Abstract
Gaucher disease (GD) is caused by mutations in the GBA1 gene, which encodes the lysosomal enzyme glucocerebrosidase (GCase). The severe forms of GD are associated with neurodegeneration with either rapid (Type 2) or slow progression (Type 3). Although the neurodegenerative process in GD has been linked to lysosomal dysfunction, the mechanisms involved are largely unknown. To identify the lysosomal alterations in GD neurons and uncover the mechanisms involved, we used induced pluripotent stem cells (iPSCs) derived from patients with GD. In GD iPSC-derived neuronal cells (iPSC-NCs), GBA1 mutations caused widespread lysosomal depletion, and a block in autophagic flux due to defective lysosomal clearance of autophagosomes. Autophagy induction by rapamycin treatment in GD iPSC-NCs led to cell death. Further analysis showed that in GD iPSC-NCs, expression of the transcription factor EB (TFEB), the master regulator of lysosomal genes, and lysosomal gene expression, were significantly downregulated. There was also reduced stability of the TFEB protein and altered lysosomal protein biosynthesis. Treatment of mutant iPSC-NCs with recombinant GCase (rGCase) reverted the lysosomal depletion and autophagy block. The effect of rGCase on restoring lysosomal numbers in mutant cells was enhanced in the presence of overexpressed TFEB, but TFEB overexpression alone did not reverse the lysosomal depletion phenotype. Our results suggest that GBA1 mutations interfere with TFEB-mediated lysosomal biogenesis, and that the action of GCase in maintaining a functioning pool of lysosomes is exerted in part through TFEB. The lysosomal alterations described here are likely to be a major determinant in GBA1-associated neurodegeneration.
Introduction
Lysosomal storage diseases (LSDs) are a group of inherited metabolic disorders that are often characterized by neurodegeneration (1). Gaucher disease (GD) is the most prevalent LSD and is caused by mutations in the GBA1 gene, which encodes the lysosomal enzyme glucocerebrosidase (GCase) (2). GBA1 mutation is also a known major risk factor for Parkinson's disease (3,4). GBA1 mutations result in decreased GCase activity and accumulation of glucosylceramide (Glc-Cer), mainly in cells of the reticuloendothelial and nervous systems (5–8). Based on neurological involvement, GD has been commonly classified into non-neuropathic (Type 1) and neuropathic forms (Types 2 and 3) (9–11). However, based on extensive clinical studies, the current view is that there is a wide spectrum of GD-associated neurological manifestations, with no distinct separation between different clinical subtypes (3,12).
In postmitotic neurons, macroautophagy (referred to as autophagy hereafter) enables lysosomes to recycle subcellular components (13,14). Autophagy also plays a central role in the removal of misfolded and toxic aggregation-prone proteins that are not efficiently degraded by the proteasome (15,16). Failure of lysosomes to carry out these clearing functions is a key alteration in many neurodegenerative diseases (16–19). Previous studies using mouse models and induced pluripotent stem cell (iPSC)-derived neurons have shown that loss of GCase activity results in decreased protein degradation by lysosomes, accumulation of α-synuclein and autophagy dysregulation (20–24). However, the mechanisms by which GBA1 mutations lead to lysosomal dysfunction remain unclear.
The basic helix–loop–helix transcription factor EB (TFEB) is an important regulator of lysosomal homeostasis and autophagy. This regulation is exerted through TFEB binding to the CLEAR (Coordinated Lysosomal Expression and Regulation) element in the promoters of autophagy and lysosomal genes (25,26). The importance of TFEB in lysosomal function and neuronal survival is underlined by reports that in many neurodegenerative disease models, enhancement of TFEB function promotes protein clearance by lysosomes and neuroprotection (27–30). It has also been shown that ectopic expression of TFEB in GD patient fibroblasts enhanced mutant GCase folding and trafficking (31), suggesting that TFEB may be a therapeutic target in GD.
The difficulty in obtaining neuronal cells from GD patients has hindered analysis of the mechanisms by which mutations in GBA1 lead to neurodegeneration. The recent advent of iPSCs generated by reprograming of adult patient cells with pluripotency transcription factors (32–34) provides a unique opportunity for disease modeling (35–37). We have previously derived iPSCs from patients with Types 1, 2 and 3 GD, and showed that GD iPSC-derived neurons had low levels of GCase enzyme and accumulated Glc-Cer substrate (5,38). It has also been reported that GD iPSC-derived neurons are responsive to chaperone treatment, which increased GCase protein level and activity (8), illustrating the relevance of GD-iPSC neurons for therapeutic development. In this study, we used GD iPSC-derived neuronal cells (iPSC-NCs) to examine the effects of GBA1 deficiency on lysosomal functions. We found that neuropathic GBA1 mutations cause widespread depletion of lysosomes, interfere with lysosomal clearance of autophagic vesicles, and that autophagy induction in the mutant cells leads to neuronal cell death. Further analysis showed that GBA1 mutations were associated with decreased expression of TFEB and downregulation of lysosomal gene expression. In addition, GCase deficiency resulted in reduced stability of the TFEB protein and also interfered with the biosynthetic pathway of lysosomal proteins. Treatment of the mutant iPSC-NCs with recombinant GCase (rGCase) reverted the lysosomal depletion and autophagy block, upregulated TFEB target gene expression and increased TFEB stability. Our results lend support to the idea that GCase is required for maintaining a functioning pool of lysosomes, and that this effect is mediated in part through TFEB.
Results
Characterization of GD iPSC-NCs
The GD iPSCs representative of all three clinical subtypes of GD used in this study have been previously described (5,38). These iPSC lines were derived from two patients with Type 2 GD harboring the biallelic mutations L444P/RecNciI and W184R/D409H (GD-2), a Type 3 patient with L444P/L444P (GD-3) and a Type 1 patient with N370S/N370S (GD-1) mutations. Supplementary Material, Table S1 shows the genotypes of all GD and control iPSC lines used in this study. Neuronal stem cells (NSCs) were obtained by directed differentiation of iPSC-derived embryoid bodies (EBs) to neuronal rosettes, which were manually picked and expanded in culture as described in Materials and Methods (Supplementary Material, Fig. S1A). Both control and GD-iPSCs were efficiently differentiated to NSCs as determined by Sox-1 and Nestin expression (Supplementary Material, Fig. S1B), and ∼90% of the cells were Sox-1 positive. NSCs were then differentiated into neuronal cells by culturing them for 3–4 weeks in Neurobasal media supplemented with glial cell-derived neurotrophic factor (GDNF) and brain-derived neurotrophic factor (BDNF). Both control and GD iPSC-NCs expressed the neuronal-specific markers class III β-tubulin (Tuj-1) and microtubule-associated protein 2 (MAP-2) (Fig. 1A), and as shown in Figure 1B, >80% of iPSC-derived neurons expressed both neuronal markers. We also detected a small percentage of cells (<10%) expressing the glial-specific marker glial fibrillary acidic protein (GFAP) in both control and GD neuronal cultures (Fig. 1C). GD iPSC neurons from all GD subtypes exhibited very low levels of GCase enzymatic activity compared with controls (Fig. 1D). In a previous study, we showed that GD iPSC neurons accumulate Glc-Cer, the substrate of GCase enzyme (5).
