Pathological roles of the VEGF/SphK pathway in Niemann Pick type C neurons

Transcription

1 Received May Accepted 8 Oct Published Nov DOI:.8/ncomms65 OPEN Pathological roles of the /SphK pathway in Niemann Pick type C s Hyun Lee,,, Jong Kil Lee,,,, Min Hee Park,,, Yu Ri Hong,, Hugo H. Marti 5, Hyongbum Kim 6, Yohei Okada 7, Makoto Otsu 8, Eul-Ju Seo 9, Jae-Hyung Park, Jae-Hoon Bae, Nozomu Okino, Xingxuan He, Edward H. Schuchman, Jae-sung Bae,, & Hee Kyung Jin, is a major storage compound in Niemann Pick type C disease (NP C), although the pathological role(s) of this accumulation have not been fully characterized. Here we found that sphingosine kinase (SphK) activity is reduced in NP C patient fibroblasts and NP C mouse Purkinje s (PNs) due to defective vascular endothelial growth factor () levels. accumulation due to inactivation of /SphK pathway led to PNs loss via inhibition of autophagosome lysosome fusion in NP C mice. activates SphK by binding to R, resulting in decreased sphingosine storage as well as improved PNs survival and clinical outcomes in NP C cells and mice. We also show that induced pluripotent stem cell (ipsc)-derived human NP C s are generated and the abnormalities caused by /SphK inactivity in these cells are corrected by replenishment of. Overall, these results reveal a pathogenic mechanism in NP C s where defective SphK activity is due to impaired levels. Stem Cell Neuroplasticity Research Group, Kyungpook National University, Daegu 7-7, Korea. Department of Laboratory Animal Medicine, Cell and Matrix Research Institute, College of Veterinary Medicine, Kyungpook National University, Daegu 7-7, Korea. Department of Physiology, Cell and Matrix Research Institute, School of Medicine, Kyungpook National University, Daegu 7-8, Korea. Department of Biomedical Science, BK Plus KNU Biomedical Convergence Program, Kyungpook National University, Daegu 7-8, Korea. 5 Institute of Physiology and Pathophysiology, University of Heidelberg, Heidelberg 69, Germany. 6 Graduate School of Biomedical Science and Engineering/College of Medicine, Hanyang University, Seoul -79, Korea. 7 Department of Physiology, School of Medicine, Keio University, Tokyo 6-858, Japan. 8 Division of Stem Cell Therapy, Center for Stem Cell Biology and Regenerative Medicine, Institute of Medical Science, University of Tokyo, Tokyo 8-869, Japan. 9 Department of Laboratory Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul 8-76, Korea. Department of Physiology, School of Medicine, Keimyung University, Daegu 7-7, Korea. Department of Bioscience and Biotechnology, Graduate School of Bioresource and Bioenvironmental Sciences, Kyushu University, Fukuoka 8-858, Japan. Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York, New York 9, USA. These authors contributed equally to this work. Correspondence and requests for materials should be addressed to J.S.B. ( jsbae@knu.ac.kr) or to H.K.J. ( hkjin@knu.ac.kr). NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

2 NATURE COMMUNICATIONS DOI:.8/ncomms65 Niemann Pick type C disease (NP C) is an inherited lipid storage disorder that affects the central nervous system. Recent studies have shown that sphingosine is a major and initiating storage compound in NP C,. However, the underlying mechanism(s) leading to sphingosine storage, as well as its role in NP C pathogenesis such as al loss, remains largely unknown. Our previous studies have shown that bone marrow mesenchymal stem cells (s) contribute to improved neurological function in the NP C mice 5,6. Furthermore, we have postulated that the prosurvival effects of s on NP C Purkinje s (PNs) are paracrine effects that restore the sphingolipid imbalance, as evidenced by decreased sphingosine and increased sphingosine--phosphate (SP) levels 7. Therefore, we speculated that sphingolipid-modulating factors derived from s are potential therapeutic agents for this disease. Sphingolipid-metabolizing enzymes control the cellular dynamic balance of bioactive lipids, including the proapoptotic compound sphingosine and the proliferative compound SP 8. kinase (SphK) is a key enzyme that converts sphingosine into SP. SphK can be activated by numerous external stimuli 9, resulting in a decrease in intracellular sphingosine and increase in SP. On the basis of these concepts and findings, we hypothesized that defects of SphK activators could be involved in the pathogenesis of NP C, and explored candidate therapeutic factors secreted by s that might influence the activation of SphK. Here we show that NPC deficiency markedly reduces vascular endothelial growth factor () expression, and that decreased levels cause impaired SphK activity in PNs. Abnormal sphingosine storage by -mediated SphK inactivity causes a decreased PN survival via disruption of autophagosome lysosome fusion. Further, replenishment of leads to restoration of SphK activity and improvement of pathology by binding to the receptor- (R) in NP C mice PNs as well as patient-specific cells, preventing sphingosine accumulation, autophagy dysfunction and abnormal calcium homeostasis. Results SphK activity is reduced in NP C patients and NP C mice. We first determined whether defects of SphK could be involved in NP C and responsible for the elevated sphingosine. SphK was significantly decreased in fibroblasts from NP C patients compared with normal control fibroblasts (Fig. a). These levels did not change as the passage numbers increased (Fig. a). SphK activity also was decreased in the cerebellum and primary cerebellar PNs from NP C mice compared with those of wild-type () mice (Fig. a). These results confirmed that SphK, a key enzyme in modulating the levels of sphingosine, is diminished in NP C, and that the reduction of this activity may influence disease progression and/or pathogenesis. -derived restores SphK activity in NP C mouse PNs. To examine whether bioactive, soluble factors released from s affected SphK activity in NP C, we cocultured BM- MSCs with PNs using an indirect coculture system (see Methods). We found that when NP C PNs were cocultured with s, their SphK activity was significantly increased (Fig. b). To identify the soluble factors that were released from the s and might be responsible for the increased SphK activity, we screened and compared the conditioned media (CM) of PNs grown with and without s using an antibody-based mouse cytokine array (Supplementary Fig. a,b). The CM of NP C PNs cocultured with s revealed stronger signals in four array spots in comparison with the CM of NP C PNs