Prediction of the Age at Onset of Spinocerebellar Ataxia Type 3 with Machine Learning.

BACKGROUND: In polyglutamine (polyQ) disease, the investigation of the prediction of a patient's age at onset (AAO) facilitates the development of disease-modifying intervention and underpins the delay of disease onset and progression. Few polyQ disease studies have evaluated AAO predicted by machine-learning algorithms and linear regression methods. OBJECTIVE: The objective of this study was to develop a machine-learning model for AAO prediction in the largest spinocerebellar ataxia type 3/Machado-Joseph disease (SCA3/MJD) population from mainland China. METHODS: In this observational study, we introduced an innovative approach by systematically comparing the performance of 7 machine-learning algorithms with linear regression to explore AAO prediction in SCA3/MJD using CAG expansions of 10 polyQ-related genes, sex, and parental origin. RESULTS: Similar prediction performance of testing set and training set in each models were identified and few overfitting of training data was observed. Overall, the machine-learning-based XGBoost model exhibited the most favorable performance in AAO prediction over the traditional linear regression method and other 6 machine-learning algorithms for the training set and testing set. The optimal XGBoost model achieved mean absolute error, root mean square error, and median absolute error of 5.56, 7.13, 4.15 years, respectively, in testing set 1, with mean absolute error (4.78 years), root mean square error (6.31 years), and median absolute error (3.59 years) in testing set 2. CONCLUSION: Machine-learning algorithms can be used to predict AAO in patients with SCA3/MJD. The optimal XGBoost algorithm can provide a good reference for the establishment and optimization of prediction models for SCA3/MJD or other polyQ diseases. © 2020 International Parkinson and Movement Disorder Society.

The association between plasma GPNMB and Parkinson's disease and multiple system atrophy.

AIMS: Parkinson's disease (PD), as the second most common neurodegenerative disorder, often presents diagnostic challenges in differentiation from other forms of Parkinsonism. Recent studies have reported an association between plasma glycoprotein nonmetastatic melanoma protein B (pGPNMB) and PD. METHODS: A retrospective study was conducted, comprising 401 PD patients, 111 multiple system atrophy (MSA) patients, 13 progressive supranuclear palsy (PSP) patients and 461 healthy controls from the Chinese Han population, with an assessment of pGPNMB levels. RESULTS: The study revealed that pGPNMB concentrations were significantly lower in PD and MSA patients compared to controls (area under the receiver operating characteristics curve (AUC) 0.62 and 0.74, respectively, P < 0.0001 for both), but no difference was found in PSP patients compared to controls (P > 0.05). Interestingly, the level of pGPNMB was significantly higher in PD patients than in MSA patients (AUC = 0.63, P < 0.0001). Furthermore, the study explored the association between pGPNMB levels and disease severity in PD and MSA patients, revealing a positive correlation in PD patients but not in MSA patients with both disease severity and cognitive impairment. CONCLUSION: This study successfully replicated prior findings, demonstrating an association between pGPNMB levels and disease severity, and also identified a correlation with cognitive impairment in PD patients of the Chinese Han population. Additionally, this study is the first to identify a significant difference in pGPNMB levels between MSA, PD, and normal controls. The data provide new evidence supporting the potential role of pGPNMB in the diagnosis and differential diagnosis of Parkinsonism.

Associations between variants in levodopa metabolic pathway genes and levodopa-induced dyskinesia in Parkinson's disease.

INTRODUCTION: Levodopa-induced dyskinesia (LID) is a common motor complication in Parkinson's disease (PD). Several genes in the levodopa metabolic pathway, such as COMT, DRDx and MAO-B, were reported associated with LID. However, there has been no systematic analyses between common variants in levodopa metabolic pathway genes and LID in a large sample of the Chinese population. METHODS: Through the whole exome sequencing (WES) and target region sequencing, we aimed to explore the potential associations between common single nucleotide polymorphisms (SNPs) in the levodopa metabolic pathway and LID in Chinese PD individuals. Five hundred and two PD individuals were enrolled in our study, among them, 348 individuals underwent WES, and 154 individuals underwent target region sequencing. We acquired the genetic profile of 11 genes including COMT, DDC, DRD1-5, SLC6A3, TH and MAO-A/B. We established a stepwise strategy to filter SNPs, which finally included 34 SNPs in our analyses. And we used a two-stage study, with discovery (348 individuals with WES) and the replication (all 502 individuals) to confirm our findings. RESULTS: Among the 502 PD individuals, 104 (20.7%) were diagnosed with LID. In the discovery stage, we found that COMT rs6269, DRD2 rs6275 and DRD2 rs1076560 were associated with LID. In the replication stage, associations between the three above-mentioned SNPs and LID were still present in all 502 individuals. CONCLUSION: We demonstrated that in the Chinese population, COMT rs6269, DRD2 rs6275 and rs1076560 were significantly associated with LID. And rs6275 was reported associated with LID for the first time.

