (2009) Defining the boundaries: rab 52. Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F GEFs and GAPs. Proc Natl Acad Sci USA 106(34):14185–14186. (2013) Genome engineering using the CRISPR-Cas9 system. Nat https​://doi.org/10.1073/pnas.09077​25106​ Protoc 8(11):2281–2308. https​://doi.org/10.1038/nprot​.2013.143 69. Zhang XM, Walsh B, Mitchell CA, Rowe T (2005) TBC domain 53. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to family, member 15 is a novel mammalian Rab GTPase-activat- ImageJ: 25 years of image analysis. Nat Methods 9(7):671–675 ing protein with substrate preference for Rab7. Biochem Bio- 54. Pons M, Ihrke G, Koch S, Biermer M, Pol A, Grewal T, Jackle phys Res Commun 335(1):154–161. https​://doi.org/10.1016/j. S, Enrich C (2000) Late endocytic compartments are major sites bbrc.2005.07.070 of annexin VI localization in NRK fibroblasts and polarized 70. Rentero C, Evans R, Wood P, Tebar F, Vila de Muga S, Cubells WIF-B hepatoma cells. Exp Cell Res 257(1):33–47. https​://doi. L, de Diego I, Hayes TE, Hughes WE, Pol A, Rye KA, Enrich org/10.1006/excr.2000.4861 C, Grewal T (2006) Inhibition of H-Ras and MAPK is compen- 55. Kanerva K, Uronen RL, Blom T, Li S, Bittman R, Lappalainen P, sated by PKC-dependent pathways in annexin A6 expressing Peranen J, Raposo G, Ikonen E (2013) LDL cholesterol recycles cells. Cell Signal 18(7):1006–1016. https:​ //doi.org/10.1016/j.cells​ to the plasma membrane via a Rab8a-Myosin5b-actin-dependent ig.2005.08.008 membrane transport route. Dev Cell 27(3):249–262. https​://doi. 71. Heid HW, Moll R, Schwetlick I, Rackwitz HR, Keenan TW (1998) org/10.1016/j.devce​l.2013.09.016 Adipophilin is a specific marker of lipid accumulation in diverse 56. Sun Q, Westphal W, Wong KN, Tan I, Zhong Q (2010) Rubi- cell types and diseases. Cell Tissue Res 294(2):309–321 con controls endosome maturation as a Rab7 effector. Proc Natl 13 2856 E. Meneses‑Salas et al. 72. Brown MS, Radhakrishnan A, Goldstein JL (2017) Retrospec- 87. Castellano BM, Thelen AM, Moldavski O, Feltes M, van der tive on cholesterol homeostasis: the central role of scap. Annu Welle RE, Mydock-McGrane L, Jiang X, van Eijkeren RJ, Davis Rev Biochem. https​://doi.org/10.1146/annur​ev-bioch​em-06291​ OB, Louie SM, Perera RM, Covey DF, Nomura DK, Ory DS, 7-01185​2 Zoncu R (2017) Lysosomal cholesterol activates mTORC1 via 73. Kristiana I, Yang H, Brown AJ (2008) Different kinetics of cho- an SLC38A9-Niemann-Pick C1 signaling complex. Science lesterol delivery to components of the cholesterol homeostatic 355(6331):1306–1311. https​://doi.org/10.1126/scien​ce.aag14​17 machinery: implications for cholesterol trafficking to the endo- 88. Tsuji T, Fujimoto M, Tatematsu T, Cheng J, Orii M, Takatori S, plasmic reticulum. Biochim Biophys Acta 1781(11–12):724–730. Fujimoto T (2017) Niemann-Pick type C proteins promote micro- https​://doi.org/10.1016/j.bbali​p.2008.08.006 autophagy by expanding raft-like membrane domains in the yeast 74. Ross AC, Go KJ, Heider JG, Rothblat GH (1984) Selective inhibi- vacuole. Elife. https​://doi.org/10.7554/eLife​.25960​ tion of acyl coenzyme A:cholesterol acyltransferase by compound 89. Johansson M, Rocha N, Zwart W, Jordens I, Janssen L, Kuijl C, 58-035. J Biol Chem 259(2):815–819 Olkkonen VM, Neefjes J (2007) Activation of endosomal dynein 75. Di Mattia T, Wilhelm LP, Ikhlef S, Wendling C, Spehner D, motors by stepwise assembly of Rab7-RILP-p150Glued, ORP1L, Nomine Y, Giordano F, Mathelin C, Drin G, Tomasetto C, Alpy and the receptor