alone (Supplementary Fig. c,d). To confirm the secretion of these factors, we performed enzyme-linked immunosorbent assays (ELISA). Of the selected cytokines, only levels were significantly elevated in the CM of NP C PNs cocultured with BM- MSCs. We also found that was significantly decreased in NP C PNs cultured alone compared with PNs (Fig. c). To confirm these effects in PNs, we performed immunostaining. was normally expressed in PNs, but the expression levels were lower in NP C PNs compared with PNs. When the NP C PNs were cocultured with s, intensity of expression was increased (Fig. d). These data identified as a potential candidate molecule that could modulate SphK and may influence pathogenesis in NP C PNs. To further examine the effects of -derived on SphK activity in NP C PNs, we used small interfering RNA ()-treated s and -overexpressing s (the latter derived from tg mice; ref. ; Supplementary Fig. a). As predicted, SphK activity was significantly increased in NP C PNs cocultured with s and tg s compared with NP C PNs alone. However, the activity did not show any changes in NP C PNs cocultured with -treated s (Fig. e). Consistent with this observation, sphingosine and SP levels in the cocultured NP C PNs were altered relative to the amount of released from the s (Supplementary Fig. b,c). We also performed SP immunostaining in PNs. SP was mainly expressed in PNs, and the expression was significantly increased in NP C PNs cocultured with normal or tg s. However, it was not increased when the cells were cocultured with -treated s (Supplementary Fig. d). binds to two tyrosine kinase receptors, known as R and (ref. 5). Among these receptors, R is highly expressed on PNs 6. To examine whether from BM- MSCs improved the sphingolipid imbalance in NP C PNs by binding to R, we treated NP C PNs with the R tyrosine kinase inhibitor PTK787 before coculturing 7. We found that SphK activity and other sphingolipid metabolites in NP C PNs were mediated by interactions of -derived and its receptor R (Fig. f; Supplementary Fig. e). These results indicated that -mediated restoration of abnormal SphK activity could be due the secreted binding to the R in NP C PNs. Next, to determine whether the -mediated SphK modulation by s promoted the survival of NP C PNs, we determined cell counts after coculture. When NP C PNs were cocultured with s or tg s, the number of PNs was significantly increased. This effect was lower when s were cocultured with the NP C PNs, although this did not reach statistical significance (Fig. g). Finally, to gain more direct insights into the relationship between and SphK activity in NP C PNs, we treated PNs with and determined the changes of sphingolipid factors. treatment of PNs strongly reduced SphK levels and led to elevation of sphingosine and reduction of SP, similar to NP C PNs (Fig. h; Supplementary Fig. f,g). The survival of PNs was also significantly decreased following transfection (Fig. i). These results suggested that inactivation of may lead to reduced SphK activity in NP C PNs. from s reduces pathology in PNs of NP C mice. To examine the in vivo effects of derived from BM- MSCs on SphK activity of PNs, we transplanted s into the cerebellum of NP C mice (Fig. a). At one day after NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

3 NATURE COMMUNICATIONS DOI:.8/ncomms65 a d f Calbindin SphK activity (%) Human fibroblast Cerebellum PN P P P5 5 PN 5 : PN PN PTK787-pretreated PN + + h 5 5 PN g PN PN PN i ized intensity in PN PN.5.5 PN PN b 5 PN 5 PN tg PN PN PN PN PN PN PN e SphK activity (%) PN density (cells mm ) PN density (cells mm ) PN PN c (pg ml ) PN PN 8 PN PN PN PN PN PN PN PN tg PN PN PN tg Figure -derived restores SphK activity in NP C mice PNs. (a) SphK activities between NP C and control were analysed in human fibroblast (n ¼ 7 per group), mouse cerebellum tissue (n ¼ 7 per group) and primary mouse PN samples (n ¼ 9 per group). SphK activity did not show passage differences between NP C and normal fibroblasts. (b) Three days after cocultures, we measured SphK activities in PNs derived from and NP C mice (n ¼ 8 per group). (c) levels were measured in CM derived from PNs with or without s by ELISA (n ¼ 7 per group). (d) Primary cultures of NP C PNs were immunostained with anti-calbindin and anti- (scale bar, 5 mm). Arrowheads indicate expression by PNs. Values represent normalized fluorescence intensities of in PNs ( PN, n ¼ 8; and NP C PN, n ¼ 9). (e) SphK activities were measured in NP C PNs alone (n ¼ 7) and NP C PNs cocultured with s, s and tg s (n ¼ 8 per group). (f) Effect of the PTK787 on s mediated SphK activation. NP C PNs were pretreated with PTK787 at mm for day and cocultured for days with s, and then SphK activity was assayed (n ¼ 7 per group). (g) Representative images of PNs stained with anti-calbindin (scale bar, mm). The mean number of PNs per squared millimetre was counted (n ¼ 8 per group). (h) Effect of knockdown on SphK activity in PNs (control, n ¼ 6; and, n ¼ 8 per group). (i) Representative images and quantification of al survival in normal and -knockdown PNs (scale bar, 5 mm; n ¼ 8 per group). a,h,i, Student s t-test. b g, one-way analysis of variance, Tukey s post hoc test. Po.5, Po., Po.5. All error bars indicate s.e.m. transplantation, SphK activity was significantly increased in the cerebellum of NP C mice compared with phosphate-buffered saline (PBS)-infused counterparts (Fig. b). transplantation also increased protein levels in the cerebellum of NP C mice (Fig. c). The elevated expression of was significant in the Purkinje cell layer (PCL) of the NP C mouse cerebellums, consistent with the decreased levels in nontreated NP C PNs compared with (Fig. d). However, BM- MSCs did not increase SphK or levels in normal cerebellums, consistent with previous reports 6,8. We also transplanted s and tg s into the cerebellum of NP C mice. As predicted, at one day after transplantation, SphK activity was significantly increased in the cerebellum of NP C mice treated with tg s. However, mice treated with s showed significantly lower SphK activity (Fig. e). The sphingosine and SP metabolites were also changed in NP C PNs in relation to SphK and levels (Supplementary Fig. a,b). Similar effects were observed when SP immunostaining was performed on the PN layer of NP C mice following transplantation with or -overexpressing BM- MSCs (Supplementary Fig. c). To further confirm these effects, we used laser capture microdissection (LCM) to selectively isolate PNs (Supplementary Fig. d). We observed that expressions of Vegf, R and Sphk mrnas were decreased in LCMcaptured PNs from NP C mice compared with that of mice. transplantation enhanced these expression levels in NP C PNs (Fig. f). We also ascertained whether R was required for the activation of SphK in NP C mice. As shown in Fig. g, SphK activity was significantly increased in the NP C mice following treatment, whereas this effect was lower in NP C mice treated with PTK787 before injecting s, NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

