Technological conception of an optical biosensor with a disposable card for use with bioluminescent bacteria

Sensors and Actuators B 122 (2007) 527–534

Technological conception of an optical biosensor with a disposable card for use with bioluminescent bacteria Habib Horry a,d,1 , Thomas Charrier a , Marie-Jos´e Durand a , Bernard Vrignaud a , Pascal Picart b , Philippe Daniel c , G´erald Thouand a,∗ a

Univ. Nantes, UMR CNRS 6144, GEPEA ERT CBAC 1052, Campus de la Courtaisi`ere-IUT, D´epartement G´enie Biologique, 18 Bd Gaston Defferre, 85035 La Roche-sur-Yon Cedex, France b UMR CNRS 6613, IAM-LAUM, Ecole Nationale des Ing´ enieurs du Mans, Universit´e du Maine, rue Aristote, 72085 Le Mans Cedex 9, France c UMR CNRS 6087, LPEC, Universit´ e du Maine, Av Olivier Messiaen, 72085 Le Mans Cedex 9, France d Biolumine, Campus de la Courtaisi` ere-IUT, D´epartement G´enie Biologique, 18 Bd Gaston Defferre, 85035 La Roche-sur-Yon Cedex, France Received 10 March 2006; received in revised form 20 June 2006; accepted 21 June 2006 Available online 22 August 2006

Abstract The detection of pollutants in the environment by chemical, physical or biological methods is a necessary part of European policy and biosensors should play a role in the monitoring of such pollution. A portable optical biosensor, namely Lumisens 2, is described for the on line detection of pollutants. Lumisens 2 features three main parts: (I) a central unit, (II) a disposable card where bacteria are immobilized and inserted into unit 1, and (III) an acquisition unit to control the device. The first card used in this study (TBTcard ) included the bioluminescent Escherichia coli strain TBT3 harboring the truncated luxAB genes from Vibrio harveyi. TBT3 specifically detects organotin compounds (tributyltin or dibutyltin). Lumisens 2 was validated according to the on line measurement of bioluminescence with a continuous flow of sample. The biosensor was able to detect 2 ␮M of TBT on line in 400 min and after a contact time of 1 h with the pollutant, the accumulation effect of the matrix allowed us to detect 1 nM of this toxic biocide. Because any bioluminescent bacteria can be immobilized in the cards, Lumisens 2 could become a multipurpose optical biosensor for the detection of either the overall toxicity of a sample or one particular pollutant. © 2006 Elsevier B.V. All rights reserved. Keywords: Biosensor; Lumisens 2; Bacterial bioluminescence; Tributyltin (TBT)

1. Introduction Biosensors are analytical devices, which use a biological material as the sensitive element in intimate contact with a transducer. A wide variety of biological elements and associated transducers have already been used for environmental monitoring, including antibodies, cell receptors, nucleic acids, enzymes, whole microorganisms and tissues [1]. With the development of the recombinant DNA technology, genetically engineered bioluminescent bacteria have been used as sensitive elements in bacterial bioluminescent biosensors for environmental monitor-

∗

Corresponding author. Tel.: +33 2 51478441; fax: +33 2 51478451. E-mail address: [email protected] (G. Thouand). 1 Present address: Solabia, ZI no. 2, Le Ther, 2 rue de l’Industrie, BP 60686, Beauvais Cedex, France. 0925-4005/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2006.06.033

ing [2]. The reaction of bioluminescence is catalyzed by the luciferase enzyme coded by the luxAB genes [3]. Three key substrates modulate the light reaction: the dissolved oxygen provided by the surrounding environment, the reduced flavine mononucleotide coming from the bacterial metabolism and a long chain aldehyde such as decanal coded by the luxCDE genes [4]. When spectral emission from the whole bacteria was recorded from 400 to 750 nm with a highly sensitive spectrometer initially devoted to Raman scattering, two peaks were clearly identified, one at 491–500 nm (±5 nm) and a second peak at 585–595 nm (±5 nm) with the Raman CCD [5]. The former peak was the only one detected with traditional spectrometers with a photomultiplier detector commonly used for spectral emission measurement due to their lack of sensitivity and low resolution in the 550–650 nm window [5]. Currently, whole cell bioluminescent biosensors can be divided into two groups. On the one hand we have the liquid-