Characterization of GD iPSC-NCs. (A) Representative immunofluorescence images of control and GD iPSC-NCs labeled with antibodies to the neuronal-specific markers Tuj-1 (green) and MAP-2 (red). Magnification, ×10; scale bar, 100 μm. (B) Quantitative analysis of neuronal marker expression in control and GD iPSC-NCs. iPSC-NCs that stained positive for Tuj-1 and MAP-2 are represented as the percentage relative to total cell number. Bars represent average ± SEM, n = 3. (C) Representative immunofluorescence images of control and GD-2 iPSC-NCs stained with antibodies to the glial marker GFAP (green) and to Nestin (red). Magnification, ×10; scale bar, 100 μm. (D) rGCase enzyme activity in cell lysates from control and GD iPSC-NCs. Data represent average fold activity relative to control in duplicate wells ± SEM. *P < 0.001 between control and all GD genotypes, assessed by one-way ANOVA.
Lysosomal depletion in neuropathic GD iPSC-NCs
To visualize the lysosomal compartment, control and GD iPSC-NCs were stained with antibody against the lysosomal-associated membrane protein 1 (LAMP1). As shown in Figure 2A, immunofluorescence images showed noticeable depletion and altered clustering pattern of LAMP1-labeled lysosomes in GD-2 and GD-3 iPSC-NCs compared with GD-1 and control cells. In contrast to the intense fluorescence signal in the supranuclear region of control cells, lysosomes in GD-2 and GD-3 cells were sparse, diffuse and weakly stained (Fig. 2A and B). The expression of lysosomal-associated membrane protein 2 (LAMP2) was similarly decreased in GD mutant cells compared with controls, and counterstaining for Tuj-1 showed no reduction of this neuronal marker in GD cells compared with control, suggesting that GCase deficiency did not impair NSC differentiation to Tuj-1-labeled neurons (Supplementary Material, Fig. S2). As shown in Supplementary Material, Figure S3, the expression of the lysosomal enzyme cathepsin D was also decreased in neuropathic GD iPSC-NCs compared with control cells. To quantitate LAMP1-labeled lysosomes, we acquired Z-stack fluorescence images from control and GD mutant cells (Fig. 2B) and performed automated morphometric analysis as described in Materials and Methods. As shown in Figure 2C, there was a significant decrease in lysosomal number and fluorescence intensity in GD-2 and GD-3 iPSC-NCs compared with control and GD-1 iPSC-NCs. Consistent with the immunofluorescence data, western blot analysis showed a significant decrease in LAMP1 protein in GD-2 and GD-3 iPSC-NCs compared with GD-1 and control cells (Fig. 2D). The LAMP1 band in GD-2 and GD-3 iPSC-NCs had an apparent lower molecular weight (MW, ∼100 kDa) than that of control and GD-1 iPSC-NCs (∼120 kDa), suggesting that post-translational modification of LAMP1 may be altered in neuropathic GD iPSC-NCs (Fig. 2D).
Lysosomal depletion in GD iPSC-NCs. (A) Representative immunofluorescence images of control and GD iPSC-NCs stained with anti-LAMP1 (red) and DAPI (blue). Magnification, ×20; scale bar, 100 μm. (B) Representative fluorescence Z-stack images of control and GD-2 iPSC-NCs stained with anti-LAMP1 (red) and DAPI (blue). Magnification, ×20; scale bar, 10 μm. (C) Quantitative analyses of LAMP1-labeled lysosomes in control, GD-1, GD-2 and GD-3 iPSC-NCs, showing automated measurement of average number of LAMP1-labeled lysosomes and average LAMP1 fluorescence intensity ± SEM. Compiled data from >1000 cells assayed in at least four different fields per genotype in at least two independent experiments. *P < 0.05 and **P < 0.005 as assessed by one-way ANOVA, error bars = SEM. (D) Left panel, representative western blot for LAMP1 protein in control, GD-1, GD-2 and GD-3 iPSC-NCs; β actin loading control is also shown. Apparent LAMP1 MW is ∼120 kDa in control and GD-1 iPSC-NCs and ∼100 kDa in GD-2 and GD-3 iPSC-NCs. Bar graph in right panel represents LAMP1 levels in GD iPSC-NCs relative to control. *P < 0.05 as assessed by one-way ANOVA. Error bars = SEM; n = 3.
We then examined whether restoring GCase activity in GD iPSC-NCs would rescue the lysosomal phenotype of the mutant cells. Control and GD iPSC-NCs were incubated with rGCase for 5 days and then assayed for GCase enzymatic activity in cell lysates. GCase enzyme activity was markedly increased in all of the treated cultures, indicating that the rGCase enzyme was taken up by GD iPSC-NCs and retained its activity (Supplementary Material, Fig. S4A). Previous studies have shown that rGCase is taken up by macrophages and neuronal cells through mannose receptors (39,40). To examine how rGCase is internalized in GD iPSC-NCs, these cells were preincubated with mannan, a polymer of mannose previously shown to block the uptake of rGCase (40), followed by incubation with rGCase. Supplementary Material, Figure S4B shows that incubation with mannan markedly decreased GCase activity in GD mutant neurons, suggesting that a significant fraction of rGCase was internalized through mannose receptors. As shown in Supplementary Material, Figure S5, rGCase treatment resulted in a 2- to 3-fold increase in lysosomal numbers in GD-2 iPSC-NCs compared with untreated cultures.
We conclude that mutant GCase caused lysosomal depletion in GD-2 and GD-3 cells, and that treatment with rGCase was able to reverse lysosomal loss in the mutant cells.
