Abstract
Multiple system atrophy (MSA) alters skin physiology, potentially impacting skin microbiota. This pilot study investigated whether skin microbiota differs in MSA and whether these differences relate to disease severity. Using 16S rRNA sequencing of cervical and axillary sites in MSA, Parkinson’s disease, and controls, we identified distinct microbial patterns among groups. These patterns allowed cohort classification and showed strong associations with clinical symptoms, suggesting disease-related microbial alterations in MSA.
Subject terms: Microbiology, Neurology
Introduction
Multiple system atrophy (MSA) is a fatal neurodegenerative disease characterized by progressive autonomic dysfunction, cerebellar ataxia, and Parkinsonism1,2. It is subclassified into MSA-P (Parkinsonian) and MSA-C (cerebellar) based on predominant motor symptoms3. Distinguishing MSA from Parkinson’s disease (PD) and other atypical Parkinsonian disorders remains clinically challenging due to overlapping clinical features and lack of sensitive and reliable biomarkers4–6. This diagnostic difficulty is compounded by the incomplete understanding of the disease’s underlying pathophysiology and molecular mechanisms, which hinders the development of effective treatments.
Emerging evidence highlights the role of the human microbiome in modulating neurological health and disease. While the gut microbiome has been implicated in neurodegenerative conditions, including MSA7–9, research on other microbial niches remains sparse. The skin, the body’s largest organ and outermost barrier, hosts a distinct microbiota that regulates immune responses, supports epithelial differentiation, and protects against pathogens10. In MSA, autonomic dysfunction frequently leads to altered thermoregulation (e.g., hyperhidrosis, hypohidrosis, and hypothermia)11,12, which is likely to disturb the skin’s microenvironment, including sweat composition, pH, and moisture levels. These changes may, in turn, reshape the skin microbial community13. Given the skin microbiome’s sensitivity to systemic physiological changes and its accessibility for non-invasive sampling, it may help reveal disease-associated alterations and contribute to a better understanding of MSA pathophysiology.
Skin microbiome dysbiosis in MSA
In this pilot study, we conducted 16S rRNA sequencing on cervical and axillary skin samples from 31 MSA patients, 20 PD patients, and 30 health controls (HC), aiming to compare the skin microbial community structures across these groups. The demographic and clinical characteristics of the subjects are summarized in Supplementary Table 1. Skin swabs were sampled from the cervical (i.e., neck) and axillary (i.e., armpit) sites of all subjects. Sequencing yielded a depth of 107,946 reads per sample. After DADA2 processing, 11,554,262 high-quality reads were obtained from 162 samples, and normalization was based on the sample with the lowest sequencing depth. The microbial diversity and composition differed significantly between the cervical and axillary sites (Supplementary Fig. 1), so all downstream analyses were conducted separately by skin sites. A total of 11,334 amplicon sequence variants (ASVs) and 5,379 ASVs were identified from the cervical and axillary samples, respectively (Fig. 1A, B). In the cervical microbiota, the most abundant genera were Cutibacterium, Staphylococcus, Paracoccus, and Enhydrobacter, whereas Corynebacterium and Staphylococcus dominated the axillary microbiota (Fig. 1C, Supplementary Table 2).
Fig. 1. Altered skin microbiota of study participants based on 16S rRNA data.
Venn diagram of observed ASVs in A cervical and B axillary microbiota across MSA, PD, and HC groups. C Average relative abundances of microbial genus in axillary and cervical regions among three groups. Cladograms generated by LEfSe showing bacterial taxa with significant differences among the three groups in D cervical and E axillary microbiota. F Relative abundance distribution of the most significantly altered taxa in the cervical and axillary microbiota of MSA patients compared with HC or PD patients. The prefix ‘n_’ denotes taxa originating from the cervical (neck) microbiota, while the prefix ‘a_’ represents taxa from the axillary (armpit) microbiota.
