www.nature.com/npjparkd ARTICLE OPEN Lysosomal lipid alterations caused by glucocerebrosidase deficiency promote lysosomal dysfunction, chaperone- mediated-autophagy deficiency, and alpha-synuclein pathology Alba Navarro-Romero1, Irene Fernandez-Gonzalez1,6, Jordi Riera1, Marta Montpeyo 1, Merce Albert-Bayo1,7, Tresa Lopez-Royo1,8, Pablo Castillo-Sanchez1, Clara Carnicer-Caceres2, Jose Antonio Arranz-Amo3, Laura Castillo-Ribelles3, Eddie Pradas1, Josefina Casas4, Miquel Vila 1,2,5 and Marta Martinez-Vicente 1 ✉ Mutations in the GBA gene that encodes the lysosomal enzyme β-glucocerebrosidase (GCase) are a major genetic risk factor for Parkinson’s disease (PD). In this study, we generated a set of differentiated and stable human dopaminergic cell lines that express the two most prevalent GBA mutations as well as GBA knockout cell lines as a in vitro disease modeling system to study the relationship between mutant GBA and the abnormal accumulation of α-synuclein. We performed a deep analysis of the consequences triggered by the presence of mutant GBA protein and the loss of GCase activity in different cellular compartments, focusing primarily on the lysosomal compartment, and analyzed in detail the lysosomal activity, composition, and integrity. The loss of GCase activity generates extensive lysosomal dysfunction, promoting the loss of activity of other 1234567890():,; lysosomal enzymes, affecting lysosomal membrane stability, promoting intralysosomal pH changes, and favoring the intralysosomal accumulation of sphingolipids and cholesterol. These local events, occurring only at a subcellular level, lead to an impairment of autophagy pathways, particularly chaperone-mediated autophagy, the main α-synuclein degradative pathway. The findings of this study highlighted the role of lysosomal function and lipid metabolism in PD and allowed us to describe a molecular mechanism to understand how mutations in GBA can contribute to an abnormal accumulation of different α-synuclein neurotoxic species in PD pathology. npj Parkinson’s Disease (2022)8:126 ; https://doi.org/10.1038/s41531-022-00397-6 INTRODUCTION alleles in PD patients worldwide5. Both mutations result in misfolded Parkinson’s disease (PD) is an age-related neurodegenerative disorder proteins and a significant reduction in GCase activity. However, the characterized by a progressive loss of dopaminergic neurons in the mechanisms by which GBA mutations contribute to the pathogenesis substantia nigra pars compacta (SNpc), causing striatal dopamine of PD are not completely understood, and many different and non- depletion and, as a consequence, a combination of motor symptoms, exclusive hypotheses have been proposed to explain the observed including resting tremors, bradykinesia, and rigidity1. The main relationship between alterations in GCase and α-syn pathology6. The neuropathological hallmark of PD is the presence of protein loss-of-function hypothesis suggests that a decrease in GCase activity inclusions, known as Lewy bodies, composed mainly of aggregated leads to substrate accumulation, which produces changes in glyco- α-synuclein (α-syn) in surviving neurons. Several lines of evidence sphingolipid homeostasis that, through different cellular mechanisms, indicate that aberrant α-synuclein misfolding, accumulation, and can ultimately affect α-syn trafficking, processing, and clearance7. The aggregation are the major pathogenic processes in PD2. The etiology gain-of-function hypothesis suggests that misfolded mutant GBA of the disorder involves an interplay of different factors, including proteins are retained in the endoplasmic reticulum (ER), promoting ER aging, genetic susceptibility, and environmental factors. Among stress8 and α-syn accumulation9. In addition, accumulated α-syn may these factors, mutations in the GBA gene are the main genetic risk impair the trafficking of GBA protein from the ER to lysosomes, factor for developing PD3,4. GBA encodes the lysosomal enzyme resulting in a positive feedback loop that propagates disease β-glucocerebrosidase (GCase), which hydrolyses glucosylceramide pathology10. Understanding the numerous and converging under- (GlcCer) to ceramide and glucose. Biallelic mutations in the GBA gene lying mechanisms leading to α-syn accumulation in PD associated cause Gaucher’s disease (GD), a lysosomal storage disorder with GBA mutations, warrant further investigation. characterized by decreased GCase activity and subsequent accumula- tion of GlcCer and glucosylsphingosine (GlcSph) in several organs. Patients with GD and mutant GBA carriers have a fivefold greater RESULTS lifetime risk of developing PD than that in the general population4. Generation of an in vitro model of PD associated with GBA L444P (c.1448T>C; p. L483P) and N370S (c.1226A>G; p. N409S) are the In the present study, we aimed to elucidate whether the lack of most common GBA mutations and account for 70% of all GBA mutant GBA or the presence of mutant GBA triggers lysosomal dysfunction 1 Neurodegenerative Diseases Research Group, Vall d’Hebron Research Institute (VHIR)-Center for Networked Biomedical Research on Neurodegenerative Diseases (CIBERNED), Barcelona, Spain. 2Catalan Institution for Research and Advanced Studies (ICREA), Barcelona, Spain. 3Biochemistry Service, Vall d’Hebron Hospital, Barcelona, Spain. 4Institut de Química Avançada de Catalunya (IQAC-CSIC), Barcelona, Spain. 5Department of Biochemistry and Molecular Biology, Autonomous University of Barcelona (UAB), Barcelona, Spain. 6Present address: Institut de Neurociènces and Facultad de Medicina, Autonomous University of Barcelona, Barcelona, Spain. 7Present address: Laboratory of Neuro-Immuno-Gastroenterology, Digestive System Research Unit, Vall d’Hebron Institut de Recerca, Barcelona, Spain. 8Present address: Department of Anatomy, Embryology and Animal Genetics, University of Zaragoza and CIBERNED, Zaragoza, Spain. ✉email: marta.martinez@vhir.org Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 2 a b GCase activity c HexCer HexSph **** *** 1.25 1600 800 **** *** N370S L444P **** GCase activity (A.U.) 1400 ** WT KO 1.00 600 1200 pmol/mg protein pmol/mg protein 0.75 1000 GBA 400 800 20 55 0.50 * 600 15 E-actin 55 * ** * 400 10 * 0.25 * ** * 200 5 * ** * 0.00 0 0 WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P d e HexCer HexSph f Viability GCase activity 1 .5 **** **** 8000 **** 3500 **** **** 100 GCase activity (A.U.) 2500 **** % viable cells 1 .0 6000 90 **** pmol/mg protein 1500 pmol/mg protein 500 * 4000 30 80 * * **** 0 .5 * ** * 20 **** 70 2000 * ** * 10 ** 60 * ** * 0 1234567890():,; 0 .0 0 WT KO N370S L444P WT KO N370S L444P 50 WT KO N370S L444P WT KO N370S L444P Lysosome Lysosome Lysosome Fig. 1 Loss of lysosomal GCase activity leads to substrate accumulation in GBA mutant cells. a Immunodetection of GBA protein; b GCase activity in total homogenate corrected by levels GBA protein, n = 3 in each group. c Quantification of total HexCer (left) and total HexSph (right) by LC/MS-MS, n = 3 in each group. d Lysosomal GCase activity in isolated lysosomal fractions, n = 4 independent isolations in each group. e Quantification of HexCer (left) and HexSph (right) by LC/MS-MS, n = 3–4 in each group. f Cellular viability, n = 6 independent samples. In all panels, data is presented as mean ± s.e.m. Statistical significance was established at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 compared to WT samples or between mutant lines when indicated after one-way ANOVA followed by Tukey’s multiple comparisons test. and affects ER stress, synuclein accumulation, and cellular viability. (Fig. 1d) by analyzing pure lysosomal fractions isolated from all To this end, we used differentiated human dopaminergic-like cell lines and validated using lysosomal markers (Supplementary neuroblastoma BE(2)-M17 cell lines with GBA knockout and two of Fig. S1b). The effect on sphingolipid accumulation was intensi- the most clinically relevant genetic variants associated with PD. fied in the lysosomal compartment, the enrichment of sphingo- First, the GBA knockout (KO) cell line was generated through the lipids was higher in this subcellular fraction and more significant CRISPR/Cas9 system, then through stable transfection, we increase in the levels of HexSph substrate was observed in all introduced human GBA constructs harboring the two most mutant differentiated cell lines (Fig. 1e). The decrease in GCase prevalent mutations, N370S and L444P, into this KO cell line. activity and increase in substrate accumulation is associated Expression of GBA was confirmed using western blot analysis, and with a decrease in the cellular viability of all three mutant cell after different clones were fully characterized, a representative lines (Fig. 1f). clone of each cell line was selected (Fig. 1a). All four cell lines (GBA WT, GBA KO, GBA N370S, and GBA L444P) were characterized Retention of mutant GBA proteins in the ER under differentiation conditions to study the consequences of loss Lysosomal enzymes follow a specific pathway to reach lysosomes; of function and gain of toxic function of GBA. As expected, the from the ER, where they are synthesized, they travel through the GBA KO line presented minimal activity, while GBA N370S line Golgi apparatus, where they undergo processing, and then reach presented 7.66% activity and GBA L444P line 15.85 % activity their final destination in the lysosomes. Several studies have (related to WT GCase activity as 100%) (Fig. 1b). The conditions indicated that mutant forms of GBA, including N370S and L444P, used to assess GCase activity exclusively ensure the detection of are retained in the ER8. the activity of GBA protein (as the product of GBA1 gene) but not We monitored the intracellular fate of GCase by assessing the cytosolic GBA2 protein11 (Supplementary Fig. S1a). The decrease processing of its N-glycans. Cell lysates were digested by in GCase activity resulted in an increase in the levels of GCase Endoglycosidase-H (Endo-H), which specifically digests high substrates determined using LC/MS-MS as monohexosylceramide mannose N-glycans added at the ER but not complex N-glycans (HexCer) and hexosylsphingosine (HexSph) species. The differ- present in proteins that have passed the mid-Golgi12. GBA L444P entiated GBA KO cell line showed a dramatic accumulation of total and GBA N370S cells were more sensitive than GBA WT cells to HexCer and HexSph. In contrast, GBA N370S and GBA L444P Endo-H restriction, showing a higher percentage of GBA EndoH- presented similar levels of total sphingolipids compare to WT, with sensitive fraction, indicating that more GBA protein is retained in the exception of GBA L444P cells that presented a discreet the ER in cell lines expressing mutant GBA proteins (Fig. 2a). No increase only in HexSph levels (5.5-fold) (Fig. 1c). Overall, total differences in the digestion were observed between groups after sphingolipids quantification suggests that the partial GCase cells were treated with PNGase F glycosidase, which digests all activity in both GBA mutant cell lines is enough to avoid the sugar types. To confirm these results, we assessed the colocaliza- massive accumulation observed in GBA KO cells. tion of GBA with calnexin, an ER marker, and LAMP-1, a lysosomal To further analyze the effect of GCase activity loss, we marker, using immunofluorescence. A lower fraction of mutant analyzed the effect of decreased lysosomal GCase activity GBA than WT GBA colocalized with LAMP-1, whereas a higher npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 3 a WT N370S L444P 80 EndoH-sensitive GBA EndoH: - + - - + - - + - **** (% of total GBA) PNGase F: - - + - - + - - + 60 **** EndoH-R (post-ER) 40 GBA EndoH-S (ER) 20 0 WT N370S L444P b GBA LAMP-1 merge GBA calnexin merge 1 .0 Pearson’s coef. 0 .8 GBA/LAMP-1 0 .6 **** WT **** 0 .4 0 .2 0 .0 N370S WT N370S L444P 1 .0 0 .8 calnexin/GCase Pearson’s coef. 0 .6 **** * L444P 0 .4 0 .2 0 .0 WT N370S L444P c IRE1D ATF6 * ** 2 .5 * 1 .5 ** ATF6(90kDa)/actin * WT KO N370S L444P * IRE 1D/actin 2 .0 1 .0 IRE1D 1 .5 * * 100 1 .0 100 0 .5 0 .5 ATF6 0 .0 0 .0 35 WT KO N370S L444P WT KO N370S L444P CHOP 25 Phospho-eIF2D CHOP 55 * ** * * ** E-actin 2 .5 * ** * 8 * ** P-eIF 2D/ eIF 2D * ** * ** CHOP/actin 2 .0 * ** * 6 eiF2D 35 1 .5 4 1 .0 P-eiF2D 35 2 0 .5 0 .0 0 WT KO N370S L444P WT KO N370S L444P Fig. 2 Mutant GBA proteins are retained in the ER. a Left: representative immunoblots of GBA before and after treatment with EndoH and PNGase F (as a positive control). b Right: quantification of GBA EndoH-sensitive (EndoH-S) percentage of total GBA protein, n = 8 independent experiments. Representative images and quantification of the colocalization of GBA with LAMP-1 (lysosomes) and calnexin (ER) determined by immunofluorescence. Colocalization was represented as Pearson’s coefficient (fraction of LAMP-1 or calnexin overlapping GBA), scale bar = 5 µM, n > 20 cells/condition. c Representative immunoblots and quantification of ER stress markers (IRE1α, ATF6, phosphorylated eIF2a, and CHOP), n = 3 independent experiments. In all panels data are presented as mean ± s.e.m, statistical significance was established at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 compared to WT samples or between mutant lines when indicated after one-way ANOVA followed by Tukey’s multiple comparisons test. Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126 A. Navarro-Romero et al. 4 fraction colocalized with calnexin (Fig. 2b). Together, these results To further investigate lysosomal dysfunction, we analyzed the suggest that trafficking of GBA L444P and GBA N370S from the ER enzymatic activity of a series of representative lysosomal enzymes, to lysosomes is impaired. Accumulation of mutant GCase in the ER including cathepsins D and B, β-hexosaminidase, acid phospha- can lead to ER stress and to the activation of the unfolded protein tase, N-acylsphingosine amidohydrolase (ASAH1), and acid sphin- response (UPR)13. We analyzed the levels of different ER stress gomyelinase (ASM). Overall, no remarkable changes were markers in differentiated mutant GBA cells; mutant N370S and observed in the activity of these lysosomal enzymes in total L444P cells but not GBA KO cells showed activation of the three lysates from the different cell lines, with some exceptions; ASM branches of the UPR system14, as indicated by upregulation of and cathepsin B activities were increased only in GBA KO cells IRE1α expression, cleavage of ATF6, and PERK-dependent (Fig. 4c). However, when we focused on the lysosomal compart- hyperphosphorylation of eIF2α, ultimately leading to overexpres- ment and analyzed enzymatic activity in the lysosome-enriched sion of the ER chaperone CHOP (Fig. 2c). fraction instead of the total homogenate, we observed a significant decrease in the activity of all lysosomal enzymes in Increase in α-synuclein species in GBA mutant cell lines all three mutant cell lines except cathepsin B (Fig. 4d), demonstrating the importance of analyzing the enzymatic activity Several studies on PD models and patient samples have revealed of lysosomal enzymes in isolated lysosomes. This decrease was an association between GCase activity loss and α-syn accumula- also confirmed by studying the live-cell enzymatic activity of tion6,10. We measured different intracellular α-syn species using cathepsin D in vivo using BODIPY-FL-pepstatin A and quantifying immunodetection; in GBA KO, L444P, and N370S cells, total and the levels of mature cathepsin D and B present in the lysosomal phosphorylated α-syn (Ser129) levels were increased in compar- fraction (Supplementary Fig. S2), corroborating that GCase activity ison with those in WT cells (Fig. 3a). In addition, other neurotoxic loss is associated with a decrease in the enzymatic activity of forms of α-syn, such as oligomeric α-syn species, were quantified several other representative lysosomal enzymes. using the specific high molecular weight β-sheet oligomer A11 We also analyzed other lysosomal features such as lysosomal pH and 5G4 antibodies. All mutant cell lines displayed increased α- and membrane permeabilization that can be altered because of syn oligomer signals compared with those of WT cells, confirm- the accumulative events triggered by the loss of lysosomal GCase. ing that loss of GCase protein and all mutant forms of GCase We analyzed the lysosomal pH in all cell lines using LysoSensor™ promote a clear increase in total α-syn levels but, more Yellow/Blue DND-160 and observed that all mutant cell lines importantly, a drastic increase in the levels of neurotoxic forms showed an increased lysosomal pH, which can contribute to of α-syn (Fig. 3b). deficient maturation of lysosomal enzymes and reduced degrada- To further study the presence of α-syn oligomers, we analyzed tion rates (Fig. 4e). In addition, the integrity of the lysosomal the levels of these α-syn species in cell lysates using the membrane was monitored by quantifying the protein level of chemical cross-linker disuccinimidyl glutarate (DSG), which galectin-3, a marker of lysosomal membrane damage19. Using forms irreversible amide bonds between interacting proteins. immunofluorescence analysis performed in differentiated KO and Immunoblotting revealed that the levels of α-syn oligomeric mutant GBA, a clear increase in the number of galectin-3 puncta species were significantly increased in GBA KO and mutant cells patterns by neurons was observed (Fig. 4f). (Fig. 3c). We also analyzed the levels of insoluble fibrillary α- synuclein aggregates using a filter retardation assay15. Results indicated that mutant cells presented increased levels of GBA mutant cell lines present a slight effect on insoluble α-synuclein aggregates retained in the acetate macroautophagy cellulose filter (Fig. 3d). As a consequence of the lysosomal dysfunction, the autophagy Analysis of the culture media of differentiated BE(2)-M17 cells process might be affected; accordingly, we analyzed both revealed increased levels of extracellular monomeric and β-sheet- macroautophagy and CMA. Macroautophagy activity has been rich oligomeric forms of α-syn in GBA KO, N370S, and L444P cells studied in several GBA models with different and contradictory (Fig. 3e). results20–22. The analysis of autophagic flux showed that in WT cells the addition of the lysosomal inhibitor can significantly Lysosomal dysfunction in GBA mutant cell lines increase LC3-II levels, however in the mutant cells including KO, although there is a slight increase of LC3-II after the inhibition, this Lysosomal dysfunction has been described as one of the main increase is not significant, indicating that there is a deficiency in mechanisms involved in PD etiology16, mainly owing to the autophagy flux (Fig. 5a). This partial impairment of autophagic flux consequences of deficits in lysosomal-dependent α-syn turnover under basal conditions was also assessed by quantifying the levels that leads to α-syn accumulation. Soluble α-syn is predominantly of p62/SQSTM1, revealing that the degradation of autophago- degraded by lysosomal pathways, particularly CMA; however, somes within lysosomes was partially compromised. We further oligomeric, aggregated, and some modified α-syn species cannot analyzed the levels of some key macroautophagy markers in all be eliminated by this pathway and are degraded by macro- mutant differentiated cell lines (Fig. 5b). Overall, we did not autophagy17. Failure in lysosomal activity can affect both observe significant changes in the levels of these markers except macroautophagy and CMA pathways and promote α-syn accu- for phosphorylated ULK1 expression, which was significantly mulation, which can trigger pathogenic α-syn aggregation. To increased in GBA mutant cells, suggesting a possible compensa- further investigate the connection between GBA and α-syn, we tory promotion of macroautophagy at the initiation step (Fig. 5b). analyzed lysosomal activity and autophagy pathways in detail A tendency for beclin 1 levels to be increased and phosphorylated using different methodologies. mTOR levels to be decreased could support this hypothesis. No First, we performed a live-cell proteolysis assay to quantify total significant changes were observed in the other macroautophagy proteolysis (Fig. 4a) and lysosomal-dependent proteolysis (Fig. 4b) markers confirming that reasonably regular macroautophagy