β-Cyclodextrin-threaded Biocleavable Polyrotaxanes Ameliorate Impaired Autophagic Flux in Niemann-Pick Type C Disease*

Background: Niemann-Pick type C (NPC) disease is associated with the impaired autophagic flux. Results: β-Cyclodextrin (β-CD)-threaded polyrotaxanes reduced autophagosome accumulation, promoted autophagic proteolysis, and facilitated the fusion of autophagosomes and lysosomes in NPC disease, whereas β-CD derivatives further disturbed autophagic flux. Conclusion: β-CD-threaded polyrotaxanes improved impaired autophagic flux in NPC disease. Significance: β-CD-threaded polyrotaxanes may be promising therapeutics for ameliorating impaired autophagy in NPC disease. Niemann-Pick type C (NPC) disease is characterized by the lysosomal accumulation of cholesterols and impaired autophagic flux due to the inhibited fusion of autophagosomes to lysosomes. We have recently developed β-cyclodextrin (β-CD)-threaded biocleavable polyrotaxanes (PRXs), which can release threaded β-CDs in response to intracellular environments as a therapeutic for NPC disease. The biocleavable PRXs exhibited effective cholesterol reduction ability and negligible toxic effect compared with hydroxypropyl-β-CD (HP-β-CD). In this study, we investigated the effect of biocleavable PRX and HP-β-CD on the impaired autophagy in NPC disease. The NPC patient-derived fibroblasts (NPC1 fibroblasts) showed an increase in the number of LC3-positive puncta compared with normal fibroblasts, even in the basal conditions; the HP-β-CD treatment markedly increased the number of LC3-positive puncta and the levels of p62 in NPC1 fibroblasts, indicating that autophagic flux was further perturbed. In sharp contrast, the biocleavable PRX reduced the number of LC3-positive puncta and the levels of p62 in NPC1 fibroblasts through an mTOR-independent mechanism. The mRFP-GFP-LC3 reporter gene expression experiments revealed that the biocleavable PRX facilitated the formation of autolysosomes to allow for autophagic protein degradation. Therefore, the β-CD-threaded biocleavable PRXs may be promising therapeutics for ameliorating not only cholesterol accumulation but also autophagy impairment in NPC disease.

Niemann-Pick type C (NPC) 2 disease is an autosomal recessive lysosomal storage disorder caused by the mutation of NPC1 (1)(2)(3). Because NPC1 is a key transmembrane protein for transferring low density lipoprotein-derived cholesterols from endosomes/lysosomes to the endoplasmic reticulum, a dysfunction in NPC1 proteins cause lysosomal accumulation of unesterified cholesterols. Due to the lysosomal accumulation of cholesterols, patients with NPC disease exhibit various clinical symptoms, including neurodegeneration and hepatosplenomegaly (4,5). Hydroxypropyl-␤-cyclodextrin (HP-␤-CD), a cyclic oligosaccharide derivative, has recently attracted considerable attention as a therapeutic for NPC disease (6 -9). The administration of HP-␤-CD has been reported to remarkably prolong the life span of Npc1 Ϫ/Ϫ mice through a reduction in cholesterol levels (6 -8). To further develop a ␤-CD-based therapy for NPC disease, we have developed ␤-CD-threaded biocleavable polyrotaxanes (PRXs), a class of polymeric supermolecules composed of cyclic compounds threaded onto a linear polymer chain (10,11), for the lysosome-specific delivery of ␤-CDs ( Fig. 1) (12). The ␤-CD-threaded PRXs bearing cleavable disulfide linkages showed less interaction with the plasma membrane than did HP-␤-CD, leading to internalization into late endosomes/lysosomes through endocytosis ( Fig. 1) (12). Due to this preferential transportation to lysosomes as well as to the local release of threaded ␤-CDs by the cleavage of terminal disulfide linkages in the intracellular reductive environment, the biocleavable PRXs effectively reduced lysosomal cholesterols to a 100-fold lower concentration than did HP-␤-CD (12). Additionally, the biocleavable PRXs were shown to induce a negligible extraction of cholesterols from the plasma membrane, whereas HP-␤-CD extracted high levels of cholesterols from the plasma membrane, indicating that the site of cholesterol reduction is different between the biocleavable PRXs and HP-␤-CD (12). Therefore, the biocleavable PRXs are considered to act as an intracellular delivery vehicle for ␤-CDs or as a type of prodrug for enhancing their therapeutic effect.
