Neuronal intranuclear inclusion disease (NIID) is an adult-onset neurodegenerative disease caused by expanded GGC repeats in the 5' untranslated region of the human-specific NOTCH2NLC gene. The high sequence similarity between NOTCH2NLC and its paralogs poses a significant challenge for precise gene editing. Here, we develop a CRISPR/spCas9-based gene-editing strategy that precisely excises the expanded GGC repeats in NOTCH2NLC without detectable off-target effects on the highly homologous NOTCH2/NOTCH2NL family genes (<2% sequence divergence at this locus). The efficacy, specificity and safety of this approach are rigorously validated across multiple experimental models, including human cell lines, NIID iPSCs, and our previously established transgenic NIID mouse model. Our results demonstrate that precise excision of the expanded GGC repeats effectively alleviates NIID-related neuropathological, molecular and behavioral abnormalities. This study establishes the proof of concept for genome editing as a therapeutic strategy for NIID and other related repeat expansion disorders.

OBJECTIVES:To investigate the prevalence of probable REM sleep behavior disorder (pRBD) in essential tremor (ET), identify associated risk factors, and evaluate its effects on motor and non-motor symptoms. METHODS:Clinical data were collected from 1297 ET patients across multicenter. pRBD was assessed using the RBD Questionnaire-Hong Kong (RBDQ-HK). Risk factors associated with pRBD were identified through multivariable logistic regression. Furthermore, a meta-analysis was conducted to synthesize existing estimates of pRBD/RBD prevalence in ET. RESULTS:In this study, pRBD was identified in 11.6% of ET patients. Meta-analysis yielded pooled prevalence estimates of 16% for pRBD (ES = 0.16, 95% CI [0.10-0.21]) and 14% for RBD (ES = 0.14, 95% CI [0.07-0.21]). ET patients with pRBD were older (59.89 ± 13.99 vs. 54.63 ± 16.59 years, p < 0.001) and had a later tremor onset (47.55 ± 15.98 vs. 43.15 ± 17.62 years, p = 0.007) compared with those without pRBD. ET-pRBD patients also exhibited a higher frequency of midline tremor (54.67% vs. 43.93%, p = 0.003), rest tremor (27.33% vs. 16.04%, p = 0.001), and elevated Non-Motor Symptom Scale (NMSS) scores (20.30 ± 18.50 vs. 10.43 ± 12.42, p < 0.001). Multivariable logistic regression identified lower educational attainment (OR = 0.93, p = 0.002) and higher NMSS scores (OR = 1.03, p < 0.001) as independent risk factors. CONCLUSIONS:pRBD is prevalent in ET and independently associated with lower education and increased non-motor symptom burden. Recognition of pRBD may help identify an ET subgroup with distinctive clinical features.

Accurate modeling of biological age has clinical value for risk stratification, personalized prevention, and intervention planning, promoting proactive healthcare for aging and age-related diseases. Because aging is multidimensional, robust and interpretable multimodal approaches are needed. We studied 908 non-dementia older adults (> 60 years), collecting data on gait, eye movements, resting-state functional connectivity (rs-FC), and plasma biomarkers (neurofilament light chain, NfL, and glial fibrillary acidic protein, GFAP). Fourteen gait features, two eye movement features, 19 rs-FC features, and plasma GFAP levels were significantly correlated with age (p < 0.05). Among single-domain models, eye movement features showed the strongest predictive performance (R2 = 0.606; MAE = 3.060). A combined multimodal model achieved markedly higher accuracy (R2 = 0.814; MAE = 1.902). These findings demonstrate that integrating physiological, neurological, and biomarker data substantially improves biological age modeling, supporting the development of comprehensive frameworks to assess aging better and guide timely, targeted preventive strategies.