by measuring the degradation rate of long-lived proteins18. No activity is maintained under basal conditions. differences were observed in total proteolysis, indicating that the lack of GBA or the presence of mutant forms of GBA under basal conditions does not directly affect the efficiency of total cellular GBA deficiency leads to lysosomal lipid modifications that degradation. However, when analyzing only lysosomal degrada- affect CMA activity tion, which was measured as proteolysis sensitive to ammonium We next questioned whether lysosomal dysfunction could affect chloride and leupeptin, a clear decrease in the lysosomal- CMA, especially considering the important role of this autophagic dependent proteolysis rate was observed. pathway in α-syn degradation and PD pathophysiology and that npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 5 a N370S L444P Total syn Phosphorylated-syn 5 WT KO syn (T) levels/actin (A.U.) 8 ** ** P-syn levels/actin (A.U.) GBA 4 * **** 6 55 **** *** 3 syn 15 4 2 2 1 p-syn (Ser129) 15 0 0 WT KO N370S L444P WT KO N370S L444P Total syn Phosphorylated-syn * b WT KO N370S L444P 15 * ** * 40 * ** * ** * MIF/cell (A.U.) MIF/cell (A.U.) ** 30 * ** * syn 10 * ** * 20 5 o-syn (A11) p-syn (S129) 10 0 0 WT KO N370S L444P WT KO N370S L444P o-syn (A11) o-syn (5G4) * 20 * 30 **** ** * ** * * * ** * 15 MIF/cell (A.U.) MIF/cell (A.U.) * ** * 20 * ** * ** * * ** * 10 o-syn (5G4) 10 5 0 0 WT KO N370S L444P WT KO N370S L444P c d e N370S L444P N370S L444P WT KO WT KO WT KO N370S L444P 100 oligo-syn 75 synuclein 50 oligo syn 37 actin 25 CD81 20 30 * * mono ponceau 15 * agg-syn 20 8 actin * Extracellular syn 10 6 * 1 00 monomer oligomer * ** * ns 80 4 olig-syn * ** * 60 0 * ** 2 WT KO N370S L444P 40 0 20 WT KO N370S L444P 0 WT KO N370S L444P lower levels of LAMP-2A and CMA activity have been reported in total LAMP-2A level was not changed in mutant or KO GBA cells. PD models and samples from PD patients23–25. We first quantified However, when analyzing the LAMP-2A levels exclusively in the the levels of the LAMP-2A isoform (Fig. 6a) since CMA activity lysosomal fraction, LAMP-2A levels were drastically decreased depends on the lysosomal receptor complex formed by the when GBA was knocked down or mutated. As expected, this oligomerization of the lysosomal protein LAMP-2A26. Interestingly, decrease in lysosomal LAMP-2A expression (Fig. 6b) correlated Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126 A. Navarro-Romero et al. 6 Fig. 3 Mutant GBA accumulates different α-syn species. a Representative immunoblots and quantification of total synuclein and phosphorylated synuclein (Ser129). b Representative confocal images and quantification of the mean fluorescence intensity (MFI) of total α-syn, phosphorylated (Ser129) α-syn, and oligomeric α-syn (A11 and 5G4 antibodies) determined by immunofluorescence, scale bar = 10 µM, n > 20 cells/condition. c Immunodetection and quantification of oligomeric α-syn by in vivo cross-linking with DSG. d Insoluble α-syn detection and quantification using the filter retardation assay in acetate cellulose membrane after in vivo cross-linking treatment. e Extracellular α-syn detected by dot blotting from an extracellular culture medium with anti-synuclein and anti-oligomeric synuclein (A11); CD81 (extracellular vesicle surface protein) was used as a marker of extracellular vesicle present in the extracellular fraction. C–E: results of at least three independent experiments. In all panels, data is presented as mean ± s.e.m, statistical significance was established at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 compared to WT samples or between mutant lines when indicated after one-way ANOVA followed by Tukey’s multiple comparisons test. with a decrease in CMA activity quantified by the proteolysis neurotoxic forms of α-syn associated with PD, including oligomeric, activity assay as ammonium chloride and leupeptin-sensitive and insoluble, phosphorylated and extracellular species (Fig. 3a–e); the 3-MA-insensitive proteolysis27 (Fig. 6c). aim of this work was to identify underlying pathways leading to α- The stability and activity of LAMP-2A are determined in part by syn accumulation in our GBA in vitro model. the lipid microdomains located in the lysosomal membrane that As expected, the GBA KO cell line accumulates GlcCer. However, are mostly enriched in sphingolipids and cholesterol. We first it more dramatically accumulates GlcSph, the main sphingolipid analyzed the transcriptional expression of the LAMP-2A isoform in accumulated in GBA deficient cells and used as Gaucher’s the different cell lines (Supplementary Fig. S3) showing no diagnosis biomarker33, while N370S and L444P cell lines maintain significant changes in the expression of LAMP-2A and suggesting a partial GCase activity that avoids this massive sphingolipid that decreases in LAMP-2A might be a consequence of its accumulation. Previous studies analyzing sphingolipid levels in increased turnover inside the lipid microdomains28,29. In addition GBA-related models showed variable results, with a clear to the high levels of GlcSph previously observed in lysosomal accumulation of sphingolipids in homozygous GBA mutant fractions (Fig. 1e), we also analyzed cholesterol levels and models34–38 and not so clear accumulation in heterozygous GBA observed an increase in mutant GBA cell lines in the lysosomal- mutant models38,39. In human studies, some authors reported enriched fraction (Fig. 6d). Similarly, when the sphingolipid GlcSph elevated levels of both substrates in the brains of neuropathic GD was added to WT GBA cell lines, we observed decreased LAMP-2A patients40,41 and significantly increased levels in the substantia levels (Fig. 6e) and increased α-syn levels associated with nigra of sporadic PD brains42,43. In contrast, other studies reported decreased CMA activity (Fig. 6f). no changes in GlcCer and GlcSph levels in different brain areas of To confirm that the loss of function of GBA was responsible, at PD patients with GBA variants44–46. This lack of a clear data on least in part, for the increase in α-syn levels through alterations in accumulation of GBA substrates in some PD-GBA models and lysosomal LAMP-2A levels and CMA activity, we attempted to samples from PD patients carrying heterozygous GBA mutations restore LAMP-2A levels in mutant cell lines through the expression might suggest that residual GCase activity may be enough to of the LAMP-2A isoform by transient transfection. Interestingly, the avoid the massive accumulation of GlcCer and GlcSph and raises attempt to overexpress LAMP-2A in the GBA WT cell line was concerns about the role of these sphingolipids in the pathogen- associated with a significant decrease in P-syn and total α-syn esis of PD. However, this apparent discrepancy might be explained levels (Fig. 6g). However, in the KO and mutant cell lines, by the types of samples analyzed in previous studies. Based on the transfection of LAMP-2A protein was not efficient, probably results of our study, many of the lysosomal-related deficits because high sphingolipid and cholesterol levels in these cell revealed in our model were only noticeable when lysosomal- lines prevent the stabilization of LAMP-2A levels in the lysosomal enriched samples instead of total homogenates were used. Thus, membrane and proper assembly of the CMA-translocation the quantification of lipids in the lysosomal fraction revealed complex. Accordingly, P-syn and total α-syn levels were not significantly increased levels of GlcSph and cholesterol (Figs. 1e significantly decreased in these cell lines (Fig. 6g). and 6d) that were not detectable in the total homogenate of the Alternatively, pharmacological activation of CMA without same samples (Figs. 1c and 6d), suggesting that accumulation of directly targeting LAMP-2A protein with AR7 and humanin sphingolipids, cholesterol, and other related lipids can occur and treatment30,31 was more efficient in promoting phosphorylated are detectable only locally. Additionally, as an aggravating effect, synuclein clearance in WT, N370S, and L444P cell lines, but not in these accumulated lipids can also be transported to other KO cells where accumulated P-syn was not significantly decreased membranes in other subcellular localizations such as mitochon- (Fig. 6h). dria22 and the ER47. Moreover, since they are bioactive lipids and components of several signaling networks, they can trigger other pathogenic effects. Thus, in addition to the consequences of DISCUSSION lysosomal accumulation, we cannot ignore the fact that increased In recent years, numerous non-exclusive mechanisms have been levels of GlcCer, GlcSph, and other lipids such as gangliosides and proposed to participate in the link between GBA and PD (reviewed GlcChol produced because of GCase impairment might alter in10). In our studies, we used differentiated BE(2)-M17 immorta- cellular homeostasis and facilitate neuronal cell death through lized cell lines of human neuronal origin, presenting a well- different mechanisms beyond the lysosome48–50. characterized dopaminergic phenotype, neurite extension, and The gain-of-toxic-function hypothesis mainly focuses on the arrest of cellular proliferation32. These cells were generated and consequences of the accumulation of mutant GBA proteins in the analyzed in parallel; thus, GBA KO was used as a model of ER. It has been extensively reported that in fibroblasts and iPSC- complete loss-of-function and maximal accumulation of sub- derived dopaminergic neurons from PD patients carrying GBA strates, and mutant GBA cell lines (L444P and N370S) as models of mutations, mutant GBA is retained in the ER8. In our neuronal gain-of-toxic function owing to the presence of aberrant mutant model, we clearly confirmed the retention of mutant GCase in the proteins as well as partial loss of enzymatic activity (Fig. 1). ER (Fig. 2a, b) and the subsequent activation of the UPR (Fig. 2c) in The link between GCase activity and α-syn may explain the differentiated L444P and N370S mutant cell lines. We hypothesize relationship between GBA mutations and PD. We could observe that that the chronic ER stress generated by the sustained