4 NATURE COMMUNICATIONS DOI:.8/ncomms65 a treatment Day d f h ELISA, LCM and UPLC analysis PBS PBS j Rota-rod performance test 5 7 Before LCM LCM-captured PN % Survival TP 5 After LCM IHC Day = weeks /calbindin Relative fold change in LCM-captured PN mrna.5.5 TP tg TP Days PN per lobes b ized intensity in PCL TP TP.5 k shrna c (pg mg protein) e Vegf R Sphk TP 5 III TP TP tg TP 9 TP TP IV & V TP TP tg TP VI TP VII VIII shrna i TP Mean score (s) TP TP tg TP 5 5 g 5 SphK activity (%) Relative fold change in LCM-captured PN mrna TP TP Days after transplantation.5.5 shrna Sphk TP PTK787 TP TP TP tg TP Figure from s reduces pathology in PNs of NP C mice. (a) Protocol of treatment in NP C mice. (b,c) SphK activity (n ¼ 7 per group; b) and levels (n ¼ 8 per group; c) were estimated in the cerebellums of and NP C mice after treatment. (d) Cerebellar sections were stained with anti-calbindin and anti- (low-magnification scale bar, 5 mm; high-magnification scale bar, mm). Values represent normalized fluorescence intensities in PCL (n ¼ 7 per group). (e) SphK activities were measured in the cerebellums of NP C mice treated with PBS (n ¼ 6), s, s and tg s (n ¼ 8 per group). (f) Left, isolation of mouse PNs using LCM (scale bar, 75 mm). Right, mrna level of Vegf, R and Sphk on LCM-captured PNs samples (n ¼ 7 per group). (g) NP C mice were treated daily with the PTK787 at mg kg or PBS, starting days before the transplantation. One day after treatment, SphK activity was estimated (NP C, n ¼ 7; NP C TP, n ¼ 8 per group). (h) Cerebellar sections were stained with anti-calbindin (scale bar, 5 mm), and the number of calbindin-positive PNs were quantified (n ¼ 7 per group). (i) Rota-rod scores of mice were averaged and plotted beginning days after transplantation (n ¼ 5 per group). (j) Survival curve of NP C mice (n ¼ 5 per group). Treatment with s and tg s resulted in significantly increased survival compared with PBS treatment (P ¼.9 and P ¼.55, respectively; log-rank test). (k) Effect of knockdown on SphK activity. Left, after intracerebellar injection of control (n ¼ 7) or shrna (n ¼ 8) in mice, SphK activities were measured in the cerebellums. Right, relative levels of Sphk mrna from LCM-captured PNs samples (n ¼ 7 per group). b i, one-way analysis of variance, Tukey s post hoc test. k, Student s t-test. Po.5, Po., Po.5. All error bars indicate s.e.m. shrna although this did not reach statistical significance. SP levels were moderately decreased with PTK787 treatment, but sphingosine did not vary between the groups (Supplementary Fig. e). Next, we evaluated the effects of on the NP C phenotype in mice. Transplantation of tg s improved NP C pathology as measured by increased number NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

5 NATURE COMMUNICATIONS DOI:.8/ncomms65 of calbindin-positive PNs on days after treatment (Fig. h), and also enhanced the Rota-rod performance (Fig. i). These effects were less in the -treated group. Rota-rod performance also diminished in the BM- MSC-treated NP C mice over time. Moreover, the lifespan of mice that had or tg transplants was extended (Fig. j). Finally, to determine whether the reduced levels in the cerebellums affected SphK activity, we injected short hairpin RNA (shrna) into the cerebellum of mice and determined the changes of sphingolipid factors. Treatment with shrna markedly reduced SphK activity and Sphk mrna levels (Fig. k; Supplementary Fig. f,g) and led to elevation of sphingosine and reduction of SP (Supplementary Fig. h). These results suggested that inactivation of may lead to reduced SphK activity in NP C mice, consistent with in vitro results. Together, these findings show a direct correlation between and SphK activity in PNs and suggest that abnormal sphingosine accumulation in NP C may be due to the dysfunction of SphK activity by inactivated expression. NPC deficiency impairs /SphK activation in PNs. We subsequently investigated the relationship between NPC and expression. NPC knockdown by markedly decreased expression in normal PNs. When NPC was knocked down in tg PNs (derived from tg mice), the decreased level of was lower than that of normal PNs (Fig. a c). Moreover, NPC deficiency markedly inactivated SphK and led to sphingolipid imbalance. In tg PNs, however, moderate changes were observed (Fig. d,e). We next tested whether NPC deficiency affected expression in the a Relative fold change (Npc).5.5 PN NPC tg PN b Relative fold change (Vegf ).5.5 c NPC (pg ml ) PN tg PN PN tg PN NPC d 5 5 NPC e (nmol mg ).5..5 PN tg PN PN tg PN PN tg PN SP (nmol mg ) NPC f Relative fold change in LCM-captured PN mrna.5.5 Npc shrna NPC shrna tg g Relative fold change in LCM-captured PN mrna.5.5 Vegf shrna NPC shrna h (pg mg protein) tg tg shrna NPC shrna i 5 5 Relative fold change in LCM-captured PN mrna.5.5 Sphk shrna NPC shrna tg tg tg tg j (nmol mg ) SP (nmol mg ) shrna NPC shrna Figure NPC knockdown reduces expression and SphK activity. (a c) Primary cultures of normal and tg PNs were transfected with control or NPC. Three days after transfection, we measured the levels of Npc (a) and Vegf (b) mrna and secreted protein (c) in PNs (n ¼ 7 per group). (d,e) SphK activity (d), sphingosine and SP (e) were estimated in PNs transfected with control (n ¼ 6) or NPC (n ¼ 7). (f j) Four-week-old and tg mice were injected with control or NPC shrna into the cerebellum. Mice were sacrificed at days after the injection. Npc (f) and Vegf (g) mrna levels were estimated in LCM-captured PNs and protein levels (h) were measured in the cerebellums (n ¼ 7 per group). (i) Left, SphK activities were measured in the cerebellums (n ¼ 7 per group). Right, relative levels of Sphk mrna from LCM-captured PNs samples (n ¼ 8 per group). (j) and SP were measured in the cerebellums (n ¼ 7 per group). a j, one-way analysis of variance, Tukey s post hoc test. Po.5, Po., Po.5. All error bars indicate s.e.m. NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms & Macmillan Publishers Limited. All rights reserved.