Moreover, autophagy, a bulk degradation system of cytoplasmic protein aggregates and subcellular organelles, is of increasing importance with respect to its relation to various diseases. Basal autophagy plays a pivotal role in the constitutive turnover of cytoplasmic components for maintaining cellular function (13)(14)(15)(16). It is known that Atg5 Ϫ/Ϫ or Atg7 Ϫ/Ϫ mice display neurodegeneration and hepatomegaly, indicating that basal autophagy is strongly related to a variety of pathological states (17,18). Indeed, impaired basal autophagy has been reported in various neurodegenerative diseases and lysosomal storage disorders (19,20), including Niemann-Pick type A disease (21), Fabry disease (22), GM1 gangliosidosis (23), multiple sulfatase deficiency (24), Dannon disease (25), Batten disease (26), and NPC disease (27)(28)(29)(30)(31)(32)(33). It has been suggested that the major pathological states of lysosomal storage disorders are attributed to impaired basal autophagy (15,20). In most of these diseases, abundant autophagosomes accumulate in the cytoplasmic compartments under basal conditions (19,20). In NPC disease, various researchers have reported autophagosome accumulation even in the basal condition (27)(28)(29)(30)(31)(32)(33). To date, various possibilities have been suggested for the mechanism of impaired autophagy in NPC disease, such as a decline in the lysosomal enzymatic activity due to lysosomal cholesterol accumulation (32) and an impaired fusion of autophagosomes with lysosomes (33). Therefore, an improvement in impaired autophagy is required for the treatment of NPC disease in addition to a chronic cholesterol accumulation in lysosomes. Jaenisch and co-workers (33) have reported that HP-␤-CD induces the decline of autophagic flux in normal and NPC1 model cells. This indicates that the simultaneous improvement of cholesterol accumulation and impaired autophagy in NPC disease is still a challenging issue.
In this study, we investigated the effect of biocleavable PRXs on autophagy impairment in NPC patient-derived fibroblasts in comparison with HP-␤-CD. We found that the biocleavable PRXs could remarkably improve the impaired autophagic flux by facilitating the formation of autolysosomes, although HP-␤-CD induced autophagosome accumulation. Notably, the biocleavable PRXs exhibited an effect opposite to that of HP-␤-CD on impaired autophagy in NPC disease, although both were able to reduce lysosomal cholesterol levels.
Synthesis of HEE Group-modified ␤-CD (HEE-␤-CD)-To a solution of ␤-CD (1.0 g, 881 mol) in dehydrated DMSO (15 ml), CDI (1.72 g, 10.6 mmol) was added at room temperature. After the reaction for 24 h, HEEA (5.25 ml, 52.9 mmol) was added to the reaction mixture and stirred for a further 24 h at room temperature. Then the PRX was purified by dialysis against methanol for 4 days (Spectra/Por 6; molecular weight cut-off of 1,000). The recovered solution was evaporated and dissolved in water. The aqueous solution was freeze-dried to obtain HEE-␤-CD (186.9 mg, 12.4% yield). The number of modified HEE groups on ␤-CD was determined to be 4.3 by the 1 H NMR. The M n,NMR of HEE-␤-CD was determined to be 1,700.
Synthesis of HEE Group-modified Biocleavable PEG/␣-CD Polyrotaxanes-The PEG/␣-CD-based PRX bearing reductioncleavable disulfide linkages capped with N-benzyloxycarbonyl-L-tyrosine (Z-Tyr-OH) was synthesized as described previously (34). The PEG/␣-CD-based PRXs with a number of threading ␣-CDs of 30.7 and an M n of PEG axle of 9,810 were used in this study. To a solution of the PRX (200 mg, 4.9 mol) in anhydrous DMSO (5 ml), CDI (196 mg, 1.21 mmol) was added at room temperature. After the reaction for 24 h, HEEA (1.2 ml, 12.1 mmol) was added to the reaction mixture and stirred for a further 24 h at room temperature. The PRX was purified as described above to obtain PEG/␣-CD HEE-SS-PRX (212.5 mg, 79.6% yield). The numbers of threading ␣-CDs and HEE groups on PRX were determined to be 30.7 and 104.5, respectively. The M n,NMR of HEE-SS-PRX was determined to be 54,200.