accumula- the loss of GCase activity and the presence of mutant forms of the tion of mutant GBA proteins in the ER probably triggers other GBA protein was clearly associated with an increase in different interconnected pathways with synergistic effects, including npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 7 a Total proteolysis b Lysosomal proteolysis c 50 ASAH1 ASM 60 * p29% p21% p 36% % proteolysis 6 20 % proteolysis 40 * * ** 5 * * ** activity (A.U.) 40 15 activity (A.U.) 30 4 20 3 10 20 2 10 5 1 0 0 0 WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P Cathepsin D Cathepsin B E-hexo Ac. Phosphatase 1.5 4.0 0 .8 2 .0 activity (A.U.) activity (A.U.) activity (A.U.) * 3.0 0 .6 activity (A.U.) 1.0 1 .5 2.0 1 .0 0 .4 0.5 0 .5 1.1 0 .2 0.0 0 .0 0.0 0 .0 WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P d Lys Cathepsin B Lys E-hexo Lys Ac. Phosphatase Lys Cathepsin D 1 .5 3 .5 1 .5 1 .5 activity (A.U.) activity (A.U.) activity (A.U.) activity (A.U.) 1 .0 3 .0 1.0 1 .0 * **** **** ** * ** * 0 .5 * ** 2 .5 0.5 0 .5 **** 0 .0 2 .0 0.0 0 .0 WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P Lysosome Lysosome Lysosome Lysosome e f Lysosomal pH WT KO Gal3 LMP (Gal3) 7.0 30 **** **** 6.0 ** ** **** ** Gal3 MFI (A.U.) 20 pH 5.0 N370S L444P 4.0 10 3.0 WT KO N370S L444P 0 WT KO N370S L444P Fig. 4 Mutant GBA exhibit lysosomal dysfunction. a Total proteolysis measured by an intracellular protein assay, n = 12. b Lysosomal proteolysis measured by an intracellular protein assay, n = 12. c Enzymatic activity of lysosomal enzymes in total homogenates, n = 3–6 independent samples. d Enzymatic activity of lysosomal enzymes in lysosomal-enriched fractions, n > 3 independent isolations in each group. e Lysosomal pH quantification, n = 6 in each group. f Representative images and quantification of galectin-3 (red) determined by immunofluorescence. Nuclei are labeled with Hoechst 33342, scale bar = 10 µm, n = 30 cells/conditions. In all panels, data are presented as mean ± s.e.m. Statistical significance was established at *p < 0.05, ** p < 0.01, ***p < 0.001, ****p < 0.0001 compared to WT samples or between mutant lines when indicated after one-way ANOVA followed by Tukey’s multiple comparisons test. autophagy alterations, oxidative stress, and mitochondrial dys- be eliminated by CMA; however, mutant GBA is retained in the function51. In addition, a new gain-of-toxic-function mechanism lysosomal membrane and blocks the multimerization of LAMP-2A, beyond the ER has been recently reported, where a fraction of impairing the normal uptake and degradation of other CMA mutant GBA that fails to fold in the ER is targeted to lysosomes to substrates such as alpha-synuclein52. Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126 A. Navarro-Romero et al. 8 a WT KO N370S L444P UT LI UT LI UT LI UT LI LC3-I 15 LC3-II 70 p62 55 E-actin 35 3 * LC3-II P= 0.058 p62 n.s n.s 2.0 * n.s n.s LC3-II/actin (A.U.) p62/actin (A.U.) 2 1.5 n.s 1.0 1 0.5 0 0.0 LI: - + - + - + - + LI: - + - + - + - + WT KO N370S L444P WT KO N370S L444P b WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P 70 LC3-I Beclin 1 LAMP-2 mTOR 15 55 100 LC3-II 150 Ponc. 70 E-actin Ponc. p62 55 130 Ponc. p-ULK1 100 p-mTOR LAMP-1 150 70 Ponc. Ponc. Atg5-12 50 Ponc. Ponc. LC3-II P-mTOR/total beclin 1 2.0 LAMP-2 1.5 2.0 2 .5 P-mTOR/Total (A.U.) LC3-II levels (AU) LAMP- 2 leve ls (AU ) beclin 1 levels (A.U.) 1.5 2 .0 1.5 1.0 1 .5 1.0 1.0 1 .0 0.5 0.5 0 .5 0.5 0 .0 0.0 0.0 0.0 WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P LAMP-1 p62 P-ULK1 1.5 Atg5-12 * ** 1.5 2 .5 1 .5 Atg5-12 levels (A.U.) P-ULK1 levels (AU) LAMP1 leve ls (AU ) 2 .0 p62 levels (A.U.) 1.0 1.0 1 .0 1 .5 1 .0 0.5 0 .5 0.5 0 .5 0.0 0 .0 0.0 0 .0 WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P WT KO N370S L444P Fig. 5 Mutant GBA causes alterations in macroautophagy. a Autophagic flux: immunodetection and quantification of LC3-II and p62 in the presence and absence of the lysosomal inhibitor (LI) 60 µM chloroquine. Results are presented as mean ± s.e.m. values of three independent experiments. Statistical significance was established at *p < 0.05 compared to untreated conditions by two-tailed unpaired Student’s t-test. b Representative immunoblots and quantification of autophagic and lysosomal markers. In all panels, results are presented as mean ± s.e.m. values of at least six independent experiments. Statistical significance was established at *p < 0.05, **p < 0.01, compared to WT samples after one-way ANOVA followed by Tukey’s multiple comparisons test. npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 9 a b c Homo Lys Homo Lys Lysosomal LAMP-2A 40 CMA proteolysis 1.5 LAMP-2A/2 total levels GBA +/+ GBA +/+ GBA +/- GBA +/- GBA -/- GBA -/- N370S L444P N370S L444P 30 WT WT 1.0 % Proteolysis * ** ** 20 GBA 55 *** 0.5 10 **** LAMP-2A 100 0.0 0 LAMP-2 100 WT KO N370S L444P WT KO N370S L444P d 1 .0 Cholesterol e f ** ctr GlcSph Cholesterol (AU/ug) CMA proteolysis 0 .8 LAMP-2A synuclein LAMP-2A n.s. * * 100 1.5 1.5 40 L2A levels (A.U.) * Syn levels (A.U.) % proteolysis CMA 0 .6 actin 30 37 1.0 * 1.0 0 .4 15 20 syn (T) 0.5 ** 0.5 10 0 .2 actin 0.0 0.0 0 37 ctr GlcSph ctr GlcSph 0 .0 ctr GlcSph WT KO N370S L444P WT KO N370S L444P Homo Lysosomes g 2 .5 ** LAMP-2A ctr 2 .0 syn levels (+L2A) s yn T WT KO N370S L444P LAMP2A levels (A.U.) n.s. +L2A P -s yn n.s. protein levels (A.U.) + LAMP2A: - + - + - + - + 2 .0 1 .5 LAMP-2A 1 .5 n.s. 100 P=0,0827 1 .0 * syn (T) 15 1 .0 ** 0 .5 0 .5 P-syn 15 0 .0 - + - + - + - + 0 .0 WT KO N370S L444P WT KO N370S L444P h ctr +AR7 +Hu P-syn 15 P-syn (+AR7) P-syn (+Hu) WT ctr ctr ponc. 1.5 1.5 +AR7 P-syn levels (A.U.) *** n.s. ** ** n.s. * +Hu * *** P-syn levels (A.U.) P-syn 15 1.0 KO 1.0 ponc. 0.5 0.5 P-syn 15 N370S 0.0 0.0 ponc. - + - + - + - + - + - + - + - + WT KO N370S L444P WT KO N370S L444P 15 P-syn L444P ponc. Fig. 6 Mutant GBA is associated with CMA impairment. a Immunodetection of LAMP-2A, total LAMP-2, and GBA in total homogenate (homo) and isolated lysosomes (Lys). b Quantification of LAMP-2A levels in the lysosomal fraction corrected using total LAMP-2, n = 4 independent isolations. c Proteolysis dependent on CMA detected as 3-MA-insensitive/leupeptin-NH4Cl-sensitive proteolytic activity, n > 4 in each group. d Total cholesterol quantification in homogenate (homo) and isolated lysosomes (Lysosomes) n = 6. e Quantification of total LAMP-2A and total synuclein levels in GBA WT cell lines after 72 h treatment with 5 mM GlcSph, n = 3. f Proteolysis dependent on CMA detected as 3-MA-insensitive/leupeptin/NH4Cl-sensitive proteolytic activity after 72 h treatment with 5 mM GlcSph, n = 3. g Representative images (left) of immunodetection of LAMP-2A and total/phosphorylated synuclein after 48 h of transient transfection. Quantification (right) of LAMP-2A transfection (LAMP-2A levels in control transfection in each cell line normalized to 1) and total and phosphorylated α-syn after transfection (syn protein levels in control transfection in each cell line normalized to 1), values of three independent experiments. h Immunodetection of phosphorylated synuclein (P-syn) and ponceau (ponc) as a loading control after 24 h treatment with AR7 and humanin (Hu) (left), and quantification of P-syn levels (left) [P-syn levels in cells treated with vehicle (ctr) in each cell line are normalized to 1]; values of six independent samples. In all panels, results are presented as mean ± s.e.m values. Statistical significance was established at * p < 0.05, **p < 0.01 after two-way ANOVA followed by Tukey’s multiple comparisons tests (b–d) or two-tailed unpaired Student’s t-test (e–h). Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126 A. Navarro-Romero et al. 10 After validating the in vitro model, the main aim of this work was studies showed that a decrease in LAMP-2A receptor expression or to analyze in detail the role of lysosomal-related mechanisms in PD- blockage of LAMP-2A receptor translocation to the lysosomal associated GBA. Lysosomal dysfunction has long been described as membrane leads to accumulation of α-syn65–68. In addition, LAMP- one of the main mechanisms involved in PD etiology16, mainly 2A levels and translocation complex assembly depend on lipid because of the impairment of α-syn turnover. In both non-neuronal microdomains on the lysosomal membrane. Higher levels of cell lines such as patient-derived fibroblasts21,53 and neuronal cell sphingolipids, gangliosides, and cholesterol, the main components lines such as dopaminergic neurons derived from iPSCs and SH- of lipid microdomains, favor an increase in the number of lipid SY5Y cells20,21,54–57, different studies have reported varied and microdomains present in the lysosomal membrane and conse- contradictory alterations in the activity of specific lysosomal quently the presence of LAMP-2A as a monomer, preventing the enzymes in total lysates33. In our study, we measured the activity assembly of the CMA-translocation complex and inhibiting CMA of different representative lysosomal enzymes in the lysosomal activity69. We confirmed that the low levels of LAMP-2A present in fraction as a readout of lysosomal function (Fig. 4d and the lysosomal membrane might be due to post-translational Supplementary Fig. S2). We detected a clear impairment of mechanisms (Supplementary Fig. S3) and the levels of GlcSph, individual enzymatic activity that was observable only in the GlcCer (Fig. 1e, f), and cholesterol (Fig. 6d) were increased in the lysosomal fraction but not in the total homogenate (Fig. 4c). The lysosomal fraction, which could explain the higher LAMP-2A discrepancies among previous studies might be explained by the turnover and consequently the decrease in CMA activity. fact that measuring activities in total homogenates does not allow The correlation between the decrease in LAMP-2A levels and α- for discriminating the enzymatic activity of the enzymes present syn turnover caused by the accumulation of GBA substrates was within the lysosome from that of those present in other subcellular confirmed by the addition of high levels of GlcSph to the GBA WT compartments such as the ER or Golgi, leading to an overestimation cell lines, which