Quantitative Transcranial Sonography Evaluation of Substantia Nigra Hyperechogenicity Is Useful for Predicting Levodopa-Induced Dyskinesia in Parkinson Disease.

Levodopa-induced dyskinesia (LID) is a common motor complication in Parkinson disease (PD). Abnormal substantia nigra hyperechogenicity (SN+), detected by transcranial sonography (TCS), plays an important role in the differential diagnosis of PD. The purpose of this study was to investigate the predictive performance of quantitative SN+ evaluations for LID. Five hundred sixty-two individuals were included in our analysis, and 198 individuals were followed up. These individuals were divided into two groups at baseline: the PD with LID (PD+LID) group and the PD without LID (PD-LID) group. The association between total hyperechogenic area of the SN on both sides (SNT) and LID was analyzed by binary logistic analysis. A binary logistic regression model including SNT was applied to establish a model for discriminating LID. At baseline, 105 (18.7%) individuals were diagnosed with LID. The PD+LID group had a longer disease duration, shorter education duration, higher levodopa equivalent doses, greater disease severity and larger SNT. A model combining clinical features and SNT was further established with better efficiency (area under the receiver operating characteristic curve = 0.839). One hundred ninety-eight individuals were followed up; individuals with a larger SNT and a higher predicted probability were more likely to develop LID in our follow-up. Our study determined that quantitative TCS evaluation of SN echogenicity is useful in predicting LID in PD.

A systematic analysis of genotype-phenotype associations with PLA2G6.

BACKGROUND: PLA2G6-associated neurodegeneration (PLAN) can be categorized into infantile neuroaxonal dystrophy (INAD), atypical neuroaxonal dystrophy (aNAD), neurodegeneration with brain iron accumulation (NBIA), and early-onset parkinsonism (EOP). OBJECTIVES: To determine the genotype-phenotype association in PLAN. METHODS: "PLA2G6" or "PARK14" or "phospholipase A2 group VI" or "iPLA2ß" were searched across MEDLINE from June 23, 1997, to March 1, 2023. A total of 391 patients were identified, and 340 patients of them were finally included in the assessment. RESULTS: The loss of function (LOF) mutation ratios were significantly different (p < 0.001), highest in INAD, followed by NBIA, aNAD, and EOP. Four ensemble scores (i.e., BayesDel, VARITY, ClinPred, and MetaRNN) were assessed to predict the deleteriousness of missense mutations and demonstrated significant differences (p < 0.001). Binary logistic regression analyses demonstrated that LOF mutations were independently associated with brain iron accumulation (p = 0.006) and ataxia (p = 0.025). CONCLUSIONS: LOF or more deleterious missense mutations are more likely to promote the development of serious phenotype of PLAN, and LOF mutations are independently associated with brain iron accumulation and ataxia.

Maillard reaction between oligopeptides and reducing sugar at body temperature: The putative anti-glycation agents.

Type 2 Diabetes mellitus (T2DM) is one of the most common chronic multifactorial diseases, which is associated with the increased concentration of glucose in the blood. Therefore, the utilization of blood lowering agents is clearly a promising approach which can lead to a suppression of the evaluated blood glucose, and thus curing T2DM and other complication. In this study, we evaluated the glucose lowering effect of a varieties of amino acids (alanine and histidine), dipeptides (carnosine and α-alanine-L-histidine), and tripeptide (glutathione) by reacting with glucose, fructose, and sucrose under 37°C and pH 7.4 to mimic their reaction in physiological condition. By measuring the reduction of reactants and the formation of Maillard reaction products over the course of 21 days' storage, we found that the glucose lowering effect of carnosine was better than the counterparts. The histidine residue in carnosine may contribute to its glucose lowing effect while ß-amino acid ß-alanine residue could facilitate the glucose lowering effect of carnosine by maintaining its chemical stability during the storage. These results may open up new avenues for the applications of bioactive peptide carnosine as a natural blood sugar lowering agent to control T2DM.

