Below is a snapshot of the Web page as it appeared on 5/7/2024 (the last time our crawler visited it). This is the version of the page that was used for ranking your search results. The page may have changed since we last cached it. To see what might have changed (without the highlights), go to the current page.
Bing is not responsible for the content of this page.
Rescue of α-synuclein aggregation in Parkinson’s patient neurons by synergistic enhancement of ER proteostasis and protein trafficking: Neuron
If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The Ken and Ruth Davee Department of Neurology, Northwestern University Feinberg School of Medicine, Chicago, IL 60611, USADepartment of Molecular Neurology, University Hospital Erlangen, Friedrich-Alexander University Erlangen-Nürnberg, Erlangen, Germany
α-Synuclein accumulation induces ER fragmentation in patient-derived midbrain neurons
•
α-Synuclein perturbs the neuron's ability to recognize and respond to misfolded proteins in the ER
•
Parkinson's neurons develop pathogenic aggregates of immature lysosomal GCase
•
GCase solubility/function is rescued by enhancing ER folding and downstream trafficking
Summary
Neurodegenerative disorders are characterized by a collapse in proteostasis, as shown by the accumulation of insoluble protein aggregates in the brain. Proteostasis involves a balance of protein synthesis, folding, trafficking, and degradation, but how aggregates perturb these pathways is unknown. Using Parkinson’s disease (PD) patient midbrain cultures, we find that aggregated α-synuclein induces endoplasmic reticulum (ER) fragmentation and compromises ER protein folding capacity, leading to misfolding and aggregation of immature lysosomal β-glucocerebrosidase. Despite this, PD neurons fail to initiate the unfolded protein response, indicating perturbations in sensing or transducing protein misfolding signals in the ER. Small molecule enhancement of ER proteostasis machinery promotes β-glucocerebrosidase solubility, while simultaneous enhancement of trafficking improves ER morphology, lysosomal function, and reduces α-synuclein. Our studies suggest that aggregated α-synuclein perturbs the ability of neurons to respond to misfolded proteins in the ER, and that synergistic enhancement of multiple proteostasis branches may provide therapeutic benefit in PD.
Parkinson’s disease (PD) is characterized by the loss of dopaminergic neurons in the midbrain and the presence of protein inclusions called Lewy bodies and Lewy neurites that are composed of α-synuclein (α-syn) (
), indicating that α-syn aggregation is critical for pathogenesis. Multiplications of wild-type SNCA also cause PD and the severity of the clinical phenotype is dependent on α-syn dosage. For example, the clinical presentation of SNCA duplication patients occurs much later and is not as severe as in SNCA triplication patients (
). Patients with SNCA multiplications display the classical Lewy pathology, indicating that overabundance of the wild-type α-syn protein leads to neurodegeneration. However, our mechanistic understanding of how α-syn aggregates induce neurotoxicity is incomplete.
The presence of α-syn aggregates suggests that proteostasis pathways, including the lysosomal clearance pathway, are disrupted in the PD brain. Genome-wide association studies in PD patients have identified several risk genes, most of which have key roles in autophagy and lysosomal function (
International Parkinson’s Disease Genomics Consortium23andMe Research Team A meta-analysis of genome-wide association studies identifies 17 new Parkinson’s disease risk loci.
International Parkinson’s Disease Genomics Consortium (IPDGC)Parkinson’s Study Group (PSG) Parkinson’s Research: The Organized GENetics Initiative (PROGENI)23andMeGenePDNeuroGenetics Research Consortium (NGRC)Hussman Institute of Human Genomics (HIHG)Ashkenazi Jewish Dataset InvestigatorCohorts for Health and Aging Research in Genetic Epidemiology (CHARGE)North American Brain Expression Consortium (NABEC)United Kingdom Brain Expression Consortium (UKBEC)Greek Parkinson’s Disease ConsortiumAlzheimer Genetic Analysis Group Large-scale meta-analysis of genome-wide association data identifies six new risk loci for Parkinson’s disease.
). Of these, loss-of-function mutations in GBA1 represent one of the strongest genetic risk factors for the development of PD and dementia with Lewy bodies (DLB) (
American Genome Center Genome sequencing analysis identifies new loci associated with Lewy body dementia and provides insights into its genetic architecture.
). Homozygous mutations in the GBA1 gene, which encodes lysosomal β-glucocerebrosidase (GCase), cause the lysosomal storage disorder Gaucher’s disease (GD) that is characterized by glycosphingolipid accumulation and neurodegeneration (
). Proper GCase maturation requires its trafficking from the endoplasmic reticulum (ER) to the Golgi portion of the secretory pathway. Previous work has shown that α-syn accumulation can impair ER-to-Golgi trafficking and disrupt protein maturation (
). Our recent studies showed that trafficking disruption occurs by α-syn-mediated inhibition of the SNARE protein ykt6, which prevents the fusion of ER-derived vesicles on the cis-Golgi, causing downstream lysosomal depletion (
Perhaps the most critical function of the proteostasis network takes place in the ER compartment, where approximately one-third of the cell’s proteome is synthesized, folded, and processed. Calcium-dependent molecular chaperones such as calnexin (CANX) are particularly important for maintaining proper protein folding and quality control of N-linked glycosylated proteins, including GCase and other lysosomal hydrolases (
). Disrupted protein trafficking and accumulation of immature proteins in the ER can overwhelm the folding machinery, leading to ER stress and initiation of the unfolded protein response (UPR). The UPR constitutes a series of pathways that transduce ER stress signals to the nucleus for transcriptional upregulation of quality control machinery and expansion of the ER to accommodate excess protein load (
). The three main UPR sensors include inositol-requiring enzyme (IRE1), double-stranded RNA-activated protein kinase (PKR)-like ER kinase (PERK), and activating transcription factor 6 (ATF6). Stimulation of these three branches acts to restore ER proteostasis by reducing global protein synthesis and stimulating gene expression of folding machinery including chaperones GRP78 and GRP94 (
). GBA1 mutations destabilize GCase structure, resulting in UPR induction, expansion of the ER compartment, and elimination of the protein through ER associated degradation (ERAD) (
), further emphasizing the importance of maintaining ER proteostasis in neuronal health. While these studies collectively suggest that ER dysfunction is associated with PD, the mechanistic links among α-syn accumulation, protein misfolding in the ER, and downstream lysosomal dysfunction have not been established. Furthermore, it is unknown whether enhancing ER proteostasis alone is sufficient to rescue lysosomal dysfunction and reduce pathological α-syn in PD patient neurons. To address these questions, we developed novel iPSC-derived midbrain dopaminergic (DA) models from PD patients that carry a triplication (3x) in SNCA. We find that SNCA-3x patient neurons exhibit severe perturbations in the ER that lead to lysosomal dysfunction and can be rescued by synergistic enhancement of protein folding in the ER and trafficking via small-molecule modulators.
). To further examine the mechanism of this process, we generated and characterized new iPSC lines from controls (Ctrls) and three distinct patients that carry a triplication in SNCA and exhibit early-onset parkinsonism and dementia (
), matured for 90 days, and analyzed for the presence of aggregated α-syn. Immunofluorescence and biochemical analysis indicated that patient lines accumulated insoluble α-syn within neurites and the cell body that were thioflavin positive compared to controls (Figures S2A–S2C). Analysis of GCase maturation by western blot showed a reduction in SNCA-3x DA neurons, indicated by the accumulation of immature, low molecular weight forms of GCase (∼55–62 kDa) (Figure 1A). We also observed a decline in GCase activity within lysosomal compartments of living SNCA-3x DA neurons (Figure 1B). Analysis of neurite degeneration by neurofilament immunostaining indicated no change at this time point, suggesting that the decline in activity is not due to cell toxicity (Figure 1C).
Figure 1Defects in GCase maturation and lysosomal function in novel SNCA-3x midbrain DA lines
(A) Western blot of GCase and α-syn in day 90 neurons. Quantification is on the right. Coomassie was used as a loading control.
(B) Live-cell lysosomal GCase activity of day 90 neurons.
(C) Neurofilament quantification of day 90 neurons.
(D) Quantitative RT-PCR analysis of SNCA mRNA expression of day 90 DA neurons from SNCA-3x and isogenic controls (iso ctrl).
(E) Western blot of soluble and insoluble α-syn (antibody C20). Quantification is shown on the right as a fold change to each parental SNCA-3x line. Irrelevant lanes were cropped, indicated by a dashed line.
(F) Western blot of GCase maturation in day 90 SNCA-3x and isogenic controls. Right: quantification from each patient line (3x-1, 2, 4) and isogenic control (n = 2 per line).
(G) GCase western blot of endoglycosidase H (Endo H) digested lysates of day 120 3x-2 neurons. Right: quantification of endo H-resistant GCase.
(H) Live-cell lysosomal GCase activity of 3x-1 and isogenic controls.
For all quantifications, values are mean ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001; ns, not significant, using Student’s unpaired t test.