promoted a decrease in LAMP-2A expression, CMA of the enzymatic activity of lysosomal enzymes in total homo- impairment, and α-syn accumulation (Fig. 6e, f). To revert the effect genates. Accordingly, although total proteolysis did not seem to be of CMA impairment, genetic overexpression of the LAMP-2A isoform affected in our differentiated neuronal model (Fig. 4a), the and the subsequent clearance of synuclein isoforms were only quantification of lysosomal-dependent proteolysis showed a effective in GBA WT cell lines (Fig. 6g). In the cell lines presenting significant decrease under basal conditions in all mutant cells lysosomal sphingolipid and cholesterol accumulation, the genetic relative to that in WT cells (Fig. 4b). overexpression of the LAMP-2A isoform was probably unable to Changes in the acidification of the lysosome can affect maintain stable levels of the LAMP-2A complex due to continuous intralysosomal processing and the efficiency of lysosomal LAMP-2A degradation within the lysosomal lipid rafts (Fig. 6g). enzymes58. Moreover, changes in the intralysosomal pH have Pharmacological activation of CMA without directly targeting been associated with different pathologies, including PD59. In the LAMP-2A protein with AR7 and Hu (Fig. 6h) was more efficient in mutant cell lines, we observed an increased intralysosomal pH and promoting phosphorylated synuclein clearance in WT, N370S, and lysosomal membrane permeabilization (Fig. 4e, f) that illustrate L444P cell lines but not in KO cells which express higher sphingolipid the consequences of GCase activity loss on lysosomal function. and cholesterol levels in the lysosomal fraction. Overall, these This lysosomal dysfunction could reciprocally affect the experiments confirmed that abnormally high levels of sphingolipids different forms of autophagy. Some preliminary macroautophagy and cholesterol in the lysosomal membrane of GCase-deficient studies have reported increased levels of the autophagosome neurons could affect CMA activity and consequently the degradation marker LC3-II and the macroautophagy substrate p62/SQSTM1 of the CMA substrate α-syn. We cannot dismiss the fact that under basal conditions in neuronal22,54,55,60 and non-neuronal61 alternative mechanisms can also directly or indirectly impair CMA in vitro models, but again other in vitro studies reported activity, as previously observed, due to high levels of α-syn, which conflicting results with no changes in macroautophagy markers33. can favor CMA impairment52,70. Finally, we also consider that CMA- In the mutant GBA cell lines, we did not detect a significant independent mechanisms based on the abnormal accumulation of increase in the levels of these two autophagy-related proteins sphingolipids and cholesterol can also influence α-syn localization, under basal conditions (Fig. 5b). Nevertheless, the analysis of autophagic flux confirmed that although new autophagosomes oligomerization, and physiological processes71.The direct interaction were formed, fused with lysosomes, and degraded, this process of α-syn with these lipids can promote the formation of oligomeric was significantly diminished but probably still partially functional and toxic forms of α-syn on the lysosomal membrane71,72, (Fig. 5a). Macroautophagy marker studies indicated normal levels suggesting that accumulated sphingolipids and cholesterol in of these proteins but with a tendency to show an activated lysosomes might act as aggregation sites for α-syn. induction of macroautophagy under basal conditions in the In conclusion, in this study, we defined a neuronal model that mutant cell lines (phosphorylated ULK1 in Fig. 5b). This suggested recapitulated the most important alterations in GBA-associated PD a compensatory response for the partial loss of lysosomal function and described a new molecular mechanism to understand the and autophagy activity as previously found in other models62. etiology of GBA-associated PD (see graphical abstract, Supple- Finally, our work focused on the CMA mechanism since this mentary Fig. S4). Importantly, most of the noticeable features of autophagy pathway plays a key role in PD. CMA is the main GBA-dependent lysosomal dysfunction are only significantly degradative pathway by which intracellular α-syn is elimi- evident in the lysosomal compartment. Our results showed that nated17,24. Consequently, a decrease in CMA activity has been the increase in lysosomal lipids associated with GCase activity loss associated with an increase in α-syn levels in PD patient samples could also lead to CMA activity impairment by destabilizing the and in vivo and in vitro models23,63–65. In our neuronal model, we LAMP-2A receptor at the lysosomal membrane. This impairment of analyzed CMA activity (Fig. 6c) and the lysosomal levels of LAMP- CMA activity promotes the abnormal accumulation of α-syn and 2A (Fig. 6a, b). Both parameters were affected in the KO and joins other synergistic and convergent pathways that contribute mutant cell lines. The observed decrease in CMA activity can be to altering α-syn levels, solubility, and conformation, ultimately attributed to decreased levels of LAMP-2A; nevertheless, the promoting neurodegeneration. previously reported dysfunction of lysosomal proteolysis and reduced enzymatic activity of several lysosomal enzymes might also have a summation effect on CMA impairment. METHODS Interestingly, in line with previous results, total LAMP-2A Culture cell lines expression was not significantly altered in homogenates but BE(2)-M17 cell line: human neuroblastoma cell line BE(2)-M17 was drastically decreased in the lysosomal fraction (Fig. 6a, b). Previous maintained in Opti-MEM supplemented with 10% fetal bovine npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 11 serum, 1% penicillin–streptomycin, and 0.5 mg/mL active genet- SQSTM1/p62 (GP62-C) from Abnova (Taipei, Taiwan), β-actin icin, all from Thermo Fisher Scientific (Waltham, MA, USA). Cells (A5441) from Sigma-Aldrich, Oligomer A11 (AHB0052) and were differentiated into dopaminergic-like neurons in Neurobasal Calnexin (MA3-027) from Thermo Fisher Scientific, α-syn medium supplemented with 10% inactivated fetal bovine serum, (610786) from BD Biosciences, and aggregated α-syn (MABN389) 1% penicillin–streptomycin, 0.5 mg/mL geneticin, B-27 Supple- from Merk Millipore (Burlington, MA, USA). Secondary HRP ment, and 10 μM retinoic acid, all from Thermo Fisher Scientific, antibodies were purchased from GE Healthcare (Chicago, IL, for at least 7 d. Generation of GBA knockout (KO) BE(2)-M17 cells by USA), and Secondary Alexa Fluor 488 and 594 were obtained from CRISPR/Cas9 nickase system: GBA KO cell line was created using the Thermo Fisher Scientific. CRISPR/cas9 nickase methodology. sgRNAs that target exon 3 of the human GBA1 genomic sequence were subcloned into the Sphingolipid quantification pSpCas9n(BB)-2A-GFP (PX461) plasmid (Plasmid #48140; Addgene, Lipid extraction was performed based on the methodology Watertown, MA, USA) according to the manufacturer’s protocol. described in ref. 73. Briefly, frozen cell pellets were dissolved in sgRNA sequences were designed using Benchling software 540 µL of IS Working Solution [MeOH:Cl3CH (2:1) solution contain- (www.benchling.com) and consisted of two pairs of oligonucleo- ing 0.01% of butylated hydroxytoluene (BHT), 1 nM internal tides: 5′-CAC CGTGTACTCTCATAGCGGCTGA-3′ and 5′-AAACTC standard glucosyl(ß)-sphingosine-d5 and 13 nM C18 glucosyl(ß)- AGCCGCTATGAGAGTACAC-3′; 5′-CACCGTACACGCAGAGGGCGAC ceramide-d5 solution (Avanti Polar Lipids Inc; 860636P, 860638P)]. GGA-3′ and 5′-AAACTCCGTCGCCCTCTGCGTGTAC-3′. Both plas- Samples were vortexed, sonicated, and centrifuged at 13,000 rpm mids were validated and co-transfected into BE(2)-M17 cell lines. for 5 min. The supernatant was evaporated under an N2 stream After 48 h, cells were sorted in a BD FACScaliburTM flow cytometer until dry and reconstituted with 500 µL of MeOH:Cl3CH (1:2). Next, (BD Biosciences, Franklin Lakes, NJ, USA), and GFP-positive cells a seven point calibration curve (from 0.0275 to 100 pmol) was were re-seeded at low density and cultured until single cells prepared by spiking a working solution containing different formed cell colonies. Thereafter, single colonies were picked and concentrations of GlcSph (Matreya LLC, State College, PA, USA), analyzed by assessing GCase levels using immunoblot, GCase GlcCer C16:0, C18:0, and C24:1 (860539 P, 860547P, 860549P; activity, and sequencing. At least five GBA KO cell lines were fully Avanti Polar Lipids Inc) standards in MeOH:Cl3CH (1:2). Analysis of characterized to confirm the complete loss of GCase protein. total hexosylsphingosine (HexSph) and hexosylceramide (HexCer) Mutant GBA plasmids, transfection, and generation of mutant GBA isoforms (C16:0, C18:0, C20:0, C22:0, C24:1, C24:0) was performed stable cell lines: site-directed mutagenesis was performed into an by liquid chromatography-mass spectrometry (LCMS-8050; Shi- original pCMV3-hGBA plasmid (HG12038-UT; Sino Biological, madzu, Kyoto, Japan) using an HPLC reversed-phase column Beijing, China) to generate pCMV3-hGBA-N370S and pCMV3- (TR-10902; Teknokroma Analítica S.A., Barcelona, Spain) with a hGBA-L444P plasmids using the primers pairs 5′-GCCACAGCA binary gradient of 0.1% formic acid in water (mobile phase A) and TCATCACGAGCCTCCTGTACCATGTGG-3′ with 5′-TGTACTGCA 0.1% formic acid in isopropanol:acetonitrile 3:1 (v/v) (mobile TCCCTCGATCCCAGGAGCCTAGCCGCA-3′ and 5′-GTCAGAAGAAC phase B). The mass analyzer was operated in positive ion GACCCGGACGCAGTGGCACTG-3′ with 5′-TGGCAACCAGCCCCACTC electrospray mode. The HexSph and HexCer species were TCTGGGAGCCCTCAG G-3′, respectively, and a Q5 Site-Directed quantified using their respective standard calibration curves using Mutagenesis Kit (New England Biolabs, Ipswich, MA, USA) LabSolutions CS Software. following the manufacturer’s instructions. Plasmids expressing N370S GBA or L444P GBA were transfected into BE(2)-M17 GBA KO cells, and stable colonies were generated after puromycin Total homogenate and subcellular fractionation from cells selection. Subsequently, single colonies were selected and Differentiated cells were harvested