528

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

phase biosensors featuring bioluminescent cells in suspension in a growth medium [6–9], and on the other hand we have the immobilized-phase biosensors featuring bioluminescent cells maintained in a hydrogel matrix [10–16]. All of the biosensors described in the literature have followed the same architectural design. Basically, the hydrogel matrix with bioluminescent bacteria is placed close to the surface of an optical fiber that has been either connected to a commercial photoncounting device [12,13,16,17] or to a dedicated photomultiplier [10,11,14,15,18]. Recent work has led to the design of integrated circuits for bioluminescence measurements leading to a huge miniaturization of bioluminescent bacterial biosensors onto a silicium chip [19,20]. Besides innovative designs, good performances and accurate applications, the bacterial bioluminescent biosensors described in the literature had to be used by technically skilled operators who often needed strong background knowledge of microbiology, or instrumentation or even signal processing in order to use the material appropriately and gain useful results. The complexity of these biosensors meant that they were confined to specialized laboratories and therefore were unavailable commercially. From this point of view, particularly for environmental applications, a bacterial biosensor should as far as possible, be easy to use and easy to handle, even by a non-microbiologist operator. Bacteria should also be easily replaced and inactivated after use and the device should perform automatic measurements for on-line monitoring purposes. We first published a biosensor namely Lumisens 1 in which cells where continuously cultivated in a small reactor, but lacking simplicity [9,21]. In this paper we can report on the design and the preliminary studies of a new versatile biosensor, namely Lumisens 2, which features characteristics allowing simple process for further environmental application. We initially employed this biosensor for the detection of the well-known biocide, tributyltin mainly used in antifouling paints [22,23]. Here we present our investigations on the position of the bacterial membrane in the disposable card, the time contact between the immobilized bacterial cells and the tributyltin and the bacterial concentration in the hydrogel matrix. Our results led us to test the feasibility of a low tributyltin concentration measurement. 2. Experimental 2.1. Bacterial strain and growth medium The strain Escherichia coli::luxAB TBT3 (Ec::luxAB TBT3) has been previously described in relation to the on line detection of tributyltin [9,24]. The strain obtained after a random insertion of the luxAB genes from Vibrio harveyi into the E. coli DH1 chromosome [F− , recA1, endA1, gyrA96, thi-1, hsdR17, supE44, relA1] is resistant to ampicillin and tetracycline [25]. Since the luxCDE genes were missing, the bioluminescence reaction, induced by organotin compounds [26] needed the addition of a long chain aldehyde, formerly n-decyl aldehyde (decanal), because of its better bioavailability for bioluminescent recombinant cells [27].

A glucose medium with a C/N/P ratio of 100/10/1 was used for the growth of microbial cells. One liter of tap water was filtered through a 0.45 ␮m filter (Sartorius) and supplemented with 1.376 g d-(+)-glucose monohydrate (Merck), 0.1919 g NH4 Cl (Merck), 0.028 g K2 HPO4 (Merck), 5 g NaCl, 0.5 g yeast extract (Merck) and 0.1 g tryptone (Biokar Diagnotics). The pH was adjusted to 7 and the medium was sterilized by autoclaving at 100 ◦ C for 30 min. After cooling, the medium then received 100 ␮L of a concentrated trace element solution SL7 [28], sterilized by filtration through a 0.22 ␮m filter. Antibiotics were sterilized by filtration through a 0.22 ␮m filter and added to the medium to reach final concentrations of 40 ␮g mL−1 ampicillin and 10 ␮g mL−1 tetracyclin. 2.2. Immobilization procedure in the disposable card In order to use the same bacteria throughout the experiments, the Ec::luxAB TBT3 strain was cultivated with the glucose medium under continuous culture conditions at 30 ◦ C under constant agitation in a 100 mL bioreactor as previously described [29]. Each continuous culture was inoculated with 5% (v/v) of a preculture grown for 15 h in the same medium at 30 ◦ C. Continuous feeding of the sterile glucose medium at D = 0.9 h−1 began when the exponential growth phase was reached (optical density at 620 nm OD620 = 0.5). Samples of the bacterial suspension were taken at steady state (OD620 = 0.3) for the immobilization procedure. All of the experiments were performed with immobilized cells in agarose hydrogels, freshly prepared without any time storage. A 2% agarose matrix (final concentration) was chosen for the bacterial immobilization according to previous results [30]. Solutions of low gelling temperature agarose (Sigma) were prepared by dissolving 2% of agarose in MgSO4 10−2 M at 90 ◦ C in a water bath for 10 min under agitation. After homogenization, the agarose solutions were slowly cooled to a temperature in the range of 38–40 ◦ C and briefly stored under agitation until the addition of microbial cells. Two concentrations of bacterial cells (OD620 = 1 and 0.1) were tested in the 2% agarose matrix. The exact volume of the bacterial suspension from the continuous cultivation was centrifugated (3000 × g, 10 min, 20 ◦ C) to reach the final cell concentration needed in agarose. After discarding the supernatant, the same volume of the agarose solution was added to the pellet and the bacterial suspension homogenized at 38–40 ◦ C for 10 min. Two hundred and fifty microliters of the homogenized bacterial cell suspension were poured into the disposable card at room temperature and the polymer hardened in less than 15 min. Given the architecture of the disposable card (see below), the thickness of the bacterial membrane did not exceed 1.5 mm. The disposable card was immediately introduced into the biosensor prior to experiments. 2.3. Chemicals and feeding media All chemicals were prepared in deionized water (Millipore) and stored at room temperature in dark flasks for 1 day. Unless