To map whole-brain dopamine transporter (DAT) depletion patterns in Parkinson’s disease (PD) using 11C-CFT PET and evaluate their utility in distinguishing tremor-dominant (TD) and postural instability/gait difficulty (PIGD) subtypes in early PD. This retrospective study included 197 PD patients (80 TD, 117 PIGD) and 21 healthy controls. 11C-CFT PET quantified DAT availability across nigrostriatal and extrastriatal regions. Differential DAT availability was assessed using FDR-corrected t-tests and Cohen’s d. LASSO-optimized features from nigrostriatal (NS), extranigral (ONS), and whole-brain (WB) regions informed three classification models (Support vector machine, Logistic Regression, Random Forest), evaluated via Receiver Operating Characteristic analysis. 197 patients (mean age 50 ± 10 years; 112 men) were evaluated. Beyond classical nigrostriatal reductions, PD patients exhibited significant DAT loss in mesolimbic (nucleus accumbens, amygdala), mesocortical (insula), and brainstem-cerebellar regions. PIGD patients demonstrated greater contralateral striatal (putamen: d = 0.58, caudate: d = 0.50; p < 0.05) and limbic (nucleus accumbens: d = 0.51; p = 0.01) DAT depletion compared to TD. The WB model incorporating striatal and extranigral features (posterior putamen, olfactory cortex, parietal regions) achieved superior subtype discrimination (AUC = 0.90) versus NS (AUC = 0.80) and ONS (AUC = 0.82) models. DAT availability in nucleus accumbens correlated with gait freezing (ρ=-0.28; p < 0.001) and autonomic scores (ρ=-0.34; p < 0.001). 11C-CFT PET reveals distinct whole-brain DAT depletion patterns in PD subtypes, extending beyond nigrostriatal degeneration. The WB model’s superior diagnostic accuracy underscores the critical role of extranigral dopaminergic circuits as biomarkers for early phenotypic stratification. These findings redefine PD as a multisystem dopaminergic disorder and provide a neuroimaging basis for precision phenotyping and therapeutic targeting.

Mesial temporal lobe epilepsy (MTLE) is known as a distributed network disorder, and regional neuroanatomical changes may significantly contribute to aberrant network formation. The major pathologies of MTLE include neuronal loss in the dentate hilus, CA3 and CA1 regions, and mossy fiber (MF) sprouting into the inner molecular layer (iML). The latter forms aberrant excitatory circuities that are considered to facilitate recurrent seizures. The subiculum is relatively preserved during hippocampal sclerosis but has been thought to play a critical role in the synchronization and propagation of epileptic activity especially to extrahippocampal brain regions. We recently identified a distinct expression of alpha-smooth muscle actin (αSMA) at the MF terminals in human hippocampus. This prompted us to explore MF sprouting in resected hippocampi (n = 20) from patients with MTLE relative to postmortem control (n = 20) using αSMA along with reference markers for pathological cross-validation. Immunolabeling of neuron-specific nuclear antigen and sortilin showed neuronal loss in the hilus/CA3 and CA1 and granule cell dispersion in all resected hippocampi relative to control. Immunolabeling of αSMA, zinc transporter 3 and β-secretase 1 displayed a widening of the iML with enhanced staining intensity and a fibrous band extending from CA2, CA1 towards subiculum in the resected hippocampi. Cytochrome c oxidase immunolabeling was increased in the iML and subiculum in the MTLE group. Taking together, the current findings suggest the possibility of long-range MF sprouting and regionally occurred hypermetabolic state in the hippocampal formation of patients with drug resistant MTLE.