in PBS and pelleted by validated by sequencing, immunodetection, and GCase activity. centrifugation at 800 × g for 5 min. Total homogenates were At least five single colonies were fully characterized, and one clone extracted in conventional lysis buffer (20 mM Tris pH 7.5, 1 mM was selected as a representative. EDTA, 150 mM NaCl, 1% triton and protease and phosphatase inhibitors (Cell Signaling, 5872). To isolate subcellular fractiona- Reagents and antibodies tions cellular pellets were resuspended in 0.25 M sucrose (pH 7.2) and disrupted in a cavitation chamber (Kontes Glass Company) Reagents: leupeptin hemisulfate salt (L2884), ammonium chloride followed by homogenization in a Teflon-glass homogenizer and (A9434), 3-methyladenine (3-MA, M9281), carbonyl cyanide centrifugation (2500 × g for 15 min). Supernatant was then 3-chlorophenyl hydrazone (CCCP, C2759), oligomycin (75351), centrifuged at 17,000 × g. The supernatant was collected and antimycin A (A8674), AR7 (SML0921), chloroquine (C6628) and ultracentrifuged 1 h at 100,000 × g to collect cytosolic fraction, conduritol β epoxide (C5424) were obtained from Sigma-Aldrich while the resulting pellet after the 17,000 × g centrifugation was (St. Louis, MO, USA). Protease/Phosphatase Inhibitor Cocktail collected as the lysosomal/mitochondrial-enriched fraction and (5872) was purchased from Cell Signaling Technology (Danvers, was then loaded in two subsequent discontinuous Nycodenz, MA, USA), and the Protease Inhibitor cocktail (A32965) and sucrose or percoll (GEHealthcare, 17-0891-01) density gradient BCA Protein Assay Kit (23225) were obtained from Thermo from which pure lysosomal fraction was collected according to74. Fisher Scientific. D-glucosyl-β1-1′-D-erythro-sphingosine (GlcSph) Protein concentration of all fractions was determined by the (860535P) was obtained from Avanti Polar Lipids Inc (Alabaster, bicinchoninic assay method with BSA as a standard protein and AL, USA) and humanin (ALX-158-011-M001) from Enzo Life samples were frozen and stored at −20 °C until used. Sciences Inc (Farmingdale, NY, USA). Antibodies: LC3 (NB-2220), ULK1 (NBP2-24738S), ATF6 (NBP1-40256S), LAMP-2 (sc-18822), LAMP-1 (sc-19992), beclin 1/BECN1 (sc-48341), CD81 (D-4) (sc- Immunoblot detection 166028), Grp78/BiP (sc-376768), Atg 5 (sc-133158), and cathepsin Cellular samples (total homogenate or isolated fractions) were D (sc-6494) were obtained from Santa Cruz Biotechnology (Dallas, analyzed by western blot, samples were resolved by SDS-PAGE on TX, USA). GBA (ab55080) and phospho-α-syn (pS129) (ab51253) different percentages of polyacrylamide gels, ranging from 7% to were purchased from Abcam (Cambridge, UK). Phospho-ULK1 15%. Resolved proteins were transferred onto nitrocellulose Ser757 (6888S), eIF2α (9722), IRE1α (14C10) (3294), Phospho-eIF2α membranes and blocked with 5% non-fat milk before incubation (Ser51) (9721), Phospho-mTOR (2971S), mTOR (2972S), and CHOP with the corresponding primary antibody (diluted in 4% bovine (2895) were obtained from Cell Signaling Technology. Cathepsin B serum albumin in PBS) and subsequently in the corresponding (219408) was obtained from Calbiochem (San Diego, CA, USA), secondary antibody coupled to horseradish peroxidase (diluted in Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126 A. Navarro-Romero et al. 12 5% non-fat milk in PBS). Proteins were visualized using either West Finally, immunodetection was conducted following the standard Pico SuperSignal Substrate or SuperSignal West Femto (Thermo immunoblot detection protocol. Fisher Scientific, 34080 and 34095, respectively) on an Image- Quant RT ECL imaging system (GE Healthcare). Immunoblots were Cellular proteolytic assays quantified by densitometry using ImageJ 1.50 (NIH). Optical The degradation of long-lived proteins was quantified following a density values for target proteins were normalized by actin and/or previously described protocol76. Radioactive pulse analysis was Ponceau staining as loading control. For each figure, all blots were performed with [3H] valine (2 µ Ci/mL) (MT1654; HARTMANN processed in parallel and derive from the same experiment. ANALYTIC GmbH, Braunschweig, Germany) for 48 h at 37 °C. When indicated, cells were supplemented with 20 mM ammonium Endoglycosidase-H sensibility assay chloride and 100 μM leupeptin to block lysosomal-dependent GCase N-glycan processing was conducted using an proteolysis and 3-MA to inhibit macroautophagy. Lysosomal Endoglycosidase-H (Endo-H) sensitivity assay. Briefly, total homo- proteolysis was calculated as the proteolysis sensitive to inhibition genates from differentiated cells were digested with Endo-H by ammonium chloride and leupeptin, and CMA proteolysis as enzyme for 2 h at 37 °C (P0702S; New England Biolabs). PNGase F that which is sensitive to ammonium chloride and leupeptin and (P0710S; New England Biolabs) was used as a control8. After insensitive to 3-MA at the indicated time77. digestion, GBA levels were analyzed by immunoblotting and percentage of EndoH-sensitive fraction was calculated. Lysosomal enzymatic activities All enzymatic assays were conducted in triplicate in 96-well plates Cellular viability using 5–10 µg of total homogenate samples or 0.5–1 µg of the Viability of differentiated cells was measured as the % of viable lysosomal-enriched fraction. Activity measurements for all cells counting the number of viable and dead cells accumulated enzymes were always corrected by mg of protein measured in during the last 72 h prior to measurement and without changing parallel with the same sample by BCA. the differentation media. Triplicates of at least three independent samples were analyzed in a Muse Cell Analyzer cytometer with the Glucocerebrosidase activity (GCase): Samples were resuspended Muse Count & Viability Assay Kit (MCH100102, Thermofisher) that in McIlvaine’s buffer (0.2 M citrate/phosphate buffer pH 5.4) stains viable and non-viable cells based on their permeability to containing 0.25% Triton X-100, 22 mM of the activator sodium two DNA binding dyes present in the reagent. taurocholate hydrate (SIGMA, 86339), and 5 mM of the fluor- escent substrate 4-Methylumbelliferyl β-D-glucopyranoside Immunofluorescence microscopy (M363; Sigma-Aldrich). A sample volume of 100 μL per well was prepared and transferred to 96-well plates in triplicate and Cells were fixed in 3% formaldehyde for 30 min at room incubated for at least 1 h at 37 °C. Next, a stop solution (25 μL of temperature (RT). Following incubation with a blocking solution 0.25 M glycine/NaOH pH 10.4) was added to terminate the (powdered milk 0.2%, glycine 0.1 M, Triton X-100 0.01%, NCS 2%, reaction, and the liberated 4-methylumbelliferone was measured and BSA 1%, in a 1× PBS solution) for 1 h at RT, the corresponding using an FLx800 spectrofluorimeter (BioTek, Winooski, VT, USA) primary antibodies were used diluted in 0.1% BSA in 1× PBS at λex = 365 nm/λem = 450 nm. When indicated cells were overnight at 4 °C. Afterward, cells were incubated for 1 h at RT treated 30 min with 1 mM conduritol β epoxide (CβE) (C5424; with the corresponding secondary antibodies, diluted in 0.1% Sigma-Aldrich) before GCase assay. BSA in 1× PBS. Thereafter, nuclei were labeled using Hoechst 33342 (62249; Thermo Fisher Scientific). Coverslips were then Hexosaminidase activity assay: Samples were dissolved in a mounted using Dako Fluorescent Mounting Medium (S3023; reaction buffer containing 0.1 M acetate buffer pH 4.4, 1 mM of Dako Denmark A/S, Glostrup, Denmark). Images were acquired fluorescent substrate 4-methylumbelliferyl-N-acetyl-B-D-gluco- using an Olympus FV1000 (Olympus Corporation, Tokyo, Japan) pyranoside (M2133; Sigma-Aldrich), and 0.125% Triton X-100. or a Zeiss LSM 980 confocal microscope (Carl Zeiss AG, Jena, Stop solution (75 μL of 2 M glycine/Na2CO3) was added to Germany). Intensity and colocalization analyses were performed terminate the reaction, and the plate was measured using ImageJ 1.50a (NIH, Bethesda, MD, USA) and Zen Blue (Carl using an FLx800 spectrofluorimeter (BioTek) at λex = 365 nm, Zeiss AG) software. λem = 450 nm. In vivo cross-linking Acid phosphatase activity assay: Samples were dissolved in citrate In vivo cross-linking of synuclein oligomers was performed buffer solution 0.09 M pH 4.8 and substarte 4-nitrophenyl phos- according to ref. 75 with disuccinimidyl glutarate (DSG). phate (1 tablet/2.5 mL buffer) (N9389; Sigma-Aldrich). Stop solution (200 μL 0.5 N NaOH) was added to terminate the reaction, Filter retardation assay and the plate was measured using an ELx800 spectrophotometer (BioTek) at λ = 405 nm. Fifty micrograms of previously DSG-cross-linked cell extracts were subjected to a filter retardation assay following a Cathepsin B activity assay: The activity was measured using a previously described protocol15 using cellulose acetate mem- Cathepsin B Activity Assay Kit (ab65300; Abcam) according to the branes (0.2-μm pore size; GE Healthcare) in a Minifold-1 Dot-Blot manufacturer’s instructions. Samples were processed without the System (Schleicher & Schuell GmbH, Keene, NH, USA). Next, cathepsin B inhibitor pepstatin A. membranes were washed twice with 1% SDS-PBS and immuno- detection of different α-syn species was performed according to Acid sphingomyelinase activity assay: Samples were dissolved in a the protocol described above. reaction buffer containing 0.1 M NaAc, 0.2% sodium taurocholate hydrate (86339; Sigma-Aldrich; pH 5.2), and 110 μM of substrate Extracellular synuclein detection HMU-PC (6-Hexadecanoylamino-4-methylumbelliferyl phosphoryl- Extracellular α -synuclein was detected from culture medium. After choline; Moscerdam Substrates, Rotterdam, Netherlands). Stop centrifugation to remove the remaining cells and debris, the solution (200 μL glycine-NaOH 0.2 M, 0.2% m/v SDS, 0.2% v/v medium was boiled and filtered in a nitrocellulose membrane Triton X-100, pH 10.6) was added to terminate the reaction, the using a Minifold-1 Dot-Blot System (Schleicher & Schuell GmbH). plate was measured using a SpectraMax M5 Multi-Mode npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 13 Microplate Reader (Molecular Devices, San Jose, CA, USA) at Statistical