We next generated and characterized isogenic controls of SNCA-3x iPSC lines by targeted disruption of the SNCA gene using previously established CRISPR-Cas9 constructs (
) (Figures S2D–S2F). We found a 50% decrease in SNCA expression in the 3x-1 isogenic control (Figure 1D), corresponding to a 70% decline in α-syn protein that is comparable to healthy controls (Figure 1E). Lines 3x-2 and 3x-4 exhibited a 75% reduction in SNCA mRNA (Figure 1D) and no detectable α-syn protein (Figure 1E). Moreover, no insoluble α-syn was detected in any of the isogenic control lines (Figure 1E). Analysis of DA neuron markers showed that α-syn reduction did not affect neural differentiation (Figure S2F), consistent with in vivo studies (
). α-Syn reduction improved GCase maturation by reducing the amount of immature GCase while promoting mature GCase (Figure 1F). Improved GCase maturation was validated by endoglycosidase H (Endo H) digestion, which only cleaves glycans from immature GCase forms (Figure 1G), and increased lysosomal GCase activity (Figure 1H). These data validate previous findings in novel iPSC-derived synucleinopathy models and indicate that wild-type GCase trafficking and activity is reduced by α-syn accumulation.
Immature GCase aggregates in the ER of patient midbrain neurons and synucleinopathy brains
The accumulation of immature proteins in the ER can overwhelm the folding machinery, leading to protein misfolding (
). Since immature forms of GCase accumulate in SNCA-3x DA neurons, we hypothesized that α-syn-induced trafficking disruptions may result in GCase instability, misfolding, and aggregation. To test this, lysates from SNCA-3x DA neurons were sequentially extracted and analyzed by western blot. We found that the proportion of aggregated, immature GCase in Triton-X-100-insoluble fractions was elevated in SNCA-3x DA neurons compared to isogenic controls (Figure 2A). This was confirmed in a distinct synucleinopathy patient model expressing A53T α-syn that was previously characterized (
(A) Western blot of Triton X-100 insoluble fractions from day 90 SNCA-3X and isogenic control DA neurons. 3x-4 is shown as a representative example. Right, quantification of % insoluble GCase from combined lines 3x-1, 3x-2, and 3x-4 (n = 3–4 from each), expressed as fold change.
(B) Insoluble GCase from a PD line expressing A53T α-syn and its isogenic control at day 90 was analyzed as in (A).
(C) GCase western blot of 1% sarkosyl-insoluble extracts from the frontal cortex of controls and synucleinopathy brains (DLB, dementia with Lewy bodies; DLB + AD, DLB with Alzheimer’s disease [AD] pathology). Right: quantification of insoluble GCase/Coomassie, grouped by similar post-mortem intervals (PMI) (gray, PMI = <10 h; black, PMI < 20 h; asterisk, PMI < 30 h). See Table S1 for details.
(D) Quantification of insoluble GCase from ER microsome fractions extracted from the cingulate cortex of controls and idiopathic PD brains. See Table S2 for details.
For all quantifications, values are the mean ± SEM, ∗p < 0.05; ∗∗p < 0.01, using Student’s unpaired t test (A, B, and D) or ANOVA with Dunnett's T3 test (C). In (B) and (C), irrelevant lanes were cropped, indicated by a dashed line.
To determine if insoluble GCase could occur from general perturbations in ER-Golgi trafficking that are independent of α-syn, we treated wild-type or isogenic control neurons with thapsigargin (Tg) to induce ER stress and perturb ER-Golgi trafficking. Although Tg induced ER stress, as shown by increased GRP78 expression, we found no evidence of insoluble GCase accumulation (Figure S3A). Additionally, we assessed GCase aggregation in the brains of LIMP2 knockout mice, since LIMP2 is required for ER-Golgi trafficking of GCase (
). While the post-ER forms of GCase were depleted as expected, immature GCase did not accumulate but was instead depleted in LIMP2−/− mice (Figure S3B). This is consistent with previous studies showing that LIMP2−/− reduces GCase levels through aberrant secretion. Collectively, our studies indicate that GCase accumulates into insoluble species selectively upon α-syn-induced inhibition of ER-Golgi trafficking, but not upon general ER stress induction or LIMP2 knockout.
To determine if GCase misfolds and aggregates in vivo, we compared the levels of GCase in 1% sarkosyl-insoluble fractions from brains of patients with either DLB or DLB with co-existing Alzheimer’s disease (AD) pathology. In age-matched healthy control brains, we detected low levels of insoluble GCase that migrated at 55 kDa likely representing the non-glycosylated immature protein, as well as GCase fragments that migrated between 42 and 48 kDa (Figure 2C). Even though we observed some variability between control brains, comparison with age and post-mortem interval (PMI)-matched synucleinopathy brain showed a 1.8-fold elevation of insoluble GCase in DLB brain and nearly a 4-fold elevation in DLB + AD brain when normalized to total protein (Figure 2C; Table S1). Analysis of ER microsomes from idiopathic PD brain indicated that GCase aggregates occur in the ER in vivo (Figure 2D; Table S2). We also assessed the solubility of two additional hydrolases to address selectivity, including cathepsin D and hexosaminidase B. We found that insoluble immature forms of cathepsin D also accumulated in DLB brain, but hexosaminidase B was only found in the soluble fraction, with no changes in the total levels observed between control and disease (Figures S3C and S3D; Table S3). These data indicate that perturbations in trafficking lead to the accumulation of aggregated, insoluble hydrolases in the ER of synucleinopathy patient brain.
ER fragmentation in SNCA-3x DA neurons that accumulate immature wild-type GCase
We hypothesized that accumulation of aggregated GCase in the ER would trigger the ER stress response. The UPR normally responds to misfolded proteins by expansion of the ER compartment and upregulation of ER chaperones to accommodate for the added protein load (
). Examination of ER morphology by electron microscopy (EM) indicated that SNCA-3x neurons unexpectedly did not exhibit ER expansion but instead showed a decrease in total ER area relative to isogenic controls (Figure 3A), with shorter, fragmented ER tubules (Figure 3A). In contrast, GD neurons that express and retain mutant GCase in the ER (GBA1 N370S/84GG) demonstrated a severely dilated ER, consistent with an activation of the UPR (Figure 3A). These data indicate that the ER fails to accommodate for the accumulation of misfolded aggregated GCase by ER expansion, suggesting that SNCA-3x DA neurons may lack the ability to initiate a UPR response.
Figure 3Characterization of ER morphology and ER stress response in SNCA-3x midbrain DA neurons
(A) Electron microscopy (EM) analysis showing representative ER ultrastructure of day 65 SNCA-3x (3x-2), mutant GBA1 (N370S/84GG), and control DA neurons. Examples of ER segments are highlighted in red. The white boxes show zoomed-in regions of the images below. N, nucleus; ∗, ER. Scale bar, 500 nm. Right: quantifications of ER length and area (n = 3–9 cells per line; each data point on the scatterplot indicates a measured ER segment).
(B) Western blot of ER chaperones in day 90 SNCA-3X, mutant GBA1 (N370S/84GG and L444P/L444P), and control DA neurons (n = 2–5 for each line). Right: quantification of SNCA-3x lines (3x-1, 3x-2, 3x-4), mutant GBA1 lines, and controls.
(C) Quantitative RT-PCR analysis of GRP78 and CANX mRNA.
(D) Quantitative RT-PCR analysis of GBA1, GRP78, and GRP94 mRNA in day 90 control and SNCA-3x (3x-2) neurons infected to overexpress L444P GBA1 (MOI = 1; dpi 14 days).
(E) Western blot of 3x and control neurons after treatment with vehicle (veh) or 30 nM thapsigargin (Tg) for 24 h. Right: quantification from 3x-1, 3x-2, 3x-4, and controls. Irrelevant lanes were cropped, as indicated by a dashed line.
For all quantifications, values are mean ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; ns, not significant, using Student’s unpaired (B and C) or paired t test (E) or ANOVA with Tukey’s post hoc test (A, B, and D).
). Compared to isogenic control lines, we observed a mild elevation in all three chaperones that ranged from 10% to 25% (Figure 3B). In comparison, GD-derived DA neurons carrying either the N370S/84GG or L444P/L444P mutation in GCase showed a more pronounced increase (∼25%–60%) in GRP78 and CANX relative to the healthy control, while GRP94 levels were elevated by ∼25% only in the GBA1 L444P/L444P mutant (Figure 3B). The increased levels of GRP78, GRP94, and CANX in GD neurons compared to SNCA-3X DA neurons are likely due to the destabilizing effect of the GCase mutations and are consistent with previous findings in fibroblast cultures (
The dramatic ER fragmentation phenotype and accumulation of aggregated immature GCase prompted us to examine UPR signaling pathways in more detail. We measured XBP1-S, a transcription factor that upregulates ER stress machinery upon IRE1 stimulation (
) and the expression of its downstream transcriptional targets. Using two independent assays, we found no increase of XBP1-S in SNCA-3x DA neurons compared to controls (Figure S4A; Figure S4B, left). Moreover, the mRNA of GRP78 was reduced and CANX was not changed in SNCA-3x DA neurons, indicating the absence of a UPR-induced transcriptional response (Figure 3C). To determine if the UPR could be triggered in SNCA-3X DA neurons by dramatic overexpression of misfolded GCase, we expressed the L444P mutant GCase by lentiviral infection. While expression of L444P GCase induced the upregulation of GRP78 and GRP94 mRNA in control neurons, SNCA-3X DA neurons showed no response (Figure 3D). This indicates that PD patient neurons fail to sense or transmit misfolded protein stress signals to initiate the UPR.