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

otherwise indicated, all chemicals were purchased from Sigma at a purity exceeding 95%. The glucose medium 0.1× (glucose medium diluted 10-fold) complemented with the SL7 solution and the antibiotics at the appropriate concentration (G medium), was used as the common feeding medium for the biosensor. Given the higher bacterial concentration in the agarose matrix (OD620 = 1), this medium was used as a regeneration medium in order to limit the growth of the bacterial cell. A stock solution of 600 ␮M decanal was made in 50 mL of the glucose medium 0.1× concentrated with 2.3 mL of isopropanol (Labogros) and 5.65 ␮L of pure n-decyl aldehyde. The freshly prepared decanal stock solution was added to 450 mL of the glucose medium to reach a final decanal concentration of 60 ␮M. This revelation medium (GD medium) fed the Ec::luxAB TBT3 cells in order to reveal the bioluminescence reaction. A stock solution of 600 ␮M tributyltin monochloride (Sigma) was made in 70% ethanol (Labogros) and conserved in sterilized amber bottles for 1 week at 4 ◦ C. 1.66 mL of the tributyltin stock solution was added to the GD medium to reach a final concentration of 2 ␮M. This assay medium (GD–TBT medium) was used to both induce and reveal the bioluminescence reaction from the Ec::luxAB TBT3 strain. The final concentrations of isopropanol and ethanol were respectively 0.46% and 0.23%, which did not affect the bacterial response (data not shown). 2.4. Design and instrumentation of the biosensor Lumisens 2 is 60 cm wide, 40 cm long and 35 cm high and includes all of the fluidic and optoelectronic equipment required for the continuous measurement of bacterial bioluminescence (Fig. 1). The biosensor has been designed so that anyone can use it. Only the disposable card, the temperature regulation con-

Fig. 1. Picture of the Lumisens 2 bioluminescent bacterial biosensor. TR, temperature regulation.

529

troller and the sample controller appear in the front of the device, whereas flasks containing the samples and the waste are at the back. Two handles have been placed at the sides so that it can be transported for experimentation in the field. An electric fan maintains a constant airflow into the biosensor to avoid any condensation. Lumisens 2 features three distinct parts: the fluidic layout, the disposable card and the optoelectronic equipment. The latter is composed of the power supply, the temperature regulation and the bioluminescence measurement device. The temperature, measured with a J thermocouple (TC Online), is regulated via a threshold value given by the controller (Chauvin–Arnoult), connected to a heater (Radiospares), inserted close to the disposable card. Variations in the temperature do not exceed ±0.2 ◦ C of the threshold value (data not shown). The bioluminescence is accurately monitored with a dedicated light-measuring device including a photon-counting photomultiplier (Hammamatsu) connected to the disposable card with a fiber optic (Oriel). The light measurements are automatically recorded with a dedicated LabView® interface throughout all of the experiments. The bacteria are immobilized in a 35 mm × 35 mm × 17 mm disposable parallelepipedic card (Fig. 2). Designed in Ertacetal® plastic, the disposable card is chemically and biologically inert, with thermal properties that enable it to be autoclaved. The card is bored through the center to let the fluids come in contact with the bacteria. A 13 mm diameter enclosure has been placed at the depth in which the bacteria are immobilized and includes a measuring chamber of 1.32 cm3 . An optical window (Edmund Optics) is hermetically stuck to the bottom of the measuring chamber on which the membrane of the bacteria is coated and maintained because of a 1 mm groove. The space between the optical window and the fiber optic does not exceed 1 mm, limiting the dispersion of the photons. This layout allows turbid samples to be analyzed, because no fluid circulates between the bacteria and the fiber optic, and also allows the card to be easily replaced. The concept of the removable card has necessitated the complete insulation of the biological material from the elements involved in the signal transduction. This is the main reason why the card is free of any electrical or electronic component and makes it entirely independent of the fixed parts

Fig. 2. Cross section of the disposable card.

530

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

Fig. 3. Hydraulic layout of the biosensor Lumisens 2. DC, disposable card; MPV, multiposition valve; OF, optical fiber; PM, photomultiplier; PP, peristaltic pump; S, sample; V, antireturn valve; W, waste.