Cognitive Performance is Associated with Altered Cerebral Hemodynamics Assessed by Transcranial Ultrasound in Parkinson's Disease.

Purpose: Cognitive impairment (CI) is a common but debilitating non-motor symptom in Parkinson's disease (PD). Although cerebrovascular functions are related to cognitive performance in healthy individuals, such a relation in PD remains elusive. This study aims to assess the association between cerebrovascular function and cognitive performance in PD individuals. Patients and Methods: Two-hundred-and-one PD individuals were retrospectively included. They were subsequently divided into two groups: PD with normal cognition (PD-NC) and PD with CI (PD-CI). Cerebral hemodynamic characteristics of the middle cerebral arteries were assessed by transcranial ultrasound. The association between scores in each cognitive domain and cerebral hemodynamic parameters was further analyzed using regression analyses. Additionally, a binary logistic regression model with backward stepwise procedure was applied to build the model for discriminating CI in PD individuals. An independent dataset of additional 46 PD individuals was used further. Results: The PD-CI group showed a relatively lower end-diastolic blood flow velocity (EDV, p < 0.05) and a higher resistive index (RI, p < 0.05) compared to the PD-NC group. RI showed significant associations with the memory item score of Montreal Cognitive Assessment (p < 0.05). A model combining clinical and hemodynamic variables was established with optimal efficiency (area under the curve, AUC = 0.651). Further replication of the model in an independent dataset yielded a great consistency (AUC = 0.704). Conclusion: In our study, cerebrovascular functions were significantly associated with the cognitive performance in PD individuals, especially with the memory task. The established model was effective in identifying CI in PD individuals, which might be a potentially useful tool to screen the cognitive decline in PD individuals at an early stage of the disease. Further studies with larger sample sizes in different populations are warranted.

The Natural History of Spinocerebellar Ataxia Type 3 in Mainland China: A 2-Year Cohort Study.

Objective: The natural history of spinocerebellar ataxia type 3 (SCA3) has been reported in several populations and shows heterogeneity in progression rate and affecting factors. However, it remains unexplored in the population of Mainland China. This study aimed to identify the disease progression rate and its potential affecting factors in patients with SCA3 in Mainland China. Participants and Methods: We enrolled patients with genetically confirmed SCA3 in Mainland China. Patients were seen at three visits, i.e., baseline, 1 year, and 2 years. The primary outcome was the Scale for the Assessment and Rating of Ataxia (SARA), and the secondary outcomes were the Inventory of Non-Ataxia Signs (INAS) as well as the SCA Functional Index (SCAFI). Results: Between 1 October 2015, and 30 September 2016, we enrolled 263 patients with SCA3. We analyzed 247 patients with at least one follow-up visit. The annual progression rate of SARA was 1.49 points per year (SE 0.08, 95% confidence interval [CI] 1.33-1.65, p < 0.0001). The annual progression rates of INAS and SCAFI were 0.56 points per year (SE 0.05, 95% CI 0.47-0.66, p < 0.001) and -0.30 points per year (SE 0.01, 95% CI -0.33∼-0.28, p < 0.001), respectively. Faster progression in SARA was associated with longer length of the expanded allele of ATXN3 (p < 0.0001); faster progression in INAS was associated with lower INAS at baseline (p < 0.0001); faster decline in SCAFI was associated with shorter length of the normal allele of ATXN3 (p = 0.036) and higher SCAFI at baseline (p < 0.0001). Conclusion: Our results provide quantitative data on the disease progression of patients with SCA3 in Mainland China and its corresponding affecting factors, which could facilitate the sample size calculation and patient stratification in future clinical trials. Trial Registration: This study was registered with Chictr.org on 15 September 2015, number ChiCTR-OOC-15007124.

Age is an important independent modifier of SCA3 phenotype severity.

