The in vitro postantibiotic effects (PAE) of aztreonam, ceftazidime, cefuroxime, imipenem, and piperacillin on Escherichia coli ATCC 25922 were studied by a bioluminescence assay of bacterial ATP. In paralel with the PAE investigation, viability and morphology studies were performed. The strain was exposed for 2 h to different concentrations of 1-lactam antibiotics. The antibiotic activity was eliminated by 10-4 dilutions, and regrowth of bacteria was monitored hourly by the bioluminescence assay of bacterial ATP. The length of PAE was dose dependent for ceftazidime (0.5 to 2.6 h), cefuroxime (0.4 to 2.6 h), and imipenem (0.3 to 4.5 h). The long PAE for these antibiotics at higher concentrations was associated with a potent initial killing and the presence of spheroplasts. Aztreonam and piperacillin produced a short, non-dose-dependent PAE (0.4 to 0.95 h). Short PAEs (below 1 h) were seen concomitantly with production of filaments, except in the case of imipenem, which only produced spheroplasts. The bioluminescence method was not jeopardized by filament formation, in contrast to the viable count assay which is normally used for PAE investigations. This makes it possible to study PAE for ,B-lactam antibiotics on gram-negative bacteria with bioluminescence.The persistent suppression of bacterial growth after short antimicrobial exposure is called the postantibiotic effect (PAE). By definition, there should be no subinhibitory concentrations of antimicrobial agent left when the PAE starts.A long PAE provides the potential for administering the antimicrobial agent with longer intervals between doses. This fact has stimulated intensified research concerning the PAE phenomenon during the last decade. P-Lactam antibiotics have consistently shown PAEs against various gram-positive cocci (1,(3)(4)(5). In contrast to gram-positive cocci, there are marked differences in PAE caused by P-lactam antibiotics on gram-negative bacteria (2,3,5,13,16). Gudmunsson and Craig did not see any PAE and frequently noted a negative PAE when Escherichia coli and Pseudomonas aeruginosa were exposed to p-lactam antibiotics (3). Similar results with differences in the PAEs for P-lactam antibiotics on gram-positive cocci and gramnegative bacteria have also been observed by other investigators (1, 4, 16). The only reported exception to the absence of PAE when 1-lactam antibiotics and gram-negative bacteria are combined is imipenem (2, 3, 10, 13).Regrowth after drug removal has generally been monitored by viable count. This is a laborious and indirect method. In this study we have used a new direct method to evaluate the PAE which is based on a bioluminescence assay of bacterial ATP. CFU/ml, and 0.5-ml aliquots of these cultures were added to the tubes, which were then incubated at 37°C. Visible growth was recorded after 24 h.
MATERIALS AND METHODSDetermination of bacterial viability. Bacterial numbers were determined before and after 2 h of exposure to the antibiotic as CFU per milliliter by plating after serial dilution (12) (see Fig. 1).Biol...