The Role Of Gut Microbiota In The Risk Of Schizophrenia

Objectives1.The difference of intestinal bacterial microbiota between schizophrenia（SCH）patients and healthy controls was analyzed to examine the relationship between intestinal bacterial microbiota and SCH.2.The difference of intestinal fungal microbiota between schizophrenia（SCH）patients and healthy controls was analyzed to examine the relationship between intestinal fungal microbiota and SCH.3.The balance of intestinal fungal microbiota and bacterial microbiota and the fungal-bacterial network was explored in patients with SCH compared with healthy controls.4.To investigate the relationship between intestinal microorganism（intestinal fungal microbiota and intestinal bacterial microbiota）and polygenic risk score（PRS）,and their role in the risk of SCH.5.The possible mechanism of gut microbiota especially the core microbial markers involved in the pathogenesis of SCH were explored using metabolomics and species function analysis.Material and methods1.Two hundred and five first episode,drug-free SCH patients enrolled to the psychiatric department of the First Affiliated Hospital of Zhengzhou University between October 2017 and October 2019.The Positive and Negative Syndrome Scale（PANSS）was used to evaluated the severity of psychiatric symptoms in SCH patients.Cognitive functions were evaluated using the MATRICS Consensus Cognitive Battery（MCCB）in Chinese.Venous blood（10 ml）was collected between 06:30am and 07:00am in the next morning after the recruitment.Fresh fecal samples were taken from participants in the morning after breakfast（8:00-10:00）provided by hospital and our research group.2.Gut bacterial and fungal microbiota was characterized by 16S and 18S ribosomal RNA gene Amplicon sequencing,respectively.Species function and metabolomics were used to identify the differential metabolic pathways between the SCH patients and healthy controls.3.The DNA was extracted from blood using magnetic bead method and used for genome-wide association studies（GWAS）.The DNA was genotyped using the InfiniumOmniZhongHua₈v Array（Illumina,Inc.）on an iScan instrument at the Bio Miao Biological Technology（Beijing）.SCH GWAS summary statistics for both the European and east Asian population were downloaded from https://www.med.unc.edu/pgc/download-results/scz/with permission.After filtering out SNPs that do not exist in data from the summary statistics,the program PRC-CS was used to compute the European-based polygenic risk score（PRS-EUR）and the East Asian-based polygenic risk score（PRS-EAS）for SCH,separately.4.SPSS 20.0 and R software（Version 3.6.3）was used for statistical analysis.The Wilcoxon tests was used to evaluate the difference in microbial abundance between the SCH patients and the HCs.The Microbial Dysbiosis Index（MD index）was determined based on the abundance of genera.Both fungal and bacterial MD index（F_MD and B_MD）was calculated.The fungi-to-bacteria a diversity ratio was determined to explore the equilibrium between fungi and bacteria diversity in the gut.The Random forest methods（randomForest package from R Version 3.6.2）were used to construct prediction models for SCH using differential fungal and bacterial markers between patients and HCs.A cross-validation procedure was employed to evaluate the generalization performance of the resultant prediction model.Correlation coefficient matrix was obtained by Spearman correlation coefficient（SCC）based on the abundance of genus.Package igraph and ggplot2 from R software（Version 3.6.3）was used to calculate the parameters of network and draw network diagrams respectively.The associations of PRS and gut microbiota with SCH were estimated by logistic regression models.Stratified sampling method was used to identify the core microbial markers eliminating bias caused by individual heterogeneity.The species function analysis and metabolomics analysis were used to explore the possible mechanism of intestinal microorganism especially the core microbial markers in regulating the pathogenesis of SCH.R software（Version 3.6.3）and GraphPad Prism（Version 8.0.2）were used for graph visualization.Results1.We found significant lover levels of intestinal Shannon and Simpson’s indices in SCH patients than in HCs（p=9.93×10-6,5.18×10-5）.When using phylogenetic diversity（PD）,we observed significant increase in this measure in the SCH patients than that of HCs（p=0.027）.We found significant differences in fungal community composition between SCH patients and HCs based on β diversity analysis.The Permutational multivariate analysis of variance（PERMANOVA）based on those four β diversity dissimilarity metrics showed that between group difference were significant than that of with-in group difference,in SCH or HCs group（all R2>0,p<0.05）.We discovered 28 microbial markers at multiple taxonomy that were significantly different between SCH patients and HCs（all p<0.05）.In addition,the B_MD index in SCH patients were significantly higher than that of HCs（p=3.23×10-15）.2.We observed significant decrease in intestinal fungal α diversity（Shannon and