6 NATURE COMMUNICATIONS DOI:.8/ncomms65 cerebellums of mice using NPC shrna. Intracerebellar injection of NPC shrna, which decreased Npc mrna expression in the LCM-captured PNs (Fig. f), reduced expression (Fig. g,h). Consistently, NPC deficiency significantly decreased SphK activity and Sphk mrna expression and led to elevation of sphingosine and reduction of SP in the cerebellums (Fig. i,j). These effects were moderated in tg mice (Fig. g j). Overall, these results indicated that knockdown of NPC may lead to reduced expression, and these reductions subsequently decreased SphK activity in PNs. overexpression ameliorates NP C pathology in mice. The -mediated SphK reduction in NP C PNs prompted us to examine possible genetic implications of this pathway. To increase in NP C mice, we generated tg / Npc / mice (Supplementary Fig. a). is widely expressed in s, glia and endothelial cells 9,, with strong expression in PNs. In NP C cerebellum, however, was mainly expressed in the granular layer and significantly decreased in the PCL. /NP C mice showed increased expression of in the PCL compared with NP C mice (Fig. a). /calbindin GCL PCL ML ized intensity in PCL Relative fold change in LCM-captured PN mrna Vegf R Sphk 5 5 PTK787 PN per lobes Mean score (s) 5 5 % Survival 5 III IV & V VI VII VIII 5 6 Weeks Days 9 ized intensity in PCL.5.5 control microsphere 5 control microsphere microsphere 5 control microsphere control microsphere microsphere PN per lobes III IV & V VI control microsphere control microsphere microsphere VII VIII Figure Replenishment of ameliorates NP C pathology in mice. (a) Cerebellar sections from 6-week-old,, NP C and /NP C mice were immunostained with anti-calbindin and anti- (low-magnification scale bar, 5 mm; high-magnification scale bar, mm). The average fluorescence intensity within the PCL was measured (, n ¼ 7;, n ¼ 7; NP C, n ¼ 9; and /NP C, n ¼ 9). (b) SphK activities were measured in cerebellums derived from 6-week-old,, NP C and /NP C mice (n ¼ 8 per group). (c) Quantitative real-time PCR for Vegf, R and Sphk mrna in LCM-captured PNs in 6-week-old,, NP C and /NP C mice (, n ¼ 6;, n ¼ 6; NP C, n ¼ 8; and /NP C, n ¼ 8). (d) /NP C mice were treated daily with the PTK787 at mg kg or PBS vehicle control for days before sacrifice (6-week-old), and SphK activity was estimated in cerebellums (n ¼ 7 per group). (e) Cerebellar sections were immunostained with anti-calbindin and the number of calbindin-positive PNs was quantified (, n ¼ 7;, n ¼ 7; NP C, n ¼ 8; and /NP C, n ¼ 8). (f) Left, beginning at weeks of age, Rota-rod scores were averaged and plotted (n ¼ 5 per group). Right, survival curves of NP C and /NP C mice (P ¼.58; log-rank test, n ¼ 5 per group). (g) Cerebellar sections from and NP C mice transplanted with -loaded or control microspheres were stained with anti-calbindin and anti-. The average fluorescence intensity within the PCL was measured (n ¼ 7 per group). (h) SphK activity was estimated in the cerebellums of and NP C mice at one day after treatment. (i) Cerebellar sections were prepared at weeks after transplantation and immunostained with anti-calbindin. The calbindin-positive PNs were counted (n ¼ 7 per group). a i, one-way analysis of variance, Tukey s post hoc test. Po.5, Po., Po.5. All error bars indicate s.e.m. 6 NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

7 NATURE COMMUNICATIONS DOI:.8/ncomms65 To examine whether genetically increasing affects SphK activity in NP C PNs, we analysed cerebellum samples derived from 6-week-old,, NP C and /NP C mice. Compared with NP C mice, /NP C mice showed significantly increased SphK activity and decreased sphingosine accumulation (Fig. b; Supplementary Fig. b). Cerebellar SP levels did not vary between the NP C and /NP C mice, although SP levels in the PCL was increased in /NP C mice (Supplementary Fig. b,d). Sphingomyelin and unesterified cholesterol levels were also significantly decreased in /NP C mice, but glycosphingolipid (GSL) levels did not vary between the groups (Supplementary Fig. c,e,f). These results revealed that genetic overexpression could reverse the SphK abnormality and abnormal lipid accumulation in NP C. To confirm mediated SphK activation within PNs, we measured the Vegf, R and Sphk mrna levels in LCM-captured PNs from these mice. LCM-captured PNs from /NP C showed slightly increased Vegf and R mrna levels and significantly enhanced Sphk mrna levels (Fig. c). Next, to further investigate the subcellular distribution pattern of SphK activity, sphingosine and SP, we isolated cytosolicenriched and lysosome-enriched fractions from the cerebellums. SphK activity was increased in /NP C-derived lysosomes and cytosol compared with NP C-derived ones, although the degree of SphK increase was greater in the cytosol than lysosome. Accumulated sphingosine in NP C was found in the lysosome. Lysosomal sphingosine levels were significantly decreased in the /NP C, whereas SP levels did not vary between the groups (Supplementary Fig. g). Taken together, these results suggested that leads to activated SphK in the lysosome and cytosol and that activated SphK decreased lysosomal sphingosine accumulation in NP C. We next observed whether the activation of R was required for the activation of SphK in / NP C mice. Increased SphK activity was lower in /NP C mice treated with the PTK787, although this did not reach statistical significance (Fig. d). levels also were moderately increased, but SP levels did not vary between the groups (Supplementary Fig. h). PN survival was significantly improved in the /NP C mice (Fig. e), and there were improvements in the Rota-rod score of 5-week-old /NP C mice compared with NP C mice (Fig. f, left). The lifespan of the /NP C mice was slightly increased (Fig. f, right). We also found that transplantation is more effective in SphK modulation than genetic replenishment of (see Fig. ). These results suggested that other factors secreted by s might also contribute to SphK activation. We next tested whether pharmacologic delivery of recombinant is beneficial to NP C pathology. Since the injected recombinant exerted a short-lived effect, to overcome this obstacle we generated a microsphere system that allows localized and sustained release (Supplementary Fig. 5a). We injected mg of -loaded microspheres or control microspheres into the cerebellum of -week-old NP C and mice. Two weeks after treatment, NP C mice transplanted with -loaded