We next determined if the UPR could be activated in SNCA-3x DA neurons by using established chemical ER stressors that induce the UPR through pleiotropic effects. Tg and brefeldin A (BFA) activate the UPR through either disturbing calcium homeostasis, or directly block ER-Golgi trafficking machinery, respectively (
Brefeldin A, thapsigargin, and AIF4- stimulate the accumulation of GRP78 mRNA in a cycloheximide dependent manner, whilst induction by hypoxia is independent of protein synthesis.
). We found that Tg and BFA induced an upregulation of XBP1-S (Figure S4B, right), and significantly increased mRNA / protein levels of ER chaperones in SNCA-3x DA neurons and α-syn overexpressing cell lines (Figures 3E and S4C–S4F). XBP1-S mRNA and ER chaperone mRNA / protein levels were increased to a similar degree in both control and SNCA-3x neurons (Figures 3E, S4B, S4C, and S4F). We next assessed the PERK pathway of the UPR by measuring eIF2α, a eukaryotic initiation factor which upon phosphorylation by PERK leads to global translational attenuation (
). We did not observe baseline elevation of phospho-eIF2α in patient neurons (Figure S4F); however, treatment with Tg increased phospho-eIF2α to a similar degree in both isogenic controls and SNCA-3x DA neurons (Figure S4F). Collectively this indicates that while the UPR is capable of activation by chemical ER stressors that broadly affect the UPR, PD neurons are specifically deficient in recognizing and responding to misfolded proteins in the ER.
Misfolded proteins in the ER are normally recognized by quality control machinery and eliminated by ERAD, which is mediated by EDEM1 (ER degradation-enhancing α-mannosidase-like protein 1). EDEM1 is a lectin-containing adaptor protein that removes misfolded glycoproteins from the CANX folding cycle and delivers them to the cytosol for proteosomal degradation (
). We measured EDEM1 expression in SNCA-3x lines and found no change in mRNA or protein levels (Figures 4A–4C). Since variability was observed in the levels of EDEM1 protein between culture samples of SNCA-3x lines, we correlated EDEM1 and α-syn protein levels in patient neurons and found a significant negative relationship (Figure 4D). This suggests that samples with abundant α-syn pathology have reduced EDEM1 levels and possibly compromised ERAD. Consistent with this, we found no change in the levels of wild-type GCase upon proteosomal inhibition of SNCA-3x DA neurons, suggesting that the misfolded/aggregated protein is not cleared through ERAD (Figure 4E). In contrast, GD neurons showed a dramatic upregulation of EDEM1 compared to both healthy controls and SNCA-3x lines (Figures 4B and 4C) and significant elevation of GCase protein upon proteosomal inhibition (Figure 4F). These data indicate that despite retaining immature misfolded GCase in the ER, the EDEM1-ERAD pathway is not activated in SNCA-3x DA neurons, while in GD neurons, elevation of this pathway leads to elimination of mutant GCase.
Figure 4Analysis of EDEM expression and ERAD in SNCA-3x midbrain DA neurons
(A) Quantitative RT-PCR analysis of EDEM1 mRNA in day 90 SNCA-3x and isogenic controls (combined 3x-1, 2, 4, and the corresponding isogenic controls; n = 4 per line).
(B) Representative western blot of EDEM1 in day 90 SNCA-3x, GBA1 mutant, and control DA neurons.
(C) Western blot quantification of EDEM1 protein in SNCA-3x (combined 3x-1, 2, and 4), GBA1 mutant, and control DA neurons.
(D) Correlation analysis of EDEM1 and α-syn protein obtained from western blots from different biological replicates of SNCA-3x patient DA neurons.
(E) Western blot of day 90 DA neurons treated with 50 nM epoxomicin (Epox) for 24 h. Right, quantification of GCase and ubiquitin (combined from 3x-1, 2 and isogenic controls; n = 3 per line). Coomassie, β3-tubulin and GAPDH were used as loading controls.
(F) GCase western blot of Epox-treated GBA1 mutant DA neurons. Quantification is shown on the right.
For all quantifications, values are mean ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; ns, not significant, using Student’s unpaired (A, C, E, and F) or paired t test (E) or ANOVA with Tukey’s post hoc test (C).
α-Syn associates with ER chaperones in SNCA-3x DA neurons
Although α-syn is known to be a synaptic protein under physiological conditions, immunofluorescence analysis indicated its accumulation at the cell body in SNCA-3x DA neurons (Figures S2A and S2B) (
). To determine if α-syn associates with the ER in SNCA-3x DA neurons, we used super-resolution imaging to examine the colocalization of α-syn with the established ER marker PDI, as well as enrichment of ER microsomes in neurons. We found that α-syn colocalizes with PDI in SNCA-3x patient neurons and is enriched within microsomal fractions (Figures S5A and S5B). In situ proximity ligation assays (PLA) and co-immunoprecipitation in α-syn-overexpressing cell lines showed that α-syn associates with CANX and GRP94 (Figures 5A, 5B, and S5C). PLA analysis validated that endogenously expressed α-syn associates with ER chaperones CANX and GRP94 in SNCA-3X neurons more so than in controls (Figure 5C). Together, these results suggest that α-syn may disrupt ER proteostasis and GCase trafficking through aberrant association and sequestration of ER chaperones.
Figure 5Increased association of α-syn with ER chaperones
(A) Proximity ligation assay (PLA) using anti-CANX and anti-α-syn antibodies in doxycycline (DOX)-inducible H4 cells overexpressing wild-type α-syn via a Tet-off system. Colocalization is indicated by a red dot. Nuclei are stained with DAPI (blue). Right: quantification of PLA signal. Scale bar, 25 μm.
(B) H4 lysates were immunoprecipitated with an anti-CANX antibody followed by western blot for α-syn and CANX. Empty lanes were cropped, as indicated by a dashed line. Normal IgG pulldown is included as a control.
(C) PLA was done in day 70 DA neurons (shown, 3x-2) to assess CANX (top) and GRP94 (bottom) associations with α-syn. The isogenic control for 3x-2 was used to assess background signal. Right: PLA signal in day 70 and 90 neurons was quantified as in (A) using n = 3–4 biological replicates per 3x line. Scale bar, 25 μm.
For all quantifications, values are mean ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001, using Student’s unpaired t test (A) or ANOVA with Tukey’s post hoc test (B).
Synergistic activation of ER proteostasis and trafficking rescues lysosomal function and reduces α-syn
We next determined if increasing ER chaperone function could rescue lysosomal GCase activity. Previous studies showed that ER proteostasis in GD can be improved by blocking ryanodine receptors (RyRs) that mediate calcium efflux from the ER, thereby increasing CANX function (
). We selected the RyR inhibitor diltiazem (DILT), since it is a US Food and Drug Administration (FDA)-approved treatment for high blood pressure and angina. Treatment of SNCA-3x neurons with 25 μM DILT abrogated the buildup of insoluble GCase while concomitantly elevating soluble GCase starting at 2 weeks and continuing to 8 weeks of treatment (Figure 6A). Although DILT mainly increased the solubility of immature forms of GCase, we observed a slight elevation in post-ER forms, indicating a mild improvement in maturation (Figure 6B). DILT also improved GCase protein levels and maturation in control DA neurons, suggesting that enhancement of the folding pathway can be achieved in neurons under physiological conditions (Figure S6A). DILT enhanced properly folded, functional GCase, as demonstrated by increased GCase activity in whole-cell lysates that include both ER and post-ER forms (Figure 6C, left). Despite this, the in situ assay that measures GCase activity within lysosomes of living neurons indicated no change, and western blot analysis showed that α-syn was also unchanged (Figure 6C right; Figure 6D). We validated that DILT could enhance chaperone function by measuring the binding activity of CANX to N-glycosylated proteins using the lectin concanavalin A (Con-A) (Figure 6E). Taken together, these results suggest that enhancing ER proteostasis with DILT can promote functional, soluble forms of GCase, but cannot improve lysosomal function in a sufficient manner to reduce α-syn.
Figure 6Enhancement of ER proteostasis improves GCase solubility in SNCA-3x DA neurons but fails to reduce α-syn
(A) GCase solubility analysis by western blot of day 90 SNCA-3X DA neurons treated with vehicle (veh) or 25 μM diltiazem (DILT) for 2 or 8 weeks. Representative blot image is from 8 week 3x-1 DA neurons. Right: quantification from combined 3x lines.
(B) Western blot of GCase in day 90 SNCA-3x DA neurons treated with DILT for 2 weeks. Right: quantification of GCase maturation from lines 3x-1, 2, and 4 (n = 3 per line).
(C) Left: whole-cell GCase activity of SNCA-3x lysates from (B). Right: live-cell lysosomal GCase activity in 2-week DILT-treated day 90 SNCA-3x DA neurons.
