The Correlation Of Preterm Birth And Vaginal Microbiome,Proteome And Metabolomics During Pregnancy

BackgroundPreterm birth is a global maternal and infant health problem.About 35%of spontaneous preterm births and 50%of preterm premature rupture of membranes(PPROM)are caused by intra-amniotic infection.Abnormal vaginal microbiome has been shown to be a risk factor for PTB,but there is a lack of effective methods to predict PTB by vaginal microbiome in clinical practice.Previous studies have used 16S rRNA gene variable region sequence amplified from vaginal swab samples to quantitatively and qualitatively analyze the information of vaginal microbiome and to perform community state types(CSTs)typing,but the results of studies on the relationship between vaginal microbiome and preterm birth are not consistent.The pathophysiological mechanisms of abnormal vaginal microbiome leading to adverse pregnancy outcomes need to be further explored.ObjectiveThe objectives of this study were as follows:(1)The first part combines the detection of a variety of microorganisms,that is,16S rRNA gene classification analysis and quantitative polymerase chain reaction(qPCR)method,By using a machine learning strategy that is able to complete the task of more complex prediction model construction,a multi-feature screening approach was attempted to evaluate vaginal microbial characteristics to construct a model to improve the accuracy of PTB prediction.(2)In the second part,the prediction model was validated and refined in a new prospective cohort.Proteomics and metabolomics were used to explore the immune and inflammatory response in vaginal secretions of preterm birth patients.Methods(1)Vaginal vault secretions from pregnant women in the first,second and third trimesters were prospectively collected.DNA was extracted and subjected to 16S rRNA and targeted qPCR for 30 species.According to the pregnancy outcome,the patients were divided into spontaneous preterm delivery group and full-term delivery group.The dynamic changes of vaginal microbiome,CST types and specific bacteria were compared between the two groups.The differential strains were obtained by comparison between the two groups,and the machine learning method was used to establish the prediction model of preterm birth.(2)Vaginal vault secretions were prospectively collected from pregnant women in the second trimester,and the same experimental methods as in the first part were applied to the spontaneous preterm delivery group and the term delivery group,and the results were put into the established preterm birth prediction model to verify its sensitivity and specificity.Liquid chromatography tandem mass spectrometry(LC-MS/MS)was used to detect the proteins in vaginal secretions of the enrolled patients,and the data independent acquisition mode proteomics quantitative method was used to analyze the classical pathway of differentially expressed proteins between the two groups,and to explore the immune and inflammatory pathways associated with preterm birth.In addition,metabonomics technology was used for chromatographic separation and analysis to analyze the differences in metabonomics between the two groups.Metabolic pathway enrichment analysis was performed on the differential metabolites to explore the metabolic pathways associated with preterm birth.Results(1)A total of 26 cases of spontaneous preterm delivery and 247 cases of term delivery were enrolled.No significant correlation was found between specific trends and dynamic changes of bacterial diversity,vaginal CST classification and sPTB.The main difference between the two groups was the higher positive rate and abundance of Ureaplasma urealyticum,Mycoplasma hominis and Bacteroides fragilis/Mobiluncus in the sPTB group.A prediction model of sPTB was established by analyzing the vaginal microbial characteristics of the first and second trimesters of pregnancy samples using a multi-machine learning strategy and multi-feature screening method.(2)The positive prediction rate was low when the microbiome data of 5 cases in the spontaneous preterm delivery group and 16 cases in the full-term delivery group were put into the prediction model for preterm birth established previously.A total of 270 differentially expressed proteins were screened by proteomics.IPA pathway analysis showed that the up-regulated differentially expressed proteins in the preterm group were related to glucocorticoid receptor signaling,B cell development,MSP-RON signaling,eIF2 signaling pathway and NRF2 mediated oxidative stress.A total of 159 differential metabolites were screened by metabolomics,and the most significant differential metabolites were lipids.KEGG pathway analysis showed that the up-regulated pathways in the preterm group were glycerophosphatidylcholine metabolism,sphingolipid metabolism,primary bile acid biosynthesis,purine metabolism,and steroid hormone biosynthesis.ConclusionIn this study,vaginal secretions were collected from the first and second trimesters of pregnancy,and vaginal microbial characteristics were analyzed using a multi-machine learning strategy and multi-feature screening method to establish a prediction model for sPTB.However,the performance of the model was evaluated by external validation in a new cohort.IPA classical pathway analysis showed that immune-inflammation-related regulatory pathways were widely activated in vaginal secretions of preterm infants.These results suggest that the potential immune inflammatory response in the vaginal region is activated in the preterm group.Metabolomics has found that some metabolic pathway disorders may stimulate inflammatory cytokines or change the properties of amniotic membrane and other mechanisms secondary to the occurrence and development of PTB,which may be the pathological principle and possible molecular mechanism of PTB.