Parkinson's Disease Phenotypes in Patient Neuronal Cultures and Brain Organoids Improved by 2‐Hydroxypropyl‐β‐Cyclodextrin Treatment

ABSTRACT Background The etiology of Parkinson's disease (PD) is only partially understood despite the fact that environmental causes, risk factors, and specific gene mutations are contributors to the disease. Biallelic mutations in the phosphatase and tensin homolog (PTEN)‐induced putative kinase 1 (PINK1) gene involved in mitochondrial homeostasis, vesicle trafficking, and autophagy are sufficient to cause PD. Objectives We sought to evaluate the difference between controls' and PINK1 patients' derived neurons in their transition from neuroepithelial stem cells to neurons, allowing us to identify potential pathways to target with repurposed compounds. Methods Using two‐dimensional and three‐dimensional models of patients' derived neurons we recapitulated PD‐related phenotypes. We introduced the usage of midbrain organoids for testing compounds. Using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/CRISPR‐associated protein 9 (Cas9), we corrected the point mutations of three patients' derived cells. We evaluated the effect of the selected compound in a mouse model. Results PD patient‐derived cells presented differences in their energetic profile, imbalanced proliferation, apoptosis, mitophagy, and a reduced differentiation efficiency to tyrosine hydroxylase positive (TH+) neurons compared to controls' cells. Correction of a patient's point mutation ameliorated the metabolic properties and neuronal firing rates as well as reversing the differentiation phenotype, and reducing the increased astrocytic levels. Treatment with 2‐hydroxypropyl‐β‐cyclodextrin increased the autophagy and mitophagy capacity of neurons concomitant with an improved dopaminergic differentiation of patient‐specific neurons in midbrain organoids and ameliorated neurotoxicity in a mouse model. Conclusion We show that treatment with a repurposed compound is sufficient for restoring the impaired dopaminergic differentiation of PD patient‐derived cells. © 2021 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society

midbrain organoids for testing compounds. Using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/CRISPR-associated protein 9 (Cas9), we corrected the point mutations of three patients' derived cells. We evaluated the effect of the selected compound in a mouse model. Results: PD patient-derived cells presented differences in their energetic profile, imbalanced proliferation, apoptosis, mitophagy, and a reduced differentiation efficiency to tyrosine hydroxylase positive (TH+) neurons compared to controls' cells. Correction of a patient's point mutation ameliorated the metabolic properties and neuronal firing rates as well as reversing the differentiation phenotype, and reducing the increased astrocytic levels. Treatment with 2-hydroxypropyl-β-cyclodextrin increased the autophagy and mitophagy capacity of neurons concomitant with an improved dopaminergic differentiation of patient-specific neurons in midbrain organoids and ameliorated neurotoxicity in a mouse model. Conclusion: We show that treatment with a repurposed compound is sufficient for restoring the impaired dopaminergic differentiation of PD patient-derived cells. Neurodegenerative diseases pose a great threat to aging populations. 1 Due to the lack of disease-modifying therapies, patients suffering from Parkinson's disease (PD) have to rely on symptomatic treatments. 2 Affected genes, reported as risk or causative factors of PD, control major cellular processes such as cell proliferation, membrane trafficking, mitochondrial homeostasis, and autophagy. 3 Among these, PINK1 is involved in regulating mitochondrial function and morphology by quarantining damaged mitochondria before their degradation as well as triggering the process of mitophagy. 4 The fact that individuals with biallelic pathogenic variants in PINK1 develop PD shows that an altered mitochondrial function, morphology, and degradation are linked to its pathogenesis. 5 One of the hallmarks of PD is the loss of dopaminergic neurons in the substantia nigra pars compacta, but other regions are also affected. 6 Reports of postmortem neuropathological studies of patients having mutations in PINK1 show that Lewy bodies, mainly composed of synuclein alpha (SNCA), are present in several brain regions of some of these patients. 7 Work in the zebrafish and organoid models previously suggested that a loss of function of PINK1 might lead to a developmental reduction in the number of dopaminergic neurons. 8,9 However, implications of PINK1 and its PD-associated mutations during the transition from neural precursor cells to differentiated neurons were not studied in depth in a human cell model. Techniques for recapitulating disease phenotypes in culture had a recent breakthrough with the introduction of organoid cultures. 10 These organ-like structures contain different cell types in a spatially organized fashion, recapitulating at least some of the main functions of the respective organ. Importantly, they have been proven valid models for human diseases. [10][11][12] In this study, the induced pluripotent stem cells (iPSCs) of patients with PD with PINK1 mutations and healthy individuals were differentiated into a neuroepithelial stem cell (NESC) state and later into dopaminergic neurons. Different features such as proliferation capacity, apoptosis, and differentiation efficiency were analyzed using computational algorithms for pattern recognition through high-content image analysis. We demonstrated that the differentiation efficiency of patient-derived NESCs is reduced while maintaining an increased proliferative activity on neuronal differentiation and exhibiting increased apoptosis of tyrosine hydroxylase positive (TH+) neurons. We performed an immuno-based protein-profiling analysis of organoids at different stages of development confirming that proteins involved in cell cycle, differentiation, apoptosis, and autophagy pathways were differentially abundant in the case of the patient. Using extracellular flux analysis and microelectrode array, we assessed the energetic profile of NESCs and the firing activity of differentiated neurons, which were altered in patient-derived cells. Moreover, using a pH-sensitive reporter tagging a mitochondrial protein, we observed a reduced autophagy and mitophagy capacity. Gene correcting the patients' mutation with the clustered regularly interspaced short palindrome repeats (CRISPR)/CRISPR-associated protein 9 (CRISPR/Cas9) system improved the mitochondrial activity, firing rate, and differentiation efficiency. Treatment with the compound 2-hydroxypropylβ-cyclodextrin (HP-β-CD) resolved the mitophagy impairment and improved the dopaminergic differentiation in patient-derived cells by modifying the levels of proteins involved in dopaminergic differentiation, autophagy, apoptosis, and neuroinflammation.