Synthesis of HEE Group-modified P123/␣-CD Biocleavable Polyrotaxanes-P123 bearing terminal cystamine (P123-SS-NH 2 ) was synthesized as previously described (12). P123-SS-NH 2 (1.0 g, 147 mol) dissolved in a small aliquot of water was added to the saturated solution of ␣-CD (Ensuiko Sugar Refining, Tokyo, Japan) (4.0 g, 4.11 mmol) (27.6 ml), and the system was stirred for 24 h at room temperature. After the reaction, the precipitate was collected by centrifugation (7,000 rpm, 10 min) and freeze-dried for 1 day to obtain pseudopolyrotaxane (2.88 g). Then, to a solution of Z-Tyr-OH (Sigma-Aldrich) (1.85 g, 5.88 mmol) and 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride (Wako) (1.63 g, 5.88 mmol) in methanol (23 ml), the pseudopolyrotaxane was added. The resulting reaction mixture was stirred for 24 h at room temperature. After the reaction, the precipitate was successively washed with N,N-dimethylformamide and water. The recovered PRX was freeze-dried to obtain P123/␣-CD PRX (226.8 mg, 7.8% yield based on P123 mol%). The number of ␣-CDs threaded onto the PRX was 13.1, as determined by 1 H NMR. Then, to a solution of the PRX (200 mg, 10.0 mol) in anhydrous DMSO (10 ml), CDI (176 mg, 1.08 mmol) was added at room temperature. After the reaction for 24 h, HEEA (1.07 ml, 10.8 mmol) was added to the reaction mixture and stirred for a further 24 h at room temperature. The PRX was purified as described above to obtain P123/ ␣-CD HEE-SS-PRX (216.8 mg, 76.4% yield). The numbers of threading ␣-CDs and HEE groups on PRX were determined to be 13.1 and 57.5, respectively. The M n,NMR of HEE-SS-PRX was determined to be 28,300.
Cell Culture-Human dermal fibroblasts derived from a Niemann-Pick type C disease patient (NPC1) (GM03123; P237S and I1061T mutations in NPC1) and normal human dermal fibroblasts (GM05659) were obtained from the Coriell Institute for Medical Research (Camden, NJ). These cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Gibco) containing 10% fetal bovine serum (FBS) (Gibco), 100 units/ml penicillin (Gibco), and 100 g/ml streptomycin (Gibco) in a humidified 5% CO 2 atmosphere at 37°C. For treatment with HP-␤-CD, HEE-SS-PRX, and Baf A, the normal and NPC1 fibroblasts were seeded into a 12-well plate (Nunc, Roskilde, Denmark) at a density of 1 ϫ 10 5 cells/well and incubated overnight. After the medium was exchanged with fresh medium (900 l), the treatment solutions (100 l) were applied to each well, followed by incubation for 24 h at 37°C. To induce autophagy via amino acid starvation, the cells were cultured in Hanks' balanced salt solution (Sigma-Aldrich) at 37°C. For microscopic observation of the treated cells, the cells were seeded into a 35-mm glass-bottomed dishes (Iwaki Glass, Tokyo, Japan) at a density of 1 ϫ 10 4 cells/dish and incubated overnight. After the medium was exchanged with fresh DMEM (270 l), the treatment solutions (30 l) were applied to the dish, followed by incubation for 24 h at 37°C.
Quantification of Total Cholesterol-After treatment, the cells were washed three times with PBS. The cells were then harvested with trypsin-EDTA, washed several times with PBS, and lysed with cell lysis buffer (100 mM phosphate buffer, 1 M NaCl, 50 mM cholic acid, and 1% Triton X-100). The total cholesterol (the sum of esterified and unesterified cholesterols) was determined using an Amplex Red cholesterol assay kit (Molecular Probes). Briefly, sample solution (50 l) was combined with the assay solution (50 l) containing cholesterol esterase (2 units/ml), cholesterol oxidase (2 units/ml), horseradish peroxidase (2 units/ml), and Amplex Red reagent (fluorescent sub- strate) (300 M) and incubated for 30 min at 37°C. The fluorescence intensities were measured on an ARVO MX multilabel counter (PerkinElmer Life Sciences) equipped with a filter set for excitation and emission at 560 Ϯ 10 nm and 590 Ϯ 10 nm, respectively. Protein content in the lysate was also determined with a Micro BCA protein assay kit (Thermo Fisher Scientific). Cellular cholesterol content was normalized to protein content and expressed as nmol/mg of protein.