Autophagy block in neuropathic GD iPSC-NCs
We then investigated whether autophagy, which is dependent on lysosomal function, was compromised in GD neuronal cells. To this end, we examined endogenous levels of the mammalian autophagy proteins Microtubule-associated protein 1A/light chain 3 (LC3) and p62/SQSTM1 in control and GD iPSC-NCs. LC3 is an autophagosome marker. LC3 protein is present in two forms, LC3-I, which has a cytoplasmic localization, and LC3-II, which associates with autophagosomal membranes; LC3-II levels and LC3 localization to intracellular puncta corresponding to autophagosomes reflect autophagosome content (41). p62/SQSTM1 (p62) is a scaffold protein targeting ubiquitinated substrates to autophagosomes (41,42). p62 is degraded by autophagy, and a decrease in p62 levels reflects efficient degradation via autophagy-lysosomal pathway. Conversely, accumulation of p62 indicates a block in autophagy flux. Thus, cellular levels of LC3-I/LC3-II and p62 reflect the autophagic activity of the cells (43). High-resolution confocal images revealed that LC3 levels are increased in neuropathic GD cells (Fig. 3A). Consistent with accumulation of autophagosomes, accumulated LC3 in mutant iPSC-NCs had punctate localization (Fig. 3A inset). Western blot analysis and band quantitation showed increase in basal levels of LC3-II and p62 proteins in neuropathic GD cells compared with GD-1 and control cells (Fig. 3B). Thus, our data indicate that under basal conditions, there was an accumulation of autophagosomes in neuropathic but not in non-neuropathic GD iPSC-NCs. Additionally, accumulation of p62 suggests that this may be due to a decrease in autophagy flux rather than increased autophagosome formation.
Autophagy phenotype in GD iPSC-NCs. (A) Representative confocal images for control, GD-1, GD-2 and GD-3 iPSC-NCs stained with anti-LC3 (green), anti-MAP-2 (red) and DAPI (blue). Magnification, ×63; scale bar, 50 μm. Inset is an enlargement of outlined area in GD-2 iPSC-NCs showing LC3-labeled cytoplasmic puncta. (B) Representative western blot showing basal levels of LC3 and p62 proteins in control, GD-1, GD-2 and GD-3 iPSC-NCs. β-Actin was used as a loading control. Bar graphs represent quantitation of LC3-II (left) and p62 (right) in control and GD iPSC-NCs. Data represent average ± SEM, n = 3 per group. *P < 0.05 between indicated groups as assessed by one-way ANOVA. (C) Representative western blot showing LC3 levels in control and GD-2 iPSC-NCs with or without 100 μm chloroquine treatment for 24 h. Bar graph represents LC3-II ratio to actin in control and GD-2 iPSC-NCs with or without lysosomal inhibitor treatment for 24 h. Data represent average ± SEM, n = 3–7. *P < 0.05 between indicated groups as assessed by Student's t-test. (D) Representative western blot showing LC3 levels in control and GD-2 iPSC-NCs treated with 100 nm rapamycin for 48 h. Bar graph represents fold LC3-II ratio to actin in control and GD-2 iPSC-NCs with and without rapamycin treatment. Data represent average ± SEM, n = 3. *P < 0.05 between indicated groups as assessed by Student's t-test. (E) Representative western blot for p62 levels in control and GD-2 iPSC-NCs with and without 100 nm rapamycin treatment for 24 h. Bar graph represents p62 ratio to actin in control and GD-2 iPSC-NCs with and without rapamycin treatment. Data represent averages ± SEM, n = 2–5. *P < 0.05 between indicated groups as assessed by Student's t-test.
As GD-2 iPSC-NCs exhibited the most pronounced lysosomal and autophagy phenotypes, we used this genotype to further investigate the mechanisms involved. To assess the autophagic flux, we used lysosomal inhibitor to block autophagosomal degradation. Under these conditions, LC3-II levels reflect autophagosome formation independent from the clearance process (43). Western blot analysis showed that inhibition of lysosomal function with chloroquine caused a significant increase in LC3-II levels in control but not in GD-2 iPSC-NCs (Fig. 3C). The presence of similar levels of LC3-II in mutant cells, regardless of the presence or absence of lysosomal inhibitors, suggests that the accumulation of autophagosomes in GD-2 iPSC-NCs is most likely due to a defect in lysosomal clearance of autophagosomes. We then examined whether autophagy induction can overcome the defect in autophagy flux and improve autophagosomal clearance. Control and GD-2 cells were treated with rapamycin, an inhibitor of mTOR and inducer of autophagy flux, and cell lysates were analyzed by immunoblotting with antibodies against LC3 and p62. As shown in Figure 3D and E, autophagy induction caused significant increase in both LC3-II and p62 levels in GD-2 iPSC-NCs but not in controls cells. As p62 protein is normally sequestered and then degraded in autophagosomes, its accumulation indicates compromised autophagic clearance (42). Thus, our data indicate that GD iPSC-NCs exhibit defective autophagic flux, which may result from decreased autophagosome clearance by the lysosomes.
Defective autophagic clearance in neuropathic GD iPSC-NCs
To further investigate the autophagy defect in GD iPSC-NCs, we infected control and GD iPSC-NCs with a lentivirus encoding a green fluorescence protein (GFP)-LC3 fusion autophagy reporter protein. Both control and GD cells were infected at similar efficiency, and there was no difference in initial GFP-LC3 fluorescence intensity between both populations. We then induced autophagy by treatment of the cells with rapamycin. As shown in Figure 4A, rapamycin treatment induced higher accumulation of GFP-LC3-labeled puncta in the cytoplasm of GD-2 iPSC-NCs than in that of control cells. To quantify the autophagosomes, Z-stack images were acquired from control and mutant cultures with and without rapamycin treatment (Fig. 4A), and automated morphometric analysis of GFP-labeled autophagosomes was performed. Morphometric analysis showed a significant increase in the average number, area and fluorescence intensity of autophagosomes in GD-2 iPSC-NCs following rapamycin treatment, compared with untreated cells (Fig. 4B). rGCase treatment reduced the increase in autophagosomal parameters induced by rapamycin in GD-2 iPSC-NCs, to levels similar to those in control cells treated with rapamycin (Fig. 4B and Supplementary Material, Fig. S6A), suggesting that the autophagy defect observed was due to GCase deficiency.
Defective autophagic clearance in neuropathic GD iPSC-NCs. (A) Representative Z-stack fluorescence images for control and GD-2 iPSC-NCs infected with GFP-LC3 virus that were either treated with 100 nm rapamycin for 24 h or were left untreated. Arrows point to GFP-LC3-labeled puncta in control and GD-2 iPSC-NCs treated with rapamycin. Magnification, ×20; scale bar, 10 μm. (B) Quantitation of GFP-LC3 puncta in experiment shown in (A). Also shown is puncta analysis in GD-2 iPSCs that were treated with rGCase enzyme for 5 days before rapamycin treatment. Data represent automated compiled measurements from >1000 cells assayed in two to five different fields in a representative experiment. Bars represent average puncti number, area and fluorescence intensity per cell ± SEM. *P < 0.05 and **P < 0.005 as assessed by one-way ANOVA. (C) Representative Z-stack fluorescence images for control and GD-2 iPSC-NCs that were infected with GFP-LC3 virus and treated with 100 nm rapamycin for 24 h. Infected cells were then stained with anti-LAMP1 antibody. The merged image of GFP-LC3 puncti (green) and LAMP1-labeled lysosomes (red) is also shown. Magnification, ×20; scale bar, 10 μm. (D) Bar graph shows quantitation of the fraction of LAMP1/GFP-LC3 puncti in experiment shown in (C). **P < 0.005 as assessed by Student's t-test.