betalll spectrin. J Cell Biol 176(4):459–471. https​ F (2018) Identification of MOSPD2, a novel scaffold for endo- ://doi.org/10.1083/jcb.20060​6077 plasmic reticulum membrane contact sites. EMBO Rep. https​:// 90. Eden ER, Sanchez-Heras E, Tsapara A, Sobota A, Levine TP, doi.org/10.15252​/embr.20174​5453 Futter CE (2016) Annexin A1 Tethers membrane contact sites 76. Wilhelm LP, Wendling C, Vedie B, Kobayashi T, Chenard MP, that mediate ER to endosome cholesterol transport. Dev Cell Tomasetto C, Drin G, Alpy F (2017) STARD3 mediates endo- 37(5):473–483. https​://doi.org/10.1016/j.devce​l.2016.05.005 plasmic reticulum-to-endosome cholesterol transport at mem- 91. Elgner F, Ren H, Medvedev R, Ploen D, Himmelsbach K, Boller brane contact sites. EMBO J 36(10):1412–1433. https​://doi. K, Hildt E (2016) The intracellular cholesterol transport inhibitor org/10.15252​/embj.20169​5917 U18666A inhibits the exosome-dependent release of mature hepa- 77. Cabukusta B, Neefjes J (2018) Mechanisms of lysosomal posi- titis C virus. J Virol 90(24):11181–11196. https:​ //doi.org/10.1128/ tioning and movement. Traffic 19(10):761–769. https​://doi. JVI.01053​-16 org/10.1111/tra.12587​ 92. Cianciola NL, Chung S, Manor D, Carlin CR (2017) Adenovi- 78. Zhao K, Ridgway ND (2017) Oxysterol-binding protein-related rus modulates Toll-Like receptor 4 signaling by reprogramming protein 1l regulates cholesterol egress from the endo-lysosomal ORP1L-VAP protein contacts for cholesterol transport from system. Cell Rep 19(9):1807–1818. https:​ //doi.org/10.1016/j.celre​ endosomes to the endoplasmic reticulum. J Virol. https​://doi. p.2017.05.028 org/10.1128/jvi.01904​-16 79. Du X, Kumar J, Ferguson C, Schulz TA, Ong YS, Hong W, Prinz 93. Balboa E, Castro J, Pinochet MJ, Cancino GI, Matias N, Jose WA, Parton RG, Brown AJ, Yang H (2011) A role for oxysterol- Saez P, Martinez A, Alvarez AR, Garcia-Ruiz C, Fernandez- binding protein-related protein 5 in endosomal cholesterol traf- Checa JC, Zanlungo S (2017) MLN64 induces mitochondrial ficking. J Cell Biol 192(1):121–135. https​://doi.org/10.1083/ dysfunction associated with increased mitochondrial cholesterol jcb.20100​4142 content. Redox Biol 12:274–284. https​://doi.org/10.1016/j.redox​ 80. Vacca F, Scott C, Gruenberg J (2016) The Late Endosome. Encycl .2017.02.024 Cell Biol 2:201–210. https​://doi.org/10.1016/B978-0-12-39444​ 94. Kumar N, Leonzino M, Hancock-Cerutti W, Horenkamp FA, Li P, 7-4.20017​-5 Lees JA, Wheeler H, Reinisch KM, De Camilli P (2018) VPS13A 81. Cuervo AM, Gomes AV, Barnes JA, Dice JF (2000) Selective and VPS13C are lipid transport proteins differentially localized degradation of annexins by chaperone-mediated autophagy. J Biol at ER contact sites. J Cell Biol 217(10):3625–3639. https​://doi. Chem 275(43):33329–33335. https​://doi.org/10.1074/jbc.M0056​ org/10.1083/jcb.20180​7019 55200​ 95. Huynh KK, Gershenzon E, Grinstein S (2008) Cholesterol accu- 82. Kaushik S, Cuervo AM (2012) Chaperone-mediated autophagy: mulation by macrophages impairs phagosome maturation. J Biol a unique way to enter the lysosome world. Trends Cell Biol Chem 283(51):35745–35755. https​://doi.org/10.1074/jbc.M8062​ 22(8):407–417. https​://doi.org/10.1016/j.tcb.2012.05.006 32200​ 83. Rodriguez-Navarro JA, Kaushik S, Koga H, Dall’Armi C, Shui G, 96. Enrich C, Rentero C, Meneses-Salas E, Tebar F, Grewal T (2017) Wenk MR, Di Paolo G, Cuervo AM (2012) Inhibitory effect of Annexins: Ca(2+) effectors determining membrane trafficking in dietary lipids on