(D) Western blot of α-syn levels (the same blot from B was reprobed for α-syn and quantified on the right). GAPDH is a loading control.
(E) CANX activity was assessed in H4 α-syn cells treated with DILT (5 days) by precipitation of N-glycosylated proteins with concanavalin A (Con-A) followed by CANX western blot. GCase and GAPDH are positive and negative controls, respectively. Right: quantification of normalized CANX levels from the Con-A pull-down.
(F) Analysis of GCase solubility was done as in (A) in day 90 SNCA-3x DA neurons treated with scrambled (scrb) or RyR3 shRNA knockdown (KD) lentivirus (MOI = 5, dpi 2 weeks),
For all quantifications, values are mean ± SEM. ∗p < 0.05; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; ns, not significant, using Student’s unpaired t test.
To confirm that enhancing ER proteostasis and wild-type GCase can be improved by RyR inhibition, we treated α-syn-overexpressing cell lines and SNCA-3x DA neurons with two additional RyR inhibitors, dantrolene (DANT) and 1,1′-diheptyl-4,4′-bipyridinium (DHBP) (
). DANT and DHBP treatment elevated soluble GCase levels in cell models, although not as robustly as DILT (Figures S6B–S6D). When higher concentrations or longer incubation periods were attempted, we observed cell toxicity, consistent with previous findings (
). We confirmed that improved GCase solubility occurred through RyR3 by knockdown (KD) with small hairpin RNA (shRNA) constructs. Quantitative RT-PCR analysis showed a 50% KD of RyR3 mRNA (Figure S6E), resulting in increased solubility of GCase in both cell lines and SNCA-3x DA neurons (Figures 6F and S6F). DILT had no effect on GCase solubility in RyR3 KD cells (Figure S6F), indicating that DILT acts to improve GCase through RyR3 receptors on the ER. Analysis of GCase maturation in RyR3 KD cells showed a mild improvement in cell lines similar to DILT treatment, and no change in SNCA-3x DA neurons (Figures 6F and S6F). DILT caused a mild elevation of GCase maturation in RyR3 KD cell lines, a result that may have occurred from the inhibition of other RyR isoforms (Figure S6F). These data show that RyR inhibition can improve GCase proteostasis in the ER but has little effect on increasing GCase trafficking in patient neurons.
The failure to rescue lysosomal GCase activity by RyR inhibition suggests that factors downstream of the ER may inhibit hydrolase trafficking. Our previous work showed that α-syn inhibits GCase trafficking by preventing ER-Golgi vesicle fusion through impeding the function of the SNARE protein ykt6 (
). Therefore, we next determined whether enhancing trafficking, together with ER proteostasis, could cooperate to rescue lysosomal GCase. We found that treatment with both FTI (LNK-754) and DILT results in a significant increase of GCase maturation compared to each compound alone (Figures 7A and S7A–S7C). This effect was not additive but synergistic, since the increase caused by FTI + DILT was greater than the sum of each individual compound alone at 4 weeks of treatment (Figure S7A). This is consistent with the notion that each compound targets a distinct portion of the proteostasis pathway. EM analysis indicated that FTI + DILT treatment substantially improved ER segment length and area, suggesting that increased movement of GCase out of the ER improves ER morphology (Figure 7B). FTI + DILT treatment also elevated functional, soluble forms of GCase in both whole-cell lysates and live-cell in situ lysosomal assays (Figure 7C) and synergistically reduced both soluble and insoluble α-syn in patient neurons and cell lines (Figures 7D, 7E, and S7B–S7D). We confirmed our findings genetically, by combining RyR3 KD with FTI, or DILT with expression of ykt6-CS that cannot be farnesylated (
). These combinations effectively enhanced GCase trafficking and reduced α-syn better than either treatment alone (Figures S7E–S7H). Finally, we sought to determine if the combination treatment could improve GCase proteostasis and reduce α-syn levels in patient iPSC neurons that express GBA1 mutations. FTI + DILT treatment of midbrain neurons from both GD (N370S/84GG) and GBA-PD (N370S/wt) patients significantly increased GCase maturation and reduced α-syn levels compared to each treatment alone (Figure S8). This suggests that combined trafficking and ER proteostasis enhancers could provide benefit in both synucleinopathies and GD.
Figure 7Rescue of ER fragmentation and lysosomal GCase by synergistic enhancement of ER proteostasis and trafficking
SNCA-3X DA neurons were treated with vehicle, 5 nM farnesyl transferase inhibitor (FTI), 25 μM diltiazem (DILT), or a combination of FTI + DILT.
(A) GCase western blot of day 90 3x-1 DA neurons treated with FTI + DILT for 2 weeks. Coomassie is shown as a loading control. Quantification of combined data from 3x-1, 2 is shown to the right.
(B) EM analysis of day 90 SNCA-3X DA neurons (shown, 3x-2) treated with DMSO vehicle or FTI + DILT for 2 weeks. Example ER segments are highlighted in red. Quantification of ER morphology is shown on the right. Scale bar, 1 μm.
(C) Left: whole-cell GCase activity of treated 3x-1 and 3x-2 lysates from (A). Right: analysis of live-cell lysosomal GCase activity of FTI + DILT-treated 3x-2 DA neurons and compared to healthy controls (Est. Ctrl).
(D) Western blot of soluble α-syn. The same membranes from (A) were sequentially probed with syn211 and then syn303 anti-α-syn antibodies. GAPDH is shown as a loading control. Right: α-syn quantification.
(E) Western blot of insoluble α-syn in day 90 SNCA-3X DA neurons treated with FTI + DILT for 2 weeks. The same membranes were sequentially probed with syn211 then syn303. Irrelevant lanes were cropped, as indicated by a dashed line. Quantification is shown to the right.
For all quantifications, values are mean ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001, using Student’s unpaired t test (B and C right) or ANOVA with Tukey’s post hoc test (A, D, C left, and E).
We identify a novel pathogenic pathway induced by α-syn accumulation at the ER characterized by severe ER fragmentation, compromised folding capacity, and aggregation of lysosomal GCase (Figure 8). Other studies using transgenic overexpression models of α-syn have documented the association of α-syn with ER components (
) and are consistent with our findings in PD patient neurons. Our data indicate that α-syn likely interacts with ER chaperones that are important for maintaining GCase folding (Figures 5 and S5). It is possible that the interaction occurs from increased abundance of α-syn at the cell body. However, our previous studies have shown that α-syn does not interact with other ER-Golgi components located at the cell body (
), suggesting that the ER chaperone association is somewhat selective. Increased α-syn at the ER may overwhelm the proteostasis capacity, sequestering chaperones away from their normal function. α-Syn can also directly perturb protein trafficking machinery downstream at the Golgi (
), which likely slows the export of cargo from the ER, resulting in the accumulation of immature proteins. These data highlight the negative effects of α-syn on multiple branches of the proteostasis pathway.
Figure 8Proteostasis dysfunction and rescue in PD patient neurons
Accumulation of α-syn at the ER is associated with ER fragmentation, accumulation and aggregation of immature β-glucocerebrosidase (GCase), and failure of UPR activation (1 and 2). Enhancing ER folding capacity promotes the formation of soluble, active GCase, while combined treatment with protein trafficking enhancers rescues lysosomal function and reduces pathological α-syn (3).
A surprising consequence of α-syn-induced trafficking disruptions was the aggregation of immature GCase into insoluble species (Figure 2). While several loss-of-function mutations in lysosomal hydrolases can cause lysosomal storage diseases (
), we document a unique example where lysosomal dysfunction occurs through the misfolding and aggregation of wild-type immature GCase. Other lysosomal diseases that are caused by mutations in trafficking machinery, including I-cell disease or LIMP2 depletion that occurs in acute myoclonus renal failure (AMRF), do not show accumulation of immature hydrolases but instead are characterized by aberrant secretion (Figure S3B) (
). Therefore, the aggregation of immature hydrolases induced by α-syn may be unique to synucleinopathies. The lack of chaperone induction and ERAD activity in SNCA-3x DA cultures (Figure 4) likely contributes to the aberrant accumulation and destabilization of GCase in the ER. GCase may also be particularly susceptible to aggregation as a membrane-associated enzyme, since aberrant exposure of hydrophobic patches during prolonged folding cycles may promote its self-association into insoluble aggregates. We also found that immature cathepsin D accumulates into insoluble species, but not hexosaminidase B (Figure S3). This indicates that while not all hydrolases are susceptible to aggregation, the effect is not specific to GCase. One other study has shown that a rare Tay-Sachs disease point mutation in β-hexosaminidase results in the accumulation of an insoluble enzyme precursor, preventing its trafficking to the lysosome (
Synthesis of beta-hexosaminidase in cell-free translation and in intact fibroblasts: an insoluble precursor alpha chain in a rare form of Tay-Sachs disease.
). It will be of interest in future studies to examine hydrolase aggregation in proteinopathies beyond PD and DLB that are characterized by lysosomal dysfunction.