Reagents and Resources Information
Detailed information about the reagents and resources used in this article are summarized in Table S7.

Detailed Description of Protocols
A summary is given of the different procedures. A detailed explanation can be found in the Supplementary Information, Extended Material and Methods.

Information About Cell Lines
Healthy controls 1 and 2 gave written informed consent at the University of Tübingen. Healthy control 3 was provided by Bill Skarnes. Patient PINK1 1 and patient PINK1 2 gave written informed consent at the University of Lübeck. Patient PINK1 3 and patient parkin samples were obtained from Coriell Institute, now held by the National Institute of Neurological Disorders and Stroke human cell and data repository. From each donor, one clone per iPSC line was derived and used in this study.

Immunocytochemistry
Fixation was done using 4% Paraformaldehyde (PFA) for 15 minutes at Room Temperature (RT). After 3Â 1 Â phosphate-buffered saline (PBS) washing steps, cells were permeabilized using 0.5% Triton X-100 in 1 Â PBS for 15 minutes at RT. Blocking was performed for 1 hour at RT. Incubation with the first antibodies was done overnight at 4 C in blocking buffer. Incubation with the secondary antibodies was for 2 hours at RT in blocking buffer.

Immunohistochemistry
Processing of organoids was performed as previously described. 14

Western Blotting
Pellets of neurons differentiated for 21 days coming from a confluent well of a 6-well plate were lysed. Lysates were then centrifuged, quantified, resolved, and transferred from the gel to polyvinylidene fluoride membranes in an iBlot2 device (Thermo Fisher [Waltham, MA], IB24001). Membranes were blocked for 60 minutes at RT. Primary antibodies were incubated at 4 C overnight. Secondary antibodies were incubated for 60 minutes at RT. Membranes were revealed using the SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher [Waltham, MA]).

Extracellular Flux Analysis (SeaHorse Measurements)
Human NESCs were seeded in a Matrigel-coated XF 96-well plate (Agilent Technologies [Santa Clara, CA], 102416-100) at a density of 65,000 cells per well. Three baseline measures and three measurements after each compound injection were performed.

Microelectrode Array Measurements
The Maestro microelectrode array (MEA; Axion Bio-Systems [Atlanta, GA]) system was used to measure the spontaneous activity of neurons. Axion Integrated Studio was used to process the raw data as previously described. 14

Rosella Mitophagy Reporter
The pH sensor fluorescent protein pH-sensitive green fluorescent protein (pHluorin) was fused to red fluorescent protein from Discosoma (DsRed) and the entire open reading frame of adenosine triphosphate (ATP) Synthase F1 Subunit Gamma (ATP5C1) or Microtubule-associated protein 1A/1B-light chain 3 (LC3) as described in Sargsyan et al 15 and Arias-Fuenzalida et al. 16

Gene Editing
Gene correction of patient's point mutation was performed as previously described. 17,18 Briefly, donor constructs with a positive selection module (PSM) and designed guide RNAs (gRNAs) targeting PINK1 were transfected into human induced pluripotent stem cells (hiPSCs). Fluorescent selection was done by cell sorting; removal of the PSM was performed with transposase piggyBac excision-only mRNA, and selection was done via cell sorting.