Cholesterol Extraction from the Plasma Membrane-After the treatment of NPC1 fibroblasts in Hanks' balanced salt solution for 2 h at 4°C, the supernatant was collected. The concentration of cholesterol in the supernatant was determined by the Amplex Red cholesterol assay kit as described above.
Plasmid DNAs and Transfection-Plasmid DNAs encoding mRFP-GFP tandem fluorescence-tagged LC3 (pmRFP-GFP-LC3) (catalog no. 21074) and mRFP-LC3 (catalog no. 21075) were obtained from Addgene (Cambridge, MA) (35). The normal and NPC1 fibroblasts were plated into 35-mm glass-bottomed dishes at a density of 1 ϫ 10 4 cells/dish and treated with transfection solutions containing 250 ng of plasmid DNA and Lipofectamine 3000 (Life Technologies, Inc.). After 24 h of incubation, the cells were washed three times with culture medium. The treatment solutions (30 l) were then applied to the dishes, followed by incubation for 24 h. The CLSM images were acquired with a FluoView FV10i imager.
Cathepsin B Activity-After treatment, the cells were washed twice with PBS and fixed with 4% paraformaldehyde for 15 min at room temperature. The cells were then stained with the Magic Red cathepsin B detection kit (Immunochemistry Technologies, Bloomington, MN) according to the manufacturer's instructions. The CLSM images were acquired with a FluoView FV10i imager. For the fluorescence intensity measurements, the cells were harvested and stained with Magic Red cathepsin B. After incubation for 1 h at 37°C, the fluorescence intensities were measured on an ARVO MX multilabel counter.
Statistical Analysis-The data are presented as the mean Ϯ S.D. Differences between the means of the individual groups were assessed by one-way analysis of variance followed by Tukey's multiple comparison test. A p value of less than 0.05 was considered as statistically significant.

Design of Biocleavable Polyrotaxanes and Their Effect on
Lysosomal Cholesterol and Autophagosomes Accumulation in NPC1 Fibroblasts-The biocleavable PRXs composed of Pluronic P123 and ␤-CD capped with (N-triphenylmethyl)glycine via intracellular cleavable disulfide linkages were synthesized (Fig. 1). HEE groups were chemically modified on the hydroxyl groups of ␤-CD in PRX to impart water solubility (denoted as HEE-SS-PRX). To confirm the cholesterol reduction ability of HEE-SS-PRX in NPC disease patient-derived fibroblasts (NPC1 fibroblasts), filipin staining for cholesterol and the quantification of total cholesterol (the sum of esterified and unesterified cholesterols) were performed in comparison with HP-␤-CD (Fig. 2, A and B). The lysosomal cholesterol in NPC1 fibroblasts was reduced to levels almost comparable with the normal levels by both HP-␤-CD (10 mM) and HEE-SS-PRX (1 mM ␤-CD) treatments, indicating that lysosomal release of threaded ␤-CDs from HEE-SS-PRX may be attributable to significant cholesterol reduction. To confirm this hypothesis, the cholesterol reduction ability of HEE-SS-PRX was evaluated in comparison with its constituent molecules, such as HEE groupmodified ␤-CD (HEE-␤-CD) and the axle polymer (Pluronic P123) (Fig. 3C). As a result, HEE-␤-CD and P123 exhibited lower ability in cholesterol reduction than HEE-SS-PRX. Accompanied with the lysosomal cholesterol reduction in NPC1 fibroblasts, HP-␤-CD concomitantly extracted cholesterols from the plasma membrane, whereas negligible cholesterol extraction from the plasma membrane was observed for the HEE-SS-PRX-treated NPC1 fibroblasts (Fig. 2D). Altogether with these results, it is obvious that HEE-SS-PRX reduced lysosomal cholesterol to a much lower concentration than HP-␤-CD without cholesterol extraction from the plasma membrane (12). In the present experimental conditions, the number of cells and the concentration of total protein were not decreased by the treatment with HP-␤-CD (10 mM) and HEE-SS-PRX (1 mM ␤-CD), suggesting the negligible toxicity of these treatments.