Unexpectedly, we did not observe activation of the UPR in SNCA-3x neurons, which normally prevents protein aggregation by expanding ER volume and upregulating folding machinery. Other synucleinopathy models generated by transgenic α-syn overexpression or patient-derived iPSC cortical models exhibited signs of UPR activation (
). The pathological stage at which the models were examined is an important consideration. In some studies, ER stress markers are only elevated during the latest stages of pathology (
). We focused on the early stages of disease to capture phenotypic events that occur prior to lysosomal dysfunction and neurodegeneration. Our findings suggest that PD neurons are deficient at detecting and responding to misfolded proteins (Figure 8), as indicated by lack of UPR induction after overexpression of L444P GCase in SNCA-3x neurons. Conversely, activation of the UPR occurs in GD neurons that endogenously express mutant GCase, or in L444P lenti-overexpressing wild-type neurons, demonstrating that the UPR sensors are capable of detected misfolded GCase. Therefore, it is likely that α-syn impedes either the ability of the UPR sensors to detect misfolded GCase or the downstream signal transduction required to activate UPR genes. While SNCA-3x patient neurons could not respond to misfolded GCase, they could react to chemical ER stress inducers that induce nonspecific, pleiotropic effects on the UPR. This indicates that while the UPR pathway is not completely disabled, SNCA-3x neurons exhibit a selective deficiency in handling misfolded proteins in the ER. The mechanisms that conceal the folding status in the ER require further study but may involve aberrant interactions of α-syn with ER chaperones and stress sensors.
Recent work has shown that the UPR can be harnessed to provide protection in neurodegenerative diseases (
). Since our data indicate that the UPR is not completely disabled, it is possible that enhancing the pathway will provide therapeutic benefit. Such strategies could restore ER proteostasis by stimulating XBP1-S-mediated ER compartment expansion and elevation of chaperones, providing a more conducive environment for GCase folding while preventing the growth of GCase aggregates. These methods would have to avoid maladaptive UPR signaling pathways that promote apoptosis from prolonged stimulation (
Our rescue studies indicate that current therapeutic strategies focused on enhancing single branches of the proteostasis pathway may be insufficient to completely rescue PD pathology. Previous work from our group and others showed that FTIs can enhance trafficking and activate lysosomes, reducing aggregated proteins in vivo (
). Our current work indicates that combining this strategy with ER proteostasis enhancers is more efficient at rescuing lysosomal GCase and reducing pathological α-syn (Figures 7 and S7). Once in the lysosome, active GCase could reduce α-syn by degrading glycosphingolipid substrates that interact and stabilize toxic α-syn (
). This strategy was also effective in patient neurons that harbor GBA1 mutations (Figure S8), indicating a potential to translate these treatments to GD and GBA-PD. Therapeutic enhancement of multiple proteostasis pathways may provide optimal benefit in PD, given the pleiotropic deleterious effects of α-syn accumulation in multiple subcellular locations. Furthermore, combining two treatments that target distinct cellular pathways may enable administration of lower doses of each drug, which would limit compound toxicity if these treatments should progress to the clinic.
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Joseph R Mazzulli ([email protected]).
Materials availability
All unique/stable reagents and cell lines generated in this study are available from the Lead Contact, Joseph R Mazzulli ([email protected]) with a completed Materials Transfer Agreement.
Experimental model and subject details
Human H4 neuroglioma cell culture
Human H4 neuroglioma cells were stably transfected to overexpress wild-type α-syn under the control of a tetracycline-inducible promoter via a Tet-off system and described previously (
). α-Syn expression was turned off by the addition of 1ug/ml doxycycline (DOX) (Sigma), a tetracycline analog, for a minimum of 3 days. Cells were cultured in Optimem media with 5% heat-inactivated FBS, 0.2 mg/ml geneticin, 0.2 mg/ml hygromycin B, and 1% penicillin / streptomycin (Thermo Fisher Scientific).
iPSC model generation, characterization and culture methods
Reprogramming and culturing of human induced pluripotent stem cells (iPSCs)
B-lymphocytes from healthy controls and PD patients that carry a triplication in the SNCA genomic region were obtained from the Coriell NINDS and NIGMS Human Genetic Cell Repositories: GM15845 (Ctrl), GM15010 (3x-1), ND00196 (3x-2), ND00139 (3x-4), ND34391 (Est. 3x). Phenotypic and genotypic data of these subjects is available on https://www.coriell.org. See key resources table for more details, including information on Est. Ctrl, SNCA A53T mutant, and GBA1 mutant iPSC lines (N370S/84GG and L444P/L444P). The B-lymphocytes were reprogrammed into iPSCs by transfection with non-integrating episomal plasmids containing Oct3/4 (Addgene: pCXLE-hOCT3/4-shp53-F), L-Myc (Addgene: pCXLE-hUL), and Sox2 and Klf4 (Addgene: pCXLE-hSK). All iPSCs were maintained in mTeSR1 media on matrigel-coated plates.
Pluripotency analysis of reprogrammed iPSC cells
I. Immunofluorescence analysis of pluripotency markers
Cells plated on glass coverslips were fixed in 4% paraformaldehyde (Polysciences, Inc.) for 15 minutes, permeabilized with 0.3% Triton X-100 (Sigma) in PBS for 30 minutes, and blocked with 2% bovine serum albumin (BSA) (Roche) in Triton-PBS for 30 minutes to prevent non-specific antibody binding. Primary antibodies (Sox2, Tra-1-60, Oct4, SSEA4, Nanog) were added overnight, followed by incubation with secondary antibodies (Alexa Fluor 488 Goat anti-rabbit IgG and Alexa Fluor 568 Goat anti-mouse IgG) for 1 hour. The cells were then washed three times with Triton-PBS and mounted onto microscope slides with DAPI mounting media.
II. PCR analysis of reprogramming factor transgenes
Forward and reverse PCR primers for each of the reprogramming factor transgenes (Oct3/4, Sox2, Klf4, L-Myc) were designed so that the PCR product will span both the transgene and the plasmid backbone, as indicated in the schematic of Figure S1C. See Table S4 for list of primers. The PCR was performed with Taq polymerase (NEB #M0273L) and 20ng of genomic DNA using the following cycling conditions: Initial denaturation 95 C 3min ; 40 cycles: Denaturation 95C 30 s, Annealing 60C 30 s, Extension 68C 30 s ; Final extension 68C 5min ; 4C hold. The PCR products were run on 1.5% agarose gel for 45min at120V and imaged on a Chemicdoc imaging system (Biorad).
Quantitative RT-PCR
Total RNA was isolated from cells in a 24 or 12 well format using an RNeasy Mini Prep kit (QIAGEN). cDNA was synthesized by reverse transcriptase PCR (RT-PCR) using the RevertAid First Strand cDNA synthesis kit (Thermo Fisher Scientific). Quantitative PCR was performed on the Applied Biosystems 7500 Fast system using the cDNA and pre-designed TaqMan-primer probes for the target genes. The target mRNA expression was quantified relative to GAPDH or β-actin using the delta-delta-Ct method, and represented as a fold change.
Copy number analysis of SNCA and puromycin
Genomic DNA was extracted from a 12 well plate of iPSCs using the DNeasy Blood and Tissue Kit (69504, QIAGEN). Quantitative PCR was performed using default cycling conditions on the Applied Biosystems 7500 Fast system with 100ng genomic DNA and pre-designed TaqMan probe for SNCA (Hs04791950_cn) or custom probe for puromycin (gi763524_CCN1F1Y). The copy number of each gene was quantified relative to reference RPPH1 copy number assay (4401631, Applied Biosystems). The analysis was performed using ddCt method and expressed as fold change.
Fluorescence in situ hybridization (FISH) analysis
To confirm the SNCA copy number in the reprogrammed iPSCs, fluorescent probes targeting SNCA (4q22.1; R: red) and a control region (4p16.3; G: green) were used for FISH analysis. The assay was performed as a service provided by Cell Line Genetics, Inc (https://www.clgenetics.com).
Differentiation of iPSCs into midbrain dopaminergic neurons
The iPSCs were differentiated into midbrain dopaminergic neurons using a well-established dual SMAD inhibition protocol (
). Neurons were cultured in neurobasal SM1 media (Thermo Fisher Scientific) containing NeuroCult SM1 supplement (StemCell Technologies), 1% penicillin / streptomycin, and 1% L-glutamine (GIBCO). Neurons were aged to 60-90 days for each experiment as indicated in the text or figure legends.