of the biosensor. Inserted into its fixed pad, the disposable card closes the hydraulic layout, and the immobilized bacteria are in contact with the fluid diffusing into the immobilization matrix. Fig. 3 shows the flow chart of the liquids from the samples through the disposable card with immobilized bacteria to the waste flask, so that the samples tested can be removed from the biosensor to avoid the release of genetically engineered bacterial cells into the environment. Three independent samples are linked to a remote controlled 10 channel multiposition valve (Valco AG) that allows their specific selection. Each channel can be periodically selected throughout the experiments with a Windows® Hyperterminal® protocol, allowing the design of sequences of liquid exposures. The liquids are drawn out of the disposable card with a peristaltic pump (Meredos, Germany) located between two anti-return valves to avoid the reflux of the liquids into the card. 2.5. Setup of the biosensor The initial bacterial concentration in agarose was adjusted to OD620 = 1 (about 1 × 109 CFU mL−1 ) according to the high cell concentrations described in the biosensor literature [12,13,16]. The concentration of decanal was chosen according to previous work (unpublished data) showing that immobilized cells had a maximum bioluminescence signal with 60 ␮M of decanal. The injection sequence of the feeding media into the biosensor was determined as follows: the sequence began with the regeneration medium (G) followed by the revelation medium (GD) for 60 min each. These two exposure steps allowed cells to reach a physiological steady state and to monitor the basal expression of the bioluminescence. Next, immobilized cells were exposed to 2 ␮M of tributyltin (GD–TBT) for a defined time and the bioluminescence reaction was monitored again with the revelation medium (GD) in order to evaluate the dynamic of the biosensor response. The injection sequence was completed by the return to the regeneration medium.

3.1. Position of the bacterial membrane in the disposable card When inserted into the disposable card, the bacterial membrane can be placed either at a lower position or at an upper position. The influence of the bacterial membrane position in the card on the bioluminescence monitoring has been evaluated according to the previously described initial biosensor setup (Fig. 4). When the bacterial membrane was at the lower position, only the exposure to 2 ␮M of tributyltin with the decanal led to a very slight increase of the signal, up to 900 RLU s−1 . This light production disappeared quickly (less than 1 h after the contact with tributyltin) for the background noise to be achieved. When the bacterial membrane was in the upper position, the exposure to the decanal alone led to an increase of the bioluminescence that corresponded to the basic activity of the cells in the range of 800–1000 RLU s−1 . With the addition of the tributyltin, the bioluminescence increased up to a maximum of 1500 RLU s−1 after approximately 90 min of contact. In contrast to the lower position, the upper position allows the monitoring of the bioluminescence from cells. This phenomenon can be explained by lower and slower diffusion rates of nutrients, tributyltin, decanal

3. Results and discussion Because of its new design, our first task was to evaluate the influence of some biosensor parameters on the bioluminescence level in order to validate the measurement of tributyltin.

Fig. 4. Influence of the position of the bacterial membrane into the disposable card. Exposition sequence: 1 h of glucose medium 0.1× concentrated (G), 1 h of glucose medium 0.1× concentrated with 60 ␮M of decanal (GD), 5 h of glucose medium 0.1× concentrated with 60 ␮M of decanal and 2 ␮M of tributyltin (GD–TBT) and 5 h of GD. Flow rate: 5 mL min−1 .

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

and dissolved oxygen from the sample stream into the agarose membrane. When inserted at the lower position, the bacterial membrane was towards the bottom of the disposable card. As a consequence, an empty space between the surface of the membrane and the sample stream was created, leading to a hydraulic inertia zone where the sample did not renew itself according to the flow rate. Thus, matter transfers from the sample to the immobilized bacteria were lowered. Progressively deprived, either of nutrient sources for the bacterial metabolism or of bioluminescence substrates and inducer, the light production dropped to the background level. However, when inserted at the upper position, the bacterial membrane was at the top of the disposable card in very close contact with the sample stream. Thus, the limitation of nutrients for the bacterial metabolism and substrates for the bioluminescence reaction did not occur. Preliminary work (data not shown) showed that the bioluminescence level was highly and positively correlated with the flow rate of the sample (up to the lowest HRT achieved) when the card was in the lower position. In contrast, when the bacterial membrane was in the upper position, the effect of the flow rate on the bioluminescence level was only effective up to 5 mL min−1 . No effect of the flow rate on the bioluminescence level was observed for higher rates, showing that matter transfers were optimum at this threshold value. More interestingly, beyond 90 min the prolonged exposure of immobilized cells to tributyltin led to a slow decrease of the signal intensity over time from 1500 to around 1100 RLU s−1 (36% loss signal). One of the hypotheses used to explain this phenomenon is that tributyltin has accumulated in the hydrogel matrix, leading to a bioluminescence inhibition. This may have occurred either because of an enzymatic inhibition of the bioluminescence reaction or because of a global metabolic inhibition leading to a loss of viable cells. Other results from bioassay experiments indicated that this loss of bioluminescence was not explained by a loss of tributyltin or decanal from the sample during the time of the experiment (results not shown). Indeed, during on line measurements, samples of the GD and GD–TBT media were collected from the disposable card outlet and used to induce the Ec::luxAB TBT3 strain in microplates as a standard bioassay experiment defined by Durand et al. [26]. Results showed that there were no statistically significant differences between the bioluminescence induced by GD and by GD–TBT, compared to control solutions made with the predicted concentrations of decanal and tributyltin. When the assay medium (GD–TBT) was replaced by the revelation medium (GD), the bioluminescence signal increased strongly to reach a maximum level of 2500 RLU s−1 . Bearing in mind the hypothesis concerning the tributyltin accumulation effect on the agarose membrane, the transition to the GD medium has resulted in a continuous washing of the chemical. The concentration therefore reached a non-toxic level and bioluminescence was recorded. Until now, we did not have any additional experimental information about the accumulation of tributyltin. Nevertheless, its accumulation can be achieved in two different ways. Tributyltin can be passively accumulated in the hydrogel matrix or it can be adsorbed at the surface of the negatively charged bacterial cell