The accumulation of autophagosomes in NPC1 fibroblasts was evaluated by immunostaining of microtubule-associated protein 1 light chain 3 (LC3), which is a soluble cytosolic protein (LC3-I form), whereas C-terminal modification with phosphatidylethanolamine (LC3-II form) leads to localization at the autophagosomal membrane (36,37). The number of cytoplasmic LC3-positive puncta in the NPC1 fibroblasts was significantly higher than in the normal fibroblasts just cultured in the growth medium, confirming that the basal autophagy was perturbed by either the mutation of NPC1 or the lysosomal cholesterol accumulation (Fig. 2, A and E) (27)(28)(29)(30)(31)(32)(33). Upon treatment with HP-␤-CD, the number of LC3-positive puncta in the NPC1 fibroblasts increased further. By contrast, the number of LC3-positive puncta in NPC1 fibroblasts was decreased to the normal level by treatment with HEE-SS-PRX. Although both HP-␤-CD and HEE-SS-PRX were effective in reducing lysosomal cholesterol in NPC1 fibroblasts, their effects on autophagy were opposite.

Improvement of Autophagosomes and p62 Accumulation in NPC1 Fibroblasts by Biocleavable
Polyrotaxanes-Further confirmation of these results was obtained by immunoblotting for LC3-II and p62/SQSTM1 (p62), a selective substrate for autophagic protein degradation (Fig. 3, A and B) (38). Note that the levels of both LC3-II and p62 increased in NPC1 fibroblasts compared with normal fibroblasts even in the basal condition, indicating that the autophagic flux in NPC1 fibroblasts was decreased compared with the normal fibroblasts. Treatment with Baf A, a specific inhibitor of the vacuolar-type H ϩ -ATPase, resulted in increased levels of both LC3-II and p62 in the normal and NPC1 fibroblasts due to the decrease in lysosomal enzymatic activity by the neutralization of the lysosomal pH (39). Similar to the Baf A treatment, HP-␤-CD induced increases in the levels of both LC3-II and p62 in a concentration-dependent manner. By contrast, HEE-SS-PRX reduced the levels of LC3-II and p62 in a concentration-dependent manner, and these protein levels reached levels comparable with those observed in normal fibroblasts at the HEE-SS-PRX concentration equivalent to 1 mM ␤-CD. HP-␤-CD treatments also increased the levels of both LC3-II and p62 in the normal fibroblasts, whereas HEE-SS-PRX treatment showed negligible effects or induced a slight reduction of LC3-II and p62 in the normal fibroblasts (Fig. 3C). In the NPC1 fibroblasts treated with HP-␤-CD or HEE-SS-PRX in the presence of Baf A, the levels of LC3-II and p62 increased to comparable levels, regardless of the concentrations of HP-␤-CD or HEE-SS-PRX (Fig.  3D).
mTORC1 Activity in NPC Fibroblasts Treated with HP-␤-CD and Biocleavable Polyrotaxanes-The mammalian target of rapamycin complex 1 (mTORC1) plays a pivotal role in the regulation of autophagy (40 -42), and the inhibition of mTORC1 activity, such as by amino acid starvation, causes the induction of autophagy. To investigate whether HP-␤-CD and HEE-SS-PRX treatments induced mTORC1-dependent autophagy, the dephosphorylation of p70 S6 kinase (p70S6K) and eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) was monitored at various time points during the treatments (Fig. 4A). When the NPC1 fibroblasts were cultured in starvation medium, the levels of LC3-II and p62 decreased with time (Fig. 4, B and C). In addition, the dephosphorylation of Thr-368 of p70S6K and the conversion of the phosphorylated ␥-form to the dephosphorylated ␣-form were observed for 4E-BP1 (Fig.  4D) (43). HP-␤-CD proportionally increased LC3-II and p62 levels during the treatment time periods, whereas negligible dephosphorylation of p70S6K and 4E-BP1 was observed (Fig. 4,  A-D). HEE-SS-PRX treatment decreased the levels of LC3-II and p62 in the NPC1 fibroblasts, which reached constant levels after 24 h of incubation, but negligible dephosphorylation of p70S6K and 4E-BP1 was observed (Fig. 4, A-D).