microspheres had higher levels of expression in the PCLs (Fig. g), exhibited increased SphK activity (Fig. h) and decreased sphingosine levels (Supplementary Fig. 5b) in their cerebellums. SP levels in cerebellum and expression in PNs were also increased by -loaded microsphere treatment (Supplementary Fig. 5b,c). Further, the -loaded microsphere-treated NP C mice showed significantly improved PN survival (Fig. i). overexpression reverses defective autophagy in NP C mice. Autophagy, a major degradative pathway of the lysosomal system, is known to be markedly impaired in NP C. These defects lead to loss of PNs in NP C. To examine whether increased PN survival in /NP C mice was related to autophagy, we first measured LC-II levels. Consistent with previous result, we found that the LC-II levels were significantly increased in PNs and cerebellum samples derived from NP C mice. This enhanced LC-II level was reduced in /NP C mice (Fig. 5a,b,d). The level of beclin- did not vary between the groups (Fig. 5a,d). The levels of cathepsin D, a lysosomal hydrolase, were slightly increased in NP C mice compared with mice (Fig. 5a,d). However, the activity of cathepsin D was not changed between the groups (Fig. 5c,e). This result indicated that the elevated levels of cathepsin D in NP C mice did not ultimately translate into a significant increase in enzyme activity. Cathepsin D levels in /NP C mice were comparable to that of NP C mice, indicating that increased in NP C mice did not influence the cathepsin D expression (Fig. 5a,c e). The level of p6 was significantly higher in NP C mice compared with mice, but was decreased in /NP C mice (Fig. 5a,d). We also performed transmission electron microscopic (EM) analysis using mouse cerebellum samples to corroborate the immunoblotting results. NP C mice brains showed massive increases of autophagic vacuoles, while brains of /NP C mice represented a reduced number of these vesicles (Fig. 5f). Next, to determine whether the endocytic pathway was affected by overexpression in NP C mice, we examined Rab5 and Rab7 expression in our animals. The levels of these proteins showed no differences between the groups (Fig. 5g). Apoptotic cells, as judged by active caspase-, did not show any differences between NP C and /NP C mice (Fig. 5h). Our results showed that endocytic pathway and apoptosis were not the main mechanisms of increased PN survival in /NP C mice. Impaired /SphK pathway causes defective autophagic flux. Improved autophagic degradation in the /NP C mice prompted us to analyze whether -mediated sphingolipid changes affect autophagy activity. First, to unravel the mechanistic link between levels and autophagic dysfunction, was depleted in the PNs by treatment. Knockdown of caused increased accumulation of LC-II and p6 (Fig. 6a,b). Beclin- expression was not affected by knockdown (Fig. 6a), indicating that the accumulation of autophagosomes was not due to the biogenesis pathway. The accumulation of autophagosomes can occur due to either an increase in their rate of formation or a reduction in their rate of degradation. To distinguish between these two events, we examined the effects of knockdown on LC-II levels in PNs in the presence or absence of NH Cl that blocks autophagic degradation but does not affect autophagosome formation. knockdown increased accumulation of LC-II. This level was not further increased by NH Cl treatment (Fig. 6c, left). In contrast, depletion in serum starvation culture resulted in a significant increase in LC-II levels (Fig. 6c, right). These observations were also supported by levels of p6 (Fig. 6c). These results suggested that depletion influences at a late step of autophagy. We also performed autophagy flux assay in, NP C and /NP C mice PNs. Under basal condition, NP C PNs showed significantly increased LC-II and p6 levels compared with PNs. NH Cl-induced lysosome inhibition led to marked increase of LC-II and p6 levels in the PNs, but this increase was significantly less in the NP C PNs (Fig. 6d). /NP C PNs showed similar pattern in LC-II and p6 increase compared with cells (Fig. 6d). Taken together, these NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms & Macmillan Publishers Limited. All rights reserved.

8 LC-II 6 kda Beclin- 6 kda 5 kda Active catd kda Cathepsin D Actin p6/actin 5 kda Pro catd PN d 5 PN PN PN 6 kda LC-I kda LC-II 6 kda Beclin- 6 kda p6 LC-II/actin c PN PN 5 kda Active catd kda Cathepsin D Actin f 5 kda Rab5 kda Rab7 kda Actin 5 GCL GCL GCL PCL PCL PCL ML ML ML Active caspase-+ cell in PCL (%) 5 g Rab5/actin Autophagic vacuoles (vacuoles μm ) h 5 5 Cerebellum 5 Rab7/actin PN Active/Pro CatD e 5 kda Pro catd 5 PN PN PN p6/actin CatD activity (%) PN PN PN PN PN PN PN PN PN PN PN p6 PN LC--puncta per cell LC-I kda b CatD activity (%) 6 kda Beclin-/actin Beclin-/actin PN PN Active/Pro CatD PN LC-II/actin a NATURE COMMUNICATIONS DOI:.8/ncomms Figure 5 replenishment reverses defective autophagy in NP C mice. (a) Western blot analysis of LC, beclin-, p6 and cathepsin D in primary cultured PNs derived from, NP C and /NP C mice (, n ¼ 5; NP C, n ¼ 6; and /NP C, n ¼ 6). (b) Immunocytochemistry of LC in, NP C and /NP C PNs (n ¼ 6 per group; scale bar, mm). (c) Cathepsin D activity in primary cultured PNs (, n ¼ 5; NP C, n ¼ 6; and /NP C, n ¼ 6). (d) Western blot analysis of LC, beclin-, p6 and cathepsin D in the cerebellums of 6-week-old, NP C and /NP C mice (, n ¼ 6; NP C, n ¼ 7; and /NP C, n ¼ 7). (e) Cathepsin D activity in the cerebellums of, NP C and /NP C mice (, n ¼ 5; NP C, n ¼ 6; and /NP C, n ¼ 6). (f) EM images and quantification data of the cerebellum (n ¼ 5 per group; low-magnification scale bar, mm; high-magnification scale bar, nm). Arrow indicates autophagic vacuole. (g) Western blot analysis of Rab5 and Rab7 levels in the cerebellum (n ¼ 6 per group). (h) Cerebellar sections were immunostained with anti-active caspase- and the number of active caspase--positive cells in PCL was quantified (n ¼ 5 per group; scale bar, 5 mm). a g, one-way analysis of variance, Tukey s post hoc test. h, Student s t-test. Po.5, Po., Po.5. All error bars indicate s.e.m. 8 NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