As was the case under basal conditions (Fig. 2A), there were also fewer LAMP1-labeled lysosomes in rapamycin-treated GD-2 iPSC-NCs compared with rapamycin-treated control cells (Supplementary Material, Fig. S6B). To examine whether the GFP-LC3 puncta were associated with lysosomes, we performed automated quantitation of colocalization between GFP-LC3 and LAMP1 fluorescence signal in GFP-LC3-infected iPSC-NCs treated with rapamycin (Fig. 4C). This quantitation showed a significant decrease in the fraction of LAMP1/GFP-LC3 double-positive puncti in GD-2 iPSC-NCs compared with control cells (Fig. 4D), suggesting a decrease in autophagosomal fusion with the lysosomes. Taken together, our data indicate that neuropathic GD iPSC-NCs exhibit autophagy block due to decreased lysosomal clearance of autophagic substrates.
Rapamycin toxicity in neuropathic GD iPSC-NCs
We then examined whether the defective lysosomal clearance in GD-2 iPSC-NCs could affect their survival. We first compared neuronal survival between GD-2 and control iPSC-NCs under basal conditions, using a propidium iodide (PI)/4′,6-diamidino-2-phenylindole (DAPI) double-staining method as described in the Materials and Methods. Under basal conditions, there was no significant difference in neuronal survival between GD-2 and control iPSC-NCs (Fig. 5A). However, prolonged treatment with rapamycin resulted in significant death of GD-2 but not control iPSC-NCs (Fig. 5A and B). After 3 days of incubation with rapamycin, up to 30% of GD-2 iPSC-NCs were dead. In contrast, rapamycin treatment appeared to improve the survival of control cells (Fig. 5B). This rapamycin toxicity was not detected in GD-1 neurons, which did not exhibit significant lysosomal depletion or an altered autophagic flux (data not shown). Taken together, our results suggest that impaired lysosomal clearance of autophagic vesicles in neuropathic GD iPSC-NCs, may predispose them to cell death in response to autophagy induction by rapamycin treatment.
Treatment with rapamycin is toxic to GD2 iPSC-NCs. (A) Representative fluorescence images from control and GD-2 iPSC-NCs untreated or treated with 100 nm rapamycin for 3 days, showing nuclei (blue) and PI-labeled nuclei (red) in the same field. Magnification, ×20; scale bar, 100 μm. (B) Quantitation of cell survival in control and the indicated GD iPSC-NCs with and without treatment with 100 nm rapamycin for 3 days. Bars represent average percentage of dead cells ± SEM, n = 3–5. Asterisks indicate significant differences between indicated groups, **P < 0.005 as assessed by Student's t-test.
Decreased TFEB levels in GD iPSC-NCs
To investigate whether the reduction in lysosomal numbers in GD iPSCs was caused by decreased lysosomal biogenesis, we examined the expression of TFEB, the master regulator of autophagy and lysosomal genes. Confocal microscopy analysis showed decreased endogenous TFEB immunofluorescence intensity in GD-2 and GD-3 iPSC-NCs compared with control cells (Fig. 6A). We also noticed that in GD-2 and GD-3 iPSC-NCs, a higher fraction of TFEB was localized to the nucleus than in control cells (Fig. 6A, merged panel). Quantitation of the number of cells with TFEB nuclear localization showed a significantly higher percentage of cells with TFEB-positive nuclei in GD-2 and GD-3 compared with control cells (Supplementary Material, Fig. S7). Western blot analysis also showed a significant decrease in TFEB protein levels in GD-2 iPSC-NCs compared with control cells (Fig. 6B). The predominant TFEB species in the mutant cells was a faster migrating band, which may correspond to the nuclear dephosphorylated form of TFEB (44). Together, the immunoblot and immunofluorescence data show that in neuropathic GD cells, there is a decrease in endogenous TFEB level, and a preferential nuclear localization. We then examined whether the decrease in TFEB protein level was due to increased proteolysis. To this end, we treated control and GD-2 iPSCs-NCs with the proteasome inhibitor (PSI) clasto-lactacystin β-lactone. After PSI treatment, the cells were stained with anti-TFEB antibodies and analyzed by fluorescence microscopy. As shown in Supplementary Material, Figure S8, treatment with PSI had a minimal effect on TFEB staining intensity in control cells, but in GD-2 iPSC-NCs, there was noticeably more TFEB in PSI-treated than in untreated cultures. These results suggest that in the mutant cells, TFEB may be subjected to increased proteasomal degradation.
Decreased TFEB levels in GD iPSC-NCs. (A) Representative confocal images for control, GD-2 and GD-3 iPSC-NCs stained with anti-Tuj-1 (red), anti-TFEB (green) and DAPI (blue); the merged image is also shown. Magnification, ×63, scale bar, 50 μm. Last panel shows enlargement of the indicated areas in the merged images in each cell line; arrows point to TFEB in the nucleus in GD-2 and GD-3 iPSC-NCs as shown by green and blue fluorescence signal colocalization. Magnification, ×63; scale bar, 25 μm. (B) Representative western blot for TFEB levels in control, GD-2 and GD-3 iPSC-NCs. Bar graph shows quantitation of TFEB levels relative to control, *P < 0.05 analyzed by Student's t-test. Error bars = SEM, n = 3. (C) qRT-PCR analysis showing fold gene expression of TFEB in control, GD-1, GD-2 and GD-3 iPSC-NCs. **P < 0.005 between control and all GD iPSC-NCs, and *P < 0.05 between GD-1 and GD-2, assessed by one-way ANOVA. Error bars = SEM, n = 2. (D) qRT-PCR analysis showing fold expression of TFEB and lysosomal genes in control and GD-2 iPSC-NCs. The lysosomal genes examined were LAMP1, GBA1, cathepsins (CATS) D and B, hexosaminidase A (HEXA) and glucosamine (N-acetyl)-6-sulfatase (GNS). *P < 0.05, **P < 0.005 and***P < 0.0005 as assessed by Student's t-test. Error bars = SEM, n = 2. (E) qRT-PCR analysis showing fold change in expression of TFEB target genes in GD-2 iPSC-NCs that were either left untreated or were treated with 0.24 U/ml rGCase for 5 days. The lysosomal genes examined were the same as in (D). *P < 0.05 and **P < 0.005 as assessed by Student's t-test. Error bars = SEM, n = 3.