analysis λex = 355 nm, λem = 460 nm. All values are expressed as the mean ± standard error of the mean (SEM). Significant differences (*p < 0.05, **p < 0.01, ***p < 0.001, Acid ceramidase activity assay: Samples were dissolved in a ****p < 0.0001) when comparing two groups were determined by reaction buffer containing 0.1 M NaAc (pH 4.5) and 40 μM of a two-tailed unpaired Student’s t-test. When comparing more than substrate RBM14-C12 (860655; Avanti Polar Lipids Inc). Next, a stop two groups, significant differences were determined by one- or solution (50 μL of methanol and 100 μL of a 2.5 mg/mL NaIO4 fresh two-way ANOVA followed by Tukey’s multiple comparisons test. solution in 100 mM glycine/NaOH buffer, pH 10.6) was added to Statistical analyses were performed using GraphPad Prism 6 terminate the reaction, after which the plate was measured using (GraphPad Software Inc, San Diego, CA, USA). a SpectraMax M5 Multi-Mode Microplate Reader (Molecular Devices) at λex = 360 nm, λem = 446 nm. DATA AVAILABILITY Cathepsin D activity assay: The activity was measured using a The data that support the findings of this study does not present restrictions and are Cathepsin D Activity Assay Kit (ab65302; Abcam) according to the available upon a request for information that is reasonably made and that can be manufacturer’s instructions. provided by the corresponding author. In situ Cathepsin D activity assay: Cells were treated with a 1 μg/ Received: 1 June 2022; Accepted: 14 September 2022; mL BODYPY-FL-pepstatin A probe (P12271; Thermo Fisher Scientific) for 1 h at 37 °C. Subsequently, flow cytometry analysis was performed using a BD LSR Fortessa Cell Analyzer (BD Biosciences); 10,000 cells were acquired, and fluorescence was determined using a 488-nm laser and 694/40 detector and REFERENCES represented as mean fluorescence intensity (MFI) using FCS 1. Tysnes, O. B. & Storstein, A. Epidemiology of Parkinson’s disease. J. Neural Transm. Express software. 124, 901–905 (2017). 2. Dehay, B. et al. Targeting α-synuclein for treatment of Parkinson’s disease: mechanistic and therapeutic considerations. Lancet Neurol. 14, 855–866 Lysosomal pH measurement (2015). Lysosomal pH was measured using LysoSensor Yellow/Blue DND- 3. Neumann, J. et al. Glucocerebrosidase mutations in clinical and pathologically 160 dye (L7545; Thermo Fisher Scientific) according to76. proven Parkinson’s disease. Brain 132, 1783–1794 (2009). 4. Sidransky, E. et al. Multicenter analysis of glucocerebrosidase mutations in Par- kinson’s disease. N. Engl. J. Med. 361, 1651–1661 (2009). Quantitative real-time PCR (qPCR) analysis of LAMP-2A mRNA 5. Lesage, S. et al. Large-scale screening of the Gaucher’s disease-related gluco- Isolation of total messenger RNA (mRNA) was performed in cell cerebrosidase gene in Europeans with Parkinson’s disease. Hum. Mol. Genet. 20, 202–210 (2011). pellets using the RNeasy Mini kit (Qiagen, Hilden, Germany), and 6. Stojkovska, I., Krainc, D. & Mazzulli, J. R. Molecular mechanisms of α-synuclein and reverse transcription was performed using 500 ng of total RNA and GBA1 in Parkinson’s disease. Cell Tissue Res. 373, 51–60 (2018). a High-Capacity cDNA Reverse Transcription kit (Thermo Fisher 7. Bae, E.-J. et al. Glucocerebrosidase depletion enhances cell-to-cell transmission of Scientific) according to manufacturer’s instructions. Real-time PCR α-synuclein. Nat. Commun. 5, 4755 (2014). was carried out using PowerUp SYBR Green Master Mix (Thermo 8. Ron, I. & Horowitz, M. ER retention and degradation as the molecular basis underlying Gaucher disease heterogeneity. Hum. Mol. Genet. 14, 2387–2398 Fisher Scientific) and a 7900HT Fast Real-Time PCR instrument (2005). (Applied Biosystems, Waltham, MA, USA) using LAMP-2A primers 9. Cullen, V. et al. Acid β-glucosidase mutants linked to Gaucher disease, Parkinson (Fw: GGGTTCAGCCTTTCAATGTG, Rev: CAGCATGATGGTGCTTGAGA), disease, and Lewy body dementia alter α-synuclein processing. Ann. Neurol. 69, LAMP-2B primers (Fw: GGGTTCAGCCTTTCAATGTG and Rev: 940–953 (2011). CCTGAAAGACCAGCACCAAC), and LAMP-2C primers (Fw: GTATT 10. Gegg, M. E., Menozzi, E. & Schapira, A. H. V. Glucocerebrosidase-associated Par- CTACAGCTGAAGAATGTTCTG and Rev: ACACCCACTGCAACA kinson disease: Pathogenic mechanisms and potential drug treatments. Neuro- biol. Dis. 166, 105663 (2022). GGAAT). HPRT (Fw: TTATGGACAGGACTGAACGTCTTG, Rev: GCAC 11. Körschen, H. G. et al. The non-lysosomal β-glucosidase GBA2 is a non-integral ACAGAGGGCTACAATGTG) and UBE2D2 (Fw: TGCCTGAGATTGCTC membrane-associated protein at the endoplasmic reticulum (ER) and Golgi. J. GGATCT, Rev: TCGCATACTTCTGAGTCCATTCC) were used as Biol. Chem. 288, 3381–3393 (2013). reference genes. Quantification was performed using the com- 12. Maley, F., Trimble, R. B., Tarentino, A. L. & Plummer, T. H. Characterization of parative CT method. All the samples were tested in triplicate and glycoproteins and their associated oligosaccharides through the use of endo- glycosidases. Anal. Biochem. 180, 195–204 (1989). the relative expression values were normalized to the expression 13. Migdalska-Richards, A. & Schapira, A. H. V. V. The relationship between gluco- value of HPRT and UBE2D2. The fold change was calculated using cerebrosidase mutations and Parkinson disease. J. Neurochem. 77–90 https:// the 2-ΔCt method. Three independent experiments with differ- doi.org/10.1111/jnc.13385 (2016). entiated cells were performed. 14. Hetz, C., Zhang, K. & Kaufman, R. J. Mechanisms, regulation and functions of the unfolded protein response. Nat. Rev. Mol. Cell Biol. 21, 421–438 (2020). 15. Recasens, A. et al. Lack of pathogenic potential of peripheral α-synuclein Cholesterol quantification aggregates from Parkinson’s disease patients. Acta Neuropathol. Commun. 6, 8 Cholesterol levels were measured using an Amplex Red Choles- (2018). terol Assay Kit (A12216; Thermo Fisher Scientific) according to the 16. Navarro-Romero, A., Montpeyó, M. & Martinez-Vicente, M. The emerging role of the lysosome in Parkinson’s Disease. Cells 9, 1–25 (2020). manufacturer’s instructions. 17. Martinez-Vicente, M. & Vila, M. Alpha-synuclein and protein degradation path- ways in Parkinson’s disease: a pathological feed-back loop. Exp. Neurol. 247, Transfection 308–313 (2013). 18. Auteri, J. S., Okada, A., Bochaki, V. & Dice, J. F. Regulation of intracellular protein Plasmid expressing human LAMP-2 cDNA (transcript variant A; degradation in IMR-90 human diploid fibroblasts. J. Cell Physiol. 115, 167–174 SC118738; OriGene, Rockville, MD, USA) was transiently trans- (1983). fected into BE(2)-M17 cell lines using Lipofectamine 3000 Reagent 19. Aits, S. et al. Sensitive detection of lysosomal membrane permeabilization by (L3000015; Thermo Fisher Scientific) according to the manufac- lysosomal galectin puncta assay by lysosomal galectin puncta assay. Autophagy turer’s protocol. 8627, 1408–1424 (2015). Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126 A. Navarro-Romero et al. 14 20. Fernandes, H. J. R. R. et al. ER stress and autophagic perturbations lead to ele- 47. Lloyd-Evans, E., Pelled, D., Riebeling, C. & Futerman, A. H. Lyso-glycosphingolipids vated extracellular α-Synuclein in GBA-N370S Parkinson’s iPSC-derived dopamine mobilize calcium from brain microsomes via multiple mechanisms. Biochem. J. neurons. Stem Cell Rep. 6, 1–15 (2016). 375, 561–565 (2003). 21. Magalhaes, J. et al. Autophagic lysosome reformation dysfunction in glucocer- 48. Korkotian, E. et al. Elevation of intracellular glucosylceramide levels results in an ebrosidase deficient cells: relevance to Parkinson disease. Hum. Mol. Genet. 25, increase in endoplasmic reticulum density and in functional calcium stores in 3432–3445 (2016). cultured neurons. J. Biol. Chem. https://doi.org/10.1074/jbc.274.31.21673 (1999). 22. Schöndorf, D. C. et al. IPSC-derived neurons from GBA1-associated Parkinson's 49. Pelled, D. et al. Enhanced calcium release in the acute neuronopathic form of disease patients show autophagic defects and impaired calcium homeostasis. Gaucher disease. Neurobiol. Dis. 18, 83–88 (2005). Nat. Commun. 5, 4028 (2014). 50. Hannun, Y. A. & Obeid, L. M. Sphingolipids and their metabolism in physiology 23. Cuervo, A. M., Stefanis, L., Fredenburg, R., Lansbury, P. T. & Sulzer, D. Impaired and disease. Nat. Rev. Mol. Cell Biol. 19, 175–191 (2018). degradation of mutant alpha-synuclein by chaperone-mediated autophagy. Sci- 51. Esmaeili, Y. et al. Targeting autophagy, oxidative stress, and ER stress for neu- ence 305, 1292–1295 (2004). rodegenerative diseases treatment. J. Control. Release. https://doi.org/10.1016/ 24. Vogiatzi, T., Xilouri, M., Vekrellis, K. & Stefanis, L. Wild type alpha-synuclein is j.jconrel.2022.03.001 (2022). degraded by chaperone-mediated autophagy and macroautophagy in neuronal 52. Kuo, S. H. et al. Mutant glucocerebrosidase impairs α-synuclein degradation by cells. J. Biol. Chem. 283, 23542–23556 (2008). blockade of chaperone-mediated autophagy. Sci. Adv. 8, eabm6393 (2022). 25. Mak, S. K., McCormack, A. L., Manning-Bog, A. B., Cuervo, A. M. & Di Monte, D. A. 53. García-Sanz, P., Orgaz, L., Bueno-gil, G. & Espadas, I. N370S-GBA1 mutation causes Lysosomal degradation of alpha-synuclein in vivo. J. Biol. Chem. 285, lysosomal cholesterol accumulation in Parkinson’s disease. Mov. Disord. 32, 13621–13629 (2010). 1409–1422 (2017). 26. Kaushik, S. et al. Chaperone-mediated autophagy at a glance. J. Cell Sci. 124, 54. Awad, O. et al. Altered TFEB-mediated lysosomal biogenesis in Gaucher disease 495–499 (2011). iPSC-derived neuronal cells. Hum. Mol. Genet. https://doi.org/10.1093/hmg/ 27. Patel, B. & Cuervo, A. M. Methods to study chaperone-mediated autophagy. ddv297 (2015). Methods 75, 133–140 (2015). 55. Mazzulli, J. R. et al. Gaucher disease glucocerebrosidase and α-synuclein form a 28. Rodriguez-Navarro, J. A. & Cuervo, A. M. Dietary lipids and aging compromise bidirectional pathogenic loop in synucleinopathies. Cell 146, 37–52 (2011). chaperone-mediated autophagy by similar mechanisms. Autophagy 8, 56. Shibata, M. et al. Regulation of intracellular accumulation of mutant Huntingtin 1152–1154 (2012). by Beclin 1. J. Biol. Chem. 281, 14474–14485 (2006). 