Alpha diversity, reflecting within-sample microbial diversity, measured by the Simpson and Shannon indices did not differ across 3 groups at either the cervical or axillary site (Supplementary Fig. 2A, B). Analysis of community composition using principal coordinate analysis (PCoA) based on weighted Bray–Curtis distances revealed group differences in the axillary microbiota, but not in the cervical microbiota (Supplementary Fig. 2C, D). Differential abundance analysis using linear discriminant analysis Effect Size (LEfSe) revealed abundance differences in microbiota composition across three groups at specific taxonomic levels, both in the neck and axillary (Fig. 1D, E). At the genus level, the MSA group exhibited significantly higher abundances of Bacillus, Prevotella_7, and Streptococcus in the cervical microbiota compared to HC. In contrast, Brachybacterium and Janibacter were significantly reduced. Additionally, Corynebacterium was more abundant in MSA patients than in those with PD. In the axillary microbiota, MSA exhibited increased levels of Corynebacterium and Sandaracinobacter compared to HC. Compared to PD patients, MSA also exhibited higher levels of Corynebacterium but lower levels of Finegoldia (Fig. 1F).
Functional adaptations in MSA
Using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2), we inferred the functional composition of the skin microbiota from 16S sequencing data. Permutational multivariate analysis of the Kyoto Encyclopedia of Genes and Genomes (KEGG) Orthology (KO) level 3 pathway abundances revealed significant alterations in metabolism-related pathways in both cervical and axillary microbiota. Differences among the three groups also involved pathways related to genetic information processing, human diseases, environmental information processing, and organismal systems (Supplementary Fig. 3).
Skin microbiota correlation with MSA
To investigate the relationship between the skin microbiota and clinical phenotypes, we performed a co-occurrence network analysis linking microbial ASVs with clinical characteristics. The results showed that cognitive scores, including the Montreal Cognitive Assessment (MoCA), the frontal assessment battery (FAB), and the mini-mental state examination (MMSE), exhibited high centrality within the axillary microbiota network, suggesting potential associations between cognitive function and the axillary microbial community. (Supplementary Fig. 4). In contrast, non-motor function scales, such as the REM Sleep Behavior Disorder Screening Questionnaire (RBDSQ) and the Scales for Outcomes in PD-Autonomic (SCOPA-AUT) questionnaire, were more prominently associated with the cervical microbiota network, indicating a potential association between the cervical microbiota and autonomic dysfunction in MSA patients. Spearman correlation analysis was performed to further investigate the relationship between differentially abundant taxa identified in LEfSe and the clinical phenotypes of MSA patients. The results revealed significant correlations between the skin microbiome and clinical phenotypes (Fig. 2A). Specifically, the relative abundances of Brachybacterium, Janibacter, Prevotella_7, and Streptococcus in the cervical microbiota were positively correlated with motor function scales including the Hoehn and Yahr Parkinson’s disease staging scheme (H&Y stage) and the Unified MSA rating scale (UMSARS), while Brachybacterium and Streptococcus were negatively correlated with cognitive scores like FAB or MoCA. In the axillary microbiota, Sandaracinobacter was negatively correlated with non-motor function scales, including Wexner score and RBDSQ. These findings suggest that alterations in the skin microbiome may reflect specific clinical features and disease progression of MSA.
Fig. 2. The potential of altered skin microbiota as a biomarker for MSA.
A Heatmap of Spearman’s correlation coefficients between clinical characteristics and the relative abundance of significantly altered skin microbiota in the cervical and axillary microbiomes. B, C Receiver operating characteristic (ROC) curve and Shapley value plot for MSA and HC classification. D, E ROC curve and Shapley value plot for MSA and PD classification. The differential diagnostic potential of microbiological markers in distinguishing MSA from HC or PD is reflected in the validation set area under the curve (AUCs). The microbial markers distinguishing the two groups are displayed in the plot on the right. The prefix ‘n_’ denotes taxa originating from the cervical (neck) microbiota, while the prefix ‘a_’ represents taxa from the axillary (armpit) microbiota.