Dual nickase CRISPR/Cas9 strategy and selection of iPSC clones
A pair of guide RNAs (guide RNA 1: 5′-AGCAGCCACAACTCCCTCCTTGG-3′; guide RNA 2: 5′- TGAGAAAACCAAACAGGGTGTGG-3′) were designed using the Optimized CRISPR design tool (https://zlab.bio/guide-design-resources), and used to direct D10A mutant Cas9 to produce nicks within Exon 2 of the SNCA gene. A PITX3-2A-eGFP-PGK-Puro plasmid (Addgene) encoding a puromycin resistance cassette driven by a phosphoglycerate kinase (PGK) promoter was used as a template for homologous recombination (HR) and as a positive selection marker. The gRNAs were cloned into a Cas9-nickase plasmid PX335 (Addgene) and transfected into iPSCs using Lipofectamine 3000 (Thermo Fisher Scientific) along with a puromycin-containing HR plasmid. Two days following the transfection, iPSCs were cultured in 1ug/ml puromycin containing media for several weeks. To confirm that the puromycin cassette was appropriately inserted in the targeted SNCA Exon 2 region, puromycin resistant clones were selected and genomic DNA was extracted and analyzed via PCR using the following primers: 5′ F: CATAAAATCTGTCTGCCCGCTCTC, 5′ R: GTGGGCTTGTACTCGGTC; 3′ F: CTTCTACGAGCGGCTCGGCTT, 3′ R: TGTGGTCATCCTCCACCTGACT. Puromycin copy number analysis and sequencing were also performed on selected clones.
Analysis of off-target effects using the T7EI cleavage assay
Genomic DNA was amplified using primers for each off-target gene (see key resource table for list of primers). The PCR products were then denatured and allowed to re-anneal using a thermal cycler with the following settings: 95°C for 10 minutes, 95-85°C (ramp rate 2°C/sec), and 85-25°C (ramp rate 0.2°C/sec). The hybridized product was then digested with T7 Endonuclease I for 1 hour at 37°C, and analyzed on an agarose gel along with a positive control (Genecopoeia).
Method details
Biochemistry and Molecular Biology
Sequential protein extraction and western blotting analysis
Cells were harvested in 1X PBS and pelleted by centrifugation at 400xg for 5 minutes. The cell pellets were extracted via homogenization in 1% Triton lysis buffer containing protease inhibitor cocktail (PIC) (Roche), phenylmethylsulfonyl fluoride (PMSF) (Sigma), sodium orthovanadate (Na3VO4) (Sigma) and sodium fluoride (NaF) (Sigma). The Triton extracted lysates were freeze-thawed three times and ultracentrifuged at 100,000xg for 30 minutes at 4°C. The Triton-insoluble pellets were further extracted in 2% SDS lysis buffer containing PIC via boiling for 10 minutes, followed by sonication and then ultracentrifugation at 100,000xg for 30 minutes at 22°C. The protein concentrations of the Triton and SDS fractions were measured via a BCA protein assay kit (Thermo Fisher Scientific) on a plate reader. Extracted protein lysates were boiled in 1X Laemmeli sample buffer containing 2% SDS, loaded on an SDS-PAGE gel, transferred onto a PVDF membrane (Millipore), and post-fixed in 0.4% paraformaldehyde. Membranes were blocked in a 1:1 mixture of 1X TBS and Intercept blocking buffer (Li-Cor Biosciences), followed by overnight incubation with primary antibodies diluted in a 1:1 mixture of 1X TBS-Tween and blocking buffer. The following day, secondary antibodies were added for 1 hour, and the membranes were scanned using a Li-Cor Biosciences infrared imaging system. Quantification of band intensity was done using the ImageStudio software and analysis was performed on Excel and GraphPad Prism. A detailed protocol of this procedure has been published (
). In some blots, irrelevant lanes were cropped out, which is indicated by a dotted line or white space between the lanes.
To quantify insoluble GCase from cell cultures, the intensity from the soluble and insoluble fractions (using Sigma antibody G4171) was normalized to total protein obtained from the Coomassie blue stained gel of the corresponding membrane. Normalized intensities of soluble and insoluble fractions were added to obtain the total GCase signal. The % insoluble GCase was calculated by dividing the insoluble intensity by the total multiplied by 100, then expressed as fold change compared to the control lines or vehicle treated samples. The proportion of insoluble GCase in healthy wild-type cells ranged between 10%–20%. For diltiazem treatment, quantification for the 2-week treatment is combined from day 90 3x-1 (n = 6) and 3x-2 (n = 3) neurons; 8-week treatment is of day 120 3x-1 neurons only (n = 3). For human brain extracts, see below under “STAR Methods, Insoluble hydrolase analysis of synucleinopathy brain tissues.”
Co-immunoprecipitation
H4 cells overexpressing a-syn were extracted in 0.3% CHAPS buffer containing 40mM HEPES pH 7.4, 120mM NaCl, 1mM EDTA, 10% vol/vol glycerol, protease inhibitor cocktail (PIC) (Roche), phenylmethylsulfonyl fluoride (PMSF) (Sigma), sodium orthovanadate (Na3VO4) (Sigma) and sodium fluoride (NaF) (Sigma). 1mg of total lysate was pre-cleared with normal mouse IgG (Santa Cruz) and protein A/G beads that were blocked in 2% BSA. Pre-cleared lysates were incubated with 3ug of CANX antibodies (clone E-10, Santa Cruz) or 3ug of normal mouse IgG rotating end over end, overnight at 4°C. Blocked protein A/G beads were added and incubated for an additional 2 hours, followed centrifugation at 1000 X G, washing 3 times in CHAPS buffer, and elution by boiling in 2X Laemmeli sample buffer. The samples were analyzed by western blot as described above.
Sequential extraction analysis of LIMP2 knock-out mice
LIMP2 knock-out mice have been previously described and characterized (
). Brain tissue was sequentially extracted as described for cell cultures in “Sequential protein extraction and western blotting analysis.” An additional extraction step was added for both Triton and SDS steps to avoid carry over between the fractions. Protein assay was performed by BCA, and 40ug of total protein was loaded per well. GCase solubility was assessed using the anti-GCase antibody from Sigma (G4171), and normalized to total protein obtained from Coomassie blue stained gels of the corresponding membranes.
Live-cell lysosomal GCase activity assay
The procedure and analysis method for the activity assay has been previously described in detail (
) Briefly, cells were plated in 96-well plates. One day prior to the assay, cells were treated with 1mg/ml cascade dextran blue (Life Technologies) for 24 hours. The next day, the cells were first treated with DMSO or 200nM bafilomycin A1 (Santa Cruz) for 1 hour at 37°C, followed by a 1 hour pulse-chase with 100ug/ml artificial fluorescent GCase substrate, 5-(pentafluoro-benzoylamino) fluorescein di-β-D-glucopyranoside (PFB-FDGluc) (Life Technologies), at 37°C. The fluorescence signal was measured every 30 minutes for the span of 3-4 hours on a plate reader (Ex = 485nm, Em = 530nm, for the GCase substrates; Ex = 400nm, Em = 430nm for cascade dextran blue). For the analysis, the GCase fluorescence signal was normalized to either lysosomal mass by using cascade dextran blue signal or total cell volume by quantifying CellTag 700 staining signal.
In vitro whole-cell lysate GCase activity assay
The procedure and analysis method for the activity assay has been previously described in detail (
). Briefly, 1% BSA and 5ug of Triton-soluble protein lysate treated with or without conduritol-β-epoxide (CBE, an inhibitor specific for lysosomal GCase) (Millipore) were added to GCase activity assay buffer (0.25% w/v sodium Taurocholate, 0.25% Triton X-100, 1mM EDTA, into a citrate/phosphate buffer pH 5.4) to a final volume of 100ul in a 96-well black bottom plate. The samples were incubated with 5mM fluorescent GCase substrate 4-methylumbelliferyl β-glucopyranoside (4-MU-Gluc) (Chem-Impex) for 30 minutes at 37°C, and the reaction was stopped using equi-volume of 1M glycine, pH 12.5. The fluorescence signal was measured on a plate reader (Ex = 365nm, Em = 445). Relative fluorescence units from CBE treated lysates were subtracted from non-CBE treated lysates to obtain the activity of GCase.
Endoglycosidase H (Endo H) digestion
These methods have been described in detail previously (
). To study the subcellular localization and trafficking of GCase between the ER and Golgi, we digested protein lysates with Endoglycosidase H (Endo H) (New England Biolabs). The experimental procedure was performed according to the manufacturer’s instructions. Briefly, 10X Glycoprotein Denaturing buffer was added to 40 μg of protein and the reaction was boiled at 100°C for 10 minutes. Following the denaturation, 10X GlycoBuffer 3 and Endo H enzyme were added, and the reaction was incubated at 37°C for 2 hours. Finally, the samples were boiled at 100°C for 10 minutes after the addition of 5X Laemmli buffer and loaded on a 10% SDS-PAGE gel, followed by western blot analysis. A positive digestion results in a downward shift in the molecular size of GCase after it is subjected to SDS-PAGE. Post-ER (70–74 kDa) and ER (55 kDa) forms of GCase were analyzed using the Endo H digested lane, and used as a measure of GCase trafficking.
