Preliminary Study On Gut Microbiota And Metabolomics In Children With Mycobacterium Tuberculosis Infection

Tuberculosis(TB)remains a serious threat to children’s health,and most of it is pulmonary tuberculosis(PTB),which is characterized by high incidence rate and high mortality.It is estimated that approximately a quarter of the world’ population have been infected with mycobacterium tuberculosis(Mtb),about 5%-10% of people develop active TB,and most of them are latent tuberculosis infection(LTBI).The occurrence of PTB is believed to be the result of the imbalance between host immunity and Mtb.Any factors that affect the immune response between Mtb and host may lead to different clinical outcomes after Mtb infection.Due to the lack of typical manifestations,the difficulty in obtaining specimens,the low bacillary load in sputum samples,and the insufficient of pathogen detection methods,it is hard to diagnose childhood PTB in early stage.Therefore,it is vital to evaluate the molecular diagnostic technology Xpert MTB/RIF Ultra in order to establish a rapid,accurate and suitable method for childhood PTB diagnosis.Simultaneously,there is an urgent need to strengthen the research on the pathogenesis of PTB and screen potential biomarkers associated with PTB to provide new ideas for the early diagnosis and treatment of PTB.Gut microbiota acts as the homeostasis regulator in a variety of physiological pathways,such as metabolism,immunity and nutrient absorption.Once the balance broken,it would inevitably cause intestinal flora dysbiosis,resulting in the occurrence of disease.Recently,it has been found that gut microbiota was closely associated with the occurrence and development of PTB.However,there was limited data concerning the relationship between PTB and the changes of gut microbiota,the exsiting studies mainly paid attention to the comparison between PTB patients and healthy controls or the changes of intestinal flora of PTB patients before and after treatment.Little was known about the diversity of gut microbiota in children with PTB and LTBI.In addition,metabonomics could comprehensively reflect the dynamic metabolic changes of the body,and has been used to screen potential disease-associated biomarkers or reveal the pathogenesis of diseases.So far,metabolomics studies on Clostridium difficile infection,irritable bowel syndrome,depression and tumor have made significant progress.However,data on the fecal metabolomics of children with PTB and LTBI was few.Based on the above reasons,we designed this research.Objective:1.To evaluate the diagnostic value of Ultra in gastric aspirate samples among children with PTB.2.To explore the overall structural characteristics of gut microbiota in PTB patients,LTBI patients and healthy controls,and to find out the differential microbiota closely related to PTB and LTBI.3.To explore the characteristics of fecal metabolites in children with PTB and LTBI,and to screen the differential metabolites,and to analyze the pathway of the differential metabolites.Methods:1.This was a prospective,multicenter,diagnostic accuracy study.Children who suspected PTB were continuously enrolled at the department of pediatrics of West China Second University Hospital and other hospitals from July 1,2018 to January31,2020 and their gastric aspirate samples were collected.All collected specimens were processed acid-fast staining smear microscopy,mycobacterium rapid culture(BACTEC MGIT960)and Ultra testing.The diagnostic value of Ultra in gastric aspirate samples of children with PTB was assessed by using the composite reference standard(CRS)and Mtb culture as the gold standard,respectively.2.Fecal specimens of the PTB patients,LTBI patients and healthy controls were collected.The diversity of gut microbiota among the three groups were explored by16 S r DNA sequencing technology and bioinformatics.3.Non-targeted metabolomics study was performed on the fecal of the PTB patients,LTBI patients and healthy controls by Ultra-performance Liquid Chromatography-Mass Spectrometry(UPLC-MS).Search the differential metabolites among the three groups using the combination of multivariate statistical analysis and univariate statistical analysis,and analyze the pathway of the differential metabolites by KEGG database.Simultaneously,a prediction model of childhood PTB was established based on the ROC curve.Results:1.A total of 109 children with PTB(31 bacteriologically confirmed cases and 78 probable cases)and 113 children with non-TB were included.Compared with CRS,the sensitivity of Ultra(54.1%,59/109)was significantly higher than that of smear(7.3%,8/109,P<0.001)and Mtb culture(25.7%,28/109,P<0.001).The