RNA Isolation, Reverse-Transcription Polymerase Chain Reaction, and Quantitative Polymerase Chain Reaction
Total RNA was isolated using miRNeasy Mini Kit (Qiagen, Hilden, Germany) and treated with the RNase-Free DNase Set (Qiagen). cDNA was reverse transcribed using the High-Capacity RNA-to-cDNA Kit (Invitrogen, Waltham, MA). Quantification of gene expression was performed using the LightCycler 480 Probes software (Roche, Basel, Switzerland).

Compound Treatment
HP-β-CD dissolved in water (Sigma, H-107) was added on every media change at the different concentrations tested and kept throughout the entire differentiation process. Untreated conditions are regular differentiation media.

Image Acquisition
Cell carrier Ultra plates were imaged in an automated manner using an Opera Quadruple Enhanced High Sensitivity (QEHS) spinning disk microscope (PerkinElmer, Waltham, MA).

Image Analysis
The image analysis was performed using MatLab (MathWorks, Natick, MA) as previously described. 16,19 Microfluidics Culture Neuroepithelial stem cells (NESCs) were seeded in an OrganoPlate (Mimetas [Leiden, the Netherlands], 9603-400-B) as explained elsewhere. 20 Immuno-Based Protein-Profiling Sample Incubation, Data Acquisition, and Analysis Organoids from control and patient-derived NESCs were treated with HP-β-CD in a 96-well ultra-low attachment (ULA) plate format for 30 days of differentiation at a concentration of 5 μM. The samples were analyzed on scioDiscover antibody microarrays (Sciomics, Baden-Wuerttemberg, Germany) targeting 1360 different proteins with 1830 antibodies. Differences in protein abundance or phosphorylation levels between different samples or sample groups are presented as log-fold changes (logFC) calculated for the basis two.

Animal Experiments
A total of 15 male, 2-month-old C57bl/6 mice were used in the study. MPTP at 30 mg/kg was intraperitoneally injected for 5 days to generate the subacute PD mouse model. Model verification was performed using both behavioral analysis and pathological assessment. TH staining was used to evaluate the dopaminergic neuronal loss in substantia nigra. HP-β-CD was injected every second day in the dose of 4000 mg/ kg subcutaneously for 15 days.

Statistical Analysis and Graphical Representation
Statistical analysis performed on each assay is mentioned in each figure legend. All of the statistical analyses were performed in R (R Foundation for Statistical Computing, Vienna, Austria). Clustering and heatmaps were produced using the complexheatmap package of R. 21