Simpson’s indices）in SCH patients than that of HCs（p=0.03 7,0.030,respectively）.The PD index was also decreased in SCH patients than that of HCs（p=5.93×10-7）.We found significant differences in fungal community composition between SCH patients and HCs based on β diversity analysis.The PERMANOVA based on those four β diversity dissimilarity metrics showed that the between group difference were significantly different than that of with-in group difference,in SCH or HCs group（all R2>0,p<0.05）.We discovered 29 microbial markers at multiple taxonomy that were significantly different between SCH patients and HCs（all p<0.05）.In addition,the F_MD index in SCH patients were significantly higher than that of HCs（p=4.68×10-10）.3.The fungi-to-bacteria diversity ratio was determined to explore the equilibrium between fungi and bacteria diversity in the gut.We analyzed fungi-to-bacteria diversity ratio of α-diversity indices.The ratio calculated by PD index showed significant lower levels in patients with SCH than in that of HCs（p=3.14×10-8）.At genus level,a predictive model based on the abundance of the bacterial markers and fungal markers can achieve a better performance than bacterial alone with AUC of 0.847,0.739,respectively（p=0.043）.The global kingdom based on correlation networks involving both the intestinal bacterial and fungal microbiota at genus level show significant difference in the degree and connectiveness between SCH patients and HCs.A disrupted and weakened fungi-bacteria network was shown in SCH patients characterized by fewer nodes and edges,lower number of neighbors compared with HCs.4.We observed significant associations between PRS and the risk of SCH.PRS-EAS showed stronger associations（OR=2.08,95%CI=1.51-2.92,p=1.22×10-5）than PRS-EUR（OR=1.73,95%CI=1.30-2.35,p=2.77×10-4）.On the observed scale,PRS-EAS and PRS-EUR explained 8.9%and 5.8%of phenotypic variance,respectively.Increased diversities measured by Shannon and Simpson indices showed a strong protective effect on SCH（Shannon:OR=0.29,95%CI=0.18-0.43,p=1.15×10-8;Simpson:OR=0.29,95%CI=0.19-0.44,p=1.25×10-8）.The proportions of phenotypic variance explained by the Shannon and Simpson index were 33%for both.Whereas,increased diversity measured by PD showed a weaker risk to SCH（OR=1.43,95%CI=1.05-2.00,p=0.03,R2=19%）.We obtained significant interactions between PRS-EAS and the Shannon,Chao1 and PD diversity indices（Shanno,p=0.05;Chao1,pp=4.43×10-3;and,PD,p=6.31×10-3）.These interaction terms explained 1.05%,3.15%and 2.91%of phenotypic variance,correspondingly.About the fungal microbiota,an increased PD index was associated with a lower risk of SCH（OR=0.82,95%CI=0.72-0.97,P=0.012,R2=19%）.We also observed significant associations between MD index and the risk of SCH.B_MD showed stronger associations（OR=1.52,95%CI=1.36-1.69,p=7.83×10-13）than F_MD（OR=1.20,95%CI=1.13-1.28,p=1.42×10-7）.The explained proportions of phenotypic variancr were significantly increased（R2=46）in the model including both PRS-EAS and the B_MD,F_MD index.5.Aspergillus and Purpureocillium was identified to be the core fungal markers.We found strong associations between fungal markers Purpureocillium,Aspergillus and the risk of SCH.The increased abundance of Purpureocillium showed a higher risk to SCH（OR=1.17,95%CI=1.09-1.25,p=8.30×10-6,R2=25%）.The increased abundance of Aspergillus showed a weaker risk to SCH（OR=0.92,95%CI=0.86-0.99,p=0.040,R2=18%）.We also obtained significant interactions between PRS-EAS and Aspergillus（p=3.68×10-3）.The interaction term explained 3.52%of phenotypic variance.After controlling for age,gender,education level and the course of disease,the abundance Purpureocillium was positively correlated with symptom severity（r=0.200,p=0.034）,and negatively correlated with cognitive function measured by visual learning test（r=-0.349,p=1.01×10-4）and Attention/vigilance（r=-0.260,p=0.002）.Patients in Purpureocillium dominant group showed more severe psychiatric symptoms and poor cognitive function compared with Purpureocillium depletion group.Inversely,high abundance of Aspergillus was associated with better cognitive function.Aspergillus spp.genes can encode the key enzyme or coenzyme of cysteine-methionine metabolic pathway.Both the species function analysis metabolomics analysis was conducted to explore the role of gut microbiota especially the core microbial markers in the pathogenesis of schizophrenia.We observed an abnormal cysteine-methionine metabolic pathway in SCH patients.Conclusions1.Present study suggests that the dysbiosis of intestinal bacterial microbiota and fungal microbiota may be associated with the pathogenesis of SCH.2.Present study also demonstrates that the fungal-bacterial correlation network may server as an important new dimension in precision psychiatry.3.Our findings clearly show that both gut microbial markers and host genetics interactively contribute to the risk of developing schizophrenia.4.Present study shows that the dysbiosis of gut microbiota may be associated with the pathogenesis of SCH via the cysteine-methionine metabolism pathway.