A biosensor was developed for the detection of tributyltin (TBT), using a bioluminescent recombinant Escherichia coli:: luxAB strain. Dedicated devices allowed the on-line measurement of bioluminescence, pH and dissolved oxygen values and the feed-back regulation of temperature. Bacterial physiology was monitored by the measurement of the cellular density, respiratory activity and the intracellular level of ATP, glucose and acetate levels. Our results showed that a synthetic glucose medium gave a better TBT detection limit than LB medium (respectively 0.02 micro M and 1.5 micro M TBT). High growth and dilution rates ( D=0.9 h(-1)) allowed maximum light emission from the bacterium. Moreover, simple atmospheric air bubbling was sufficient to provide oxygen for growth and the bioluminescence reaction. Real-time monitoring of bioluminescence after TBT induction occurred with continuous addition of decanal up to 300 micro M, which was not toxic throughout a 7-day experiment. The design of our biosensor and the optimization of the main parameters that influence microbial activity led to the capacity for the detection of TBT.

A biosensor (LUMISENS I), based on the inducible bioluminescence of the Escherichia coli strain TBT3 (Ec::luxAB TBT3), was developed for the detection of the biocide tributyltin. LUMISENS I was set up with a minibioreactor and additional equipment for growth monitoring and light acquisition. The 100-mL minibioreactor has allowed us to establish a stable and reproducible environment for the bacteria (regulation of the growth rate, temperature, pH, and oxygenation), as well as for in situ contact with the xenobiotic. The optical components of the transducer were chosen according to the spectral emission of the strain being studied using a highly sensitive spectrophotometer that was initially devoted to Raman scattering. LUMISENS I was patented according to the in situ, automatic, and simultaneous measurement of the cell density and bioluminescence in the bioreactor. The first results showed that cells cultivated in a synthetic glucose medium provided a better detection limit than did those cultivated in a complex Luria-Bertani (LB) medium (0.02 and 1.5 microM of tributyltin, respectively). Cells maintained at a high growth rate (0.9 h(-1)) led to maximum bioluminescence. Moreover, air bubbling was efficient enough to provide suitable quantities of oxygen for both growth and light emission. When the TBT3 strain used the luxAB genes on its own, decanal, a long-chain aldehyde, had to be added to obtain the bioluminescence reaction. We found that the continuous addition of decanal was the most effective means of obtaining this reaction. The monitoring of the bioluminescence after tributyltin induction showed that the aldehyde was not toxic up to 300 microM during a 7-day experiment. Measurement of tributyltin with LUMISENS I was performed, which showed significant response up to 0.125 microM without any effect on optical density. Even though optimization of the performance of LUMISENS I is still under development, because of its original design, this biosensor is already in use as a warning system for the online monitoring of tributyltin.

No abstract