| Type 2 diabetes（T2DM）was a complex disease caused by the combined action of environmental and genetic factors.As an internal environmental factor,human gut microbiotas were affected by a variety of external environments such as heavy metals,pesticides and particulates.It played a key mediating role in the process of insulin resistance or DM related metabolites caused by environmental factors such as PM2.5,arsenic and cadmium.The related research was of great value to clarify the mechanism of environmental factors of T2DM and its complications.Vascular complications were the main cause of death in T2DM.Diabetes patients with coronary heart disease（DM＿CHD）was the more common macrovascular complication,while the more common microvascular complication was diabetic peripheral neuropathy（DPN）.However,the pathogenesis of vascular complications in T2DM was not yet fully understood.Current epidemiological studies had shown inconsistent results in terms of altered gut microbiotas diversity,differential gut microbiotas and metabolite species between T2DM,T2DM with vascular complications,and healthy populations due to differences population characteristics.As well as differences in the association between gut microbiotas and metabolites and vascular complications.At the same time,most previous studies had used only single histological techniques for the study of gut microbiotas or plasma metabolites,and few had studied a single complication of T2DM including DM＿CHD,or DPN.The results of which were vulnerable to confounding by other coexisting complications.In addition,metagenomic birdshot sequencing techniques allowed for a more accurate analysis of colony species and functional differences.However,no studies have been reported on the use of metagenomic birdshot sequencing technology for metagenomic analysis of the gut microbiotas of DM＿CHD and DPN,especially in combination with metabolomics for exploring the etiology and mechanisms of diabetic vascular complications.Based on the above research background,this study adopted a case-control study design,combined metagenomic and metabolomic technologies to explore the structure,metabolite levels,and related functions of gut microbiotas in the T2DM with vascular complications（DM＿CHD,DPN）population.To reveal the association of gut microbiotas and metabolites with the development of T2DM with the vascular complications.Then provided a scientific basis for further research on the pathogenesis and personalized prevention and treatment of T2DM with cardiovascular complications.This study was carried out in three parts:PartⅠStructure and functional pathways of gut microbiotas in population among type 2 diabetes mellitus with vascular complicationsObjective:To investigate the diversity,structure and functional pathway characteristics of fecal gut microbiotas in DM＿CHD and DPN populations and provide a scientific basis for the study of the pathogenesis of T2DM with vascular complications.Methods:53 T2DM cases hospitalized in two hospitals in Shiyan,Hubei from 2015 to 2016 were selected as the discovery population,included 20 in the T2DM without complications group,13 in the DM＿CHD group,and 20 in the DPN group.The validation population was collected at the same hospitals and a total of 84 T2DM were included.Included 30 in the T2DM without complications group,24 in the DM＿CHD group and 30 in the DPN group.An additional 20 and 49 from the non-DM population were selected as controls for the discovery and validation populations respectively for further comparative analysis.A uniform questionnaire,physical examination,blood and fecal samples were collected from all study subjects.Blood samples were completed for clinical indicators such as blood glucose and lipids according to standard methods,and fecal samples were assayed by metagenomic birdshot sequencing using the Illumina platform（Novaseq 6000）.After quality control filtering,the data were annotated for species and function.Mann-Whitney test and multiple response permutation procedure were used to analyze the differences of species diversity and relative abundance of species and functional genes between groups.Binary logistic regression model was used to calculate odds ratio and 95%confidence intervalthe between significantly changed gut microbiotas and DM＿CHD and DPN.Results:No significant difference was found in alpha diversity index at genus and species levels for DM＿CHD and DPN groups in both of discover and validate population（P>0.05）.At the genus and species level,the mean relative abundance of Butyrate-producing bacterium was significantly lower in the DM＿CHD group in both of the discovery and validation populations（P<0.05）.When combined with the results of species abundance comparisons and disease risk analyses,after adjusting the age,sex and body mass index（BMI）,these Butyrate-producing bacterium could reduced the risk of DM＿CHD（OR<1,P<0.05）.The most significant reductions were found for the Butyrate-producing bacterium such as Dysosmobacter（P=0.041）,Dysosmobacter welbionis（P=0.044）,Lachnospira（P=0.039）and Lachnospira eligens（P=0.038）.The validation population showed similar results（FDR<0.05）.The results of the DPN group were similar to those of the DM＿CHD group in terms of changes in the Butyrate-producing bacterium such as Phocaeicola salanitronis（P<0.001）,Bifidobacterium angulatum（P=0.019）,Bifidobacterium angulatum（P=0.030）and Prevotella ruminicola（P=0.032）were significantly lower.The