Reduced TH+ Neuron Differentiation, Increased Proliferation, Astrocyte Activation, and Apoptosis in Patient-Specific Cells
To evaluate whether impaired mitochondrial function could affect dopaminergic differentiation, hiPSCs were derived from three patients carrying a mutation in PINK1: two carrying p.Q456X (rs45539432) 22 and one carrying p.I368N (rs774647122) and three agematched and sex-matched controls (Fig. S1A). Human 220) were collected over three independent replicates using all of the lines. (G) Images representing the median values of a 21-day differentiation neuronal two-dimensional culture of control and patient groups. Raw images of the markers glial fibrillary acidic protein (GFAP) and S100 calcium binding protein B (S100b) are presented with their respective perimeter masks and zoomed regions (scale bar = 100 μm). (H) Quantification of colocalization between GFAP, S100B, and TH normalized to nuclear area at time points 7, 14, and 21 after the induction of differentiation in a two-dimensional culture. (I) Images representing the median values of a 21-day differentiation neuronal two-dimensional culture of control and patient groups. Raw images of the markers synuclein alpha (SNCA) and TH are presented with their respective perimeter masks and zoomed regions (scale bar = 100 μm). (J) Quantification of colocalization between SNCA and TH normalized to nuclear area at time points 7, 14, and 21 after the induction of differentiation in a two-dimensional culture. (K) Heatmap clustering different phenotypes during the process of differentiation between control and patient-derived neurons in a two-dimensional culture. Normalized scale within category of phenotype. (L) Western blot analysis of TH, TUBB3, and GFAP proteins extracted from two patient-derived PINK1 p.Q456X-mutant neurons (patients 1 and 2) and their respective isogenic gene-corrected (GC) controls (patient 1 GC and patient 2 GC) after 30 days of differentiation (two-dimensional cultures). β-actin was used as loading control. (M) Quantitative immunoblot analysis of data presented in C. Histogram bars represent the mean values (AE standard deviation) of TH, TUBB3, and GFAP signals in at least three independent experiments using the two isogenic pairs. Data were normalized against β-actin levels and expressed as fold change. Except for panels L and M, all control and patient lines were used. Statistical analysis was performed using Kruskal-Wallis and Dunn's tests for multiple comparisons. Adjustment of the P-value for multiple tests was performed using Benjamini-Hochberg. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. [Color figure can be viewed at wileyonlinelibrary.com] iPSCs were further differentiated into a stable neural precursor state (NESCs) following a previous report 13 and used as a starting population for studying dopaminergic neuron differentiation efficiency (Fig. S1B). Using an automated image analysis algorithm, the proportion of Tubulin Beta 3 Class III (TUBB3)-positive signal that colocalizes with TH was quantified. Patient-derived NESCs showed a reduced capacity to differentiate into TH+ neurons (worsened at later time points) while maintaining the same level of overall neuronal differentiation compared with controls ( Fig. 1A,B, Table S8). To confirm the impaired dopaminergic differentiation, we differentiated NESCs in a three-dimensional microfluidic environment observing the same pattern (Fig. S1C,D). Assessment of the proliferation marker Ki67 after 7, 14, and 21 days of differentiation FIG. 2. Differential abundance of proteins between control and patient-derived organoids at day 30 of neuronal differentiation. (A) Volcano plot of proteomics data. The x axis represents the log fold change (logFC) between patient-derived and control organoids, with positive logFC indicating that the protein is more abundant in patient data than in control data, and the opposite for negative logFC. The y axis represents the P-value of the comparison adjusted for multiple testing using Benjamini-Hochberg. Proteins with adjusted P-value <0.05 and absolute logFC >0.5 were considered differential. (B) Network of the protein-protein interactions among the differential proteins obtained from the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database. Interactions obtained from all data sources and with a confidence score >0.9 (high) were considered. Differential proteins that are not reported to interact with other differential proteins are not represented. ( Fig. 1C-F) showed that patient-derived cells maintained a higher proliferative capacity after induction of differentiation. The apoptotic marker cleaved poly adenosin phosphate (ADP)-ribose polymerase (cPARP) showed an increased signal in patient-derived TH+ neurons at time point 14 (Fig. 1E,F). The amount of glial fibrillary acidic protein (GFAP) and S100 calcium binding protein B (S100b)-positive cells (markers of activated astrocytes) was significantly higher at day 21 of differentiation ( Fig. 1G,H) the lowest TH levels in patients. During differentiation, there is a reduction of the GFAP marker at day 14 (controls being higher) followed by an increase at day 21 (higher increase in patient cells). An increase in astrocytic markers have been previously linked to oxidative stress. 23 Total levels of SNCA colocalizing with TH significantly increased in the patients through the different stages of differentiation (Fig. 1I,G). A cluster analysis considering all the different phenotypic features analyzed through the different time points of differentiation showed a discerning pattern in TH proportion and morphology at later time points, days 14 and 21 of differentiation (Fig. 1K).

Gene Correction-Restored Energetic Profile and Differentiation Efficiency
To evaluate the effect of the point mutation in cells derived from patients with PD, homozygous correction of the g.20655C>T (p.Gln456Ter) mutation in PINK1 was performed in two patient lines and homozygous correction of the c.1103T>A (p.Ile368Asn) in one patient line using Fluorescence-activated cell sorting (FACS)-assisted CRIPSR/Cas9 editing. 17,18 After gene correction, the reduced TH levels were improved, and the increased levels of astrocytes were reduced in neurons cultured in twodimensional conditions (Fig. 1L,M). The increased levels of TH after gene correction were also confirmed by immunofluorescence (Fig. S2A,B). Extracellular flux analysis (Seahorse, Agilent Technologies [Santa Clara, CA]) showed that gene correction reduced significantly the higher glycolytic activity of patient-derived NESCs (Fig. S2C-E,F).
MEA measurements showed that the firing activity and the network burst firing in neurons cultured in twodimensional were increased after gene correction (Fig. S2F-H). To assess the gene correction of PINK1 not only at the genomic level but also at a functional level, we analyzed the expression of PINK1-regulated mitophagy markers in differentiated neurons cultured in two-dimensional. Monoubiquitination of Voltage Dependent Anion Channel 1 (VDAC1), performed by parkin in a PINK1-dependent manner, 24 was restored in the two gene-corrected lines tested. This was accompanied by decreased parkin levels in Carbonyl cyanide m-chlorophenyl hydrazone (CCCP)treated neurons after PINK1 gene correction as a result of mitophagy-induced parkin degradation 25 (Fig. S2I,J).

