WO2003100084A1 - A new biosensor - Google Patents

Abstract

The present invention relates to a whole-cell bioluminescent biosensor, PpF1G4, which contains a chromosomally-based sep-lux transcriptional fusion. The biosensor showed significant induction of the sepABC genes by a wide variety of aromatic solvents, including benzene, toluene, ethylbenzene, and xylenes (BTEX), naphthalene, and complex mixtures of hydrocarbons containing trichloroethylene and/or limonene. PpF1G4 represents a second-generation biosensor that is not based on a catabolic promoter but is nonetheless inducible by aromatic pollutants.

Description

A NEW BIOSENSOR

BACKGROUND OF THE INVENTION

(a) Field of the Invention

The invention relates to a new whole-cell bioluminescent biosensor showing significant induction of the sepABC genes by a wide variety of aromatic solvents, including benzene, toluene, ethylbenzene, xylenes (BTEX), naphthalene, and complex mixtures of hydrocarbons containing trichloroethylene (TCE) and/or limonene.

(b) Description of Prior Art

Pseudomonas putida F1 (FpF1 ) is an indigenous soil bacterium that has been well-characterized for its ability to degrade monoaromatic hydrocarbons and its chemotactic response to toluene, benzene, ethylbenzene, trichloroethylene (TCE), and naphthalene. While most gram-negative microbes are susceptible to the toxic effects of hydrophobic compounds accumulating in the membrane structure, some strains, including PpF1 , have evolved resistance mechanisms (Huertas, M.-J., et al., Appl Environ Microbiol 64: 38-42, 1998). This tolerance involves 1 ) modification of the composition of the membrane to increase its rigidity, and 2) an energy-dependent active efflux pump for solvents that is analogous to the multidrug efflux pumps found in antibiotic-resistant microbes (Kieboom, J., et al., J Bacte ol 180: 6769-6772, 1998). Solvent efflux pumps in other P. putida strains are known to respond to a wide variety of aromatic hydrocarbons (Ramos, J.L., et al., J Bacteriol 180: 3323-3329, 1998, and Bugg, T., et al., Appl Environ Microbiol 66: 5387- 5392, 2000).

To date, biosensors developed for the detection of organic compounds have all been based on fusions with promotors from specific catabolic pathways. These include biosensors for naphthalene (King, J.M.H., et al., Science 249: 778-781 , 1990), toluate (de Lorenzo, V., et al., Gene 130: 41-46, 1993), phenols (Shingler, V. and Moore, T. J Bacteriol 176: 1555-1560, 1994), BTEX compounds (Applegate, B.M., et al., J Ind Microbiol Biotechnol 18: 4-9, 1997), isopropylbenzene and related compounds (Selifonova, ON. and Eaton, R.W. Appl Environ Microbiol 62: 778-783, 1996), middle-chain alkanes (Sticher, P., et al., Appl Environ Microbiol 63: 4053-4060, 1997), polychlorinated biphenyls (Layton, A.C., et al., Appl Environ Microbiol 64: 5023-5026, 1998), as well as 2,4-D and 2,4- dichlorophenol (Hay, A.G., et al., Appl Environ Microbiol 66: 4589-4594, 2000.

Organic solvents pose a continuing threat to the health of the environment and biodiversity at large. It would be highly desirable to be provided with a biosensor capable of monitoring the presence of environmental pollutants.

SUMMARY OF THE INVENTION

One aim of the present invention is to provide a biosensor capable of monitoring the presence of environmental pollutants.

In accordance with the present invention there is provided a biosensor for detecting at least one target effectors selected from the group consisting of aromatic compounds, a hydrocarbon mixture containing trichloroethylene and a hydrocarbon mixture containing limonene, said biosensor comprising a first nucleic acid molecule including an operator/promoter sequence of sepR and a sequence encoding sepR, or a functional fragment thereof, and a second nucleic acid molecule including a sequence encoding a reporter molecule having a detectable activity, such as generation of light, said second nucleic acid molecule being under the control of said first nucleic acid molecule acting as an inducible regulator whereby sepR, in absence of the target effector binding thereto, binds to its operator/promoter sequence preventing transcription of said second nucleic acid molecule encoding said reporter molecule, and whereby in presence of the target effector binding to sepR, sepR detach from its operator/promoter sequence allowing transcription of said second nucleic acid molecule encoding the reporter molecule.

In accordance with one aspect of the invention, the operator/promoter sequence of sepR is sepABC.

In accordance with another aspect of the invention, the second nucleic acid molecule is the lux gene or a functional fragment thereof, preferably obtained from a lux operon of a bioluminescent microorganism. The microorganism is preferably of the genera Vibrio, Xenorhabdus, Photorhabdus or Photobacterium, and is more preferably Photorhabdus luminescens. The biosensor of the present invention can be inserted into a host cell. Accordingly, the biosensor can either be incorporated in a plasmid to be inserted in the host cell, or alternatively, the biosensor can be integrated in the chromosome of the host cell.

The host cell can be a bacterial cell, a yeast cell, a fungal cell, a plant cell or an animal cell, and more preferably a bacterial cell.

In accordance with one aspect of the invention, the biosensor is capable of detecting all three isomers of xylene. In another aspect of the invention, the biosensor is capable of detecting aromatic compounds such as benzene, toluene, ethylbenzene, and xylenes (BTEX), naphthalene, or complex mixtures of hydrocarbons.