validation population also found the similar results（FDR<0.05）.In addition,the Butyrate-producing bacterium was more significantly reduced in abundance in the DM＿CHD group compared to the DPN group,especially for Dysosmobacter（P=0.042）and Dysosmobacter welbionis（P=0.045）.After further analyzing the functional genes of flora,it was found that the significantly changed flora functions of DM＿CHD and DPN group were mainly reflected in the up regulation of insulin resistance,Glycerol metabolism functional genes,biosynthesis genes of Lysine,Phenylalanine and the down regulation of propionic acid metabolism,folate synthesis,Arginine functional genes（P<0.05）.Conclusion:There was no significant changes in gut microbiotas diversity in DM＿CHD and DPN populations.There were significant changes in both structural and functional pathways of the gut microbiotas.The changes of gut microbiotas mainly showed that the abundance of Butyrate-producing bacteria and Prevotella decreased significantly,and was more pronounced in the DM＿CHD population.These altered gut microbiotas may increase the risk of vascular complications in T2DM by promoting aromatic amino acid and peptidoglycan biosynthesis and inhibiting carbohydrate and vitamin metabolism.Part Ⅱ Characteristics and metabolic pathways for metabolites in population among type 2 diabetes mellitus with vascular complicationsObjective:To reveal the metabolites and metabolic pathways of DM＿CHD and DPN populations,and to provide a scientific basis for the etiology of T2DM with vascular complications through untargeted and targeted metabolomic assays.Methods:In the first stage,84 T2DM cases hospitalized in two hospitals in Shiyan,Hubei from 2015 to 2016 were selected as the discovery population,included 30 in the T2DM without complications group,24 in the DM＿CHD group,and 30 in the DPN group.Untargeted metabolomics testing of fecal and plasma samples was executed respectively.In the second stage,the major between-group differential metabolite classes in the untargeted metabolomic results were tested.And the study population remained consistent with the population found in the first part.An additional 20 non-DM population were selected to further analyse.The fecal and plasma samples of the above population were examined separately using targeted metabolomics techniques.After quality control of the data,t-tests,difference multiples and partial least squares discriminant analysis and corrected partial least squares discriminant analysis were used to screen differential metabolites.Multivariate logistic regression models were used to calculate the risk ratios and 95%confidence intervals for the prevalence of significantly altered fecal and plasma metabolites with DM＿CHD and DPN,respectively.Result:Combined univariate and multivariate differential metabolite screening results and risk of disease analysis,results for both of the untargeted and targeted metabolism groups showed significantly lower levels of Histidine(P untargeted=0.013,P targeted=0.004)and Arginine(P untargeted=0.010,P targeted=0.015)in plasma among the DM＿CHD group compared with T2DM without complications group.Compared to the non-DM group,the DM＿CHD group had significantly higher levels of Ursodeoxycholic acid(P fecal=0.039,P plasma=0.009)andβ-Ursodeoxycholic acid(P fecal=0.025,P plasma=0.021)in both feces and plasma,except for significantly lower levels of Histidine（P<0.001）and Arginine（P=0.001）in plasma（FDR<0.05）.After adjusting for age,sex and BMI,both of Histidine（OR=0.973,P=0.023）and Arginine（OR=0.856,P=0.037）could reduce the risk of the DM＿CHD.Compared to the T2DM without complications group,both of tested populations,the levels of Taurocholic acid(P untargeted=0.007,P targeted=0.030)in fecal and Arginine(Puntargeted=0.015,P targeted<0.001),Glycine(P untargeted=0.026,P targeted<0.001)and Lysine(P untargeted=0.034,P targeted<0.001)in plasma among the DPN group were significantly elevated.But only Lysine could significantly increase the risk of DPN（OR=1.092,P=0.017）.Compared with the non-DM group,the levels of Octadecanedioic acid（P=0.023）in fecal and the levels of six amino acids,included Deoxycholic acid（P=0.007）,Ursodeoxycholic acid（P=0.020）and Histidine,Arginine and Lysine in plasma,were significantly higher in the DPN group.The levels of five free fatty acids,included Azelaic acid and cis-5-Tetradecenoic acid in plasma were significantly lower（P<0.05,FDR<0.05）.There were also significant differences in metabolite levels between the DM＿CHD and DPN groups:Histidine(P untargeted=0.021,P targeted<0.001)and Arginine(P untargeted=0.034,P targeted<0.001)in plasma,which had significantly lower levels in the DM＿CHD population,were significantly higher in the DPN group（FDR<0.05）.Metabolic pathway analysis showed that,the highest enrichment factors for metabolite pathways both of fecal and plasma in the DM＿CHD and DPN populations were the vascular endothelial growth factor signaling pathway（maximal enrichment factor=0.176,P=0.054）and the Arginine biosynthesis pathway（maximal enrichment factor=0.429,P=0.031）.Conclusion:Both of levels and pathways for metabolite were significantly altered in DM＿CHD and DPN populations in fecal and plasma,and the changes of metabolites in plasma were more obvious.Levels ofβ-Ursodeoxycholic acid and Ursodeoxycholic acid were significantly elevated in both of feces and plasma among the DM＿CHD population,while levels of Histidine and Arginine in plasma were