specificity of smear,Mtb culture and Ultra were 100%(113/113),100%(113/113)and 99.1%(112/113),respectively.Compared with Mtb culture,the sensitivity of Ultra were 89.3%(25/28).When the three methods were combined,the rate of children diagnosed with definite PTB rosed from 28.4%(31/109)to 57.8%(63/109).In addition,we also found that the sensitivity of smear,Mtb culture and Ultra in children < 5 years old was 14.7%,26.5%,73.5%,respectively,and 4.0%,25.3% and 45.3% in children ≥ 5 years old.The sensitivity of Ultra in children < 5 years old was significantly higher than that in children ≥ 5 years old(73.5% vs 45.3%,P=0.006).In PTB group and PTB complicated with extrapulmonary tuberculosis group,the sensitivity of smear was 8.8%and 5.8%,Mtb culture 24.6% and 26.9%,and Ultra 51.9% and 61.5%,respectively.There were no significant differences in the sensitivities of the three detection methods between PTB group and PTB complicated with extrapulmonary tuberculosis group(P > 0.05).2.Overall,22 PTB patients,34 LTBI patients and 24 healthy controls were included in this study.There were no significant differences in age,sex and BMI among the three groups of children(P>0.05).A total of 5307534 useful gene sequences were obtained from 80 samples,with an average of 66344 ± 5073 per case.The Observed species index(P<0.05)and Chao1 index(P<0.05)of the PTB group were significantly lower than those of the LTBI group and the healthy controls.However,the Shannon index and Simpson index showed no remarkable differences among the three groups(all P values>0.05).The PCo A analysis based on the unweighted unifrac distance matrix have showed that dramatic differences of gut microbiota composition in PTB group were found compared with LTBI group and healthy controls group.The relative abundance ratio of Firmicutes to Bacteroides increased in PTB patients,while it decreased in LTBI patients.We also observed that Firmicutes,Enterobacteriales,Enterobacteriaceae,Bacteroidaceae,Bacteroides and Faecalibacterium were enriched in children with PTB.Prevotellaceae,Prevotella＿9,Prevotella＿2 and Alloprevotella increased dramatically in children with LTBI.Lachnospiraceae and Fusobacterium were abundant in healthy controls.Furthermore,5 differential bacteria as biomarkers with AUC more than 0.7 was screened based on receiver operating characteristics(ROC)curve,the sensitivity was 72.7% to 86.4%,and the specificity was 67.6% to 82.4%.3.The metabolites of fecal specimens were changed significantly between PTB patients,LTBI patients and healthy controls.A total of 21 different metabolites were screened and identified,such as 11beta-Hydroxyprogesterone,12,13-Ep OME,14,15-Di HETr E,16(R)-HETE,9,10,13-Tri HOME,9,12,13-Tri HOME,Chenodeoxycholic acid,Cholic acid,Eicosapentaenoic acid,Estrone 3-glucuronide,Calcitriol and so on.The 21 different metabolites were mainly involved in linoleic acid metabolism,primary bile acid biosynthesis,steroid biosynthesis and steroid hormone biosynthesis.Furthermore,on the basis of the binary logistic regression model,7 fecal metabolites with AUC greater than 0.7 as biomarkers were screened.The AUC of 7 metabolites combined to predict PTB was 0.985,the sensitivity was 95.5%,and the specificity was 91.2%.Conclusion:1.Xpert MTB/RIF Ultra has better diagnostic value in children with PTB compared with traditional microbiological detection methods,especially in younger children(under 5 years old).2.The intestinal flora of children with PTB was significantly different from children with LTBI and healthy controls.The richness of gut microbiota in PTB patients decreased significantly while the relative abundance ratio of Firmicutes to Bacteroides increased.Enterobacteriales,Enterobacteriaceae,Bacteroidaceae,Bacteroides and Faecalibacterium were enriched in childhood PTB.Prevotellaceae,Prevotella＿9,Prevotella＿2 and Alloprevotella play a vital role in LTBI patients.The changes of these specific bacteria may have a potential impact on the occurrence and development of PTB.Moreover,5 differential bacteria could act as PTB-associated biomarkers which provided new ideas for childhood PTB diagnosis.3.There were significant differences in fecal metabolic spectrum among PTB patients,LTBI patients and healthy controls.21 disease-related metabolites were identified.The combined diagnostic model of PTB fecal metabolites was established and its efficacy needs to be evaluated in the future.Abnormal metabolites were mainly enriched in linoleic acid metabolism,primary bile acid biosynthesis and steroid biosynthesis.The present study provided a laboratory basis for exploring the pathogenesis and treatment of PTB.