We then examined the levels of mRNA expression of TFEB and some of its target lysosomal genes. TFEB expression was significantly decreased in all GD iPSC-NCs compared with control cells, and the extent of the decrease correlated with the severity of the GBA1 mutations (Fig. 6C). The TFEB target genes LAMP1, GBA1, cathepsins D and B, hexosaminidase A and glucosamine (N-acetyl)-6-sulfatase were also downregulated in GD-2 iPSC-NCs compared with control cells (Fig. 6D). As shown in Figure 6E, rGCase treatment partially upregulated their expression in GD-2 neuronal cells. These results lend support to the idea that the reduced lysosome numbers in GD cells may be caused by impaired lysosomal biogenesis, and that this effect may be mediated by interference of mutant GCase with TFEB mRNA expression and TFEB protein stability.
Finally, we examined whether GD NSCs also exhibited lysosomal depletion and a decrease in endogenous TFEB protein. Immunofluorescence analysis showed that in GD-2 NSCs, both LAMP1 and TFEB levels were significantly lower than in control cells (Supplementary Material, Fig. S9). Incubation of NSC cultures with rGCase increased both LAMP1 and TFEB levels (Supplementary Material, Fig. S9), indicating that the observed alterations were due to GCase deficiency.
TFEB overexpression does not restore lysosomal numbers in neuropathic GD iPSC-NCs
To determine whether overexpression of TFEB could rescue the lysosomal phenotype in mutant cells, we infected control and GD-2 iPSC-NCs with a lentivirus encoding a TFEB-GFP fusion protein, and stained TFEB-infected cultures with anti-LAMP1 antibodies. As shown in Figure 7A, TFEB overexpression was unable to reverse the lysosomal depletion in the mutant cells. Quantitative analysis of lysosomal numbers in TFEB-GFP-expressing GD-2 iPSC-NCs showed a significant decrease in both, number and fluorescence intensity of LAMP1-labeled lysosomes compared with control cells (Fig. 7B). Although the infection efficiency in control and GD-2 mutant cultures was similar, the fluorescence intensity of TFEB-GFP was decreased in GD-2 mutant cells. Quantitative analysis showed that TFEB-GFP intensity was 3-fold lower in the mutant cells compared with controls (Fig. 7C). This suggests that the stability of ectopically expressed TFEB in mutant cells is compromised.
TFEB-GFP overexpression in GD iPSC-NCs. (A) Representative Z-stack fluorescence images of control and GD-2 iPSC-NCs infected with TFEB-GFP virus and labeled with anti-LAMP1 (red) and DAPI (blue). Merged image is shown at the right panel. Magnification, ×20; scale bar, 10 μm. (B) Automated quantitation of number (left) and fluorescence intensity (right) of LAMP1-labeled lysosomes in control and GD-2 iPSC-NCs expressing TFEB-GFP. Compiled data from >1000 cells, assayed in at least four different fields (×20), in two independent experiments. Data represent average ± SEM, ***P < 0.0005. Assessed by Student's t-test. (C) Automated quantitation of TFEB-GFP fluorescence intensity in control and GD-2 iPSC-NCs infected with TFEB-GFP virus. Compiled data from >1000 cells, assayed in at least four different fields in two independent experiments. Error bars = SEM, ***P < 0.0005. Assessed by Student's t-test. (D) Representative Z-stack fluorescence images of control and GD-2 iPSC-NCs infected with TFEB-GFP virus and colabeled with anti-LAMP1 and DAPI. Red, LAMP1; green, TFEB-GFP; blue, DAPI. The infected cultures were treated with 0.24 U/ml rGCase for 5 days or left untreated as indicated. Magnification, ×60; scale bar, 10 μm. (E) Automated quantitation TFEB-GFP fluorescence intensity in GD-2 iPSC-NCs infected with TFEB-GFP virus that were either treated with 0.24 U/ml rGCase for 5 days or left untreated. Compiled data from >1000 cells, assayed in at least 4 fields in two independent experiments. Error bars = SEM, *P < 0.05, Assessed by Student's t-test. (F) Automated quantitation of lysosomal number in control and GD-2 iPSC-NCs that were infected with TFEB-GFP virus and then stained with anti-LAMP1 antibody. Lysosomes in TFEB-GFP-infected cells (GFP+) and in uninfected cells (GFP−) were quantitated in the same high-power fields (×60). Compiled data from >200 cells assayed in at least four high-power fields in a representative experiment. Data represent average ± SEM. *P < 0.05 and **P < 0.005, assessed by one-way ANOVA.
TFEB enhances GCase effect on lysosomal biogenesis
To obtain more insight into the functional interactions between GCase and TFEB, control and GD-2 iPSC-NCs cultures infected with TFEB-GFP were treated with rGCase for 5 days, and high-magnification Z-stack images were acquired. We then performed automated quantitation of the number of LAMP1-labeled lysosomes in both, the TFEB-GFP-infected cells (GFP+), and uninfected cells (GFP−) that were present in the same high-power fields. As shown in Figure 7D, rGCase treatment resulted in upregulation of LAMP1-labeled lysosomes in both control and GD-2 iPSC-NCs infected with TFEP-GFP. This treatment also increased the TFEB-GFP fluorescence signal in GD-2 iPSC-NCs, indicating that rGCase stabilized the ectopically expressed TFEB protein (Fig. 7D and E). Quantitation of lysosomal numbers in GFP+ and GFP− cells showed that in the absence of rGCase, there was no significant difference in lysosomal numbers between GFP+ and GFP− cell populations in control or GD-2 iPSC-NCs (Fig. 7F). However, rGCase treatment resulted in a significant increase in lysosomal numbers in control GFP+ but not control GFP− cells (Fig. 7F). Figure 7F also shows that rGCase treatment significantly increased lysosomal numbers in both GFP+ and GFP− GD-2 cells. The presence of significantly more lysosomes in response to rGCase treatment in TFEB-GFP-expressing than in non-expressing cells in both mutant and control cells suggests that the GCase effect on lysosomal biogenesis is enhanced in the presence of TFEB. Taken together, our results suggest that an enzymatically active GCase is required for lysosomal biogenesis and that its effects may be mediated in part through TFEB.
Discussion
Using GD iPSC-NCs as a model system, we found that neuropathic GBA1 mutations result in lysosomal depletion, a block in clearance of autophagosomes and neuronal cell death following autophagy induction. Analysis of the mechanisms involved showed that GBA1 mutations interfered with the expression of TFEB and its target genes, the stability of the TFEB protein, and with post-translational modification of proteins destined for the lysosome. Our results lend support to the idea that an enzymatically active GCase is required for maintaining a functioning pool of lysosomes, and that this effect of GCase is mediated in part through TFEB.