significantly lower.Levels of Taurocholic acid in fecal and Arginine,Lysine in plasma were significantly higher in the DPN population,while levels of free fatty acids such as Azelaic acid in plasma were significantly lower.Levels of Lithocholic acid in fecal and Histidine,Arginine in plasma were higher in the DPN population compared to the DM＿CHD population.These altered metabolites may increase the risk of T2DM with vascular complications by affecting pathways such as the vascular endothelial growth factor signaling pathway and Arginine biosynthesis.Part Ⅲ Association analysis of gut microbiotas and metabolites in population with type 2 diabetes mellitus with vascular complicationsObjective:To integrate the metagenome and metabolome for the association analysis.And then constructing the random forest model to screen the characteristic gut microbiotas and metabolites for distinguishing and identifying the T2DM with vascular complications.Exploring the possible related mechanisms of T2DM with vascular complications.Methods:Based on the metagenome and targeted metabolome data of the same population discovered in Part I,the association analysis of gut microbiotas and metabolites was performed.The Spearman correlation was used to analyze the gut microbiotas-functional pathway-metabolism-clinical phenotype of feces and plasma samples from different groups.A random forest model was constructed to classify each group,and the importance points of the predictive model indicators were plotted,and the receiver operating characteristic curve（ROC）was used to assess the efficacy of the predictive model.Results:Compared with non-DM group,Butyrate-producing bacteria such as Lachnospira and Leuconostoc bacteria significantly decreased in DM＿CHD group,which were significantly positively correlated with the bacterial functional pathway,and significantly negatively correlated with the Nutritional cholic acid of the plasma.Then they significantly positively correlated with the diabetic course(|r|max=0.602,|r|min=0.364,P<0.05).Compared with T2DM without complications group,the significantly reduced Bifidobacteria in DPN group were significantly positively correlated with the significantly reduced functional pathway,and significantly negatively correlated with the significantly increased Taurocholic acid in feces and Alanine glycine,Citrulline in plasma.And further significantly positively correlated with blood glucose,duration of diabetes and other clinical indicators(|r|max=0.586,|r|min=0.329,P<0.05).Compared with non-DM group,the significantly altered flora in the DPN group was significantly positively correlated with the functional pathways and the metabolites such as Histidine,Arginine and Indoacetic acid in this group.And then significantly correlated with blood glucose and duration of diabetes(|r|max=0.709,|r|min=0.324,P<0.05).When distinguishing from non-DM group,in fecal samples,DM＿CHD and DPN groups were treated with 9（4 gut microbiotas and 5 metabolites）and 16（12 gut microbiotas and 4 metabolites）had the highest accuracy of prediction and recognition.Their accuracy were 68.9%and 70.3%respectively.The area under ROC curve could reach 0.751 and 0.797 respectively.The most influential discriminant characteristics of each group wereβ-Ursodeoxycholic acid and Duncaniella respectively.In plasma samples,the DM＿CHD and DPN groups were treated with 15（5 gut microbiotas and 10 metabolites）and 33（12 gut microbiotas and 21 metabolites）have the highest accuracy of prediction and recognition.Their accuracy were 76.7%and 95.0%respectively.The area under ROC curve could reach 0.905 and 0.989respectively.The most influential distinguishing features of each group were Histidine and Arginine respectively.Conclusion:The altered gut microbiotas of the DM＿CHD and DPN populations may affect the levels of Lithocholic acid in fecal and Histidine,Nutritional cholic acid in plasma and other metabolites through functional pathways such as peptidoglycan biosynthesis,insulin resistance and Glycerol metabolism.So as to directly or indirectly affect the blood glucose and clinical phenotypes such as the course of T2DM.In addition,specific gut microbiotas（Butyrate-producing bacteria）and metabolites in plasma（Histidine,Arginine,etc.）have higher recognition and prediction value in distinguishing the population of T2DM with vascular complications.Overall Conclusion:1.There were significant changes in both structural and functional pathways of the gut microbiotas in the T2DM with vascular complications such as DM＿CHD and DPN populations.The changes of gut microbiotas mainly showed that the abundance of Butyrate-producing bacteria and Prevotella decreased significantly.2.Both of levels and pathways for metabolite were significantly altered in T2DM with vascular complications in fecal and plasma such as DM＿CHD and DPN populations,and the changes of metabolites in plasma were more obvious.The changes of metabolites mainly showed that the levels of fecal Cholic acid,plasma Histidine and Arginine in DPN population increased significantly,while the results in DM＿CHD population were just the opposite.3.The altered gut microbiotas of the T2DM with vascular complications such as DM＿CHD and DPN populations may affect the levels of Lithocholic acid in fecal and Histidine,Nutritional cholic acid in plasma and other metabolites through functional pathways.So as to affect the clinical phenotypes such as blood glucose. |
PDF Full Text Request