Insoluble hydrolase analysis of synucleinopathy brain tissues
Sequential protein extraction was performed on post-mortem frontal cortex brain tissues (obtained from the Northwestern University Alzheimer’s disease pathology core) obtained from controls, DLB, and DLB+AD patients. We employed a 5-step extraction protocol using high salt buffer, 1% Triton X-100, 1% Triton + 30% sucrose (Sigma), 1% sarkosyl (Sigma), and sarkosyl-insoluble extracts. Brain tissues were homogenized in high-salt buffer (HSB) (50 mM Tris-HCl pH 7.4, 750 mM NaCl, 10 mM NaF, 5 mM EDTA) with protease and protein phosphatase inhibitors, incubated on ice for 20 minutes and centrifuged at 100,000 x g for 30 minutes at 4 °C. The pellets were then re-extracted with HSB, followed by sequential extractions with 1% Triton X-100-containing HSB and 1% Triton X-100-containing HSB with 30% sucrose. The pellets were then resuspended and homogenized in 1% sarkosyl-containing HSB, rotated at 4 °C overnight and centrifuged at 100,000 x g for 30 min. The resulting sarkosyl-insoluble pellets were washed once with PBS and resuspended in PBS by brief sonication. This suspension was termed the ‘sarkosyl-insoluble fraction’, which was analyzed by western blot. GCase was probed using Sigma antibody G4171, and the total intensity of the immunoreactive signal from ca. 45kDa to 60kDa was normalized to Coomassie blue staining of the corresponding gel.
Insoluble GCase analysis of ER microsome-enriched idiopathic PD brain tissues
ER microsomes were enriched using subcellular fractionation and the purity of the fractions have been assessed previously (
). Post-mortem cingulate cortex brain tissues obtained from idiopathic PD patients were lysed and homogenized in 0.25M sucrose buffer containing 10mM HEPES (pH 7.4) and 0.01M EDTA, and centrifuged at 6,800 x g for 5 minutes at 4°C to remove nuclei and unbroken cells. The extraction was repeated to wash the pellet. The final supernatants were combined and further centrifuged at 17,000 x g for 10 minutes at 4°C to remove mitochondria. Further centrifugation of the resulting supernatant at 100,000 x g for 1 hour was done to pellet the ER microsome components. Sequential extraction of soluble and insoluble protein from this final pellet was performed using 1% Triton and 2% SDS lysis buffer, respectively, as described above. Insoluble fractions were analyzed via western blot. GCase was probed using Sigma antibody G4171, and the total intensity of the immunoreactive signal shown (from ca. 50 to 64kDa) was normalized to Coomassie blue staining of the corresponding gel.
GBA1 mutation genotyping of human brain samples
Genomic DNA was extracted from 50mg human brain tissue (frontal / temporal cortex) using the PureLink genomic DNA kit (Invitrogen). To amplify the GBA1 gene, 25ng genomic DNA was used as a template for PCR using the following forward and reverse primers, respectively: 5′-TGTGTGCAAGGTCCAGGATCAG-3′ and 5′-ACCACCTAGAGGGGAAAGTG-3′. The PCR products were run on a 1% agarose gel to confirm amplification of the GBA1 gene and to rule out accidental amplification of the GBA1 pseudogene (GBAP). Sequencing of the most common GBA1 mutations (L444P, N370S, E326K) was performed using primers listed in the Key Resource Table, and analysis was done using the Snapgene software.
ER microsome-enrichment of iPSC-derived neurons
ER microsomes were enriched using subcellular fractionation. SNCA-3X and healthy and isogenic control iPSC-derived neurons were gently homogenized in sucrose HEPES buffer (SHB). The homogenate was centrifuged at 6,800 x g for 5 minutes at 4C to remove nuclei and unbroken cells. Following removal of the supernatant (S1), the extraction was repeated using SHB buffer and the second supernatant (S2) was combined with S1. The combined supernatants (S1+S2) were further centrifuged at 17,000 x g for 10 minutes at 4C to remove mitochondria. Further centrifugation of the resulting supernatant (S3) at 100,000 x g for 1 hour at 4C removes the cytosolic components (supernatant S4), leaving the ER microsomes in the third and final pellet, termed P3. The P3 pellet was extracted in 1% Triton lysis buffer and analyzed by western blot.
Semiquantitative RT-PCR analysis of XBP1 mRNA
Using cDNA as the template, human XBP1 mRNA was detected using PCR primers (forward: TTACGAGAGAAAACTCATGGCC; reverse: GGGTCCAAGTTGTCCAGAATGC) specific for both spliced (S; product size 263 bp) and unspliced (U; product size 289) isoforms. The PCR product was analyzed on an agarose gel along with a brefeldin A positive control.
Assessment of calnexin activity by Concanavalin-A pulldown
H4 neuroglioma cells were treated with vehicle or 25uM Diltiazem (Sigma) for 4 days, harvested, and extracted in 0.3% CHAPS lysis buffer (0.3% CHAPS, 40mM HEPES pH 7.4, 120 mM NaCl, 1mM EDTA, 10% v/v glycerol). For pulldown of total N-linked glycosylated proteins, 1500 μg lysate was mixed with 20 μg/ml biotinylated Concanavalin A (CON-A) (Vector Laboratories) and the reaction mixture was incubated overnight at 4°C under gentle rotation. To recover CON-A bound proteins, 25 μl neutrAvidin agarose beads (Thermo Fisher Scientific) were added to the reaction mix and samples were incubated at 4°C for 1 hour. The beads were collected by centrifugation at 2500 x g for 2 min, followed by three washes with lysis buffer. N-glycosylated proteins were eluted by boiling the samples at 95°C for 10 min in 2X Laemmli sample buffer. Samples were analyzed by western blot for calnexin (CANX), GCase, and total N-glycosylated proteins by Coomassie brilliant blue staining. Calnexin activity was indirectly assessed by quantifying CANX levels in CON-A pulled down samples.
Imaging analysis
Immunofluorescence analysis of midbrain neuron differentiation efficiency, α-synuclein accumulation, and thioflavin staining
Neurons were fixed in 4% paraformaldehyde for 15 minutes, permeabilized with 0.1% Triton X-100 in PBS for 30 minutes, and blocked with 2% BSA and 5% normal goat serum (NGS) (Jackson Immuno Research) in Triton-PBS for 30 minutes to prevent non-specific antibody binding. Primary antibodies (anti-α-synuclein LB509, anti-tyrosine hydroxylase (TH), anti-FoxA2, anti-β3-tubulin) were added overnight, followed by incubation with secondary antibodies (Alexa Fluor 488 Goat anti-mouse IgG and Alexa Fluor 568 Goat anti-rabbit IgG) for 1 hour. The cells were then washed three times with Triton-PBS and mounted onto microscope slides with DAPI mounting media. For thioflavin S (Thio S) co-staining, following primary incubation with a-syn, 0.05% Thio S was directly added to cells and incubated for 15 min at RT. Next, cells were washed with a sequence of ethanol steps (twice with 50% ethanol for 20 min each, then once with 80% ethanol for 20 min) and then with Triton-PBS prior to mounting. The Thio S and α-synuclein staining has been described in detail (
). All images were obtained on a Leica confocal microscope, and image analysis was performed using ImageJ.
Measurement of Neuron viability through neurofilament quantification
For this assay, the same cultures used in the live cell activity were used from a 96 well plate. Following the live-cell lysosomal GCase activity assay, the cells were fixed in 4% paraformaldehyde in PBS for 15 minutes, and stained with an anti- neurofilament antibody overnight at 4°C (refer to (
) for details). The next day, IRdye 800CW goat anti-mouse IgG secondary antibody and CellTag 700 stain were added to the wells and incubated for 1 hour, and the plate was scanned on a Li-Cor infrared imaging system.
Electron Microscopy (EM) analysis
Neurons were fixed in 2.5% glutaraldehyde (Electron Microscopy Sciences) in PBS for 30 minutes, and then washed six times with PBS for 5 minutes. Cells were post-fixed with 1% osmium tetroxide (OsO4) (Electron Microscopy Sciences) in PBS for 1 hour, and then washed three times with H2O. Next, cells were dehydrated with ethanol (twice with 50% ethanol for 5 minutes, then twice with 70% ethanol for 10 minutes) and stained with 1% uranyl acetate (Electron Microscopy Sciences) in 70% ethanol for 45 minutes. Cells were further dehydrated with ethanol (once with 70% ethanol, then twice with 90% ethanol for 10 minutes, then three times with 100% ethanol for 10 minutes). To evaporate the ethanol, 100% ethanol was mixed at a 1:1 ratio with an LX112 resin mix containing LX112 (Ladd Research Industries), DDSA (Electron Microscopy Sciences), and NMA (Electron Microscopy Sciences), and added to the cells for 1 hour with the lid off. Next, LX112 resin mix alone was added to the cells for 1 hour. Finally, cells were embedded by combining LX112 resin mix with DMP-30 (Electron Microscopy Sciences) and allowing the resin to solidify overnight at 60°C. Samples were then thin sectioned (∼70nm width) on a UC7 ultramicrotome, as a service provided by the Northwestern University Center for Advanced Microscopy, and viewed on a FEI Tecnai Spirit G2 TEM. For each cell that was imaged via EM, all clearly defined ER regions were analyzed for both length and area using the ‘Measure’ function in ImageJ. The length and ER area (in micrometers) of each individual ER profile were plotted on a graph using GraphPad Prism. 3 to 9 cells per line were quantified and each data point on the scatterplot indicates a measured ER segment.