The Maturation of Autophagosomes and the Lysosomal Enzymatic Activity in NPC Fibroblasts-To verify whether the HP-␤-CD and HEE-SS-PRX treatments affected the autolysosome formation, autophagic flux in the NPC1 fibroblasts was monitored using an expression vector encoding mRFP-GFP tandem fluorescence-tagged LC3 (mRFP-GFP-LC3) (Fig. 5, A  and B) (35). When the mRFP-GFP-LC3 is localized to the autophagosomes, it emits both mRFP (red) and GFP (green) signals and appears as yellow. By contrast, when the mRFP-GFP-LC3 is localized to the acidic autolysosome, it emits only an mRFP signal and appears as red because the GFP signal quenches immediately under acidic conditions (pK a of GFP, 6.0) (44). Compared with the untreated normal fibroblasts, the number of mRFP ϩ -GFP ϩ yellow puncta (autophagosomes) increased, and the number of mRFP ϩ -GFP Ϫ red puncta (autolysosomes) decreased remarkably in the untreated NPC1 fibroblasts. HP-␤-CD treatment slightly increased the number of mRFP ϩ -GFP ϩ yellow puncta, but a negligible change was observed in the number of mRFP ϩ -GFP Ϫ red puncta compared with the untreated NPC1 fibroblasts. Note that HEE-SS-PRX treatment decreased the number of mRFP ϩ -GFP ϩ yellow puncta and increased the number of mRFP ϩ -GFP Ϫ red puncta in the NPC1 fibroblasts to levels comparable with those observed in the normal fibroblasts.
There are two possible reasons to explain the results of HP-␤-CD and HEE-SS-PRX treatments in the mRFP-GFP-LC3 reporter experiments; these treatments (i) inhibited or promoted the fusion of autophagosomes and lysosomes or (ii) changed the lysosomal pH or lysosomal enzymatic activity. To discriminate between these two possibilities, the lysosomal enzymatic activity in the treated NPC1 fibroblasts was investigated by assessing the activity and the maturation of cathepsin B, one of the major lysosomal cysteine proteases (45,46). The enzymatic activity and the maturation of cathepsin B were suppressed by Baf A treatment due to the neutralization of the acidic lysosomes (Fig. 6A-D) (47, 48). By contrast, both HP-␤-CD and HEE-SS-PRX treatments caused negligible effects on the enzymatic activity and the maturation of cathepsin B in the NPC1 fibroblasts, indicating that the results of the mRFP-GFP-LC3 reporter experiments were attributable to the fusion of autophagosomes and lysosomes.
The Effects of HP-␤-CD and Biocleavable Polyrotaxanes on the Colocalization of Autophagosomes and Lysosomes-To further confirm the facilitated maturation of autophagosomes in NPC1 fibroblasts by HEE-SS-PRX, the colocalization of the transiently expressed mRFP-LC3 with LAMP1-positive puncta was investigated (Fig. 7, A and B). Consistent with the mRFP-GFP-LC3 reporter experiments, the colocalization percentage of LAMP1 and mRFP-LC3 in the NPC1 fibroblasts was lower than in the normal fibroblasts, indicating that the mutation of NPC1 perturbed the maturation of autophagosomes. HP-␤-CD treatment further decreased the colocalization percentage of mRFP-LC3 with LAMP1 in the NPC1 fibroblasts. HEE-SS-PRX treatment increased the colocalization percentage, indicating that the HEE-SS-PRX facilitated the formation of autolysosomes in the NPC1 fibroblasts. We also confirmed the colocalization of endogenous LC3 and LAMP1 in the treated NPC1 fibroblasts by immunostaining. Consistent with Fig. 7, most of the LC3-positive puncta were colocalized with LAMP1-positive puncta after the treatment with HEE-SS-PRX (data not shown). These results further confirm that the HEE-SS-PRX facilitated the maturation of accumulated autophagosomes in NPC1 fibroblasts.
The Effects of ␣-CD-threaded Polyrotaxanes and Other ␤-CD Derivatives-To clarify the effects of the ring size of cyclodextrins on cholesterol reduction and autophagy response in NPC1 fibroblasts, two sets of ␣-CD-threaded PRXs composed of PEG/ ␣-CD or Pluronic P123/␣-CD were synthesized. Because the binding affinity of ␣-CD (six glucose units) to cholesterol is lower than for ␤-CD (seven glucose units) (49), the lysosomal cholesterols in NPC1 fibroblasts were only minimally reduced by treatment with the two sets of ␣-CD-threaded PRXs or ␣-CD alone (Fig. 8, A and B). Additionally, the levels of LC3-II and p62 in NPC1 fibroblasts remained unchanged by the two sets of ␣-CD-threaded PRXs or ␣-CD alone (Fig. 8C). These results suggest that the effect of biocleavable PRXs on facilitating the autophagic flux in NPC1 fibroblasts is related to the ability to reduce the lysosomal cholesterol level.