531

wall [31]. These two phenomena can be achieved together. The physico-chemical properties of organotin compounds [22,32] enable a strong adsorption behavior on charged surfaces, and some authors suggested that it was a quantitatively relevant phenomenon for bacterial membranes [31,33]. Further experiments have to be performed in order to localize and quantify the amount of tributyltin in the bacterial membranes. 3.2. Influence of the length of exposure to tributyltin Our previous results showed that the tributyltin measurement was proof of concept during on line experiments with immobilized bacteria, but also highlighted the need to limit both the potential of its accumulation in the membrane and as a consequence the length of the measurement. Considering the fact that tributyltin was uploaded from the liquid sample to the membrane over time, one of the ways to reduce this phenomenon was to reduce the length of exposure (Fig. 5). Cells were exposed to the assay medium (GD–TBT) for 1 h, followed by a 5 h contact time with the revelation medium (GD). In both cases, the pattern of the bioluminescence measurement was identical (initial light production with decanal, increase of bioluminescence with tributyltin and return to the baseline). The bioluminescence peak was also preserved in both cases, showing that one hour was enough to induce the bioluminescence to its highest level, bearing in mind the tributyltin concentration. In this case, not only the amount of time needed to reach the maximum light signal was reduced, but the dynamic of the light response seemed to be faster. Indeed, before the light intensity reached its maximum, close to 2500 RLU s−1 , it slowly decreased, suggesting a progressive loss of induction by tributyltin. Moreover, the 1 h exposure seemed to limit the accumulation phenomenon of tributyltin in the agarose membrane because no bioluminescence inhibition occurred. These results underlined the need to have a strong knowledge of the bioluminescence induction at the molecular level.

Fig. 5. Influence of the contact time between tributyltin and immobilized bacteria. Exposition sequence: 1 h of glucose medium 0.1× concentrated (G), 1 h of glucose medium 0.1× concentrated with 60 ␮M of decanal (GD), 1 h or 5 h of glucose medium 0.1× concentrated with 60 ␮M of decanal and 2 ␮M of tributyltin (GD–TBT) and 5 h of GD. Flow rate: 5 mL min−1 .

532

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

The Ec::luxAB TBT3 strain was selected after the random insertion of the luxAB genes into the bacterial chromosome. To our knowledge, the luxAB genes were inserted close to the stpA gene (DuBow, personal communication), which codes for an HNS-like protein StpA identified as a pleiotropic regulator of gene expression, recombination and genome stability [34], and induced by stress conditions [35]. However, compared to the stpA gene, the luxAB transcription pathway was inverted, which did not allow the expression of the luxAB genes under the control of the stpA promoter. This is coherent with the sensitivity to only organotin compounds in the strain, as demonstrated by Durand et al. [26]. Thus, the bioluminescence induction cannot be correlated with the global stress response governed by the stpA gene. Lately, we isolated the promoter involved in the TBT induction that was cloned close to the ␤-galactosidase reporter gene in order to verify a possible interaction between the TBT and luciferase (for example, a possible non-specific inhibitor). After TBT addition, ␤-galactosidase expressed as much as signal than the bioluminescence with the same detection limit, showing no interaction between TBT and luciferase (H. Gueun´e, personal communication). Nevertheless, the regulation of the luxAB genes still remains under development in our laboratory to elucidate the induction mechanism. We also need to understand the behavior of tributyltin in the agarose membrane, because this phenomenon can be a major limitation for the measurements. Indeed, the “adsorption/desorption” rates in agarose have to be known to predict the bioavailable part of the compound for bacteria. The interaction between tributyltin and bacteria has also to be studied because this can be a major limitation for the induction response of bacterial cells. 3.3. Influence of the bacterial concentration The bioluminescence level of the whole bacterial membrane depends greatly on the bacterial cell concentration. In order to evaluate the impact of the bacterial concentration on the response of Lumisens 2, we tested the effect of a low concentration (OD620 = 0.1) (Fig. 6). For both bacterial loads in the membrane, the pattern of the bacterial response matched all the data previously obtained. Nevertheless, a slight difference in the signal intensity was observed between the two bacterial concentrations, a weaker intensity being found for the lower bacterial load (about 250 RLU s−1 ). If the bioluminescent signal was normalized with the bacterial concentration, the differences have to be completely inversed with a 10-times higher intensity for the lower bacterial load. Two phenomena occurred in the agarose membrane. On the one hand, the light intensity is positively correlated with the bacterial concentration, because the whole bioluminescent signal was the sum of the signals emitted by each cell. The more bacteria were concentrated, the more the signal was intense. On the other hand, the increasing concentration of bacteria in the membrane enhanced the absorption or even the quenching of photons emitted by cells from surrounding congeners. The more the bacteria were concentrated, the less the signal was intense. The loss of light production in the case of a high bacte-

