RESUMO
A yearlong study was performed to examine the effect of antibiotic administration on the bacterial gut flora. Gram-negative facultative anaerobic bacteria were recovered from the feces of healthy adult volunteers administered amoxicillin, minocycline or placebo, and changes determined in antimicrobial resistance (AMR) gene carriage. Seventy percent of the 1039 facultative anaerobic isolates recovered were identified by MALDI-TOF as Escherichia coli. A microarray used to determine virulence and resistance gene carriage demonstrated that AMR genes were widespread in all administration groups, with the most common resistance genes being bla TEM, dfr, strB, tet(A), and tet(B). Following amoxicillin administration, an increase in the proportion of amoxicillin resistant E. coli and a three-fold increase in the levels of bla TEM gene carriage was observed, an effect not observed in the other two treatment groups. Detection of virulence genes, including stx1A, indicated not all E. coli were innocuous commensals. Approximately 150 E. coli collected from 6 participants were selected for pulse field gel electrophoresis (PFGE), and a subset used for characterisation of plasmids and Phenotypic Microarrays (PM). PFGE indicated some E. coli clones had persisted in volunteers for up to 1 year, while others were transient. Although there were no unique characteristics associated with plasmids from persistent or transient isolates, PM assays showed transient isolates had greater adaptability to a range of antiseptic biocides and tetracycline; characteristics which were lost in some, but not all persistent isolates. This study indicates healthy individuals carry bacteria harboring resistance to a variety of antibiotics and biocides in their intestinal tract. Antibiotic administration can have a temporary effect of selecting bacteria, showing co-resistance to multiple antibiotics, some of which can persist within the gut for up to 1 year.
RESUMO
Tuberculosis is a major human and animal disease of major importance worldwide. Genetically, the closely related strains within the Mycobacterium tuberculosis complex which cause disease are well-characterized but there is an urgent need better to understand their phenotypes. To search rapidly for metabolic differences, a working method using Biolog Phenotype MicroArray analysis was developed. Of 380 substrates surveyed, 71 permitted tetrazolium dye reduction, the readout over 7 days in the method. By looking for ≥5-fold differences in dye reduction, 12 substrates differentiated M. tuberculosis H37Rv and Mycobacterium bovis AF2122/97. H37Rv and a Beijing strain of M. tuberculosis could also be distinguished in this way, as could field strains of M. bovis; even pairs of strains within one spoligotype could be distinguished by 2 to 3 substrates. Cluster analysis gave three clear groups: H37Rv, Beijing, and all the M. bovis strains. The substrates used agreed well with prior knowledge, though an unexpected finding that AF2122/97 gave greater dye reduction than H37Rv with hexoses was investigated further, in culture flasks, revealing that hexoses and Tween 80 were synergistic for growth and used simultaneously rather than in a diauxic fashion. Potential new substrates for growth media were revealed, too, most promisingly N-acetyl glucosamine. Osmotic and pH arrays divided the mycobacteria into two groups with different salt tolerance, though in contrast to the substrate arrays the groups did not entirely correlate with taxonomic differences. More interestingly, these arrays suggested differences between the amines used by the M. tuberculosis complex and enteric bacteria in acid tolerance, with some hydrophobic amino acids being highly effective. In contrast, γ-aminobutyrate, used in the enteric bacteria, had no effect in the mycobacteria. This study proved principle that Phenotype MicroArrays can be used with slow-growing pathogenic mycobacteria and already has generated interesting data worthy of further investigation.