Proteomics Analysis in Organoids Confirm Dysregulated Pathways at Different Time Points
We performed an immuno-based protein analysis of three-dimensional midbrain organoids derived from control and patient cells at different time points of differentiation. Differentially abundant proteins were detected at days 10 (182 proteins), 20 (302), and 30 (267) of differentiation ( Fig. 2A,  Fig. S3-S5, and Supplemental Datasets 1-3). Selected protein-protein interaction analysis using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database 26 and Cytoscape 27 revealed several direct as well as indirect interactions of the differential proteins corresponding with pathways involved in apoptosis, necroptosis, protein synthesis, metabolism, and cell cycle (Fig. 2B-F). Also, proteins , and TH+ and their respective colocalizations, normalized to nuclear area at different chloroquine concentrations. Except for panels A to E, all control and patient lines were used. Statistical analyses for panels B, G, and I were performed using Kruskal-Wallis and Dunn's tests for multiple comparisons. Statistical analysis for panel D was performed using a nonparametric test for repeated measures in factorial design (nparLD). Adjustment of the P-value for multiple tests was performed using Benjamini-Hochberg. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns, not significant. baf, bafilomycin; CCCP, carbonyl cyanide mchlorophenyl hydrazone; chlo, chloroquine; DFP, deferiprone; dmso, Dimethyl sulfoxide; EBSS, Earle's Balanced Salt Solution; oligo, oligomycin; rapa, rapamycin; thap, thapsigargin; val, valinomycin. [Color figure can be viewed at wileyonlinelibrary.com] involved in inflammation and autophagy showed dysregulation between controls and patients during the process of differentiation (Fig. 2E,F and  Fig. S3).

