Molecular Characteristics Of Vancomycin - Resistant Enterococci And Metabolomics Analysis Of Enterococcus Infection In Mice

Objective To screen vancomycin non-susceptible enterococci from 325 strains of clinical enterococci isolated from 2006 to 2012, and to further conduct the resistance phenotype, resistance genotype and homology analysis of the vancomycin non-susceptible Enterococci; To establish the mouse sepsis model infected by E. faecalis ATCC 29212 and E. faecalis ATCC 51299; To investigate a large scale of sepsis subjects with LC-MS based untargeted and targeted analysis methods to find reliable potential biomarkers for diagnosis and pathways related to sepsis.Methods Screening of the vancomycin non-susceptible enterococci and detection of the susceptibility to teicoplanin were conducted by blood agar dilution method or broth microdilution method. Resistance genotypes were determined by PCR method. Homology analysis was carried out by ERIC-PCR, PFGE and MLST technology. Mouse sepsis models infected with E. faecalis ATCC 29212 and E. faecalis ATCC 51299 were established and confirmed by cultivation of the blood samples from infected animals. To monitor the changes during sepsis development, blood samples were collected before and after infections for white blood cell counting as well as C-reactive protein and procalcitonin assays. LC-MS based untargeted metabonomics strategy was established and applied to find biologically significant metabolite biomarkers in mice infected with E. faecalis ATCC 29212. Blood samples for metabonomic analysis were collected before the injection (0 h), and at 4 h,8 h and 24 h after infection. To explore more reliable potential biomarker candidates, multivariate statistics was applied and an independent t-test was conducted. Discriminating metabolites were selected according to variable importance in projection values (VIP), S-plot, jack-knifed-based confidence intervals and P value. Metabolite set enrichment analysis (MSEA) and metabolite pathway analysis were applied to find pathways which may be affected by sepsis. The LC-MS/MS based targeted metabonomic study was developed based on the results of the untargeted metabonomics and literature reports. The method was intensively validated for precision and sensitivity, and used in targeted metabonomic study of mice infected with E. faecalis ATCC 29212 and ATCC 51299. Potential biomarkers were screened out based on independent t-test and nonparametric test results.Results Totally 17 vancomycin non-susceptible enterococci were screened out from 325 strains of clinical enterococci, including 1 E. faecalis,11 E. faecium, and 5 E. gallinarum. The 1 E. faecalis and 11 E. faecium strains showed high level resistance to vancomycin, and resistance or intermediate resistance to teicoplanin, PCR results suggested vanA genotype. The 5 E. gallinarum strains demonstrated intermediate resistance to vancomycin, and susceptibility to teicoplanin, PCR results suggested vanCl genotype. ERIC homology analysis results showed that most of isolates with the same genotype demonstrated similar ERIC profiles. Pulsed-field gel electrophoresis results of the 17 vancomycin non-susceptible enterococci showed differences in PFGE type. The 1 E. faecalis strain displayed ST4 in multilocus sequence typing analysis, and the 11 E. faecium isolates were allocated into six different sequence types (STs), including ST17, ST78, ST203, ST323, ST571, ST262. Four of the E. faecium isolates belonged to ST78, which was the most prevalent ST type in E. faecium of this study. The LC-MS based untargeted metabonomics analysis was used to find biologically significant metabolite biomarkers for mice infected with E. faecalis ATCC 29212. The levels of 69,69 and 61 metabolites were found to be significantly altered in 4h group,8h group and 24h group respectively in comparison to the control group in positive ion mode; and 99,91 and 90 metabolites respectively in negative ion mode. Based on the pairwise analysis results,95 metabolites were screened out to be significant metabolites, with 21 metabolites showing significant changes during the entire infection course. The metabolic correlation and pathway analysis studies suggested the effects of sepsis on amino acid metabolism, phospholipid metabolism, TCA cycle, linoleic acid and linolenic acid metabolism, urea cycle. The targeted metabonomics with 40 discriminating metabolites were further conducted to screen potential biomarkers for mice infected with E. faecalis ATCC 29212 and E. faecalis ATCC 51299, and 17 and 12 metabolites were identified as potential diagnostic biomarkers for mice infected with E. faecalis ATCC 29212 and ATCC 51299, respectively.Conclusions Vancomycin resistance rate is low in E. faecalis, but vancomycin resistance is relatively a serious problem in E. faecium in China. And the resistant isolates usually carry the vanA gene favoring resistance transfer between different isolates. Homology analysis of vancomycin resistant isolates suggested the possibility of homologic transfer of vancomycin resistance. Among the 17 potential biomarkers in mice infected with E. faecalis ATCC 29212, the levels of 8 metabolites (Taurine, Thymidine, Creatine, Glutamine, Citramalic acid, Cytidine, Riboflavin and Citric acid) increased, and those of 9 metabolites (Cortisone, Hippuric acid, Indole, Tryptophan, Gentisic acid, PE (P-16:0/0:0), Methionine, Insoline, Pyruvate) decreased.9 potential biomarkers in mice infected with E. faecalis ATCC 51299 were screened out, including significant decrease of the levels of 9 significant metabolites (Tryptophan, Pyruvate Indole, Cortisone, Citrulline, Hippuric acid, PE, Gentisic acid and p-Coumaric acid), and increase of the level of cytidine, Thymidine, L-Histidine.9 biomarkers in mice infected with E. faecalis ATCC29212 and E. faecalis ATCC 51299 were the same. Significant changes of the levels of these metabolites indicated that amino acid metabolism, phospholipid metabolism, TCA cycle, and energy metabolism were disturbed in the infected mice.