Skin microbiota distinguishing MSA
To explore whether skin microbiota profiles could aid in distinguishing MSA from controls, we constructed two classification models using a 7:3 training-validation split. Least absolute shrinkage and selection operator (LASSO) regression was used for feature selection (Supplementary Fig. 5), and a support vector machine (SVM), suitable for small-sample, high-dimensional classification, was employed. The model effectively distinguished between MSA and HC (training AUC = 0.903, validation AUC = 0.852), as well as MSA and PD (training AUC = 0.838, validation AUC = 0.753). To interpret feature importance, features were ranked by Shapley Additive exPlanations (SHAP) values, illustrating their impact on model predictions (Fig. 2B–E).
Emerging evidence suggests that the human microbiome plays a critical role in the onset and progression of various diseases, including neurodegenerative disorders14,15. Yet its role in MSA remains largely uncharacterized. As the largest organ, the skin serves as a critical immune and metabolic interface, with resident microbes playing key regulatory roles16. In MSA, α-synuclein accumulation may disrupt the skin environment, reshaping microbial communities and contributing to disease through immune modulation, metabolic byproducts, and neuroinflammation17. These changes may contribute to MSA pathophysiology through metabolic byproducts, immune modulation, and neuroinflammation18–20. Consistent with this, our study identified significant microbial changes in the cervical and axillary regions of MSA patients. A defining feature of MSA is early and severe autonomic dysfunction, including impaired thermoregulation, abnormal sweating, and altered skin moisture, largely driven by α-synuclein deposition in autonomic nerves21. These physiological disturbances may directly affect skin microenvironment, creating conditions that favor microbial imbalance.
Dysbiosis of the skin microbiota in MSA may not merely reflect disease pathology but could also potentially influence the host in a reciprocal manner. Altered microbial communities could influence host physiology by producing pro-inflammatory metabolites and modulating immune responses22. Notably, enrichment of Corynebacterium in the axillary microbiota of MSA patients likely reflects underlying autonomic dysfunction, where abnormal sweat gland activity and skin moisture create a niche favoring its growth. In turn, Corynebacterium can metabolize sweat components into volatile compounds13, which may further impair neurological function and exemplify the bidirectional interplay between host pathology and microbial ecosystems. Functional predictions via PICRUSt2 showed enrichment of metabolic and environmental adaptation pathways in MSA-associated microbiota, though such findings require cautious interpretation due to the limits of 16S rRNA analysis.
Co-occurrence network and Spearman correlation analyses revealed significant associations between skin microbiota alterations and MSA clinical manifestations. The abundance of certain bacterial taxa was correlated with disease severity, suggesting that altered skin microbiota may reflect aspects of disease status. Machine learning classification models based on skin microbiome profiles effectively distinguished MSA patients from HCs (validation AUC = 0.852) and PD patients (validation AUC = 0.753). While the model showed strong performance, the lower discriminatory power between MSA and PD likely reflects overlap in clinical features and microbial profiles. While these preliminary findings are promising, further validation in larger, independent cohorts is required before any clinical utility can be established.
In conclusion, our study provides preliminary evidence that skin microbiota in MSA patients differs from that of HCs and PD patients, suggesting a disease-specific profile. These findings highlight the potential associations between skin microbiota composition and MSA. However, several limitations should be noted. The present study cannot establish causal relationships between skin microbiota and MSA, which requires further experimental validation. The small sample size may also limit statistical power and generalizability. Additionally, 16S rRNA sequencing does not resolve species or strain-level differences. Future studies integrating metagenomics and metabolomics with larger MSA cohorts are warranted to clarify host–microbe interactions and their relevance to disease pathology.
Methods
Patients and study design
In this cross-sectional study, 31 Multiple system atrophy (MSA) patients were recruited from Xiangya Hospital of Central South University. All patients were diagnosed with clinically established or clinically probable MSA according to the 2022 Movement Disorder Society Criteria2, which include autonomic failure, Parkinsonism, or cerebellar ataxia in conjunction with supportive clinical and imaging findings. As disease controls, 20 Parkinson’s disease (PD) patients were enrolled in the study based on established diagnostic criteria23. Diagnoses were established by two experienced neurologists, with an independent blind diagnosis from a third movement disorders expert included in cases of discrepancy. Additionally, 30 gender and age-matched healthy controls (HCs) from spouses or siblings of patients were included. Exclusion criteria for all participants encompassed a history of stroke, brain surgery, or unrelated cerebrovascular disease; a history of dermatitis, psoriasis, vitiligo, or other dermatological diseases; and usage of antibiotics or steroids. The study received approval from the Ethics Committee of Xiangya Hospital of Central South University (No. 2024060699) and all participants provided informed consent.