PINK1 Patient-Specific Neurons Present a Reduced Mitophagy Capacity
To understand the mitochondrial dynamics in these cells, we generated lines expressing the Rosella construct bound to microtubule-associated proteins 1A/1B light chain 3B (LC3) or to adenosine triphosphate (ATP) synthase F1 subunit gamma (ATP5C1) to evaluate the autophagy status and mitochondrial degradation by mitophagy (Fig. 3). Already at the hiPSC level, classification of the different stages of autophagy 16 showed that autophagy was reduced in patient-derived cells (PINK1, p.I368N) compared with a control line (Fig. 3A,B). To confirm with a different assay, we analyzed the ratio of LC3 membrane-bound form/LC3 cytosolic form (LC3-II/LC3-I) by Western blot, which was reduced in 21-day differentiated neurons in two-dimensional cultures derived from patients (Fig. S6A,B). Measurements throughout differentiation revealed significantly fewer mitophagy events in patients' neurons carrying the Rosella construct (Fig. 3C,D and Fig. S6C). The mitophagy events observed in PINK1 patient-derived cells might be mediated by PINK1-parkin independent mitophagy pathways. 28 Rapamycin inhibits the mammalian target of rapamycin (mTOR), which modulates autophagy by reducing transcription factor EB (TFEB)'s nuclear translocation. 29 Patient-derived cells treated with rapamycin showed an increase in the frequency of phagophores and autophagic vacuoles to similar levels as those observed in controls, which presented a similar response to the rapamycin treatment ( Fig. 3E and  Fig. S6D). In this study, one patient and control line were genetically modified to stably carry the Rosella construct, hence caution is needed in interpreting the mitophagy results detected with the reporter. We then assessed the effect of chloroquine, a known blocker of the autophagic flux by reducing the fusion between autophagosomes and autolysosomes, 30 during the neuronal differentiation process. The total levels of LC3 were increased in the patient-untreated group and reduced after treatment with chloroquine (Fig. S6E). No difference was observed in the total levels of lysosomal associated membrane protein 1 (LAMP1) in untreated samples. When evaluating the levels of colocalization of LC3 and LAMP1, no difference in the overall colocalization was observed, but when considering the colocalization with TH as well, control TH+ neurons presented a higher level of colocalized LC3-LAMP1 in the soma as well as in the neurites in untreated conditions (Fig. 3F,G). This occurred concomitantly with a significant reduction in the levels of TH+ cells at a concentration of 10 nM of chloroquine and increased GFAP particularly in controls (Fig. 3H,I  and Fig. S6F). Higher concentrations of chloroquine showed an increase of the TH+ levels compared with the treatment with 10 nM chloroquine, which could be explained by the reported increase of transforming growth factor beta (TGFβ) superfamily receptors in the plasma membrane after chloroquine treatment. 31 TGFβ plays an important role in the specification of dopaminergic neurons. [32][33][34] Treatment with HP-β-CD Increased TFEB Nuclear Translocation in Neurons, Modified the Abundance of Proteins Related to Dopaminergic Neurons Differentiation, and Increased Dopaminergic Neuron Differentiation Efficiency of Brain Organoids Due to the observed altered autophagy and mitophagy pattern, we explored ways for modulating autophagy with repurposed compounds. Treatment with the  (SQSTM1 or p62), LAMP1 and TH+, normalized to nuclear area at different HP-β-CD treatment concentrations. (E) Protein abundance measured in control-derived, patientderived, and HP-β-CD-treated patient-derived organoids at time 10, 20 and 30 days of the neuronal differentiation was scaled for each protein separately. A k-means partitioning (k = 8) was performed in order to obtain clusters of proteins with similar expression dynamics. The proteins differentially abundant between control and patient-derived organoids, and between patient and treated patient-derived organoids are also shown (Benjamini-Hochberg, BH-adjusted P-value<0.05 and absolute logFC>0.5). (F) Normalized expression of selected proteins that show differential expression between patient and treated patient-derived organoids is reported. Three data points were collected for each condition and time point. Whiskers represent one standard deviation from the median of the measurements. (G) Representative images of control, patient and patient treated derived organoids at 30 days of differentiation (scale bar = 200 