29. Kaushik, S., Massey, A. C. & Cuervo, A. M. Lysosome membrane lipid micro- 57. Yang, S., Gegg, M., Chau, D. & Schapira, A. Glucocerebrosidase activity, cathepsin domains: novel regulators of chaperone-mediated autophagy. EMBO J. 25, D and monomeric α- synuclein interactions in a stem cell derived neuronal model 3921–3933 (2006). of a PD associated GBA1 mutation. Neurobiol. Dis. 104620. https://doi.org/ 30. Anguiano, J. et al. Chemical modulation of chaperone-mediated autophagy by 10.1016/j.nbd.2019.104620 (2019). retinoic acid derivatives. Nat. Chem. Biol. 9, 374–382 (2013). 58. Wolfe, D. M. et al. Autophagy failure in Alzheimer’s disease and the role of 31. Gong, Z. et al. Humanin is an endogenous activator of chaperonemediated defective lysosomal acidification. Eur. J. Neurosci. 37, 1949–1961 (2013). autophagy. J. Cell Biol. 217, 635–647 (2018). 59. Dehay, B. et al. Lysosomal dysfunction in Parkinson disease: ATP13A2 gets into 32. Filograna, R. et al. Analysis of the catecholaminergic phenotype in human SH- the groove. Autophagy 8, 1389–1391 (2012). SY5Y and BE(2)-M17 neuroblastoma cell lines upon differentiation. PLoS ONE 10, 60. Li, H. et al. Mitochondrial dysfunction and mitophagy defect triggered by het- 1–18 (2015). erozygous GBA mutations. Autophagy 15, 113–130 (2019). 33. Muñoz, S. S., Petersen, D., Marlet, F. R., Kücükköse, E. & Galvagnion, C. The 61. Kim, M. J., Jeon, S., Burbulla, L. F. & Krainc, D. Acid ceramidase inhibition ame- interplay between Glucocerebrosidase, α-synuclein and lipids in human models liorates a -synuclein accumulation upon loss of GBA1 function. Hum. Mol. Genet. of Parkinson’s disease. Biophys. Chem. https://doi.org/10.1016/j.bpc.2020.106534 27, 1972–1988 (2018). (2021). 62. Massey, A. C., Kaushik, S., Sovak, G., Kiffin, R. & Cuervo, A. M. Consequences of the 34. Enquist, I. B. et al. Murine models of acute neuronopathic Gaucher disease. Proc. selective blockage of chaperone-mediated autophagy. Proc. Natl Acad. Sci. USA Natl Acad. Sci. USA 104, 17483–17488 (2007). 103, 5805–5810 (2006). 35. Farfel-Becker, T. et al. Neuronal accumulation of glucosylceramide in a mouse 63. Mak, S. K., McCormack, A. L., Manning-Bog, A. B., Cuervo, A. M. & Di Monte, D. A. model of neuronopathic Gaucher disease leads to neurodegeneration. Hum. Mol. Lysosomal degradation of α-synuclein in vivo. J. Biol. Chem. 285, 13621–13629 Genet. 23, 843–854 (2014). (2010). 36. Sun, Y. et al. Ex vivo and in vivo effects of isofagomine on acid β-glucosidase 64. Martinez-Vicente, M. et al. Dopamine-modified α-synuclein blocks chaperone- variants and substrate levels in Gaucher disease. J. Biol. Chem. 287, 4275–4287 mediated autophagy. J. Clin. Investig. 118, 777–778 (2008). (2012). 65. Xilouri, M. et al. Impairment of chaperone-mediated autophagy induces dopa- 37. Sun, Y. et al. Substrate compositional variation with tissue/region and Gba1 minergic neurodegeneration in rats. Autophagy 12, 2230–2247 (2016). mutations in mouse models-implications for Gaucher disease. PLoS ONE. https:// 66. Orenstein, S. J. et al. Interplay of LRRK2 with chaperone-mediated autophagy. doi.org/10.1371/journal.pone.0057560 (2013). Nat. Neurosci. 16, 394–406 (2013). 38. Polinskii, N. K. et al. Decreased glucocerebrosidase activity and substrate accu- 67. Kabuta, T., Furuta, A., Aoki, S., Furuta, K. & Wada, K. Aberrant interaction between mulation of glycosphingolipids in a novel gba1 d409v knock-in mouse model. Parkinson disease-associated mutant UCH-L1 and the lysosomal receptor for PLoS ONE 16, e0252325 (2021). chaperone-mediated autophagy. J. Biol. Chem. 283, 23731–23738 (2008). 39. Sardi, S. P. et al. CNS expression of glucocerebrosidase corrects alpha-synuclein 68. Cuervo, A. M., Stafanis, L., Fredenburg, R., Lansbury, P. T. & Sulzer, D. Impaired pathology and memory in a mouse model of Gaucher-related synucleinopathy. degradation of mutant α-synuclein by chaperone-mediated autophagy. Science Proc. Natl Acad. Sci. USA 108, 12101–12106 (2011). 305, 1292–1295 (2004). 40. Nilsson, O. & Svennerholm, L. Accumulation of glucosylceramide and glucosyl- 69. Rodriguez-Navarro, J. A. et al. Inhibitory effect of dietary lipids on chaperone- sphingosine (psychosine) in cerebrum and cerebellum in infantile and juvenile mediated autophagy. Proc. Natl Acad. Sci. USA 109, E705–E714 (2012). Gaucher disease. J. Neurochem. 39, 709–718 (1982). 70. Klionsky, D. J. et al. Guidelines for the use and interpretation of assays for 41. Orvisky, E. et al. Glucosylsphingosine accumulation in tissues from patients with monitoring autophagy (4th edition). Autophagy. https://doi.org/10.1080/ Gaucher disease: correlation with phenotype and genotype. Mol. Genet. Metab. 15548627.2020.1797280 (2021). 76, 262–270 (2002). 71. Man, W. K. et al. A role of cholesterol in modulating the binding of α-synuclein to 42. Huebecker, M. et al. Reduced sphingolipid hydrolase activities, substrate accu- synaptic-like vesicles. Front. Neurosci. https://doi.org/10.3389/fnins.2020.00018 mulation and ganglioside decline in Parkinson’s disease. Mol. Neurodegener. 14, (2020). 1–21 (2019). 72. Jakubec, M. et al. Cholesterol-containing lipid nanodiscs promote an α-synuclein 43. Rocha, E. M. et al. Progressive decline of glucocerebrosidase in aging and Par- binding mode that accelerates oligomerization. FEBS J. https://doi.org/10.1111/ kinson’s disease. Ann. Clin. Transl. Neurol. 2, 433–438 (2015). febs.15551 (2020). 44. Gegg, M. E. et al. No evidence for substrate accumulation in Parkinson brains with 73. Gorrochategui, E., Casas, J., Porte, C., Lacorte, S. & Tauler, R. Chemometric strategy GBA mutations. Mov. Disord. 30, 1085–1089 (2015). for untargeted lipidomics: Biomarker detection and identification in stressed 45. Kurzawa-Akanbi, M. et al. Altered ceramide metabolism is a feature in the human placental cells. Anal. Chim. Acta 854, 20–33 (2015). extracellular vesicle-mediated spread of alpha-synuclein in Lewy body disorders. 74. Storrie, B. & Madden, E. A. Isolation of subcellular organelles. Methods Enzymol. Acta Neuropathol. 142, 961–984 (2021). https://doi.org/10.1016/0076-6879(90)82018-W (1990). 46. Clark, L. N. et al. Gene-wise association of variants in four lysosomal storage 75. Dettmer, U., Newman, A. J., Luth, E. S., Bartels, T. & Selkoe, D. In vivo cross-linking disorder genes in neuropathologically confirmed Lewy body disease. PLoS ONE. reveals principally oligomeric forms of α-synuclein and β-synuclein in neurons https://doi.org/10.1371/journal.pone.0125204 (2015). and non-neural cells. J. Biol. Chem. 288, 6371–6385 (2013). npj Parkinson’s Disease (2022) 126 Published in partnership with the Parkinson’s Foundation A. Navarro-Romero et al. 15 76. Dehay, B. et al. Loss of P-type ATPase ATP13A2/PARK9 function induces general COMPETING INTERESTS lysosomal deficiency and leads to Parkinson disease neurodegeneration. Proc. The authors declare no competing interests. Natl Acad. Sci. USA 109, 9611–9616 (2012). 77. Kaushik, S. & Cuervo, A. M. Methods to monitor chaperone-mediated autophagy. In Methods in Enzymology. Vol. 452 (Elsevier Inc., 2009). ADDITIONAL INFORMATION Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41531-022-00397-6. ACKNOWLEDGEMENTS The authors wish to thank Dr. Arango (VHIR) for the PX461 vector and all the Vila lab Correspondence and requests for materials should be addressed to Marta Martinez- members for their support. This work was supported by the Fondo de Investigación Vicente. Sanitaria-Instituto de Salud Carlos III (Spain)-FEDER (PI17/00496 and PI20/00728), the Michael J. Fox Foundation, the Silverstein Foundation (MJFF 16182), and the BBVA Reprints and permission information is available at http://www.nature.com/ Foundation (NanoERT). M.M. was supported by an FPU doctoral fellowship (FPU18/ reprints 05595) from MINECO (Spain); J.R. was supported by a PERIS fellowship (Generalitat de Catalunya); E.P. was supported by a VHIR doctoral fellowship (VHIR, Barcelona). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. AUTHOR CONTRIBUTIONS A.N.-R.: performed lipids analysis, GCase activity, RT-PCR and contributed to the conception of the manuscript, the discussion of results and writing; I.F.-G.: Open Access This article is licensed under a Creative Commons participated in the characterization of α-syn species and ER stress experiments; J.R.: Attribution 4.0 International License, which permits use, sharing, provided experimental support to all the experimental tasks, performed all lysosomal adaptation, distribution and reproduction in any medium or format, as long as you give isolations, cell culture maintenance, proteolysis assays and autophagy characteriza- appropriate credit to the original author(s) and the source, provide a link to the Creative tion; M.M.: performed GCase activity assays and viability assays; M.A.-B. and P.C.-S.: Commons license, and indicate if changes were made. The images or other third party performed the ER retention experiments; T.L.-R.: participated in the generation of the material in this article are included in the article’s Creative Commons license, unless GBA mutant cell lines; E.P.: participated supporting experimental work; C.C.-C., indicated otherwise in a credit line to the material. If material is not included in the J.A.A.-A. and L.C.-R. performed GlcCer and GlcSph quantification and interpretation; article’s Creative Commons license and your intended use is not permitted by statutory J.C.: performed ASAH1 and ASM activities; M.V.: contributed to the support of the regulation or exceeds the permitted use, you will need to obtain permission directly project, the discussion of results and revising of the article; M.M.V.: performed the from the copyright holder. To view a copy of this license, visit http:// lysosomal enzymatic activities, lysosomal pH, LMP assay, macroautophagy and CMA creativecommons.org/licenses/by/4.0/. characterization, cholesterol quantification. Conceived, designed, and supervised the project, analyzed results and was responsible for data interpretation and writing and revising the manuscript. All authors read and approved the final manuscript. © The Author(s) 2022 Published in partnership with the Parkinson’s Foundation npj Parkinson’s Disease (2022) 126