Clinical assessments
A thorough neurological examination was conducted on all MSA patients by two experienced neurologists. The unified MSA rating scale (UMSARS) served as the primary assessment tool for evaluating the disease severity of MSA patients, including four parts (I. Activities of Daily Living; II. Motor Examination Scale; III. Orthostatic hypotension; and IV. Disability Scale)24. The total UMSARS score was defined as the sum of parts I and II, with higher scores indicating a more severe condition. Hoehn and Yahr Parkinson’s disease staging scheme (H&Y stage) was used to grade the overall disease severity25. In terms of autonomic nervous system assessment, the scales for outcomes in PD-autonomic (SCOPA-AUT) questionnaire was used to evaluate MSA patients from six domains: gastrointestinal, urinary, cardiovascular, thermoregulatory, pupillomotor, and sexual, while the Wexner score was used to evaluate the anal incontinence and constipation specifically26. Meanwhile, the rapid-eye movement sleep behavior disorder screening questionnaire (RBDSQ) was used for the diagnosis and recruitment of patients with RBD27. Moreover, the mini-mental state examination (MMSE) and the Montreal cognitive assessment (MoCA) score was used for overall cognitive function evaluation, and the frontal assessment battery (FAB) was used for frontal lobe function evaluation in patients28,29. In addition, the patient health questionnaire-9 (PHQ-9) and the seven-item general anxiety disorder scale (GAD-7) were used to assess the frequency of recent depression-related symptoms in patients30.
Sample collection
Sterile swabs were used to collect microbiota from the cervical and axilla. Participants avoided antiperspirants and deodorants before sampling. Each swab, pre-soaked in saline with 0.1% Tween 20, was applied to the axillary or cervical region for 10 seconds, then placed in a 1.5 ml Eppendorf tube with preservation buffer. Samples were refrigerated and transported promptly. Negative controls, consisting of sterile swabs that did not touch the skin, were included and processed identically to the skin swab samples. In the lab, swabs were vortexed for 5 minutes and then removed with sterile forceps. The microbial suspension was centrifuged at 13,000 rpm for 5 minutes, after which the pellet was used for DNA extraction.
DNA extraction and 16S rRNA sequencing
Total genomic DNA samples were extracted using the OMEGA Soil DNA Kit (M5635-02) (Omega Bio-Tek, Norcross, GA, USA), following the manufacturer’s instructions. The quantity and quality of extracted DNAs were measured using a NanoDrop NC2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Due to the low DNA yield from negative controls, these samples were not subjected to the downstream library preparation. Primer pair (27F: 5’- AGAGTTTGATCCTGGCTCAG-3’/ 533 R: 5’- TTACCGCGGCTGCTGGCAC-3’) was used to amplify 16S rRNA V1-V3 region. Sample-specific 7-bp barcodes were incorporated into the primers for multiplex sequencing. The PCR components contained 5 μl of buffer (5×), 0.25 μl of Fast pfu DNA Polymerase (5U/μl), 2 μl (2.5 mM) of dNTPs, 1 μl (10 uM) of each Forward and Reverse primer, 1 μl of DNA template, and 14.75 μl of ddH2O. Thermal cycling conditions included 98 °C for 2 min and 25 cycles (98 °C for 15 s, 55 °C for 30 s, and 72 °C for 30 s), and a final extension of 5 min at 72 °C. Amplified PCR products were purified using Vazyme VAHTSTM DNA Clean Beads (Vazyme, Nanjing, China). All purified PCR products were further quantified with the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, USA), and sequenced on the Illumina MiSeq platform. Pair-end (2 × 300 base pair) sequencing was conducted according to NovaSeq 6000 SP Reagent Kit (500 cycles) (Illumina, San Diego, USA).