μm). (H) Quantification of the markers TH, Tubulin Beta 3 Class III (TUBB3) and Hoechst. Each dot represents one section analyzed. Sections analyzed: control (7, 9, 5, 14, and 9) and patient (9, 25, 15, 18, and 19) respectively for the different treatments (Un, 500 nM, 1 μM, 5 μM, and 10 μM) collected over three independent replicates. Control 1, patient 1, and patient 3 lines were used. For panels A to D, all control and patient lines were used. For panels E and F, control line 1 and patient line 1 were used. For panels B, D, and H, statistical analyses were performed using Kruskal-Wallis and Dunn's tests for multiple comparisons. Adjustment of the P-value for multiple tests was performed using Benjamini-Hochberg (BH). The adjusted significance are represented in red. Comparisons between control untreated and patient untreated are presented with #. Comparison between the patient-untreated condition and the different treatment concentrations are represented with *. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns, not significant. Significance hashtag represent: #P < 0.05, ##P < 0.01, ###P < 0.001; ns stands for not significant. Un, untreated. 1433Z, tyrosine 3-monooxygenase/tryptophan 5-monooxygenase; BIRC7, baculoviral inhibitor of apoptosis protein (IAP) repeat containing 7; CCL7, C-C motif chemokine ligand 7; CD15, fucosyltransferase 4; DLK1, Delta like non-canonical notch ligand 1; DLX4, distal-less homeobox 4; ECHM, enoyl coenzyme a hydratase short chain 1 mitochondrial; EGFR, epidermal growth factor receptor; IFNA1, interferon alpha 1; compound HP-β-CD was reported to modulate autophagy by increasing TFEB nuclear translocation in human neuroglioma cells. 29,35,36 Treatment with HPβ-CD during the entire differentiation process was able to increase the proportion of nuclei that colocalized with TFEB in a two-dimensional neuronal culture at 21 days of differentiation (Fig. 4A,B). Moreover, HP-β-CD upregulated the gene expression of FK506 binding protein 8 (FKBP8) and FUN14 domain-containing protein 1 (FUNDC1), outer mitochondrial membrane-anchored proteins that present the LC3 interacting region motif mediating mitophagy (Fig. S7A). Treatment with HPβ-CD led to an increase in mitophagy and in overall autophagy assessed with the Rosella construct (Fig. S7B, C). We also observed that this increased autophagy occurs in TH+ neurons by assessing the colocalization of sequestosome 1 (SQSTM1 or p62), LAMP1, and TH at concentrations above 5 μM reaching a plateau (Fig. 4C,D).
The effect of HP-β-CD during neuronal differentiation was assessed at three different time points in a three-dimensional environment using midbrain organoids. Protein abundance analysis (Fig. 4E,F [LAMP2], tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein zeta [14-3-3ζ], and baculoviral IAP repeat containing 7 [BIRC7]) (Fig. 4F). The reduced proportion of dopaminergic neurons in patient-derived brain organoids was increased after treatment with HPβ-CD increased the proportion of TH+ cells without changing the amount of TUBB3+ neurons (Fig. 4G,H,  Fig. S8A, Table S8). HP-β-CD treatment in patientspecific neurons, carrying the homozygous p.R275W (rs34424986) parkin RBR E3 ubiquitin protein ligase (PRKN) mutation, increased the amount of dopaminergic neurons (Fig. S8B) concomitant with an increase in nuclei pixels. Parkin is known to be a downstream effector of PINK1 induction of mitophagy, and alterations can lead to mitochondrial alterations and TH+ neuronal loss. 37 Because HP-β-CD composition is a mixture of cyclodextrin rings substituted with different degrees of hydroxypropyl, 38 we further characterized the composition of the used HP-β-CD mixture in these experiments and determined that the most frequent isomer is one with six degrees of substitution ( Fig. S9A-C). Knowledge of the cyclodextrin composition will leverage application in future clinical trials and facilitate comparisons of treatment results between different neurodegenerative diseases. 38 Treatment with HP-β-CD Reduces the Toxicity Generated by -Methyl-4-Phenyl-1,2,3,6-Tetrahydropyridine (MPTP) Treatment in Mice To assess the effect of HP-β-CD in in vivo conditions, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) induced subacute PD mice models were generated by intraperitoneal injection for 5 days of 30 mg/kg MPTP. A preventive treatment with HP-β-CD (Fig. 5A) was applied subcutaneously for 15 days. This treatment scheme showed a positive effect of HP-β-CD in reducing the loss of TH+ neurons in the midbrain of mice caused by MPTP (Fig. 5B,C).