Fig. 6. Influence of the bacterial concentration into the agarose matrix. Exposition sequence: 1 h of glucose medium 0.1× concentrated (G), 1 h of glucose medium 0.1× concentrated with 60 ␮M of decanal (GD), 1 h of glucose medium 0.1× concentrated with 60 ␮M of decanal and 2 ␮M of tributyltin (GD–TBT) and 5 h of GD. Flow rate: 5 mL min−1 .

rial load can also be due to a higher competition for nutrients and substrates, which are in excess in the case of the low bacterial load. 3.4. Detection of a low tributyltin concentration Due to their extensive use in numerous areas of human activity, large amounts of tributyltin have been introduced into various ecosystems. The range of environmental concentrations of tributyltin varied from 10−3 to 1 ␮g L−1 in seawater and freshwater and were about 1 ␮g kg−1 in some sediments and up to 103 ␮g kg−1 in some harbor sediments in the early 90 s [32]. The extensive characterization of the sensitivity to tributyltin of the Ec::luxAB TBT3 strain during bioassay experiments featured a detection limit of around 0.08 ␮M (26 ␮g L−1 ) [26]. As a consequence, it seemed that the Ec::luxAB TBT3 strain was only able to detect environmentally relevant concentrations of tributyltin in heavily contaminated sediments, but not in water. Considering our previous results, we exploited the membrane accumulation capacity for tributyltin in order to detect lower concentrations of pollutant compared to bioassay experiments. We tested the measurement of 1 nM (325 ng L−1 ) of tributyltin using a selected membrane with a low bacterial load (OD620 = 0.1) that was exposed to the pollutant during an extended time contact of 5 h (Fig. 7). When exposed to the revelation medium (GD), the light signal increased from the background noise to reach the standard value of 800 RLU s−1 . After 5 h of contact with tributyltin (GD–TBT), the results showed a smooth but significant increase in the signal intensity to approximately 1000 RLU s−1 . After the tributyltin was removed from the medium (GD), the signal leveled off during 400 min to reach the baseline. Thanks to its accumulation effect in the agarose membrane, it was possible to detect an 80-times lower concentration of tributyltin compared to those previously described with the same bacterial strain.

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

[6]

[7] [8]

[9]

[10]

Fig. 7. Detection of 1 nM (325 ng L−1 ) of tributyltin. Exposition sequence: 1 h of glucose medium 0.1× concentrated (G), 1 h of glucose medium 0.1× concentrated with 60 ␮M of decanal (GD), 5 h of glucose medium 0.1× concentrated with 60 ␮M of decanal and 1 nM of tributyltin (GD–TBT) and 5 h of GD. Flow rate: 5 mL min−1 .

4. Conclusions

[11]

[12]

Lumisens 2 has been designed to be a simple device. The separation between the immobilized bacteria and the device, achieved with the disposable card, means that the system can be used by anyone for on line monitoring purposes. Our results show the proof of concept of the device but further study has to be carried out for the accurate characterization and optimization of its performances and in-field experiments. The main function of Lumisens 2 is to monitor only one class of pollutant, tributyltin. It is possible to detect several classes of pollutants by changing the disposable card—loading it with some different natural or genetically modified bacteria. This work finally suggests that the simple, automatic and continuous pollution detection system we have developed is suited to the real-time monitoring of water pollution.

[13]

Acknowledgements

[19]

The authors wish to thank Sulivan Jouanneau for Fig. 2. This research was supported by grant CER 2000–2006, Action no. 15 (section I), research program no. 18035 (Ville de La Roche-surYon, Conseil G´en´eral de Vend´ee, Conseil R´egional des Pays de Loire, Minist`ere franc¸ais charg´e de la Recherche) and Biolumine SA.