Types 2 and 3 neuropathic GBA1 mutations resulted in a reduced pool of lysosomes, as determined by immunostaining of LAMP1 and LAMP2, which together make up 50% of the lysosomal membrane proteins (45). Besides their function in maintaining lysosomal integrity, LAMP proteins are involved in fusion of lysosomes with other cellular compartments including autophagosomes (46). Thus, this reduction in LAMP proteins may contribute to the failure of autophagosomes to fuse with lysosomes in GD neurons. Interestingly, the LAMP1 protein from Types 2 and 3 GD neurons had a higher electrophoretic mobility than that of Type 1 and control cells. LAMP proteins undergo glycosylation, a post-translational modification that occurs during transport through the ER and Golgi to the lysosome and is also important for its stability (45). This suggests that LAMP1 glycosylation may be altered in GD neurons. The nature of the potential defects in LAMP1 biosynthesis and transport in GD neurons will be analyzed in future studies. A recent study using iPSCs from Pompe disease, another lysosomal storage disorder, reported that LAMP proteins in iPSC-derived mutant cardiomyocytes also had higher electrophoretic mobility than that in control cells, and it was shown that this was due to Golgi-based glycosylation deficiencies (47). It will be interesting to examine whether interference with glycosylation and transport of lysosomal proteins also occurs in other LSDs. As the expression of TFEB and lysosomal genes was also downregulated in GD neurons, it appears that neuropathic GBA1 mutations interfere with lysosomal biogenesis by several mechanisms.
The lysosomes in mutant neurons were impaired in their ability to clear autophagic vesicles. GD neurons exhibited marked accumulation of autophagosomes and p62/SQSTM1 protein under both basal and autophagy-inducing conditions, and there was a reduction in the fraction of autophagosomes fused with lysosomes. A similar decrease in autophagosome/lysosome fusion was also reported in other LSD models (48,49). rGCase treatment reversed the decrease in lysosome numbers and restored autophagosome clearance, demonstrating that the abnormal phenotypes observed were caused by GCase deficiency. Our data showed that failure of lysosomes to clear autophagic vesicles predisposed GD iPSC-NCs to cell death. Although autophagy induction improved the survival of control neurons, it had the opposite effect on neuropathic GD cells, with up to 30% cell death induced by prolonged rapamycin treatment. Although it has been proposed that autophagy induction might be an effective strategy to treat neurodegeneration by forced clearance of protein aggregates (15,50), in GD neurons, autophagy induction led to neuronal cell death. This finding has important implications for the use of autophagy inducers as potential therapeutic agents in GD, because pharmacological induction of autophagy without improvement of lysosomal function may be detrimental to the patients.
One of the most interesting findings of this study was that mutant GCase interferes with TFEB functions by both transcriptional and post-translational mechanisms. GD neuronal cells had decreased mRNA levels of TFEB, a critical regulator of lysosomal homeostasis (25,51,52), and there was reduced stability of the TFEB protein. Blocking of proteasomal degradation resulted in increased TFEB protein levels in GD neurons (Supplementary Material, Fig. S8), suggesting that mutant GCase may cause destabilization of the TFEB protein. In the mutant cells, endogenous and ectopic TFEB had a preferential nuclear localization. As inhibition of lysosomal function has been shown to result in TFEB nuclear localization (25,53), the higher proportion of nuclear TFEB we observed in the mutant neurons may be a compensatory mechanism to induce lysosomal biogenesis and restore lysosomal function. The mechanisms by which abnormal GCase alter TFEB regulation will be the object of future studies, but as TFEB-dependent lipophagy may be regulated by lipid sensing (25,51,52), it is tempting to speculate that the abnormal sphingolipid profile in GD neurons is also sensed by the TFEB regulatory machinery, and that TFEB responds to the changes in lipid profile by modulating its phosphorylation status and localization. In GD patient fibroblasts, TFEB overexpression and activation have been shown to enhance mutant GCase folding and trafficking through the upregulation of proteostasis genes involved in folding and lysosomal trafficking (31). However, TFEB overexpression alone was not sufficient to increase lysosomal numbers in GD iPSC-NCs. Yet, in the presence of ectopically expressed TFEB, the effect of normal GCase in increasing lysosomal numbers was significantly enhanced in both mutant and control neurons. The presence of significantly more lysosomes in response to rGCase treatment in TFEB-GFP-overexpressing cells than in uninfected cells in both, mutant and control neurons, suggests that the effects of GCase on lysosomal biogenesis may be mediated in part through TFEB. Thus, TFEB may be a valuable therapeutic target in the treatment of neurodegeneration caused by GCase deficiency. The phenotypes uncovered in this study will be used as the basis of functional assays to identify new effective treatments for GBA1-associated neurodegeneration.
Previous observations also point to an important role of GCase and glucosylsphingolipids in lysosomal homeostasis. In Saposin C-deficient fibroblasts, it was shown that cathepsins B and D were reduced, causing impaired autophagic clearance (54). It has also been shown that lowering Glc-Cer levels by inhibition of Glc-Cer synthetase in primary neurons increased the size and number of lysosomes and enhanced autophagy flux, an effect that was mediated through mTOR signaling (55). In another study using GD patient fibroblasts, treatment with the chaperone Ambroxol, which helps mutant GCase reach the lysosome, resulted in upregulation of TFEB mRNA and an increase in lysosomal numbers (56). Using iPSC dopaminergic neurons from Parkinson's disease patients with GBA1 mutations, Schondorf et al. (21) reported a decrease in autophagic vesicle fusion with lysosomes and calcium dysregulation.
In summary, we have shown that GBA1 mutations have a deleterious effect on lysosomal biogenesis and function. GCase deficiency caused a downregulation of TFEB mRNA and protein, interfered with the expression and biosynthetic pathway of lysosomal proteins, and the lysosomes in the mutant cells were dysfunctional. The lysosomes in GD neurons were not able to clear autophagic vesicles, and autophagy induction in these cells led to cell death. We conclude that the combined effects of lysosomal depletion and alteration of lysosomal functions caused by GCase deficiency are likely to be major factors leading to neurodegeneration in GD. Our results also suggest that an enzymatically active GCase is required to maintain a functional pool of lysosomes, and that this effect of GCase is mediated in part through TFEB.
Materials and Methods
iPSC lines
The iPSCs from patients with Types 1, 2 and 3 GD and from control subjects used in this study have been previously described (5,38) and are listed in Supplementary Material, Table S1. These included two Type 2 acute neuropathic GD iPSCs harboring the biallelic genotypes L444P/RecNciI and W184R/D409H (GD-2); a Type 3 iPSC with an L444P/L444P genotype (GD-3); a non-neuropathic Type 1 iPSC with an N370S/N370S genotype (GD-1); MJ, a control iPSC line derived from foreskin fibroblasts; and the control DF4-7T.A iPSC, which was purchased from the WiCell Repository.