Data preprocessing
Raw sequence reads were quality-filtered and truncated before amplicon sequence variants (ASVs) were inferred using DADA2 (v.1.32.0)31. The V1–V3 region amplicon was approximately 506 bp in length. Sequencing was performed using a 2 × 300 bp paired-end protocol, providing sufficient overlap for accurate merging of forward and reverse reads. After quality filtering and merging, the average length of merged reads was approximately 460 bp. After chimera removal via the “consensus” strategy, quality-controlled reads were merged and clustered into ASVs. Taxonomic classification was performed using a naïve Bayes classifier trained on the SILVA v138 database32. Potential contaminant ASVs were identified using the frequency-based method from the decontam package (v. 1.24.0), which detects contaminants based on their inverse relationship with total DNA concentration of each sample33. The ASV abundance table, taxonomic classifications, phylogenetic tree, and metadata were compiled into a phyloseq (v.1.48.0) object for downstream analyses.
Statistical analysis
For clinical data, the Shapiro-Wilk test and Levene test were respectively used for normal distribution and the homogeneity of variance test of continuous data. Continuous data are presented as mean ± standard deviation or median (interquartile range), depending on the distribution of the data. All categorical data are presented as counts and percentages. To improve robustness to outliers, the Mann–Whitney U test or Kruskal–Wallis test was used for comparisons of continuous variables between groups, while the Chi-square test or Fisher’s exact test was applied for categorical variables.
For 16S rRNA sequencing data, alpha- and beta-diversity analyses were used to examine differences in skin microbial community structures. Alpha diversity was measured by the Shannon and Simpson indices using the phyloseq package. Beta diversity was estimated using Bray–Curtis distances and visualized using Non-metric Multidimensional scaling (NMDS) and principal coordinate analysis (PCoA). Permutational multivariate analysis of variance (PERMANOVA) was performed on distance matrices to assess microbial community differences among sample groups. The microeco package (v.1.9.0) and Linear discriminant analysis Effect Size (LDA Effect Size, LEfSe) were used to identify significantly different microbial taxa (LDA > 3) between MSA, PD, and HC groups. The functional profiles of microbial communities were predicted by Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) and categorized based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) Orthology (KO), with ANOVA and FDR correction applied for group comparisons.
A microbial co-occurrence network was constructed using Spearman correlation analysis to explore interactions between microbial data with clinical variables. Taxa with abundance > 0.01% and presence in ≥10% of samples were included. Significant correlations (|r|≥ 0.6, p < 0.01) at the genus level were included for network construction and visualized using the igraph (v.2.0.3) and Gephi v0.1.0. In these networks, high centrality indicates that a node (e.g., a clinical variable) has numerous and/or influential connections with other nodes, implying a central or integrative role in the network’s structure. In MSA patients, Spearman correlations linked differentially abundant microbes to clinical variables, with heatmaps displaying correlation matrices and significance levels.
Support vector machines (SVM) suited for small and nonlinear datasets were used to evaluate the predictive potential of skin microbiota profiles in MSA classification. Variable selection was performed using the Least absolute shrinkage and selection operator (LASSO) regression, with the optimal λ determined by cross-validation to minimize the mean squared error for optimal model fitting. Two SVM models were trained to differentiate MSA from HC and PD using differentially abundant microbes. A 7/3 split was used for training and validation, with genus-level abundances transformed via centered log-ratio before model training. To interpret feature importance, we applied the Shapley Additive exPlanations (SHAP) method, which quantifies each feature’s contribution to the classification outcome based on game theory. Two-tailed test was applied in the study, with a significance level of p < 0.05. Statistical analyses were performed using the R software (v. 4.1.0).