[14]

[15]

[16]

[17]

[18]

[20]

[21]

References [1] U. Bilitewski, A.P.F. Turner, Biosensors for Environmental Monitoring, Harwood Academic Publishers, Amsterdam, 2000. [2] S.F. D’Souza, Microbial biosensors, Biosens. Bioelectron. 16 (2001) 337–353. [3] L.J. Blum, Bio- and Chemi-luminescent Sensors, World Scientific Publishing, Singapore, 1997. [4] E.A. Meighen, Molecular biology of bacterial bioluminescence, Microbiol. Rev. 55 (1991) 123–142. [5] G. Thouand, P. Daniel, H. Horry, P. Picart, M.J. Durand, K. Killham, O.G. Knox, M.S. DuBow, M. Rousseau, Comparison of the spectral emission

[22] [23] [24]

[25]

533

of lux recombinant and bioluminescent marine bacteria, Luminescence 18 (2003) 145–155. S.P. Rupani, M.B. Gu, K.B. Konstantinov, P.S. Dhurjati, T.K. Van Dyk, R.A. LaRossa, Characterization of the stress response of a bioluminescent biological sensor in batch and continuous cultures, Biotechnol. Prog. 12 (1996) 387–392. M.B. Gu, G.C. Gil, J.H. Kim, A two-stage minibioreactor system for continuous toxicity monitoring, Biosens. Bioelectron. 14 (1999) 355–361. M.B. Gu, G.C. Gil, J.H. Kim, Enhancing the sensitivity of a two-stage continuous toxicity monitoring system through the manipulation of the dilution rate, J. Biotechnol. 93 (2002) 283–288. H. Horry, M.J. Durand, P. Picart, L. Bendriaa, P. Daniel, G. Thouand, Development of a biosensor for the detection of tributyltin, Environ. Toxicol. 19 (2004) 342–345. A. Heitzer, K. Malachowsky, J.E. Thonnard, P.R. Bienkowski, D.C. White, G.S. Sayler, Optical biosensor for environmental on-line monitoring of naphthalene and salicylate bioavailability with an immobilized bioluminescent catabolic reporter bacterium, Appl. Environ. Microbiol. 60 (1994) 1487–1494. Y. Ikariyama, S. Nishiguchi, T. Koyama, E. Kobatake, M. Aizawa, M. Tsuda, T. Nakazawa, Fiber-optic-based biomonitoring of benzene derivatives by recombinant E. coli bearing luciferase gene-fused TOLplasmid immobilized on the fiber-optic end, Anal. Chem. 69 (1997) 2600– 2605. G.C. Gil, R.J. Mitchell, S.J. Chang, M.B. Gu, A biosensor for the detection of gas toxicity using a recombinant bioluminescent bacterium, Biosens. Bioelectron. 15 (2000) 23–30. M.B. Gu, S.T. Chang, Soil biosensor for the detection of PAH toxicity using an immobilized recombinant bacterium and a biosurfactant, Biosens. Bioelectron. 16 (2001) 667–674. B. Polyak, E. Bassis, A. Novodvorets, S. Belkin, R.S. Marks, Optical fiber bioluminescent whole-cell microbial biosensors to genotoxicants, Water Sci. Technol. 42 (2000) 305–311. B. Polyak, E. Bassis, A. Novodvorets, S. Belkin, R.S. Marks, Bioluminescent whole cell optical fiber sensor to genotoxicants: system optimization, Sens. Actuators B: Chem. 74 (2001) 18–26. S.T. Chang, H.J. Lee, M.B. Gu, Enhancement in the sensitivity of an immobilized cell-based soil biosensor for monitoring PAH toxicity, Sens. Actuators B: Chem. 97 (2004) 272–276. G.C. Gil, Y.J. Kim, M.B. Gu, Enhancement in the sensitivity of a gas biosensor by using an advanced immobilization of a recombinant bioluminescent bacterium, Biosens. Bioelectron. 17 (2002) 427–432. H. Tani, K. Maehana, T. Kamidate, Chip-based bioassay using bacterial sensor strains immobilized in three-dimensional microfluidic network, Anal. Chem. 76 (2004) 6693–6697. M.L. Simpson, G.S. Sayler, B.M. Applegate, S. Ripp, D.E. Nivens, M.J. Paulus, G.E. Jellison Jr., Bioluminescent-bioreporter integrated circuits form novel whole-cell biosensors, Trends Biotechnol. 16 (1998) 332– 338. M.L. Simpson, G.S. Sayler, G. Patterson, D.E. Nivens, E.K. Bolton, J.M. Rochelle, J.C. Arnott, B.M. Applegate, S. Ripp, M.A. Guillorn, An integrated CMOS microluminometer for low-level luminescence sensing in the bioluminescent bioreporter integrated circuit, Sens. Actuators B: Chem. 72 (2001) 134–140. L. Bendriaa, P. Picart, P. Daniel, H. Horry, M.-J. Durand, G. Thouand, Versatile device for on-line and in situ measurement of growth and light production of bioluminescent cells, Sens. Actuators B: Chem. 103 (2004) 115–121. K. Fent, Ecotoxicology of organotin compounds, Crit. Rev. Toxicol. 26 (1996) 1–117. C. Alzieu, Tributyltin: case study of a chronic contaminant in the coastal environment, Ocean Coast. Manage. 40 (1998) 23–36. G. Thouand, H. Horry, M.J. Durand, P. Picart, L. Bendriaa, P. Daniel, M.S. DuBow, Development of a biosensor for on-line detection of tributyltin with a recombinant bioluminescent Escherichia coli strain, Appl. Microbiol. Biotechnol. 62 (2003) 218–225. A. Guzzo, M.S. DuBow, Construction of stable, single-copy luciferase gene fusions in Escherichia coli, Arch. Microbiol. 156 (1991) 444–448.

