Antibiotic Residues and R-Plasmid Selection: Are in vitro Methods Good Models?

Minimum antibiotic levels for selecting a resistance plasmid in a gnotobiotic animal model

“…We tested four possible effects of low antibiotic concentrations on the in vivo equilibrium between plasmid-free and plasmid-bearing strains (the antibiotic is represented in the following equations by AB, as a function of concentration). For this purpose, it was alternatively assumed that the drug would kill some plasmid-free cells, thus changing equation 1 to: dNsldt = N,(4S -p -AB) (6) reduce the ability of susceptible cells to use a limiting substrate, hence increasing the saturation constant and changing equation 4 to: = [S LI(Ks, + AB) + L) (7) reduce the maximum growth rate of plasmid-free cells (I. ), thus changing equation 4 to: *, = (, -AB) -LI(Kss + L) (8) or enhance the rate of R plasmid transfer from resistant to susceptible strains, leading to a more complex model that we also tested (equations not shown).…”

“…However, most R-plasmid-bearing strains are disadvantaged compared with their plasmid-free counterparts, probably because their in vivo growth rate is reduced (28) by the rise in the saturation constant for the limiting substrate (11). Then, the dominance of RP4 and R270 is likely due to a high rate of plasmid transfer leading to the spread of the plasmid to susceptible cells, as previously shown in gnotobiotic mice colonized by E. coli K-12 strains with chromosomal markers, one of which bore plasmid R270 (6). We chose here to study isogenic strains, because of the detrimental effect of chromosomal mutations on strains in vivo (6,21).…”

“…Then, the dominance of RP4 and R270 is likely due to a high rate of plasmid transfer leading to the spread of the plasmid to susceptible cells, as previously shown in gnotobiotic mice colonized by E. coli K-12 strains with chromosomal markers, one of which bore plasmid R270 (6). We chose here to study isogenic strains, because of the detrimental effect of chromosomal mutations on strains in vivo (6,21). It was therefore not possible to detect in vivo plasmid transfer, since the parent strains had no markers.…”

An evaluation of methods to assess the effect of antimicrobial residues on the human gut flora

“…The experimental system mimics in some ways the digestive transit, and a kind of simplified barrier effect is obtained from the competition between E. coli strains. This model however lacks the ecological conditions in vivo (antimicrobial metabolism, anaerobiosis, substrate limitation, growth rate), and was later invalidated in vivo, by testing the same strains in the gut of gnotobiotic mice (Corpet and Lumeau, 1987).…”

“…* The minimum antibiotic levels for selecting a resistance plasmid was determined in dixenic mice, i.e. germfree mice harboring two isogenic E. coli strains, one of which was carrying an R-plasmid (Corpet and Lumeau, 1987). A strong correlation between antimicrobial dose and selection of the resistant E. coli clone was observed in 70 groups of 3 dixenic mice.…”