SUMMARYWe studied the effect of oral selective antibiotic decontamination (SD) on the faecal endotoxin content and microflora in individual C57BL mice. Suppression of the coliform count was associated with an initial rise in faecal endotoxin concentration from 0-1 to 3 1 mg/g wet faeces during the first week of SD, which fell to 0 04 mg/g during the second week of treatment. Cessation of SD resulted in an immediate sharp increase in coliform count followed by its decline and gradual recovery to pre-treatment counts. Faecal endotoxin levels followed a parallel course. SD did not effect significantly the counts of lactobacilli, bacteroides and enterococci.It appears that the coliform population is responsible for the overall level of faecal endotoxin, and that during the initial period of SD endotoxin levels are elevated, an effect which may be mediated by antibiotic-enhanced release of endotoxin.

Polymyxins are cyclic polypeptide antibiotics. In addition to their bactericidal activity they bind lipid A and neutralize the biological effects of bacterial endotoxin. We have studied the three available polymyxin preparations: polymyxin B sulphate (PB), colistin sulphate (CS) and colistin sulphomethate sodium (CMS), and compared their endotoxin binding capacity at equivalent therapeutic dosage. Each polymyxin was bound to a column of Sepharose 4B and challenged with 5 micrograms of endotoxin from Escherichia coli O127:B8. Recovery of endotoxin in the eluate was measured by a quantitative Limulus lysate microassay. PB and CS bound 94% of the challenge dose, CMS 89% and the control column (Sepharose alone) 24%. These results suggest that parenteral CMS (the least toxic polymyxin) retains useful anti-endotoxin capacity, and that in neutropenic patients, oral polymyxin may exert both anti-endotoxin and antimicrobial effects.

S U M M A R YTransfer of anti biotic resistance between Bacteroides organisms and E. coIi in mixed culture under optimal anaerobic conditions was attempted. Donor strains used were E. coli with R factors of a number of compatibility groups and Bacteroides with unusually high antibiotic resistance. Recipient strains included E. coli strains with characteristics favouring conjugal transfer. Although control experiments verified that conjugal transfer of derepressed R factors could occur at a high frequency between E. coli in an anaerobic environment, transfer of antibiotic resistance between E. coli and Bucteroides was never demonstrated.

Differences between minocycline and tetracycline transport were demonstrated in an Escherichia coli strain with and without an R factor (R46) which confers moderate Tc resistance. Minocycline uptake was similar in R+ and Rorganisms, whereas tetracycline uptake was decreased in the R+ as compared to the R-organism. Sodium azide had little effect on tetracycline uptake by either strain but minocycline uptake was completely inhibited by azide. p-Chloromercuribenzoate greatly decreased tetracycline uptake in both strains while minocycline uptake was markedly augmented. Both minocycline and tetracycline were effective inducers of decreased tetracycline transport, but only tetracycline had an effect on minocycline uptake. Mutual inhibition of uptake of one antibiotic by the other could not be demonstrated. These studies indicate different mechanisms of transport of minocycline and tetracycline in E. coli.

No abstract