| Objective Parkinson’s disease is the second largest neurodegenerative disease noly to Alzheimer’s disease.No cure for Parkinson’s disease untill today,and the etiology is not yet clear.there is still a lack of early diagnosis and effective biomarkers to guide the treatment of Parkinson’s disease.In terms of clinical research on Parkin’s disease,most of them use a single omics or a single biological sample to conduct two-omics studies.At present,the joint study of gut microbiota and blood metabolomics is still lacking.We believe that the most important blood circulation phenotypic link is missing in the study of the gut-brain axis.The purpose of this study is to conduct a combined study of two omics based on the characteristics of the gut microbiota and plasma metabolism of Parkinson’s disease,and to provide basic datas of circulating metabolic phenotypes for the study of the microbiota-gut-brain axis.The characteristics of gut microbiota in the constipation and olfactory dysfunction groups of Parkinson’s disease patients were preliminarily investigated which will provide indications for the study of the early pathogenesis of Parkinson’s disease.Methods(1)A total of 121 subjects were included,including 65 in Parkinson’s disease group(PD group)and 56 in healthy control group(Con group).The Parkinson’s disease group was divided into 38 cases of constipation group(CO group)and 27 cases of non-constipation group(NCO group)according to the presence or absence of constipation;and 26 cases of olfactory dysfunction group(OL group)and 39 cases of non-olfactory dysfunction group(NOL group)according to the presence or absence of olfactory dysfunction,and the clinical data of all subjects were collected.Collect fresh stool samples in the morning,extract total gut microbiota DNA,conduct PCR amplification on the V3-V4 region of 16 S r RNA gene,use NEXTFLEX Rapid DNA-Seq Kit for Miseq library construction,and use Miseq PE300/Nova Seq PE250 platform for sequencing.OUT classification,species annotation,gut microbiota composition,differential microbiota between groups,prediction model construction and correlation analysis of microbiota and clinical characteristics were carried out through relevant bioinformatic analysis methods.(2)42 cases of Parkinson’s disease group(PD group)and 34 cases of healthy control group(Con group)who collected stool samples and plasma samples at the same time were included.Morning stool were collected for gut microbiota detection,and morning plasma was collected for LC-MS untargeted omics assays.We used the UHPLC-Q Exactive HF-X system for liquid-phase mass spectrometry detection,and the off-machine data used the metabolomics software Progenesis QI for peak extraction,alignment,and identification.The main differential metabolites between the two groups of plasma samples were identified by combining PCA and OPLS-DA,and metabolic function analysis was carried out according to the metabolite database information.Finally,spearman correlation was performed between the screened plasma differential metabolites and the differential gut microbiota screened by LEf Se linear discriminant analysis at family level.Results(1)In terms of alpha diversity of microbiota: the PD group and CO group were compared with the corresponding control group,the p values of sobs,chao,and ace index were all less than 0.05,and the p values of shannon and simpson were all less than 0.05;OL Compared with NOL group,the p values of sobs,chao,and ace index were all more than 0.05,and the p values of shannon and simpson were all less than 0.05.(2)In terms of species composition: At different classification levels,the proportion of comparatively dominant bacteria in PD group,CO group and OL group was different from the corresponding control group.(3)Comparison of sample composition: Partial least squares discriminant analysis found that the composition of microbiota was significantly different between PD group and Con group,CO group and NCO group,and OL group and NOL group.(4)Difference analysis of gut microbiota between groups:Though the analysis of the significance of differences between each groups,PD group and Con group,CO group and NCO group,and OL group and NOL group had significantly different microbiota between groups and at each classification level.(5)In the differential microbiota of the constipation group and the olfactory dysfunction group to their asymptomatic group,there were common microbiota with increased abundance including Actinobacteriota,Bifidobacteriaceae,Anaerovoracaceae etc.,and the decreased abundance microbiota was norank＿o＿＿Gastranaerophilales.(6)Correlation analysis of microbiota and clinical characteristics: Bifidobacteriaceae were positively correlated with H-Y staging of Parkinson’s disease,UPDRS II score,UPDRS III score,UPDRS IV score,benzhexol and amantadine drugs,and the p values were less than 0.05.The rest of the microbiota was related to H-Y stage,UPDRSⅠ,Ⅱ,Ⅲ,Ⅳ scores and some anti-Parkinson’s disease drugs.(7)Random forest and ROC curve prediction model: The ROC curve drawn by the top 7 families that are most important for the classification of the Con group and PD group at the family level,the area under the curve AUC was 0.73,and the specificity was68%,the sensitivity was 75%.(8)Plasma differential metabolite analysis:According to the OPLS-DA multivariate statistics,the variable VIP value that contributed the most to the classification was greater than 1.5 and p<0.05,and52 named differential metabolites were screened out,and 28 of them had an increased relative expression in the PD group,and 24 metabolites were reduced.Most of them belong to lipids,amino acids,peptides,purine nucleotides and carbohydrates.(9)KEGG pathway and enrichment analysis: There were 9differential metabolites involved in 8 secondary metabolic pathways in the KEGG pathway,and the primary pathways are mainly involved in the metabolic pathways.Four pathways were enriched on the KEGG pathway,including galactose metabolism,purine metabolism,caffeine metabolism and retinol metabolism.(10)Prediction of potential biomarkers for Parkinson’s disease:Screening out 6 metabolites with area under the ROC curve AUC greater than0.85,among which alanyl-methionine,P-Mentha-1,3,8-triene and4-Vinylguaiacol,these 3 metabolites combined as a biomarker predictor,AUC was 0.983,the sensitivity was 95.24%,the specificity is 94.12%,the confidence interval was 0.923-0.999.(11)Correlation analysis of differential microbiota and differential metabolites: 32 differential metabolites were screened according to the VIP value greater than 2 and p < 0.05,which contributed the most to the classification,and 14 differential microbiota were screened according to the LDA threshold was greater than 3 by LEf Se linear discriminant analysis at the family level.Correlation analysis was carried out for 14 different families and32 differential metabolites,and 11 metabolites were found to be correlated with7 families.Conclusion(1)The PD group,the CO group and the OL group had significant different gut microbiota compared with the corresponding control group.and the CO group and the OL group had the same different microbiota.There are 7bacterial families of with good predictive performance as potential biomarkers to distinguish Parkinson’s disease from healthy person.(2)Patients with Parkinson’s disease have obvious disorder of plasma metabolites,mainly manifested as energy metabolism dysfunction.(3)The combination of plasma alanyl-methionine,P-Mentha-1,3,8-triene and 4-vinylguaiacol as biomarkers to differentiate Parkinson’s disease from healthy person had highly predictive performance.(4)There were significant correlations between 7 stool differential bacterial families and 11 plasma differential metabolites. |
PDF Full Text Request