OBJECTIVE: This study aimed to investigate factors modulating spinocerebellar ataxia type 3 (SCA3) phenotype severity besides the expanded CAG repeats (ExpCAG) of ATXN3. METHODS: Data regarding CAG trinucleotide repeats, age at onset (AO), duration, age, sex, transmitting parent, and scale scores of SCA3 patients were collected. Multiple linear regression analysis was performed to identify influential independent variables. Age, AO, ExpCAG, and duration were considered control variables to analyze the correlation between independent variables and scale scores. RESULTS: Duration, age, and ExpCAG were screened as influential independent variables (P = 0.000). Age had the greatest impact on multiple linear regression models (P<5E-8). ExpCAG and SARA/ICARS/INAS/Barthel index were not correlated (P > 0.05); considering only age as the control, ExpCAG was slightly-to-moderately correlated with all aforementioned scores except INAS (P < 0.05). Age and all scores, except INAS, were positively correlated (P < 0.05); considering duration, AO, or ExpCAG as controls, their correlations did not change significantly. On controlling age, AO was negatively correlated with all scores (P < 0.05), except for the Barthel index (P > 0.05). Furthermore, the interaction model revealed that the interaction between age, duration, and ExpCAG was significantly associated with SCA3 disease severity (P < 0.05). CONCLUSION: Age is a potentially important modifier of SCA3 phenotype severity, through the interaction between ExpCAG and aging factors.

Corrigendum: Alterations of the Gut Microbiota in Multiple System Atrophy Patients.

[This corrects the article DOI: 10.3389/fnins.2019.01102.].

Association of serum neurofilament light and disease severity in patients with spinocerebellar ataxia type 3.

OBJECTIVE: To investigate serum neurofilament light protein (sNfL) levels in patients with spinocerebellar ataxia type 3 (SCA3) and to determine whether they are associated with disease severity. METHODS: This cross-sectional study enrolled 185 healthy controls and 235 ATXN3 mutation carriers (17 asymptomatic stage, 20 preclinical stage, and 198 ataxic stage). We measured sNfL levels with the single molecule array (Simoa) platform. Clinical disease severity was assessed using the Scale of Assessment and Rating of Ataxia (SARA) and the Inventory of Nonataxia Signs (INAS). In a subgroup of 50 ataxic stage patients, we further evaluated the gray matter volume and the integrity of white matter fibers by MRI. RESULTS: sNfL concentrations were elevated in asymptomatic, preclinical, and ataxic ATXN3 mutation carriers compared to controls (12.18 [10.20-13.92], 21.84 [18.37-23.45], 36.06 [30.04-45.90], and 8.24 [5.92-10.84] pg/mL, median [interquartile range], respectively, p < 0.001). sNfL correlated with SARA (r = 0.406, 95% confidence interval [CI] 0.284-0.515, p < 0.0001) and INAS (r = 0.375, 95% CI 0.250-0.487, p < 0.0001), and remained significant after adjustment for age and CAG repeats. In addition, we observed negative correlations of the sNfL with gray matter volume in the left precentral gyrus and the left paracentral lobule as well as with the mean diffusivity in widespread white matter tracts. CONCLUSION: Our results demonstrate that sNfL levels are increased in SCA3 and are associated with clinical disease severity, which supports sNfL as a biomarker for disease severity in SCA3. CLASSIFICATION OF EVIDENCE: This study provides Class II evidence that in patients with SCA3, sNfL elevations are associated with clinical disease severity.

RNA Expression Profile and Potential Biomarkers in Patients With Spinocerebellar Ataxia Type 3 From Mainland China.

Long non-coding RNAs (lncRNAs) play an important role in growth, development, and reproduction and undoubtedly contribute to the pathogenesis and progression of diseases. Emerging evidence suggests the involvement of lncRNAs as regulatory factors in pathological conditions, including some neurodegenerative diseases. Spinocerebellar Ataxia Type 3/Machado-Joseph Disease (SCA3/MJD) has a prominent prevalence in China. Because the role of lncRNAs in SCA3/MJD pathogenesis has not yet been investigated, we conducted a pilot study to investigate the expression profile of lncRNAs by high-throughput sequencing in 12 patients and 12 healthy individuals. The sequencing analysis detected 5,540 known and 2,759 novel lncRNAs. Six lncRNAs were confirmed to be differentially expressed in peripheral blood mononuclear cells between SCA3/MJD patients and healthy individuals and were further validated in cerebellar tissue. Based on these results, NONHSAT022144.2 and NONHSAT165686.1 may be involved in the pathogenesis of SCA3/MJD and may be potential biomarkers for SCA3/MJD. Together with NONHSAT022144.2 and NONHSAT165686.1, the other four novel lncRNAs increase our understanding of lncRNA expression profile.