We also demonstrated the autophagy response of the NPC1 fibroblasts after the treatment with DM-␤-CD, which has a higher binding affinity for cholesterols than does HP-␤-CD (50). DM-␤-CD showed a greater ability in reducing lysosomal cholesterols in the NPC1 fibroblasts (Fig. 8D) and induced a higher level of cholesterol extraction from the plasma membrane than did HP-␤-CD (Fig. 8E). The levels of LC3-II and p62 in the NPC1 fibroblasts treated with DM-␤-CD were higher than in those treated with HP-␤-CD (Fig. 8F). According to these results, it is plausible that the effect of ␤-CD derivatives on the inhibition of autophagic flux is related to the cholesterol binding affinity of ␤-CD derivatives or the ability of cholesterol extraction from the plasma membrane.

DISCUSSION
Impaired autophagy in lysosomal storage disorder is of significant importance, not only for an understanding of the symptoms but also for finding a therapeutic method. In NPC disease, increased levels of LC3-positive puncta was observed even under basal conditions. This result can be attributed to various possibilities, such as the alteration of lysosomal enzymatic activities (32), the activation of Beclin-1 (30), and the impairment of autophagosome maturation (33). Our present results suggest that the perturbation of autophagosome-lysosome fusion is a major issue in NPC disease because negligible change in lysosomal enzymatic activities and Beclin-1 expression was observed in the human models of NPC1 fibroblasts (Figs. 3A, 5, and 7). It is clear that basal autophagic flux decreased in the NPC1 fibroblasts compared with the normal fibroblasts, but it is not completely inhibited because the levels of LC3-II and p62 in the untreated NPC1 fibroblasts are lower than in the Baf A-treated cells (Fig. 3).
There are various reports on the effects of HP-␤-CD or ␤-CD derivatives on autophagy. Fujimoto and co-workers (51) have reported that methyl-␤-CD increases the number of autophagosomes, presumably due to the extraction of cholesterols from the plasma membranes. Segatori and co-workers (52) have reported that HP-␤-CD induces autophagy through the activation of transcription factor EB. Arima and co-workers (53) have reported that folate-appended methyl-␤-CD induces autophagy in folate receptor-expressing cancer cells. Jaenisch and co-workers (33) have reported that treatment with a high concentration of HP-␤-CD induces the cytoplasmic accumulation of autophagosomes and perturbs the protein degradation in Npc1 Ϫ/Ϫ mouse embryonic fibroblasts. We confirmed that HP-␤-CD treatment increased the levels of both LC3-II and p62 in a time-dependent manner in the NPC1 fibroblasts, consistent with the report of Jaenisch and co-workers (Figs. 2 and 3) (33). Also, our results indicate that HP-␤-CD treatment only minimally induced dephosphorylation of p70S6k and 4E-BP1, indicating that the levels of LC3-II and p62 increased by HP-␤-CD were not attributable to an mTOR-dependent autophagy (Fig. 3). Additionally, HP-␤-CD did not decrease lysosomal enzymatic activity (Fig. 6). These results suggest that HP-␤-CD perturbs autophagic flux without affecting mTORCa activity and lysosomal enzymatic activity.
Notably, the ␤-CD-threaded PRXs improved the impaired autophagic flux in NPC disease. The HEE-SS-PRX treatment showed negligible induction of mTOR-dependent autophagy, and a long incubation time was required to induce a decline in the levels of LC3-II and p62 (24 h) compared with the mTORdependent autophagy induced by the amino acid starvation (2 h) (Fig. 3). This suggests that for reducing LC3-II and p62 levels, the mechanism of action by which HEE-SS-PRX works is different from the starvation-induced autophagy. The mRFP-GFP-LC3 reporter experiments and the result of mRFP-LC3 and LAMP1 colocalization revealed that HEE-SS-PRX may promote the fusion of autophagosomes and lysosomes (Figs. 5 and 7). Because the HEE-SS-PRX also had no effect on the lysosomal enzymatic activity (Fig. 6) and the reduction of LC3-II and p62 was not observed in the presence of Baf A (Fig.  3D), it is reasonable to consider that the reduction of LC3-II and p62 levels by HEE-SS-PRX could be due to the promotion of autophagosome maturation.