9 NATURE COMMUNICATIONS DOI:.8/ncomms65 a c NH Cl 6 kda kda 6 kda kda g LC-II/actin p6/actin 6 kda kda 6 kda 6 kda kda j 5 Curcumin LC-I LC-II Beclin- p Actin Starvation LC-I 6 kda LC-II kda p6 6 kda Actin kda LC-II/actin p6/actin 6 LC-II/actin k Fluorescence intensity (F/F) Basal Starvation + + PN + PN density (cells mm ) Basal NH Cl 5 PN mcherry-egfp-lc+ structures per cell Beclin-/actin d e f NH Cl LC-I 6 kda LC-II kda p6 6 kda Actin kda LC-II/actin p6/actin 6 Autolysosomes Autophagosomes GPN ( μm) Time (s) h LC-I 5 kda LC-II p6 kda Actin kda m n o p6/actin Basal NH Cl 5 5. b TFEB/actin LAMP/actin. l + PN + control PN + TFEB LAMP Actin + Curcumin μm Amplitude change (F/F) (nmol mg ) NS P<. P<. PN + control + SP (nmol mg ) i PN + Curcumin 6 kda kda 6 kda kda + LC--puncta per cells Relative mean fluorescence intensity LC-I LC-II p6 Actin 5 5 LC-II/actin p6/actin Curcumin. Counts PN + control PN + H O NH Cl 8 FL-H (nmol mg )... + H O NH Cl mcherry-egfp-lc+ structures per cell 8 6 Autolysosomes Autophagosomes APG/APGL Autophagosomes (APG) Autolysosomes (APGL) PN density (cells mm ) 5 Figure 6 /SphK inactivity impairs autophagic flux. (a) Western blots of LC, beclin- and p6 in PNs after knockdown (n ¼ 6 per group). (b) Immunocytochemistry of LC in PNs after knockdown (n ¼ 6 per group; scale bar, mm). (c) Autophagic flux assay. Western blots of LC and p6 in PNs (n ¼ 6 per group). (d) Western blots of LC and p6 in cultured PNs in the presence of NH Cl (n ¼ 5 per group). (e) Western blots of TFEB and Lamp in -knockdown PNs (control, n ¼ 5 and, n ¼ 6). (f) Effect of knockdown on lysosomal ph. PNs stained with LysoTracker red (n ¼ 5 per group; scale bar, mm). (g) Fluorescence analysis of autophagosomes and autolysosomes (control, n ¼ 7 and, n ¼ 8; scale bar, mm). (h) Sphk activity, sphingosine and SP levels in PNs after knockdown in the presence of curcumin (n ¼ 8 per group). (i) Western blot analysis of LC and p6 in -knockdown PNs treated with curcumin (control, n ¼ 7;, n ¼ 8; and /curcumin, n ¼ 8). (j) Survival of -knockdown PNs treated with curcumin (n ¼ 8 per group). (k) Left, representative traces showing intracellular [Ca þ ] changes monitored in single fluo--loaded PNs. Right, maximal peak fluorescence changes were determined as the differences between basal and the maximum fluorescence (n ¼ cells per group). (l) Sphk activity and sphingosine levels were measured in cultured PNs (n ¼ 8 per group). (m) Quantification of autophagosomes and autolysosomes in primary cultured PNs (, n ¼ 7; NP C, n ¼ 7; and /NP C, n ¼ 8). (n) EM analysis of the PNs (n ¼ 5 per group; low-magnification scale bar, mm; high-magnification scale bar, nm). (o) Survival of primary cultured PNs (, n ¼ 6; NP C, n ¼ 8; and /NP C, n ¼ 8). a,b,e,g, Student s t-test. c,d,f,h o, one-way analysis of variance, Tukey s post hoc test. Po.5, Po., Po.5. All error bars indicate s.e.m. NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms & Macmillan Publishers Limited. All rights reserved.

10 NATURE COMMUNICATIONS DOI:.8/ncomms65 results indicated that reduced levels in NP C PNs caused a defect of autophagic degradation, but not induction. Our findings that depletion of affects autophagic degradation prompted us to more closely examine how might influence in defective autophagic degradation. We first examined the transcription factor EB (TFEB), which coordinates lysosomal formation. depletion in PNs did not affect the levels of TFEB and Lamp, indicating that did not impair lysosome biogenesis (Fig. 6e). Next, we assessed alteration in lysosomal ph using the acidotropic dye LysoTracker red. H O - and NH Cl-treated cells were used as positive and negative controls, respectively. -treated PNs exhibited a similar fluorescence to control -treated PNs, indicating that did not affect lysosomal acidification (Fig. 6f). Following the initiation of the phagophores, autophagosomes undergo a stepwise maturation process from early to late autophagosomes, which ultimately fuse with lysosomes to form autolysosomes. To study the effect of depletion on the maturation of autophagosomes, we used mcherry-egfp-lc reporter 5. Before fusion with lysosomes, the LC-II-positive autophagosomes are shown by both GFP and mcherry signals as yellow puncta, and after fusion, autolysosomes are shown by only mcherry signals as red-only puncta because GFP loses its fluorescence in acidic ph. Compared with control PNs, -treated cells showed significantly increased yellow puncta (autophagosomes) and decreased mcherry-only puncta (autolysosomes), indicating that depletion inhibited the autophagosome lysosome fusion (Fig. 6g). accumulation can induce defective calcium release from the acidic compartment such as lysosome, which inhibits fusion of lysosome with other organelles. As shown in Supplementary Figs g and h, decreased levels caused sphingosine accumulation in PNs. Thus, we hypothesized that defective lysosomal calcium release by -mediated sphingosine accumulation disturbs autophagosome lysosome fusion and evokes the abnormal autophagosomes amassment. To test this hypothesis, we used the weaker sarcoplasmic reticulum ATPase antagonist curcumin, a natural product derived from turmeric 6, which correct sphingolipid imbalance by increasing the cytosolic calcium release. Importantly, abnormal sphingolipid levels in the -treated PNs were normalized after curcumin treatment (Fig. 6h). We also observed significantly decreased protein level of abnormal autophagic markers (Fig. 6i) and increased al survival (Fig. 6j) in the -knockdown PNs after curcumin treatment compared with PNs with knockdown alone. Similar results were observed in the shrna-treated mice after curcumin injection (Supplementary Fig. 6). To further examine these effects, we analysed calcium homeostasis in the primary cultured PNs derived from, NP C and /NP C mice. To specifically assess the lysosomal calcium content, we used Gly Phe b-naphthylamide (GPN), which osmotically lyses cathepsin-containing lysosome. We observed a reduction in NP C PNs calcium release from lysosome compared with PNs, consistent with our previous study 7. Notably, this reduction was corrected in /NP C PNs (Fig. 6k). As expected, abnormal sphingosine accumulation was reduced in /NP C cells by restoration of SphK activity (Fig. 6l). Moreover, /NP C PNs showed decreased autophagosome (yellow LC puncta) accumulation (Fig. 6m). Decreased autophagosomes in the /NP C PNs were further confirmed by EM analysis (Fig. 6n). The survival of PNs was also significantly increased in the /NP C (Fig. 6o). Together, these findings show that inactivated /SphK pathway in NP C PNs causes sphingosine accumulation and this amassment inhibits autophagosome lysosome