The gut-brain axis is increasingly recognized as a critical contributor to Parkinson's disease (PD) pathogenesis, yet the therapeutic impact of microbiota modulation remains unclear due to lack of clinical trials in drug-naïve patients. We conducted a randomized, double-blind, placebo-controlled phase 2 trial to evaluate the safety, tolerability, and efficacy of repeated donor fecal microbiota transplantation (dFMT) in de novo PD. FMT was administered for seven days (200 mL on days 1-3; 50 mL on days 4-7) per 4-week cycle. Seventy-two patients were randomized 1:1 to receive dFMT or autologous FMT (aFMT), and 66 completed the trial. At 35 weeks, the dFMT group showed significant improvement in motor symptoms (mean change in Unified Parkinson's Disease Rating Scale [UPDRS] III: -3.8 vs. +0.1; p = 0.0001) and a substantially greater reduction in constipation severity (dFMT vs. aFMT: -6.5 vs. -0.7; p < 0.0001), accompanied by improved quality-of-life scores. Microbiome profiling revealed greater similarity to donor composition and a marked reduction in Escherichia-Shigella, correlating with decreased colonic α-synuclein aggregation (r = 0.3775, p = 0.0277), supporting a gut-brain mechanistic link. Biochemical analyses showed elevated fecal dopamine and 3,4-dihydroxyphenylacetic acid levels, while histological assessments demonstrated strengthened epithelial barrier integrity with increased E-cadherin expression. All adverse events were mild and self-limited; no serious treatment-related events were observed. These findings demonstrate that repeated dFMT is safe, well tolerated, and yields clinically meaningful motor and gastrointestinal improvements in drug-naïve PD, providing integrated mechanistic and clinical evidence that microbiota-targeted modulation represents a promising nonpharmacologic therapeutic strategy for neurodegenerative disease. Trial registration: Chinese Clinical Trial Registry, ChiCTR2200064151.

BACKGROUND:Beyond genetic factors, demographic and environmental associations with onset and severity in spinocerebellar ataxia type 3 (SCA3) remain incompletely understood. OBJECTIVE:The aim of this study was to identify the association of nongenetic factors with onset and severity of SCA3. METHODS:Associations of nongenetic factors with age at onset (AAO) or Scale for the Assessment and Rating of Ataxia (SARA) scores were examined using generalized linear models. RESULTS:In the cohort of 324 patients with SCA3, higher education level was associated with lower SARA scores but earlier AAO. Smoking was associated with earlier AAO. Tea consumption was inversely associated with SARA scores. Body mass index, coffee consumption, alcohol consumption, as well as other assessed variables, were not associated with AAO or SARA scores. CONCLUSIONS:These findings support the associations of nongenetic factors and onset or severity in SCA3, while highlighting the need for cautious interpretation given the observational design. © 2026 International Parkinson and Movement Disorder Society.

INTRODUCTION:Certain observational studies have indicated a link between gut microbiota, metabolites, and synucleinopathies, but the causal relationship mediated by metabolites remains un-proven. METHODS:To examine potential causal relationships among gut microbiota, plasma metabolites, and synucleinopathies, we utilized Mendelian randomization (MR), with the inverse variance weighted (IVW) approach serving as the main analytical method. Additionally, we investigated whether plasma metabolites function as mediators via two-step/mediation MR analysis. RESULTS:We observed indirect effects of the taurine-to-cysteine ratio in associations between genus DefluviitaleaceaeUCG011 and the risk of Parkinson's disease (PD), with a mediated proportion of 26.4% (p = 0.043). Besides, the pregnenediol sulfate (C21H34O5S) mediated the causal pathway from species Odoribacter_laneus to dementia with Lewy bodies (DLB), with a mediated proportion of 28.4% (p=0.028). DISCUSSION:Our findings indicate that the taurine-to-cysteine ratio and pregnenediol sulfate (C21H34O5S) serve as key mediating factors connecting the gut microbiota and synucleinopathies. Given that cysteine, taurine, and neurosteroids exert neuroprotective effects, targeting these meta-bolic pathways emerges as a potential therapeutic strategy for synucleinopathies. CONCLUSIONS:In the present study, we revealed that gut microbiota is essential to the development and progression of synucleinopathies with the involvement of cysteine, taurine, and the pregnenedi-ol sulfate (C21H34O5S), establishing an axis among gut microbiota, metabolites, and synucleinopa-thies. The comprehensive exploration uncovers new insights into synucleinopathy mechanisms and could support efforts in prevention and treatment.