Ballabio and co-workers (54) have reported that lysosomal cholesterol accumulation inhibits the fusion of endosomeslysosomes and endosomes-autophagosomes in two model cells of lysosomal storage disorders. These authors have shown that the soluble N-ethylmaleimide-sensitive factor attachment protein (SNAP) receptors (SNAREs), such as VAMP7, Vti1b, and syntaxin 7, are preferentially sequestered to these cholesterolenriched compartments. Enrich and co-workers (55) have reported the mislocalization of syntaxin 6 to endosomes instead of to the trans-Golgi network in NPC1-mutated Chinese hamster ovary cells. Accordingly, lysosomal cholesterol accumulation may cause the mislocalization of these transmembrane proteins to inhibit membrane fusion. Therefore, it is hypothesized that the localization of syntaxin 17 or VAMP8, essential factors in the fusion of autophagosomes with lysosomes (56,57), might be altered by excess cholesterols, perturbing the fusion of autophagosomes with lysosomes. If this is true, then the reduction in lysosomal cholesterol by HEE-SS-PRX would be expected to improve the localization of these fusion factors. When we demonstrated the autophagy response of ␣-CDthreaded PRXs, which induced lower lysosomal cholesterol reductions than the ␤-CD-threaded PRXs, negligible change in the levels of LC3-II and p62 was observed ( Figs. 2 and 8, A and  B). Although the detailed mechanism is still controversial, lysosomal cholesterol reduction by ␤-CD-threaded biocleavable HEE-SS-PRXs played a pivotal role in the improvement of autophagic flux in the NPC1 fibroblasts.
Nevertheless, it is not known why the treatment with ␤-CD did not improve the impaired autophagic flux in NPC1 fibroblasts, although it was able to reduce lysosomal cholesterols at a high concentration (10 mM). Note that HP-␤-CD increased the levels of LC3-II and p62, even in the normal fibroblasts (Fig.  3C). Additionally, DM-␤-CD induced higher levels of LC3-II and p62 than did HP-␤-CD (Fig. 8, D and E). The critical difference between HP-␤-CD and HEE-SS-PRX is the ability to extract cholesterols at the plasma membrane (Fig. 2C). Because the hydrophobic cavity of ␤-CD is occupied with the polymer chain in the PRX structure, this unique structure does not permit the further inclusion of other molecules into the cavity of ␤-CD. In reality, HEE-SS-PRX has shown negligible interaction with the plasma membrane and becomes internalized into cells via endocytosis (12). By contrast, HP-␤-CD interacts with the plasma membrane to extract cholesterols due to the exposure of its hydrophobic cavity (Fig. 2C). It is known that the cholesterol extraction from the plasma membrane via ␤-CD perturbs the formation of clathrin pit budding, leading to an inhibition of endocytosis (58,59). It is reported that the clathrin is a key factor in autophagy-related events, such as autophagic lysosome reformation (60). It is hypothesized that the extraction of cholesterols from the plasma membrane via HP-␤-CD affects the process of autophagic flux.
In this study, we have shown that HEE-SS-PRX can simultaneously improve lysosomal cholesterol accumulation and autophagy impairment in NPC disease. To date, no drugs have been reported that can improve both lysosomal cholesterol accumulation and autophagy impairment in NPC disease. For example, rapamycin is reported to improve autophagy impairment in NPC disease but has negligible effects on lysosomal cholesterols (33). N-Butyldeoxynojirimycin improves the clinical symptoms of NPC disease (61) but has a negligible effect on autophagy in NPC disease model cells (30). Therefore, our designed ␤-CD-threaded biocleavable PRXs have great potential for the treatment of NPC disease. However, it is important to note that delivery of the therapeutics to the brain is required for ameliorating the neurodegeneration in NPC disease. The body disposition and the delivery efficacy of biocleavable PRXs to the brain are currently being investigated in our laboratory as well as the molecular design of PRXs to cross the blood brain barrier.