534

H. Horry et al. / Sensors and Actuators B 122 (2007) 527–534

[26] M.J. Durand, G. Thouand, T. Dancheva-Ivanova, P. Vachon, M. DuBow, Specific detection of organotin compounds with a recombinant luminescent bacteria, Chemosphere 52 (2003) 103–111. [27] E.A. Meighen, Genetics of bacterial bioluminescence, Annu. Rev. Genet. 28 (1994) 117–139. [28] M. Mergeay, D. Nies, H.G. Schlegel, J. Gerits, P. Charles, F. Van Gijsegem, Alcaligenes eutrophus CH34 is a facultative chemolithotroph with plasmidbound resistance to heavy metals, J. Bacteriol. 162 (1985) 328–334. [29] P. Picart, L. Bendriaa, P. Daniel, H. Horry, M.J. Durand, L. Jouvanneau, G. Thouand, New bioreactor for in situ simultaneous measurement of bioluminescence and cell density, Rev. Sci. Instrum. 75 (2004) 747–755. [30] M. Pernetti, D. Poncelet, G. Thouand, M.C. Annesini, C. Merli, Characterization of agarose as immobilization matrix model for a microbial biosensor, Chem. Ind. 57 (2003) 600–604. [31] G.M. Gadd, Bioremedial potential of microbial mechanisms of metal mobilization and immobilization, Curr. Opin. Biotechnol. 11 (2000) 271–279. [32] M. Hoch, Organotin compounds in the environment—an overview, Appl. Geochem. 16 (2001) 719–743. [33] J.J. Cooney, S. Wuertz, Toxic effects of tin compounds on microorganisms, J. Ind. Microbiol. 4 (1989) 375–402. [34] D.W. Ussery, J.C. Hinton, B.J. Jordi, P.E. Granum, A. Seirafi, R.J. Stephen, A.E. Tupper, G. Berridge, J.M. Sidebotham, C.F. Higgins, The chromatinassociated protein H–NS, Biochimie 76 (1994) 968–980. [35] A. Free, C.J. Dorman, The Escherichia coli stpA gene is transiently expressed during growth in rich medium and is induced in minimal medium and by stress conditions, J. Bacteriol. 179 (1997) 909–918.

Biographies Habib Horry received his PhD in microbiology at the Universit´e de Nantes, France in 2004. He worked on biosensors development mainly for tributyltin detection in the CBAC laboratory. He is now in the Solabia company (France) working on microbiology.

Thomas Charrier is a PhD student in microbiology at the Universit´e de Nantes, France. His thesis is focused on the development of microbial biosensor array for pollutants detection. Marie-Jos´e Durand received her PhD degree in biology from the Universit´e de Metz, France, in 1993. She is assistant professor in ecotoxicology at the University of Nantes (IUT de La Roche-sur-Yon, France). Her research interest includes environmental toxicology bioassay and risk assessment in marine and terrestrial environments. ´ Pascal Picart graduated in optics from the Ecole Sup´erieure d’Optique in 1992 and received his PhD in physics from the University of Paris XI-Orsay in 1995. ´ In 1996 he joined the Ecole Nationale Sup´erieure d’Ing´enieurs du Mans and the Laboratoire d’Acoustique de l’Universit´e du Maine, Le Mans, France. His research interests are with the development of digital holography for mechanics and acoustics metrology. He also contributes to the development of biosensors in collaboration with the Laboratoire CBAC, La Roche-sur-Yon, France. He is member of IEEE, OSA and SPIE. Philippe Daniel received his PhD in physics of materials in 1990. He is professor at the Universit´e du Maine, Le Mans, France, since 2000 and assistant director of the UFR Sciences et Techniques of Universit´e du Maine, since 2001. His research interest includes Raman diffusion applied to the crystallized solids presenting of structural instabilities and applied to the biological systems. G´erald Thouand received his PhD degree in Microbiology from the Universit´e de Nancy, France, in 1993. He is professor in microbiology at the University of Nantes (IUT de La Roche-sur-Yon, France). His research interest includes environmental monitoring of biodegradation and biotechnology. He is mainly involved in the development of biosensors for chemical pollutants detection and pathogenic bacteria. He is member of the French Society of Microbiology (SFM, France) and the Society for Applied Microbiology (SFAM, UK).