BACKGROUND:Multiple system atrophy (MSA) is a fatal neurodegenerative disease with highly variable progression and poor prognosis. This study aimed to characterize survival patterns, identify prognostic factors, and develop an interpretable machine learning model for individualized survival prediction in MSA. METHODS:In this multicenter longitudinal cohort study, 391 MSA patients were analyzed. Prognostic factors were identified using Kaplan-Meier and Cox regression analyses. Six survival models were trained with five-fold cross-validation, with Shapley Additive Explanations (SHAP) being used for feature selection and interpretability. RESULTS:During a median follow-up of 4.9 years, 149 deaths occurred, with a median survival of 6.9 years. The random survival forest model achieved the best performance (C-index 0.769; mean time-dependent area under the curve [AUC] 0.815) with 11 predictors. An interactive web-based platform and a risk score were developed for clinical application. CONCLUSION:This interpretable machine learning model accurately predicted individualized survival in MSA and identified key prognostic factors. © 2026 International Parkinson and Movement Disorder Society.

Neuronal intranuclear inclusion disease (NIID) arises from GGC repeat expansions in NOTCH2NLC. These expanded repeats produce polyglycine (polyG) proteins, and the accumulation of these polyG proteins in neuronal nuclei serves as the characteristic pathological hallmark of NIID. However, the native cellular ultrastructure of polyG and its contribution to pathology remain poorly understood. Here, using a transgenic NIID mouse model, we extract polyG assemblies from diseased brain and characterize their architecture by cryo-electron tomography (cryo-ET). We further examine their native organization by tracer-guided in situ cryo-ET in vitrified mouse brain. We find that polyG forms highly branched ∼5 nm fibrils that laterally coalesce into densely packed ribbons, which represent the predominant polyG state within neuronal nuclei in situ. In parallel, proximity-dependent labeling coupled to mass spectrometry reveals selective enrichment of proteostasis factors-including proteasome subunits and molecular chaperones-at polyG assemblies in mouse brain. Consistent with this, cryo-ET visualizes proteasome-like particles decorating ribbon-shaped surfaces and edges in cells. Together, these findings uncover an unexpected ribbon-shaped supramolecular architecture for a low-complexity disease protein and suggest that nuclear polyG ribbons act as scaffolds that engage proteostasis machinery, providing mechanistic insight into NIID.

Abstract Background Understanding disease natural history is important for the development of potential treatments for people with MSA. We describe the natural progression of early MSA in a Chinese population. Methods Observational, 12-month study conducted in 8 sites across China . Eligible participants were aged 40–75 years, with possible or probable MSA of the parkinsonian (MSA-P) or cerebellar (MSA-C) subtype, and anticipated survival of ≥ 3 years. Disease progression was analyzed using a linear mixed model of Total UMSARS (Part I + II) progression, including baseline, Month 6 and Month 12 data. Results A total of 89 participants with a mean ± SD time since diagnosis of 0.4 ± 0.6 years were enrolled. Of these 52% had MSA-C and 48% participants had MSA-P. The mean ± SE [95%CI] rate of Total UMSARS progression was 1.27 ± 0.13 [1.01, 1.53] points per month. Participants showed a progression of 0.64 ± 0.06 [0.51, 0.76] points/month on UMSARS Part I and 0.62 ± 0.07 [0.47, 0.77] points/month on UMSARS Part II. Differences in the rates of UMSARS progression between patients with MSA-P and MSA-C were not statistically significant ( p > 0.05). Conclusions This is the first multicenter natural history study of MSA progression conducted in China. While prior studies have indicated a predominance of MSA-C in Asian populations, we found a more even split of MSA-C and MSA-P subtypes. In this early population, patients showed an average progression rate of ~ 15 Total UMSARS points/year; rates of progression were similar between the two subtypes and were in alignment with previous studies that assessed disease progression using UMSARS in Western populations. Trial registration Clinicaltrals.gov, NCT05453058 (registered June 16, 2022).