Proximity Ligation Assay (PLA)
Inducible H4 cells overexpressing a-syn were treated with 1ug/ul DOX for 3 days to turn off α-syn expression. iPSC derived neurons were cultured to day 70 or 90 and separate batches were used for analysis. Cells were plated on coverglass were fixed with 4% paraformaldehyde for 20 minutes at RT. The cells were then washed three times with PBS, permeabilized with 0.3% Triton X-100 in PBS for 1 hour at 4°C, and then blocked with 2% BSA (Roche) and 5% NGS (Jackson Immuno Research) in Triton-PBS for 30 minutes at RT. Interaction between α-syn and ER chaperones was determined via the Duolink In Situ Red Starter Kit Mouse/Rabbit (Sigma). Cells were incubated with primary antibodies (anti-α-synuclein syn211, anti-α-synuclein C20, anti-CANX, anti-GRP94) overnight followed by a 1 hour, 37°C incubation with the PLA probes (secondary antibodies labeled with distinct oligonucleotides) provided in the kit. If the PLA probes are in proximity, the addition of ligase and DNA polymerase results in rolling circle amplification. For the ligation step, cells were washed twice with 1X wash buffer A (provided in the PLA kit) for 5 minutes each, and incubated with ligase (1:40 dilution) for 30 minutes at 37°C. For the amplification step, cells were washed twice with 1X wash buffer A for 2 minutes each, and incubated with polymerase diluted (1:80) in an amplification buffer containing fluorescently labeled complementary nucleotide probes for 100 minutes at 37°C. After the incubation, the cells were washed twice with 1X wash buffer B (provided in the PLA kit) for 10 minutes each followed by a quick wash with 0.01X wash buffer B. Finally, the cover glass was mounted onto microscope slides with DAPI mounting media. All images were obtained on a Leica confocal microscope (PLA excitation: 488nm). Counting of PLA particles was automated using ImageJ using the ‘Measure’ function. To determine the level of interaction, the number of PLA particles were normalized to the number of nuclei within an acquired field of view (n = 5 FOV per condition), and expressed as a fold change to the +DOX condition.
iPSC neurons were plated on coverglass, fixed in 4% paraformaldehyde for 15 minutes, permeabilized with 0.1% Triton X-100 in PBS for 30 minutes, and blocked with 2% BSA and 5% NGS in Triton-PBS for 30 minutes. Primary antibodies (anti-α-synuclein syn211, anti-PDIA6) were added overnight, followed by incubation with secondary antibodies (Alexa Fluor 488 Goat anti-mouse IgG and Alexa Fluor 568 Goat anti-rabbit IgG) for 2 hours. The cells were then washed three times with Triton-PBS and mounted onto microscope slides with DAPI mounting media. Imaging was performed using an oil immersion 100X objective lens on a Nikon structured illumination microscope (N-SIM) at the Northwestern University Center for Advanced Microscopy. Images were captured and slices were reconstructed using the Nikon NIS Elements program.
Pharmacological treatment of cell cultures
ER stress induction of H4 cells or iPSC neurons
To induce ER stress and activate the UPR, H4 cells or iPSC neurons were treated with 30nM thapsigargin (Tg) (Sigma) or 50ng/ml brefeldin A (BFA) for 24 hours prior to harvesting, and analysis of mRNA and/or protein expression of known ER stress markers was performed.
Proteasomal inhibition of iPSC neurons
iPSC neurons were treated with 50 nM epoxomicin (Fisher) for 24 hours. Analysis of GCase levels following treatment was performed via western blot analysis. Successful proteasomal inhibition was confirmed by blotting for ubiquitin.
Treatment of H4 cells or iPSC neurons with ER proteostasis and trafficking enhancers
H4 cells or iPSC neurons were treated with vehicle or either 25 μM diltiazem (DILT) (Sigma), 40 μM dantrolene (DANT) (Sigma) or 1 μm DHBP (Sigma), and media was changed every other day for the duration of the experiment. For the combination compound treatments, H4 cells or iPSC neurons were treated with vehicle, 25uM DILT, 5nM farnesyl transferase inhibitor (FTI) (gift of Peter T. Lansbury, Jr.), or FTI+DILT combination, and media was changed every other day for the duration of the experiment. For the combination of genetic manipulation and compound treatment (e.g., RyR3 KD + FTI, DILT + ykt6-CS), cells were infected and treatment was begun at the same time, with media change every other day for the duration of the experiment.
Lentiviral treatment of cell cultures
Lentiviral preparation and transduction of H4 cells and iPSC neurons
In combination with a packaging vector (psPAX2) and an envelope vector (VSV-G), lentiviral plasmids were used to create lentiviral particles by transfecting HEK293FT cells using X-tremeGENE transfection reagent (Roche). The lentiviral particles were collected and concentrated using Lenti-X concentrator (Clontech) and titered with a HIV1-p24 antigen ELISA kit (Zeptometrix). For RyR3 shRNA knock-down and ykt6-CS overexpression, H4 cells and iPSC neurons were infected at a multiplicity of infection (MOI) of 3-5 and were harvested 5 days or 2 weeks post-infection, respectively. For GBA1 L444P overexpression, neurons were infected at MOI of 1 and were harvested 2 weeks post-infection.
Ryanodine receptor RyR3 knockdown using shRNA constructs
MISSION shRNA sequences targeting human RyR3 were obtained from Sigma and tested for efficiency in HEK293T cells by quantitative RT-PCR analysis using RyR3 TaqMan probes. Clone ID #TRCN0000053349 was found to achieve the most efficient knock-down and was therefore used in subsequent experiments. This lentiviral plasmid was used to create lentiviral particles, as described under “Lentiviral preparation and transduction of H4 cells and iPSC neurons.”
Generation of the GBA1 L444P plasmid and transduction in iPSC-neurons
GBA1 L444P was generated by site-directed mutagenesis (SDM) of the pER4-wild-type GBA1 lentiviral plasmid. Mutagenesis primers (5′-GTGCCACTGCGTCCGGGTCGTTCTTCTGA-3′ and 5′-TCAGAAGAACGACCCGGACGCAGTGGCAC-3′) were created using the Agilent tool. SDM was performed using the materials and procedures from the QuikChange XL Site-Directed Mutagenesis Kit (Agilent). The L444P mutation was confirmed by sequencing. The pER4 GBA1 L444P plasmid was then packaged into lentiviral particles, as described under “Lentiviral preparation and transduction of H4 cells and iPSC neurons.”
Quantification and statistical analysis
Quantification methods of western blots and images have been described above. In each quantification, a single plot point indicates a separate biological replicate (individual culture well), taken from at least two distinct iPSC passages / differentiation batches. The value of n and what n represents is indicated in each figure legend. Analyzed data was plotted and tested for statistical significance using the GraphPad Prism software. Statistical significance between two samples was determined using a paired or unpaired t test with Welch’s correction. For more than two conditions, significance was determined using a one-way ANOVA with Tukey’s multiple comparisons mulmu test, except in Figure 2C where a Dunnett’s T3 multiple comparison test was used with a Brown-Forsythe and Welch’s correction . A p value of < 0.05 was considered to be significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). For each quantification, the type of error bar used and statistical test is specified in the figure legends.
Data and code availability
•
Data availability: All data reported in this paper will be shared by the lead contact upon request. Raw image files are avialable at Mendeley Data: http://doi.org/10.17632/29t6p4wrs7.1.
•
Code: This paper does not report original code.
•
Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
Acknowledgments
This work was supported by the National Institute of Neurological Disorders and Stroke of the National Institutes of Health (grants R01NS092823 and R21NS107770 awarded to J.R.M.) and the National Science Foundation Graduate Research Fellowship (grant no. DGE1324585 awarded to I.S.). We thank Sohee Jeon for technical support and Dr. Eileen Bigio for providing human brain samples
Author contributions
Conceptualization, I.S. and J.R.M; methodology, validation, formal analysis, and investigation, I.S., W.Y.W., F.Z., N.R.B., E.P., N. M., K.S., L.F., and J.R.M.; writing – original draft, I.S. and J.R.M.; writing – review & editing, I.S., W.Y.W. F.Z., N.R.B., E.P., N. M., K.S., L.F., and J.R.M.; visualization, I.S. and J.R.M.; supervision, J.R.M.; project administration, J.R.M.; funding acquisition, J.R.M.
Declaration of interests
J.R.M. is a scientific co-founder of Lysosomal Therapeutics.
Brefeldin A, thapsigargin, and AIF4- stimulate the accumulation of GRP78 mRNA in a cycloheximide dependent manner, whilst induction by hypoxia is independent of protein synthesis.
Synthesis of beta-hexosaminidase in cell-free translation and in intact fibroblasts: an insoluble precursor alpha chain in a rare form of Tay-Sachs disease.
00862-X&id=fx1.jpg){kind=link}
00862-X&id=gr1.jpg){kind=link}
00862-X&id=gr2.jpg){kind=link}
00862-X&id=gr3.jpg){kind=link}
00862-X&id=gr4.jpg){kind=link}
00862-X&id=gr5.jpg){kind=link}
00862-X&id=gr6.jpg){kind=link}
00862-X&id=gr7.jpg){kind=link}
00862-X&id=gr8.jpg){kind=link}