Supplementary information
Acknowledgements
The authors thank the High Performance Computing Center of Central South University for their resources and all participants for their involvement in this study. This study was funded by the National Key R&D Program of China (No. 2021YFA0805200 to H Jiang), the National Natural Science Foundation of China (82171254 to H Jiang; 82371272 to Z Chen; 82301628 to L Wan; 82401496 to L Peng), the Natural Science Foundation of Hunan Province (No. 2024JJ3050 to H Jiang; 2024JJ6638 to L Wan; No. 2024JJ6493 to L Sun), the Major Scientific Research Project for High-level Health Talent in Hunan Province(No.R2023047 to H Jiang), Furong Lab Research Project (No. 2023SK2084 to H Jiang), the Central South University Research Program of Advanced Interdisciplinary Study (No. 2023QYJC010 to H Jiang), the Science and Technology Innovation Program of Hunan Province (2022RC1027 to Z Chen), Postdoctoral Fellowship Program of CPSF(GZB20230870 to L Peng), and China Postdoctoral Science Foundation (2024M753690 to L Peng).
Author contributions
Concept and design: L.S., D.C.; Acquisition, analysis, or interpretation of data: L.S., D.C.; J.P., R.O., X.L., K.D., X.D., X.W., X.X., R.H.; Drafting of the manuscript: D.C.; L.S.; Administrative, technical, or material support: R.Q., B.T., HJ.; Supervision: L.S., L.W., Z.C., H.J. All authors participated in the protocol revision, contributed to the interpretation of the results, critically revised the manuscript for important intellectual content, read and approved the final version of this manuscript.
Data availability
Sequence data that support the findings of this study are available at NCBI (SRA accession no. PRJNA1284078).
Code availability
All scripts for analyses were deposited on GitHub at https://github.com/Daji1998/16sAnalyse.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Daji Chen, Lang Sun.
Supplementary information
The online version contains supplementary material available at 10.1038/s41531-025-01121-w.
References
- 1.Gilman, S. et al. Second consensus statement on the diagnosis of multiple system atrophy. Neurology71, 670–676 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wenning G. K. et al. The movement disorder society criteria for the diagnosis of multiple system atrophy. Mov. Disord. (2022). [DOI] [PMC free article] [PubMed]
- 3.Krismer, F. & Wenning, G. K. Multiple system atrophy: insights into a rare and debilitating movement disorder. Nat. Rev. Neurol.13, 232–243 (2017). [DOI] [PubMed] [Google Scholar]
- 4.Jellinger, K. A. et al. Accuracy of clinical diagnosis of Parkinson disease: A systematic review and meta-analysis. Neurology87, 237–238 (2016). [DOI] [PubMed] [Google Scholar]
- 5.Lin, D. J., Hermann, K. L. & Schmahmann, J. D. Multiple system atrophy of the cerebellar type: clinical state of the art. Mov. Disord.29, 294–304 (2014). [DOI] [PubMed] [Google Scholar]
- 6.Poewe, W. et al. Multiple system atrophy. Nat. Rev. Dis. Prim.8, 56 (2022). [DOI] [PubMed] [Google Scholar]
- 7.Wan, L. et al. Alterations of the gut microbiota in multiple system atrophy patients. Front Neurosci.13, 1102 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tan, A. H. et al. Altered gut microbiome and metabolome in patients with multiple system atrophy. Mov. Disord.33, 174–176 (2018). [DOI] [PubMed] [Google Scholar]
- 9.Engen, P. A. et al. The potential role of gut-derived inflammation in multiple system atrophy. J. Parkinson's Dis.7, 331–346 (2017). [DOI] [PubMed] [Google Scholar]
- 10.Harris-Tryon, T. A. & Grice, E. A. Microbiota and maintenance of skin barrier function. Science376, 940–945 (2022). [DOI] [PubMed] [Google Scholar]
- 11.Fanciulli, A. & Wenning, G. K. Multiple-system atrophy. N. Engl. J. Med.372, 249–263 (2015). [DOI] [PubMed] [Google Scholar]
- 12.Stefanova, N. & Wenning, G. K. Multiple system atrophy: at the crossroads of cellular, molecular and genetic mechanisms. Nat. Rev. Neurosci.24, 334–346 (2023). [DOI] [PubMed] [Google Scholar]