Epigenetic dysregulation plays an essential role in autism spectrum disorder (ASD), but the parent-of-origin effects (POEs) of DNA methylation remain unknown. Here, we applied PacBio HiFi sequencing with haplotype-phased methylation profiling in 124 individuals (31 ASD quartets) to systematically dissect POE-dependent methylation. Comparative analysis of phased methylomes between probands and unaffected siblings identified 114 paternal- and 106 maternal-specific differential methylated cytosines (DMCs), 45 and 46 differential methylated regions (DMRs), and 2425 and 2693 methylation outliers (MOs), respectively. These POE methylation alterations were enriched in ASD-relevant gene categories but exhibited distinct genomic distributions and functional pathways between parental haplotypes. Furthermore, genome-wide parent-of-origin DMR analysis identified 443 allele-specific methylation (ASM) regions, from which we detected 34 differential expression ASMs and 62 ASM outliers, showing pronounced enrichment within PWS/AS locus and ASD-associated genes. Collectively, this study provides a comprehensive evidence of pervasive POE-dependent methylation imbalance and aberrant ASM at imprinting regions underlying ASD pathogenesis, offering insights into epigenetic mechanisms of complex neurodevelopmental disorders.

BACKGROUND:Parkinson's disease (PD) remains challenging to diagnose at early stages owing to subtle and heterogeneous clinical manifestations and the lack of reliable biomarkers. Metabolomics offers a powerful approach to capture disease-related biochemical alterations that reflect underlying pathophysiology. This study aimed to identify robust plasma metabolic signatures for PD diagnosis and to elucidate metabolic alterations associated with clinical severity. METHODS:We performed untargeted plasma metabolomic profiling using ultra-high performance liquid chromatography-tandem mass spectrometry in a large Chinese population comprising two independent early-stage PD groups, one of which consisted of drug-naïve, de novo patients. Integrative statistical analyses, pathway enrichment, and machine learning-based diagnostic modelling were applied to identify discriminative metabolites and characterise disease- and treatment-related metabolic changes. FINDINGS:In the two case-control datasets, 111 metabolites were consistently altered in early-stage PD, among which 12-hydroxyeicosatetraenoic acid, spermine, and niacinamide emerged as key differential metabolites. Pathway enrichment analysis highlighted sphingolipid metabolism as a major dysregulated pathway in early-stage PD. Using machine learning-based models, a classification model based on six metabolites achieved strong performance (area under the receiver operating characteristic curve [AUC] = 0.976), while individual metabolites also demonstrated good discriminative ability, with the highest AUC reaching 0.916. We further observed that antiparkinsonian medication was significantly associated with metabolic alterations in tyrosine, tryptophan, and polyamine pathways. In addition, gut microbiota-derived metabolites, particularly phenylacetylglutamine and p-Cresol glucuronide, were markedly elevated in PD and associated with both motor and non-motor symptom severity, suggesting a potential contribution to clinical heterogeneity. INTERPRETATION:These findings indicate reproducible plasma metabolic differences associated with early PD and suggest the potential diagnostic value of internally validated classifiers for disease diagnosis. Alterations in gut microbiota-derived metabolites correlate with clinical severity, highlighting the need for further mechanistic and translational research. FUNDING:This study was supported by the National Natural Science Foundation of China (82271281 and 82471267), the Science and Technology Major Project of Hunan Provincial Science and Technology Department (2021SK1010), and the National Key Research and Development Program of China (2021YFC2501204).