fusion by disturbance of calcium homeostasis. rescues autophagic defects in patient-specific cells. To further validate our observation regarding treatment in NP C mice, we studied effects of on SphK activity in human NP C fibroblasts. Human NP C fibroblasts cocultured with human s, tg s or treated with recombinant showed significantly increased SphK activity, decreased sphingosine and elevated SP (Supplementary Fig. 7a c). Increased LC-II levels and p6 accumulation in NP C fibroblasts were reduced by treatment (Supplementary Fig. 7d,e). treated NP C fibroblasts also showed increased calcium release and decreased autophagosome accumulation (as judged by yellow LC puncta) compared with non-treated NP C fibroblasts (Supplementary Fig. 7f,g). Lysosomal exocytosis is necessary to affect clearance of stored intracellular lipids and ameliorates the endolysosomal lipid storage phenotype in NP C cells 7. To determine whether directly induced lysosomal exocytosis, the culture media of normal and NP C fibroblasts treated with or without were analysed for the presence of the lysosomal enzyme b-hexosaminidase as a marker for lysosomal content secretion. In all groups, the activity of b-hexosaminidase was not significantly elevated at the indicated times (Supplementary Fig. 7h). The low-level appearance of b-hexosaminidase in the culture media of fibroblasts is not a result of generalized cell lysis, since the levels of lactate dehydrogenase (LDH) in the media remained unchanged in all groups for the duration of the assay (Supplementary Fig. 7h). These results suggested that the ability of Figure 7 ameliorates sphingolipid imbalance in NP C ipsc s. (a) Left, normal, and -7 ipscs generated nestin-positive neuroprogenitor cells (scale bar, 5 mm). Right, representative images of immunocytochemical staining the b-iii tubulin following neural differentiation (scale bar, 5 mm). (b) and -7 s were treated with human s, tg s or recombinant ( ng ml ). Three days after treatment, levels were measured in cell lysates. (c f) SphK activity (c), sphingosine (d), SP (e) and sphingomyelin (f) were measured in normal ipsc s and s with or without treatment. (g) Filipin staining of unesterified cholesterol in s with or without treatment of recombinant for days (scale bar, 5 mm). Quantification of filipin fluorescence intensities normalized to normal s. Unesterified cholesterol levels in normal ipsc s and s with or without treatment were measured (n ¼ 6 per group). (h) Effect of the R inhibitor on mediated sphingolipid modulation. s were pretreated with PTK787 at mm for day and were treated for days with ng ml and then assayed for SphK activity, sphingosine and SP (n ¼ 7 per group). (i,j) Effect of NPC knockdown on sphingolipid factors in normal ipsc s. (i) Three days after NPC transfection, we measured the levels of NPC mrna and expression. (j) SphK activity, sphingosine and SP were measured in normal ipsc s treated with control or NPC (control, n ¼ 7; NPC, n ¼ 9). (k) Effect of knockdown on sphingolipid factors in normal ipsc s. Three days after transfection, we measured the levels of mrna, SphK activity, sphingosine and SP in normal ipsc s (n ¼ 7 per group). (l) Effect of a specific SphK inhibitor on sphingolipid factors in normal ipsc s. s were treated with or without mm SK-I for 6 h. Lipids were extracted and sphingosine, SP and unesterified cholesterol levels were determined (n ¼ 6 per group). b h, one-way analysis of variance, Tukey s post hoc test. i l, Student s t-test. Po.5, Po., Po.5. All error bars indicate s.e.m. NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

11 NATURE COMMUNICATIONS DOI:.8/ncomms65 a ipsc -derived NSCs Nestin/DAPI -derived NSCs -7 -derived NSCs ipsc -derived s β-iii tubulin/dapi -derived s -7 -derived s b (pg mg protein) 8 6 Human tg Human tg c SphK activity (%) 5 5 Human tg Human tg d (nmol mg )..5 Human tg Human tg e SP (nmol mg )... Human tg Human tg -7-7 f Sphingomyelin (nmol mg ) + g Filipin + Filipin intensity (normalized) + Unesterified cholesterol (μg mg protein) 6 + h j NPC (nmol mg ) PTK787.. (nmol mg ) NPC l SP (nmol mg ) (nmol mg ) PTK787 NPC k.. SP (nmol mg ) Relative fold change () SP (nmol mg ) PTK787 Unesterified cholesterol (μg mg protein) 5 5 i.5.5 None SK-I None SK-I None SK-I Relative fold change (NPC) (nmol mg ).. NPC (pg ml ) SP (nmol mg )... NPC NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

12 NATURE COMMUNICATIONS DOI:.8/ncomms65 to reduce sphingosine storage in NP C cells was not due to lysosomal exocytosis. The recent developments in induced pluripotent stem cells (ipscs) and ipsc-derived s have allowed investigation of pathogenesis of neurological diseases in vitro. To explore whether the observed effects of we describe above were similar in NP C human s, we established human NP C ipscs (, 6, 7) by transduction of human NP C fibroblasts with retroviruses encoding OCT, SOX, KLF and c-myc similar to previous studies 8. Analysis of NP C ipscs () revealed typical characteristics of pluripotent stem cells: similar morphology to embryonic stem cells (ES cells), expression of pluripotent markers including SSEA-, Tra--6 and Tra--8, normal chromosomal number and genomic structure, silencing of retroviral transgene and reactivation of genes indicative of pluripotency (Supplementary Fig. 8a c). The differentiation ability of NP C ipscs was also confirmed in vivo by teratoma formation (Supplementary Fig. 8d). We analysed SphK activity and sphingolipid levels in the normal ipsc and NP C ipsc lines. NP C ipsc lines exhibited decreased SphK activity, increased sphingosine accumulation and decreased SP levels compared with normal ipscs (Supplementary Fig. 8e). Next, human s were induced from the, - 7 and normal ipsc. Early-differentiating cells expressed nestin and differentiated cells expressed -specific b-iii tubulin (Fig. 7a). These NP C s also exhibited phenotypes seen in human NP C samples, including abnormal levels and sphingolipid metabolism (Fig. 7b e). To confirm the effects of in human NP C s, the - or -7- derived s were cocultured with human or tg BM- MSCs, or treated with recombinant. We found that all treated groups exhibited increased, elevated SphK activity, decreased sphingosine accumulation and increased SP levels (Fig. 7b e). Sphingomyelin and unesterified cholesterol levels were also significantly decreased in -treated NP C s (Fig. 7f,g). We also pretreated NP C s with PTK787 before treatment. We found that