ABSTRACT
Tetracycline and beta-lactam resistances among others are used as selection markers in the production of recombinant proteins. The beta-lactam resistance is based on degradation, i.e. the selection pressure gradually disappears from the culture, whereas tetracycline resistance is based on active efflux. We have studied the kinetics of the stability of antibiotic selection pressure in culture using a simple model system (pBR322 in Escherichia coli). Concentrations of ampicillin, carbenicillin and tetracycline were measured with novel sensor cells developed in our lab. These cells are specifically induced to produce light in the presence of the drugs and here their performance was shown to be excellent in monitoring antibiotic concentrations in cell culture. The sensor cells are cheap to produce and use and a high number of samples can be analysed simultaneously. To our surprise, ampicillin and carbenicillin were completely degraded after 2.5-3.0 h of culture, although it has been widely claimed that especially carbenicillin is a good selective agent, whereas tetracycline was stable in culture. beta-lactamase activity in culture was found to correlate with the kinetics of ampicillin degradation.
Subject(s)
Biosensing Techniques/methods , beta-Lactam Resistance , beta-Lactams/metabolism , Anti-Bacterial Agents/pharmacology , Biosensing Techniques/trends , Culture Media/chemistry , Drug Stability , Escherichia coli/drug effects , Escherichia coli/enzymology , Escherichia coli/genetics , Tetracycline/analysis , Tetracycline/pharmacology , Tetracycline Resistance/genetics , Time Factors , beta-Lactamases/analysis , beta-Lactamases/metabolism , beta-Lactams/analysisABSTRACT
We report here the construction of a bacterial reporter system for high-throughput screening of antimicrobial agents. The test organism is the Escherichia coli K-12 strain carrying luciferase genes luxC, luxD, luxA, luxB, and luxE from the bioluminescent bacterium Photorhabdus luminescens in a runaway replication plasmid. The replication of the plasmid can be induced, resulting in a change of the plasmid copy number from 1-2/cell to several hundreds per cell within tens of minutes. This increase in plasmid copies is independent of the replication of the host cells. The system will therefore amplify the effects of antibiotics inhibiting bacterial replication machinery, such as fluoroquinolones, and the inhibitory effects can be measured in real time by luminometry. The biosensor was compared with a strain engineered to emit light constitutively, and it was shown to be much more sensitive to various antibiotics than conventional overnight cultivation methods. The approach shows great potential for high-throughput screening of new compounds.
Subject(s)
Anti-Bacterial Agents/pharmacology , Biosensing Techniques/methods , Drug Evaluation, Preclinical/methods , Escherichia coli/metabolism , Cell Division , Dose-Response Relationship, Drug , Inhibitory Concentration 50 , Light , Luciferases/metabolism , Photometry/methods , Photorhabdus/metabolism , Plasmids/metabolism , Temperature , Time FactorsABSTRACT
A group-specific bioluminescent Escherichia coli strain for studying the action of beta-lactam antibiotics is described. The strain contains a plasmid, pBlaLux1, in which the luciferase genes from Photorhabdus luminescens are inserted under the control of the beta-lactam-responsive element ampR/ampC from Citrobacter freundii. In the presence of beta-lactams, the bacterial cells are induced to express the luciferase enzyme and three additional enzymes generating the substrate for the luciferase reaction. This biosensor for beta-lactams does not need any substrate or cofactor additions, and the bioluminescence can be measured very sensitively in real time by using a luminometer. Basic parameters affecting the light production and induction in the gram-negative model organism E. coli SNO301/pBlaLux1 by various beta-lactams were studied. The dose-response curves were bell shaped, indicating toxic effects for the sensor strain at high concentrations of beta-lactams. Various beta-lactams had fairly different assay ranges: ampicillin, 0.05-1.0 microg/ml; piperacillin, 0.0025-25 microg/ml; imipenem, 0.0025-0.25 microg/ml; cephapirin, 0.025-2.5 microg/ml; cefoxitin, 0.0025-1.5 microg/ml; and oxacillin, 25-500 microg/ml. Also, the induction coefficients (signal over background noninduced control) varied considerably from 3 to 158 in a 2-hour assay. Different non-beta-lactam antibiotics did not cause induction. Because the assay can be automated using microplate technologies, the approach may be suitable for higher throughput analysis of beta-lactam action.