Parkinson's disease (PD) is the second most common neurodegenerative disease and the fastest-growing disability-causing neurological disorder worldwide. Based on the CPDR cohort from 19 clinical centers, we summarized the mortality information and characteristics of patients with PD, and analyzed the related factors affecting their survival. After a 6-year follow-up period, 562 of the 3,148 patients died, with a mortality rate of 3.03 deaths per 100 person-years, and a median survival time from disease onset of 23.33 years. The most common cause of death was cardiovascular disease, followed by cerebrovascular disease and respiratory disease. Older age at onset, carriers of GBA1 gene variants, type 2 diabetes, higher LEDD, late H Y stage (especially H Y stage 4 and H Y stage 5), higher UPDRS part Ⅲ scores, a history of falls, depression, and cognitive dysfunction were associated with increased mortality. In contrast, undergoing deep brain stimulation (DBS) surgery and higher educational attainment was associated with a lower risk of death. Our findings contributed to further expanding the survival data of PD and advocated for early identification of high-risk patients for timely intervention to improve prognosis.

INTRODUCTION:Progressive supranuclear palsy (PSP) is a disabling movement disorder disease characterized by prominent synaptic loss in its early stage. The present study utilized 18F-SynVest-1 positron emission tomography (PET) in patients with PSP to assess the expression of synaptic vesicle glycoprotein 2A (SV2A), a marker of synaptic density, with the aim of diagnosing PSP patients. METHODS:A total of 13 patients with PSP and 11 healthy controls were enrolled, and they underwent 18F-SynVest-1 PET scan and brain MRI scan. Voxel-based and region-of-interest (ROI)-based analysis were performed to compare synaptic density between PSP patients and controls. Receiver operating curve (ROC) analysis was used to evaluate the diagnostic efficacy of synaptic density. Spearman correlation analysis was employed to explore the correlation between synaptic density, disease duration, and disease severity. RESULTS:PSP patients exhibited significantly lower synaptic density in the midbrain, basal ganglia, thalamus, and subregions of frontal lobes compared to controls (p < 0.05). The synaptic density of left caudate effectively distinguished PSP patients from controls with an AUC of 0.94. Additionally, synaptic density showed a strong positive correlation with the midbrain to pons ratio in PSP patients (R = 0.58, p = 0.04). CONCLUSION:We applied the 18F-SynVest-1 PET in PSP patients to display synaptic density changes in vivo for the first time, and the 18F-SynVest-1 PET showed promise in early detection and monitoring disease severity of PSP.

BACKGROUND:A variable number tandem repeat (VNTR) expansion in HSF1 has recently been linked to essential tremor (ET). METHODS:We analyzed the VNTR in an existing HiFi cohort (n = 159) and subsequently in an expanded prospective case-control cohort (n = 2121) using fluorescent polymerase chain reaction (PCR). Ten size-matched case-control pairs were newly HiFi-sequenced to further characterize sequence details. A gene-burden analysis of HSF1 used retrospective whole-genome sequencing data (n = 5147). RESULTS:Fluorescent PCR genotyping showed no case-control difference in VNTR length distribution (range: 135-847 bp; P = 0.63) and no association with disease status (P = 0.93). HiFi sequencing of existing (n = 159) and new (n = 20) samples identified two core VNTR motifs (13-bp [CCGCNCCGCCTCC]n and 8-bp [CCGCCTCC]n), but no significant case-control differences in fine-scale sequence composition. Gene-burden analysis showed no enrichment of rare damaging coding variants in ET. CONCLUSION:Our data do not support HSF1 VNTR or rare damaging coding variants as major risk factors for ET. © 2026 International Parkinson and Movement Disorder Society.