Generation of iPSC-NCs
To initiate neuronal differentiation, iPSC-derived EBs were maintained in EB culture medium for 10 days. This was followed by 4 days in the presence of 5 μm Dorsomorphin and 10 μm SB431542 (Sigma–Aldrich). EBs were then transferred to Petri dishes coated with Matrigel (BD Biosciences) and maintained in Dulbecco's modified Eagle's medium/F12 media (Invitrogen) plus 1× (Vol/Vol) N2 supplement and 20 ng/ml bFGF (Stemgent). Neural rosettes were visible after 7–10 days in adherent culture and were manually picked and dissociated into single cells using StemPro Accutase (Life Technologies). NSCs were expanded in Neurobasal medium (Life Technologies), containing 1× (Vol/Vol) MEM, non-essential amino acids (Life Technologies), 1× (Vol/Vol) GlutaMAX-I CTS (Life Technologies), 1× (Vol/Vol) B27 supplement (Life Technologies), 1× (Vol/Vol) penicillin/streptomycin and 20 ng/ml bFGF (Stemgent). NSCs were maintained in culture at >80% confluency with media change every other day (57). Neuronal differentiation was induced by plating NSCs on culture dishes or glass cover slips coated with 20 μg/ml poly-l-ornithine (Sigma–Aldrich, cat. no. P3655-50MG) and 10 μg/ml laminin (Life Technologies, cat. no. 23017-015). Neuronal differentiation medium composition: Neurobasal medium (Life Technologies) supplemented with 1× (Vol/Vol) MEM non-essential amino acids (Life Technologies), 1× (Vol/Vol) GlutaMAX-I CTS, 1× (Vol/Vol) B27 supplement, BDNF (10 ng/ml) (eBioscience cat. No. 14-8365), GDNF (10 ng/ml) (eBioscience cat. No. 14-8506), cyclic adenosine monophosphate (100 nm) (Sigma–Aldrich cat. No. D-0260) and ascorbic acid (200 μm) (Sigma–Aldrich cat. No. A-4403). Neuronal cells were maintained in culture for 3–4 weeks with half media change every 2–3 days (57,58).
Immunofluorescence analysis
For marker expression analysis, NSCs were plated on chamber slides (Lab-Tek) and differentiated as described above. For autophagy studies and confocal microscopy imaging, NSCs were plated and differentiated on coverslips or glass-bottom culture dishes (MatTrek). Neuronal cultures were fixed in 4% (Vol/Vol) paraformaldehyde for 15 min and blocked in phosphate-buffered saline (PBS) containing 8% fetal bovine serum (Vol/Vol) for 1 h. Primary antibodies or matching isotype controls were diluted in PBS containing 2 mg/ml saponin and incubated for 2 h at room temperature, followed by 1 h incubation with the corresponding fluorochrome-conjugated secondary antibodies. The following antibodies were used: rabbit anti-Sox-1 (Chemicon cat. No. AB15766) 1:200; mouse anti-Nestin (BD Transduction laboratories, cat. no. 611658) 1:200; mouse anti-Tuj-1 (Neuromics, cat. No. MO15013) 1:200; rabbit anti β3-tubulin (Cell Signaling, cat. No. 5568) 1-200; mouse anti-MAP-2 (Millipore, cat. No. MAB3418) 1:200; rabbit anti-LC3 (Invitrogen, cat. No. L10382) 1:500; rabbit anti-GFAP (Millipore, cat. No. AB5804) 1:200; mouse anti-LAMP1 (U. Iowa Developmental Hybridoma Bank, H4A3) 1:100; mouse anti-LAMP2 (Santa Cruz) 1:200; rabbit anti-TFEB (Millipore) 1:50; goat anti-cathepsin D (Santa Cruz) 1:100 and rabbit anti-GCase, as we described (5). The secondary antibodies used were dyLight 488- or 549-conjugated mouse or rabbit (Jackson ImmunoResearch Laboratories) and Alexa fluor 488- or 594-conjugated mouse or rabbit (Life Technologies), all at 1:200 dilution. Cell nuclei were labeled using DAPI-containing mounting medium (Vectashield; Vector Laboratories). For TFEB studies, neuronal cultures were treated with the PSI clasto-lactacystin β-lactone (Cayman Chemical) at a concentration of 0.5 μg/ml for 18 h, before staining with anti-TFEB antibody as described above. For NSC and neuronal marker expression analysis, cells were plated in duplicate wells, and fluorescence microscopy images were acquired from different fields. The percentage of neuronal cells expressing specific markers was calculated as the number of immunolabeled cells divided by the number of DAPI-positive nuclei in the same vision field. The percentage of neuronal cells with TFEB-positive nuclei was calculated as the number of cells with TFEB immunofluorescence signal colocalized with DAPI, divided by the number of DAPI-positive nuclei in the same vision field, using ImageJ software [National Institutes of Health (NIH)].
Construction of GFP-LC3 and TFEB-GFP lentiviruses
The 1600 bp GFP-LC3 insert of pEGFP-C1-LC3 (59,60) was subcloned at the NheI–EcoRI sites of the pLESIP vector using standard recombinant DNA technology (pLESIP-GFP-LC3). pLESIP is a derivative of pWPT (61), containing a modified EF-1 α promoter and was obtained from J. S. Gutkind (NIH). To generate pLESIP-TFEB-GFP, the 2200 TFEB-GFP insert of pCMV-TFEB-GFP was subcloned into pLESIP. The preparation of high-titer lentiviral stock and infection was as previously described (5,38).
GFP-LC3 lentiviral infection and autophagy analysis
To label autophagosomes, NSCs were plated on glass-bottom culture dishes at ∼70% confluency and were differentiated into neurons as described above. Differentiated neuronal cells were infected with lentivirus encoding a GFP-LC3 fusion protein as described above. Three days after infection, cultures were examined under fluorescence microscopy for GFP expression. To determine the infection efficiency, we calculated the percentage of GFP-expressing cells relative to total cell number determined by DAPI staining. To induce autophagy, iPSC-NCs were treated with 100 nm rapamycin (Sigma–Aldrich) for 18–24 h. Cultures were then fixed in 4% paraformaldehyde and stained with anti-LAMP1 antibody, followed by fluorescence image acquisition. For biochemical studies, NSCs were differentiated in 6- or 12-well plates. The differentiated neurons were then treated with rapamycin or with the lysosomal inhibitors Chloroquine diphosphate (100 μm) (Invitrogen) or E64D (5 μg/ml) (Sigma–Aldrich) for 18–48 h as indicated, before cell lysis and protein quantitation.
Morphometric analysis of autophagosomes and lysosomes
Images for morphometric analysis were acquired using a fluorescent Nikon Ti-E inverted microscope (CFI Plan APO VC ×20 NA 0.75 WD 1 mm) magnification; the emission wavelengths were 460 nm (DAPI), 535 nm (GFP-LC3) and 620 nm (Alexa Flour 594). All images were acquired as Z-stacks and focused using an Extended Depth of Focus (EDF) module of Elements software (Nikon). Background for all images was subtracted using Elements Software. All images were quantitated using Elements; nuclei were identified based on DAPI staining using a Spot Detection algorithm. We also performed automated nuclear size exclusion for small cells (undifferentiated NSCs) to ensure comparison between homogenous cell populations. Cells expressing GFP-LC3 and positive for the lysosomal marker LAMP1 were identified using a Detect Regional Maxima algorithm, followed by global thresholding. Intracellular puncta were detected using Spot Detection and normalized to the number of GFP-LC3 cells imaged. More than 300 cells were quantitated per cell line per experiment.