SphK activity and other sphingolipid metabolites in NP C s were mediated by interactions of and its receptor R in these ipscderived NP C s (Fig. 7h). To reconfirm the in vitro mechanism whereby there is a direct relationship between NPC, and SphK activity in human s, we treated normal ipsc s with NPC and (Fig. 7i k) and determined changes in various sphingolipid factors. NPC decreased expression and SphK activity (Fig. 7i,j). also strongly inactivated SphK levels (Fig. 7k). Both treatments led to changed levels of sphingosine and SP, similar to NP C s (Fig. 7j,k). To determine whether reduction in SphK activity affected sphingolipid factors and unesterified cholesterol in ipsc s similar to those in classical NP C cells, we treated the normal ipsc s with a specific SphK inhibitor, SK-I. We found that inhibition of SphK activity increased sphingosine and unesterified cholesterol accumulation and decreased cellular SP (Fig. 7l). We also examined whether NP C s exhibited abnormal autophagy. NP C s had significantly higher abnormal autophagic markers than normal s (Fig. 8a). treatment significantly decreased the protein level of abnormal autophagic markers in NP C s (Fig. 8a). Similar to previous results (Supplementary Fig. 7), -treated NP C s showed increased calcium release and decreased autophagosome accumulation, suggesting that elevates autophagosome lysosome fusion (Fig. 8b d). Consistent with the restored autophagy flux, cell survival was also significantly improved in -treated NP C s (Fig. 8e). Collectively, these results confirm that defective autophagy by abnormal /SphK pathway and sphingosine levels in NP C mice and human fibroblasts also occur in NP C patient s, and replenishment of is able to ameliorate autophagy defect by correction of sphingolipid imbalance in the NP C patient cells. Discussion NP C patients and mice exhibit progressive al loss, mainly of cerebellar PNs, but the mechanism is largely unknown. Recent studies have shown that inactivation of NPC caused abnormal autophagy and the defect may contribute to PN loss in NP C. Loss of NPC function leads to trapping of lipids within aberrant membrane compartments, and this may induce a lipid-starvation response analogous to the well-characterized autophagic response to amino-acid deprivation. In addition, destructive autophagy in NP C PNs may also be stimulated hormonally via neurosteroids. Neurosteroids might inhibit autophagy in PNs and when their synthesis is severely decreased, as in NP C, autophagic cell death might ensue. Similar to previous results,, we found that the impaired autophagic flux in NP C was associated with decreased autophagosome lysosome fusion, and that this defect led to PNs loss. Recent studies have also demonstrated that cholesterol, sphingomyelin and GSL storage are downstream events in NP C disease pathogenesis caused by sphingosine storage, leading to altered acidic compartment calcium levels. They Figure 8 rescues the autophagic defects in NP C ipsc s. (a) Western blot analysis of LC and p6 in normal and NP C ipsc-derived s treated with or without ng ml recombinant (normal, n ¼ 6;, n ¼ 7; and -treated, n ¼ 7). (b) Left, representative traces showing intracellular [Ca þ ] changes monitored in single fluo--loaded normal and NP C ipsc s treated with or without recombinant ( ng ml ). Right, maximal peak fluorescence changes were determined as the differences between basal and the maximum fluorescence, on addition of mm GPN (n ¼ cells per group). (c) Fluorescence staining and quantification of autophagosomes (mcherry þ -EGFP þ -LC) and autolysosomes (mcherry þ -EGFP -LC) in normal and NP C ipsc s after recombinant treatment (normal, n ¼ 7;, n ¼ 8; and -treated, n ¼ 8; scale bar, mm). (d) EM images and quantification data of normal and NP C ipsc-derived s after ng ml treatment (n ¼ 5 per group; low-magnification scale bar, mm; high-magnification scale bar, nm). (e) Quantification of cell viability (normal, n ¼ 5;, n ¼ 6; and -treated, n ¼ 6). (f) Model of -mediated SphK activation in NP C s. (A,B) In NP C cells, sphingosine accumulation is increased due to defective SphK activity together with decreased caused by mutated NPC and defective uptake via R. (C) Abnormal sphingosine accumulation decreases calcium release from lysosomes and the reduction in calcium release causes an autophagic defect by inhibiting autophagosome lysosome fusion. Eventually, these defects cause loss of cerebellar s. (D,E) When NP C s are exposed to s or pure, the cells exhibit elevated intracellular levels of (F), which induces R-mediated activation of SphK in the cytosol and lysosome. (G) This activation leads to decreased sphingosine accumulation and increased SP levels. (H) Reduced sphingosine accumulation results in improved autophagosome lysosome fusion by correction of calcium homeostasis. Finally, this restoration prevents al loss in NP C. a e, one-way analysis of variance, Tukey s post hoc test. Po.5, Po., Po.5. All error bars indicate s.e.m. NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.

13 NATURE COMMUNICATIONS DOI:.8/ncomms65 NPC disease pathogenesis that causes altered calcium homeostasis, leading to the secondary storage of sphingolipids and cholesterol, although additional studies are required. Similarly, we found that -mediated sphingosine modulation also significantly decreased sphingomyelin and unesterified cholesterol levels. Therefore, we suggest that Fluorescence intensity (F/F) GPN ( μm) P<.5 P< NS Actin 6 kda + p6 p6/actin 6 kda + LC-II LC-I kda 6 kda LC-II/actin Amplitude change (F/F) have determined the chronology of events after inactivation of NPC. In a drug-induced NP C cellular model, sphingosine storage in the acidic compartment led to calcium depletion in these organelles, which then resulted in cholesterol, sphingomyelin and GSL storage in these compartments. Therefore, sphingosine storage might be an initiating factor in Time (s) Autolysosomes Autophagosomes + mcherry-egfp-lc mcherry-egfp-lc+ Structures per cell Cell viability (%) 8 6 Autophagosomes (APG) Autolysosomes (APGL) APG/APGL / (D) NPC mutant (A) R (B) SphK/ Autolysosome Fusion SphK/ SPH (C) Ca+ (F) Activated (E) R replenishment NPC mutant (G) SphK/ Autolysosome SphK/ Fusion SPH SP (H) Ca+ Ca+ Lysosome Lysosome Autophagosome Neuronal loss SP Ca+ Autophagosome Neuronal survival SPH: sphingosine NATURE COMMUNICATIONS 5:55 DOI:.8/ncomms65 & Macmillan Publishers Limited. All rights reserved.