Lysosomal defects are closely linked to Parkinson’s disease (PD). Mutations in the GBA1 gene, encoding the lysosomal enzyme glucocerebrosidase (GCase), are major genetic risk factors for PD. GBA1 deficiency causes lysosomal dysfunction, leading to α-synuclein (α-syn) accumulation and PD progression. However, the underlying mechanisms remain unclear. In this study, we identified a novel GBA1-KAT8 regulatory pathway that controls lysosomal activity. GBA1 overexpression enhances lysosomal enzyme expression, regulates histone H4 acetylation at K16 via KAT8, and promotes lysosome-associated gene expression, highlighting an epigenetic mechanism in lysosomal biogenesis. Furthermore, GBA1 upregulated KAT8 expression, increased lysosomal enzyme levels, and decreased PFF-induced α-syn accumulation both in vitro and in vivo. The involvement of KAT8 as a critical acetyltransferase that modulates nuclear–lysosomal signaling pathways provides a mechanistic explanation for GBA1 deficiency-induced lysosomal dysfunction in association with PD pathology.

BACKGROUND:Autosomal recessive cerebellar ataxias (ARCA) are clinically and genetically heterogeneous, often presenting as sporadic cases that pose a significant diagnostic challenge. OBJECTIVES:The aim was to characterize the clinical and mutational landscape of ARCA in the largest Chinese cohort using whole-genome sequencing (WGS). METHODS:We performed WGS on a cohort of 187 patients with suspected hereditary ataxia. Clinical severity was assessed using the Scale for the Assessment and Rating of Ataxia and the International Cooperative Ataxia Rating Scale. Pathogenicity was determined according to American College of Medical Genetics and Genomics guidelines, and variants were validated via Sanger sequencing and co-segregation analysis. RESULTS:We identified 109 variants across 56 genes, with a rate of 79.8% (87/109) being novel. Definitive molecular diagnoses were achieved in 21 cases across 11 distinct subtypes. SYNE1 and SETX were the most frequent genetic contributors, with SETX truncating mutations (c.22_23insTA) associated with higher clinical burden. Multisystemic features were identified, including conjunctival telangiectasia in ATM, progressive action myoclonus in SCARB2, and spastic paraplegia in CAPN1. CONCLUSIONS:In this study with, to date the largest Chinese cohort of ARCA, 87 novel variants were classified, which significantly expanded the genomic and phenotype landscape of ARCA in the Chinese population. © 2026 International Parkinson and Movement Disorder Society.

Parkinson's disease (PD) is a progressive neurodegenerative disorder influenced by complex genetic and environmental factors. We report that biallelic variants in hexose-6-phosphate dehydrogenase (H6PD), which encodes a key enzyme in the endoplasmic reticulum (ER) pentose phosphate pathway, contribute to PD and investigate its role in maintaining mitochondrial homeostasis. Through whole-exome sequencing of 2223 patients with PD and 1229 controls, together with whole-genome sequencing of 4010 patients and 6072 controls, we found 13 biallelic H6PD variants in eight probands, including two homozygous and six compound heterozygous cases (six early-onset PD, two late-onset PD). Functional studies were conducted using cultured cells, Drosophila, and AAV-shRNA-mediated H6PD knockdown mice. Mitochondrial function and redox status were assessed using confocal imaging, flow cytometry, and Seahorse metabolic flux analysis. ER-mitochondria contacts, Ca2⁺ dynamics, and mitophagy were evaluated using SPLICS sensors, calcium imaging, and PINK1-Parkin pathway assays. Our study revealed that H6PD depletion impaired NADPH generation, disrupted ER-mitochondria coupling, caused abnormal Ca2+ release, mitochondrial fragmentation, reduced respiratory capacity, and suppressed PINK1-Parkin-dependent mitophagy. PD-related H6PD variants lost the ability to maintain NADPH/redox balance and mitochondrial protective function. In Drosophila, H6PD loss induced dopaminergic neurodegeneration, locomotor deficits, and shortened lifespan, all partially rescued by human H6PD. Similarly, H6PD knockdown in mice aggravated MPTP-induced neuronal loss and mitochondrial abnormalities. In conclusion, our study identifies biallelic variants in H6PD as a novel cause of PD. H6PD maintains ER NADPH/redox homeostasis, stabilizes ER-mitochondria communication, and preserves mitochondrial function and mitophagy, thereby supporting dopaminergic neuron survival.