Alterations of the Gut Microbiota in Multiple System Atrophy Patients.

Multiple system atrophy (MSA) is a fatal neurodegenerative disease, and the pathogenesis is still quite challenging. Emerging evidence has shown that the brain-gut-microbiota axis served a pivotal role in neurological diseases; however, researches utilizing metagenomic sequencing to analyze the alteration in gut microbiota of MSA patients were quite rare. Here, we carried out metagenomic sequencing in feces of 15 MSA patients and 15 healthy controls, to characterize the alterations in gut microbial composition and function of MSA patients in mainland China. The results showed that gut microbial community of MSA patients was significantly different from healthy controls, characterized by increased genus Akkermansia and species Roseburia hominis, Akkermansia muciniphila, Alistipes onderdonkii, Streptococcus parasanguinis, and Staphylococcus xylosus, while decreased genera Megamonas, Bifidobacterium, Blautia, and Aggregatibacter and species Bacteroides coprocola, Megamonas funiformis, Bifidobacterium pseudocatenulatum, Clostridium nexile, Bacteroides plebeius, and Granulicatella adiacens. Further, functional analysis based on the KEGG database revealed aberrant functional pathways in fecal microbiome of MSA patients. In conclusion, our findings provided evidence for dysbiosis in gut microbiota of Chinese MSA cohorts and helped develop new testable hypotheses on pathophysiology of MSA.

Polymorphisms in DNA methylation-related genes are linked to the phenotype of Machado-Joseph disease.

DNA methylation has been reported as an important regulator of genomic structure stability, including large tandem repeats. To test the modulation effect of variants in DNA methylation-related genes on distribution of expanded (CAG)n alleles and age at onset (AO) of patients with Machado-Joseph disease (MJD), we conducted an association analysis on 23 selected SNPs in these genes in 613 patients with MJD and 581 controls. There were significant differences in the distribution of rs12957023 between patients and controls (OR = 1.296, p = 0.007 and OR = 1.206, p = 0.008, for genotype and alleles, respectively). The distribution of (CAG)n size was also different between patients carrying a CC and the other genotypes (TT and TC, p = 0.011 for expanded (CAG)n and p = 0.012 for normal size alleles), indicating that DNA methylation might modulate the (CAG)n instability. We found also that rs13420827 in DNMT3A and rs7354779 in DNMT3L contribute to AO of MJD (p = 0.019 and p = 0.008, respectively). In conclusion, our data provide the first evidence that SNPs in DNA methylation-related genes may contribute to (CAG)n instability and modulate the AO of this disease.

Investigation on modulation of DNA repair pathways in Chinese MJD patients.

It has been reported that DNA repair pathways could modify age at onset (AO) in Huntington disease (HD) and spinocerebellar ataxias. We genotyped 22 SNPs from DNA repair pathways in a large cohort of 798 Chinese Machado-Joseph disease patients to investigate the association with AO, and no significant finding was observed. Our findings did not provide a strong evidence for the modulatory effect of DNA repair pathways on the AO of Chinese Machado-Joseph disease patients. Further analyses with more representative DNA repair-related SNPs in different populations are needed to identify new potential genetic modifiers.

Identifying SYNE1 Ataxia With Novel Mutations in a Chinese Population.

Objective: Variants in SYNE1 have been widely reported in ataxia patients in Europe, with highly variable clinical phenotype. Until now, no mutation of SYNE1 ataxia has been reported among the Chinese population. Our aim was to screen for SYNE1 ataxia patients in China and extend the clinicogenetic spectrum. Methods: Variants in SYNE1 were detected by high-throughput sequencing on a cohort of 126 unrelated index patients with unexplained autosomal recessive or sporadic ataxia. Pathogenicity assessments of SYNE1 variants were interpreted according to the ACMG guidelines. Potential pathogenic variants were confirmed by Sanger sequencing. Clinical assessments were conducted by two experienced neurologists. Results: Two Chinese families with variable ataxia syndrome were identified (accounting for 1.6%; 2/126), separately caused by the novel homozygous SYNE1 mutation (NM_033071.3: c.21568C>T, p.Arg7190Ter), and compound heterozygous SYNE1 mutation (NM_033071.3: c.18684G>A, p.Trp6228Ter; c.17944C>T, p.Arg5982Ter), characterized by motor neuron impairment, mental retardation and arthrogryposis. Conclusions: SYNE1 ataxia exists in the Chinese population, as a rare form of autosomal recessive ataxia, with a complex phenotype. Our findings expanded the ethnic, phenotypic and genetic diversity of SYNE1 ataxia.