Discussion
Our findings suggest that a loss-of-function mutation in PINK1 39 affects the transition between a neural precursor state and a differentiated TH+ neuron (Fig. 1). Previous reports showed no difference between a patient line carrying the p.Q456X mutation and control lines; however, these differences could be explained by the different differentiation protocol 22 and/or the low number of patient lines used in these studies. 40 Patientderived cells remained highly proliferative upon induction of differentiation (Fig. 1C,D). This matches previous reports showing that the loss of PINK1 activity triggers an increase in glycolysis via a reactive oxygen species-mediated stabilization of hypoxia-inducible factor-1α (HIF1α), associated with the Warburg effect. 41,42 Cells that manage to differentiate into TH+ neurons presented higher levels of apoptosis as previously reported. [43][44][45] Patients' derived neurons also presented an increased proportion of activated astrocytes (Fig. 1G,H) as well as of total SNCA in TH+ cells (Fig. 1I,J). Gene correction of PINK1 allowed neural precursor cells to reduce their dependence on glycolysis and increase the firing activity in differentiated neurons as well as increase the proportion of TH+ neurons upon differentiation (Fig. 1L,M, Fig. S2). Gene correction seems to trigger a metabolic change that allows the switch from stem cells to differentiated cells. 46 Proteomics analysis of control and patient-derived organoids at different stages of differentiation confirmed that the altered cell cycle, increased apoptosis, and reduced differentiation capacity observed coincided with dysregulation of these pathways (Fig. 2C-F, and  Fig. S3A-D). It also showed dysregulation of the autophagy pathway, a common process impaired in PD as well as in other neurodegenerative diseases. 47,48 Impaired mitochondrial turnover, due to reduced mitophagy activity, was observed in patient cells (Fig. 3C,D) in accordance with a previous report. 49 This altered mitophagy balance occurred simultaneously with an overall impaired autophagy (Fig. 3A,B, G). Regulation of autophagy with rapamycin led to a clustering of patient's together with control's cells (Fig. 3E), matching previously reported effects. 50 Modulating autophagy with chloroquine showed that a reduced colocalization of LC3 and LAMP1 led to a reduction TH+ neurons without altering the overall amount of neurons and increased the amount of GFAP levels (Fig. 3H,I). Treatment with different concentrations of HP-β-CD increased the presence of TFEB in the nuclei of patient-derived neurons (Fig. 4A,B). Further studies need to be performed to establish the indirect link between TFEB nuclear translocation and the upregulation of PINK1 independent mitophagy pathways. HP-β-CD treatment increased the amount of autophagy events as well as the colocalization of LAMP1-p62 in TH+ neurons at concentrations above 5 μM, plateauing at 10 and 50 μM, concomitantly to the increase of TFEB nuclear translocation increase (Fig. 4C,D). Using cells edited with the Rosella construct tagging LC3, we observed similar effects (Fig. S7).
Differentially abundant protein analysis between untreated and treated patient organoids showed that the HP-β-CD has an effect not only in autophagy but also in proteins regulating dopaminergic neuronal differentiation (Fig. 4E,F). Levels of SOX9 and DLK1 were significantly reduced after treatment. SOX9 maintains the multipotency characteristics of neural precursor cells, reducing neuronal differentiation as well as increasing astrogliogenesis and astrocyte differentiation. [51][52][53] Downregulation of DLK1 through paired like homeodomain 3 (PITX3) has been shown to be necessary for the differentiation of A9/substantia nigra pars compacta dopaminergic neurons. 54-56 Levels of DLK1 increased over time in untreated patient organoids and were reduced after HP-β-CD treatment (Fig. 4F). Increased levels of RA51C, reported to help mtDNA replication and integrity, were seen after treatment.
This coincided with reduced levels of IFNA1, known to reduce the autophagy clearance of mtDNA. 57,58 Reduced levels of LAMP2 in untreated patient organoids match previous reports in cerebrospinal fluid concentrations in patients with PD. 59,60 Similar results were reported in the brain extracts of patients with PD. 61 LAMP2 is known to increase autophagic flux, reduce SNCA levels, and reduce the degeneration of dopaminergic neurons. 62,63 Treatment with HP-β-CD significantly increased the levels of LAMP2 in patientderived organoids (Fig. 4E,F). Moreover, HP-β-CD significantly reduced the levels of BIRC7, a protein known for inhibiting autophagy by reducing the levels of LC3II, autophagy related 5 (ATG5), and beclin 1 (BECN1). 64 Levels of another autophagy-related protein, 14-3-3ζ, were increased in untreated patient organoids (Fig. 4F). 14-3-3ζ negatively regulates the early stages of autophagy. 65 Plus, 14-3-3 proteins are known regulators of the localization of TFEB between the nucleus and the cytoplasm, maintaining TFEB in the cytoplasm when it is phosphorylated. 66 A significant reduction of 14-3-3ζ was observed after HP-β-CD treatment of patient organoids (Fig. 4F), suggesting that its downregulation facilitates the translocation of TFEB to the nucleus. Furthermore, treatment with HP-β-CD improved the impaired dopaminergic differentiation observed in the patient organoids (Fig. 4G,  H). The induction of the translocation of TFEB is significant at concentrations of 5 and 10 μM (Fig. 4B), whereas the effect of HP-β-CD in the differentiation of the organoids was observed only at 5 μM. This would lead one to think that the beneficial aspects of the TFEB translocation when treating organoids with HP-β-CD for 30 consecutive days occur in a narrow concentration window.
Interestingly, it has been reported that increasing autophagy via expression of TFEB has positive effects in other neurodegenerative diseases, such as Alzheimer's disease, Niemann Pick disease, and Gaucher disease by improving the degradation of protein aggregates. [67][68][69] Moreover, treatment with HPβ-CD is currently in a phase 2b/3 clinical trial for treating Niemann Pick disease (NPD, NCT02534844). Although the permeability of HP-β-CD to the bloodbrain barrier is low (≈0.2% of blood circulating HPβ-CD), [70][71][72] previous reports have shown a positive effect in reducing neurodegeneration in mice treated only systemically. 70 We observed protection against the toxicity generated by MPTP after intraperitoneal treatment with HP-β-CD (Fig. 5). It has been reported that MPTP treatment increases brain-barrier permeability, 73 and this could have boosted the beneficial effects observed by HP-β-CD treatment. It is also possible that the pretreatment with HP-β-CD blocked MPTP-induced dysfunction of the blood-brain barrier as it has been reported for other treatments. 74 The present work demonstrates that HP-β-CD ameliorates the dopaminergic neuronal loss phenotype, confirming the therapeutic potential of modulating the autophagy/lysosomal pathway in the context of PD.