chaperone-mediated autophagy. Proc Natl Acad the late endocytic compartment. Adv Exp Med Biol 981:351–385. Sci USA 109(12):E705–E714. https:​ //doi.org/10.1073/pnas.11130​ https​://doi.org/10.1007/978-3-319-55858​-5_14 36109​ 97. Gruenberg J, Stenmark H (2004) The biogenesis of multivesicu- 84. Sobo K, Le Blanc I, Luyet PP, Fivaz M, Ferguson C, Parton RG, lar endosomes. Nat Rev Mol Cell Biol 5(4):317–323. https​://doi. Gruenberg J, van der Goot FG (2007) Late endosomal cholesterol org/10.1038/nrm13​60 accumulation leads to impaired intra-endosomal trafficking. PLoS 98. Rintala-Dempsey AC, Rezvanpour A, Shaw GS (2008) S100- One 2(9):e851. https​://doi.org/10.1371/journ​al.pone.00008​51 annexin complexes–structural insights. FEBS J 275(20):4956– 85. Fraldi A, Annunziata F, Lombardi A, Kaiser HJ, Medina DL, 4966. https​://doi.org/10.1111/j.1742-4658.2008.06654​.x Spampanato C, Fedele AO, Polishchuk R, Sorrentino NC, Simons 99. Tong J, Tan L, Chun C, Im YJ (2019) Structural basis of human K, Ballabio A (2010) Lysosomal fusion and SNARE function ORP1-Rab7 interaction for the late-endosome and lysosome tar- are impaired by cholesterol accumulation in lysosomal storage geting. PLoS One 14(2):e0211724. https​://doi.org/10.1371/journ​ disorders. EMBO J 29(21):3607–3620. https​://doi.org/10.1038/ al.pone.02117​24 emboj​.2010.237 86. Koga H, Kaushik S, Cuervo AM (2010) Altered lipid content Publisher’s Note Springer Nature remains neutral with regard to inhibits autophagic vesicular fusion. FASEB J 24(8):3052–3065. jurisdictional claims in published maps and institutional affiliations. https​://doi.org/10.1096/fj.09-14451​9 13 Annexin A6 modulates TBC1D15/Rab7/StARD3 axis to control endosomal cholesterol export in… 2857 Affiliations Elsa Meneses‑Salas1,2 · Ana García‑Melero1 · Kristiina Kanerva3,4 · Patricia Blanco‑Muñoz1,2 · Frederic Morales‑Paytuvi1,2 · Júlia Bonjoch1,2 · Josefina Casas5 · Antonia Egert6 · Syed S. Beevi6 · Jaimy Jose6 · Vicenta Llorente‑Cortés7,8,9 · Kerry‑Anne Rye10 · Joerg Heeren11 · Albert Lu12 · Albert Pol1,2,13 · Francesc Tebar1,2 · Elina Ikonen3,4 · Thomas Grewal6 · Carlos Enrich1,2 · Carles Rentero1,2 1 8 Departament de Biomedicina, Unitat de Biologia Cel·lular, CIBERCV, Institute of Health Carlos III, Madrid, Spain Facultat de Medicina i Ciències de la Salut, Universitat de 9 Biomedical Research Institute of Barcelona-CSIC, Barcelona, 08036 Barcelona, Spain Barcelona, Spain 2 Centre de Recerca Biomèdica CELLEX, Institut 10 School of Medical Sciences, University of New South Wales, d’Investigacions Biomèdiques August Pi i Sunyer Sydney, NSW 2052, Australia (IDIBAPS), 08036 Barcelona, Spain 11 3 Department of Biochemistry and Molecular Biology Faculty of Medicine, Anatomy, University of Helsinki, II: Molecular Cell Biology, University Medical Center 00014 Helsinki, Finland Hamburg-Eppendorf, Martinistraße 52, 20246 Hamburg, 4 Minerva Foundation Institute for Medical Research, Germany 00290 Helsinki, Finland 12 Department of Biochemistry, Stanford University School 5 Research Unit on BioActive Molecules (RUBAM), of Medicine, Stanford, USA Department of Biological Chemistry, Institute for Advanced 13 Institució Catalana de Recerca i Estudis Avaçats (ICREA), Chemistry of Catalonia (IQAC-CSIC), Barcelona, Spain 08010 Barcelona, Spain 6 School of Pharmacy, Faculty of Medicine and Health, University of Sydney, Sydney, NSW 2006, Australia 7 Lipids and Cardiovascular Pathology Group, Biomedical Research Institute Sant Pau (IIB Sant Pau), Barcelona, Spain 13