Is the High Frequency of Machado-Joseph Disease in China Due to New Mutational Origins?

Machado-Joseph disease (MJD, also known as spinocerebellar ataxia 3 or SCA3) is the most common dominant ataxia worldwide, with an overall average prevalence of 1-5/100,000. To this date, two major ancestral lineages have been found throughout the world. In China, the relative frequency of MJD among the SCAs reaches as high as 63%, however, little is known about its mutational origin in this country. We analyzed 50 families with MJD patients in two or more generations to study the hypothesis that new mutational events have occurred in this population. Haplotypes based on 20 SNPs have shown new genetic backgrounds segregating with MJD mutations in our cohort from China. We found the "Joseph-derived" lineage (Joseph lineage with a G variant in rs56268847) to be very common among Chinese MJD patients. Moreover, we estimated the time for the origin of this MJD SNP background based on STR diversity flanking the (CAG)n of ATXN3. It was surprising to find that the Chinese MJD population originated from 8,000 to 17,000 years ago, far earlier than the previous literature reports, which will be an important evidence to explain the origin, spread and founder effects of MJD.

Cerebellar lncRNA Expression Profile Analysis of SCA3/MJD Mice.

Spinocerebellar ataxia type 3 (SCA3) or Machado-Joseph disease (MJD) is the most common autosomal dominant spinocerebellar ataxia in China with highly clinical heterogeneity, such as progressive cerebellar ataxia, dysarthria, pyramidal signs, external ophthalmoplegia, dysphagia, and distal muscle atrophy. It is caused by the abnormal expansion of CAG repeats in a coding region of ATXN3. However, by focusing on the ATXN3 itself cannot fully explain the heterogeneous clinical features of SCA3/MJD. With the discovery of the increasing number of long noncoding RNAs (lncRNAs) that are believed to be involved in spinocerebellar ataxia type 8 (SCA8) and Huntington disease (HD), we wonder whether the lncRNAs are differentially expressed in the SCA3/MJD patients compared to the nonpatients. As the first step, we used lncRNA-Seq to investigate differential expression of the lncRNAs in the SCA3/MJD mice. Two known lncRNAs, n297609 and n297477, and a novel lncRNA TCONS_00072962 have been identified in SCA3/MJD mice with abnormal expression. The first discovery of the novel lncRNA TCONS_00072962 enriched the lncRNA expression profile in the SCA3/MJD mouse model.

Novel mutations in ADSL for Adenylosuccinate Lyase Deficiency identified by the combination of Trio-WES and constantly updated guidelines.

Whole-exome sequencing (WES), one of the next-generation sequencing (NGS), has become a powerful tool to identify exonic variants. Investigating causality of the sequence variants in human disease becomes an important part in NGS for the research and clinical applications. Recently, important guidelines on them have been published and will keep on updating. In our study, two Chinese families, with the clinical diagnosis of "Epilepsy", which presented with seizures, psychomotor retardation, hypotonia and etc. features, were sequenced by Trio-WES (including the proband and the unaffected parents), and a standard interpretation of the identified variants was performed referring to the recently updated guidelines. Finally, we identified three novel mutations (c.71 C > T, p.P24L; c.1387-1389delGAG, p.E463-; c.134 G > A, p.W45*; NM_000026) in ADSL in the two Chinese families, and confirmed them as the causal variants to the disease-Adenylosuccinate Lyase Deficiency. Previous reported specific therapy was also introduced to the patients after our refined molecular diagnosis, however, the effect was very limited success. In summary, our study demonstrated the power and advantages of WES in exploring the etiology of human disease. Using the constantly updated guidelines to conduct the WES study and to interpret the sequence variants are a necessary strategy to make the molecular diagnosis and to guide the individualized treatment of human disease.