Online Detection of Metals in Environmental Samples: Comparing Two Concepts of Bioluminescent Bacterial Biosensors Sulivan Jouanneau, Marie Jose ́ Durand, and Ge ́ rald Thouand* University of Nantes, UMR CNRS 6144 GEPEA, France ABSTRACT: In this study, we compared two bacterial biosensors designed for the environmental monitoring of metals: Lumisens III and Lumisens IV. These two biosensors are based on the same bacterial sensors (inducible or constitutive bacterial strains) but with a different conservation mode. The results showed that the biosensor Lumisens III using immobilized cells in agarose hydrogel, allowed to detect artificial mercury contaminations on the limited period of 7 days in laboratory conditions with a reproducibility of 40%. With environmental samples, bioluminescence of the immo- bilized bacteria inside the biosensor was strongly limited by the environmental microflora because of the lack of oxygen, limiting the use of the biosensor to 2 days. The biosensor of the last generation, Lumisens IV, using freeze-dried bacteria in a disposable card allowed a stable detection during 10 days with 3% of reproducibility of the bioluminescence signal both in laboratory conditions and environmental samples. One analysis was performed in only 90 min against 360 min for Lumisens III. Nevertheless, the lack of specificity of the promoter, which regulates the bioluminescent reporter genes, limits the metal detection. We addressed the problem by using Lumisens IV and a data analysis software namely Metalsoft, developed in previous works. Thanks to this analytical software, Lumisens IV was a reliable online biosensor for the multidetection of Cd, As, Hg, and Cu. 1. INTRODUCTION For the improvement of the water resource monitoring, the European Community published, in 2000, the Water Frame- work Directive (WFD, 2000/60/EC). One of these objectives aims at estimating and verifying the quality of these resources in particular through the monitoring of 33 priority substances (Appendix I, guideline 2008/105/CE). Several heavy metals (lead, cadmium, or mercury) are included in the compounds to be detected. Conventional tools (physicochemical methods) used for the detection and quantification of metals in an aqueous environ- ment are particularly sensitive and specific but do not appear to be adequate means for online environmental pollution monitoring (cost, complexity, pretreatement). So, to address the WFD’s objectives, several international research teams developed innovative alternative tools dedicated to metal detection and based on biological sensors: 1,2 enzymes, 3,4 antibodies, 5 or engineered bacteria. 6-16 In this study, we were particularly interested in the last evoked category of biological sensors. For metal detection, bacteria are engineered after the transformation of host strains (e.g., Escherichia coli, Staph- ylococcus aureus) with bioluminescence genes (e.g., lux genes from Aliivibrio f ischeri) controlled by an inducible promoter, which is involved in the intracellular mechanism of metal resistance. 17 Thanks to this genetic modification, these strains are able to emit bioluminescence in the presence of some metals 10,11,18 more or less specifically. 11,19 For online metal measurements, several biosensors were designed featuring two designs: bacteria in a liquid phase (reactor) 20-22 or immobilized (entrapped in a solid matrix) in a dedicated card or chip. 15,16,23-26 With the former, it is easy to control most of the parameters influencing the bacterial activity (temperature, pH, oxygen, etc.). Nevertheless, these systems are complex to use 27 and are not really suited for environmental applications. The systems based on immobilized bacteria seem to be easier to manage but are more difficult to control since microsensors are not yet simple to integrate. For the first time, the recent works of Elad et al. 26 demonstrated the use of entrapped bacteria for the online monitoring of several pollutants, including arsenic, during 10 days in laboratory conditions with a reproducibility of nearly 20%. In our study, we were interested in determining the limits of the bacterial immobilization in biosensors through two major parameters: the reliability of the immobilized bacterial sensors after 10 days in the biosensor and the influence of environmental samples. For that, we compared two biosensors designed in our laboratory: Lumisens III 27,28 and Lumisens IV (this study). These systems differ in the conservation of the cells. Indeed, in Lumisens III, bacteria were maintained active and entrapped in a solid matrix of agarose hydrogel, 27,29 while Received: June 27, 2012 Revised: September 12, 2012 Accepted: September 14, 2012 Published: September 18, 2012 Article pubs.acs.org/est © 2012 American Chemical Society 11979 dx.doi.org/10.1021/es3024918 | Environ. Sci. Technol. 2012, 46, 11979-11987