- 13.Swaney, M. H., Nelsen, A., Sandstrom, S. & Kalan, L. R. Sweat and sebum preferences of the human skin microbiota. Microbiol Spectr.11, e0418022 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Arikan, M. et al. Axillary microbiota is associated with cognitive impairment in Parkinson’s disease patients. Microbiol Spectr.10, e0235821 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee, K. A. et al. Characteristics of skin microbiome associated with disease severity in systemic sclerosis. J. Microbiol63, e2409018 (2025). [DOI] [PubMed] [Google Scholar]
- 16.Weiglein, A., Gaffal, E. & Albrecht, A. Probing the skin-brain axis: new vistas using mouse models. Int. J. Mol. Sci.23, 7484 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Doppler, K. et al. Distinctive distribution of phospho-alpha-synuclein in dermal nerves in multiple system atrophy. Mov. Disord.30, 1688–1692 (2015). [DOI] [PubMed] [Google Scholar]
- 18.Blake, K. J. et al. Staphylococcus aureus produces pain through pore-forming toxins and neuronal TRPV1 that is silenced by QX-314. Nat. Commun.9, 37 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kashem, S. W. et al. Nociceptive sensory fibers drive Interleukin-23 production from CD301b+ dermal dendritic cells and drive protective cutaneous immunity. Immunity43, 515–526 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pinho-Ribeiro, F. A. et al. Blocking neuronal signaling to immune cells treats streptococcal invasive infection. Cell173, 1083–1097.e1022 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kleinz, T. et al. The association of vagal atrophy with parameters of autonomic function in multiple system atrophy and progressive supranuclear palsy. Ther. Adv. Neurol. Disord.17, 17562864241267300 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Cha, J., Kim, T. G. & Ryu, J. H. Conversation between skin microbiota and the host: from early life to adulthood. Exp. Mol. Med.57, 703–713 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Postuma, R. B. et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov. Disord.30, 1591–1601 (2015). [DOI] [PubMed] [Google Scholar]
- 24.Geser, F. et al. Motor progression of multiple system atrophy (MSA): A prospective study using the unified MSA rating scale (UMSARS). Mov. Disord.20, S105–S105 (2005). [Google Scholar]
- 25.Goetz, C. G. et al. Movement Disorder Society Task Force report on the Hoehn and Yahr staging scale: status and recommendations. Mov. Disord.19, 1020–1028 (2004). [DOI] [PubMed] [Google Scholar]
- 26.Damon-Perriere, N. et al. Evaluation of the Scopa-Aut questionnaire in MSA patients: Relation to UMSARS scores and progression over time. Mov. Disord.26, S342 (2011). [DOI] [PubMed] [Google Scholar]
- 27.Stiasny-Kolster, K. et al. The REM sleep behavior disorder screening questionnaire-a new diagnostic instrument. Mov. Disord.22, 2386–2393 (2007). [DOI] [PubMed] [Google Scholar]
- 28.Appollonio, I. et al. The Frontal Assessment Battery (FAB): normative values in an Italian population sample. Neurol. Sci.26, 108–116 (2005). [DOI] [PubMed] [Google Scholar]
- 29.Hoops, S. et al. Validity of the MoCA and MMSE in the detection of MCI and dementia in Parkinson disease. Neurology73, 1738–1745 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Erkoreka, L. et al. Reported severity of psychotic, depressive and anxiety symptoms in relation to bilingual language profile: An exploratory study and the validation of Basque versions of the PQ-B, DASS-42, PHQ-9 and GAD-7. PLoS One20, e0314069 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods13, 581–583 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Quast, C. et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res.41, D590–D596 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Davis, N. M., Proctor, D. M., Holmes, S. P., Relman, D. A. & Callahan, B. J. Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome6, 226 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Sequence data that support the findings of this study are available at NCBI (SRA accession no. PRJNA1284078).
All scripts for analyses were deposited on GitHub at https://github.com/Daji1998/16sAnalyse.