TFEB-GFP lentiviral infection
iPSC-NCs were infected with lentivirus encoding a TFEB-GFP fusion protein as described above. Where indicated, culture plates were incubated with 0.24 U/ml rGCase for 5 days. For LAMP1 and TFEB-GFP fluorescence quantitation, Z-stack images were acquired with a fluorescent Nikon Ti-E inverted microscope (CFI Plan APO VC ×20 NA 0.75 WD 1 mm). Nuclei were identified based on DAPI staining using Spot Detection algorithm; cells expressing TFEB-GFP were identified using Detect Regional Maxima algorithm, followed by global thresholding, and analysis was carried out as described above. In order to obtain better resolution for lysosomal quantitation in GFP-expressing and non-GFP-expressing cells, images were acquired at high magnification using CFI Plan APO VC ×60 NA 1.4 Oil, and analyzed as described above.
Western blot analysis
Whole cell lysates of neuronal cultures were prepared using Radioimmunoprecipitation assay buffer supplemented with protease inhibitor (Roche) and phosphatase inhibitors (Pierce) followed by sonication. Protein extracts were denatured in loading buffer at 95°C for 5 min, loaded onto a 4–20% polyacrylamide gel (Bio-Rad), and the gel was transferred to a polyvinylidene difluoride membrane (Millipore). The membrane was blocked with 5% non-fat milk and probed with either rabbit anti-LC3 antibody, 1:1000 (Novus Biologicals); mouse anti-p62 lck ligand, 1:1000 (BD Bioscience); anti-LAMP1, 1:500 (University of Iowa Developmental Hybridoma Bank, H4A3); goat anti-cathepsin D (Santa Cruz), 1:500; rabbit anti-TFEB, 1:1000 (Millipore) or mouse anti-β actin (Sigma–Aldrich), followed by incubation with the corresponding horseradish peroxidase-conjugated secondary antibodies. Blots were developed using a chemiluminescence kit (Pierce) and visualized using Chemi-doc system (Bio-Rad). Bands were quantitated and analyzed using Image Lab software (Bio-Rad).
rGCase treatment
Recombinant human GCase (rGCase) (Cerezyme®, Genzyme, Cambridge, MA) was added to neuronal cultures at a concentration of 0.24 U/ml for 5 days and replenished with each media change. Cerezyme was obtained from patient infusion remnants.
GCase assay
GCase enzyme activity was assayed in neuronal cell lysates using fluorescence-conjugated substrate, 4-methylumbelliferyl β-d-glucopyranoside, as previously described (5,38). Released 4-methylumbelliferone was measured using a fluorescence plate reader (excitation 365 nm, emission 445 nm). The assay was performed in duplicate or triplicate wells with the addition of 1 mm conduritol B epoxide (CBE) to control for non-GCase enzymatic activity. Background fluorescence was subtracted in each condition. For measuring GCase activity after blocking the mannose receptor, control and GD-iPSC-NCs grown in 24-well plates were preincubated with 2 mg/ml mannan (Sigma–Aldrich). After 1 h, cells were treated with 0.24 U/ml rGCase and after another hour, an additional dose of mannan was added to the medium. Sixteen hours later, the cells were washed and GCase enzyme activity was assayed in duplicate wells in the presence or absence of CBE, as described above.
Cell survival assay
Neuronal viability was determined using PI/DAPI double-staining method (62). The indicated neuronal cultures were pretreated with 100 nm rapamycin for 72 h, followed by incubation with PI (5 μg/ml) for 10 min. Neurons were washed in PBS three times for 5 min, then fixed in 4% paraformaldehyde for 15 min and washed again. DAPI-containing mounting medium (Vectashield; Vector Laboratories) was added to culture plates and cover slips were applied. Labeled neurons were visualized by fluorescence microscopy, and images were acquired for quantitation using ×20 magnification. Experiments were done in duplicate cultures, and at least three random fields were acquired for each cell line in each treatment condition. The percentage of dead cells was calculated as the number of PI-positive nuclei divided by the number of DAPI-positive nuclei, in the same vision field. Data represent the mean values from at least three independent experiments for each cell line.
Real-time polymerase chain reaction
Control and GD iPSC-NCs in 12-well plates were cultured in neuronal differentiation media for 3 weeks in duplicate wells. Where indicated, the neuronal cultures were treated with rGCase as described above. mRNA was isolated using an RNA isolation kit (Qiagen), and cDNA was synthesized using the iScript kit (Bio-Rad). Gene expression was determined by quantitative polymerase chain reaction (PCR, 7900 HT; Applied Biosystems) in duplicate or triplicate wells using the SYBR Green method (38). The relative mRNA expression of each gene tested was normalized to the values of Glyceraldehyde 3-phosphate dehydrogenase mRNA for each reaction and then normalized to the mRNA levels for the corresponding genes. Supplementary Material, Table S2 lists all the primers that were used in this study.
Imaging
Fluorescence and phase images were captured using an inverted Nikon Eclipse TE-2000 microscope with Nikon Imaging Systems-Elements AR 3.0 collection software. High-resolution images were captured using a Zeiss LSM-510 confocal microscope (Carl Zeiss) and an AxioCam digital microscope camera at ×63 magnification. For autophagosome and lysosome analysis, Z-stack images were captured at ×20 and ×60 magnification using a fluorescent Nikon Ti-E inverted microscope and analyzed using an EDF module of Elements software (Nikon), as described above.
Statistical analysis
Results are expressed as mean ± standard error of the mean (SEM). Data were analyzed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc test to determine statistical differences between multiple groups. Two-tailed unpaired Student's t-tests were used for comparison between two groups when appropriate. P-values <0.05 were considered statistically significant. The confidence level for significance was 95%. Data were analyzed using Prism software version 4.0c (GraphPad Software).
Funding
This work was supported by grants from the Maryland Stem Cell Research Fund (MSCRF) 2009-MSCRFII-0082-00 (R.A.F.), 2007-MSCRFE-0110-00 (R.A.F.) and the March of Dimes [6-FY10-334 (R.A.F.)]. O.A. was a recipient of a MSCRF postdoctoral fellowship.
Acknowledgement
We thank Dr Andrea Ballabio (Telethon Institute of Genetics and Medicine, Italy, Baylor College of Medicine, Houston, Texas) for generously providing the TFEB-GFP plasmid used in this study.
Conflict of Interest statement. None declared.
References