In accordance with the present invention, there is also provided the whole-cell construct PpF1G4.

Still in accordance with the present invention there is provided a cell containing the biosensor as defined above.

In accordance with one embodiment of the present invention, there is now provided a whole-cell bioluminescent biosensor, constructed with a modified strain of Pseudomonas putida, to be known as Pseudimonas putida F1 G4, hereinafter referred to as PpF1G4, which contains a chromosomally-based sep-lux transcriptional fusion. The biosensor showed significant induction of the sepABC genes by a wide variety of aromatic solvents, including benzene, toluene, ethylbenzene, and xylenes (BTEX), naphthalene, and complex mixtures of hydrocarbons containing TCE and/or limonene. PpF1G4 represents a second-generation biosensor that is not based on a catabolic promoter but is nonetheless inducible by aromatic pollutants.

It will be appreciated that any reporter molecule having a detectable activity will be suitable for the present invention. One of the distinct advantage of the biosensor of the present invention is its capacity to function in a rich medium, i.e. the biosensor is not subject to catabolic repression. In a commercial aspect, the biosensor of the present invention would ideally be inserted in the chromosome of the host organism. Such integration ensure stability. Alternatively, the biosensor of the present invention could be also placed on a plasmid, or a combination of chromosomally inserted elements or plasmid-borne constructs could be used.

In accordance with the present invention, the term "functional fragment" of the lux gene is intended to mean any fragment or part of the gene that can still generate detectable light.

In the present application, the expression "functional fragment of sepR" is intended to mean a fragment of sepR that can still bind to its operator/promoter sequence, such as sep4BC, preventing transcription of the reporter molecule such as the lux gene.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 illustrates DNA from PpF1 encoding a solvent efflux pump and a restriction map on a 13.36 kb DNA segment, showing the fragments that were used for constructing the plasmids or probes listed below the restriction map;

Fig. 2 illustrates a photograph of a gel illustrating the expression of [³⁵S] methionine-labelled polypeptides in E. coli K38/pGP1-2 containing the control espression vector pT7-6 (lane 1 ), the SepC carrying plasmid pT-sepC (lane 2), the SepB carrying plasmid pT-sepβ (lane 3), and the SepA carrying plasmid pT-sep.4 (lane 4), with respect to standard protein weight markers indicated on the left and the major labeled protein bands (in the case of SepA and SepB) indicated on the right;

Fig. 3 illustrates growth of wildtype Ppf1 and some of its derivatives in presence of toluene as sole carbon source in mixed liquid phase (0.1 % v/v);

Figs. 4A and 4B illustrate survival in response to toluene shock of wildtype PpF1 and other null mutants of the same strain in LB media at 30°C, after sudden addition of toluene for cells that had not been exposed to toluene before the shock (Fig. 4A) and after sudden addition of toluene for cells that had been pregrown in toluene (0.3%v/v) before the shock (Fig. 4B), the number of viable cells being determined just before toluene addition and at 30 minute intervals for a period of 3.5 hours;

Fig. 5 illustrates β-galactosidase activity expressed from sepABC-lacZ transcriptional fusion (pPsep-/acZ-Sm^r), todX-lacZ (pMR149), and control plasmid pHRP311 in E. coli RFM443 strain;

Fig. 6 illustrates overexpression and purification of MBP-SepR;

Figs. 7A to 7C illustrate the sequence of the 241 -bp sepR- sepABC intergenic region (Fig. 7A), different segments of the DNA region (Fig. 7B) that were used as probes in the gel retardation assays with MBP- SepR protein (Fig. 7C);

Fig. 8 illustrates the β-galactosidase activity expressed from sepABC-lacZ transcriptional fusion (pPsep-/acZ-Sm^r) and control plasmid pHRP311 mobilized in PpF1 wildtype and sepR null mutant PpF1 (sepR: :Km^r);

Fig. 9 illustrates a dot-blot analysis of sepA RNA extracted from PpF1 and the corresponding null mutant of sepR, PpF1 (sepR: :Km^r), at 0, 10, 30, and 60 minutes after induction with toluene (2.17mM);

Fig. 10 illustrates the β-galactosidase activity of a chromosomally encoded sepABC-lacZ transcriptional fusion (pPsep-lacZ- Km^r) in PpF1APR1 , in the presence of different concentrations of toluene;

Fig. 11 illustrates the β-galactosidase activity of a chromosomally encoded sepABC-lacZ transcriptional fusion (pPsep-lacZ- Km^r) in PpF1APR1 , in the presence of different concentrations of inducers;.

Fig. 12 illustrates expression of β-galactosidase activity from a chromosomal sepABC-lacZ transcriptional fusion (pPsep-/acZ) in PpF1l2.1 and PpF1 (todS::Km^r)L2.1 in presence or absence of toluene (0.05%) added at the mid-log phase;

Fig. 13 illustrates the response of PpFiG4 to a selection of organic compounds (log P values are given in brackets) used at concentrations of 1 mM, except for sparingly soluble compounds, which were a 1 :2 dilution of a saturated solution;

Fig. 14 illustrates the specific bioluminescent response of PpF1G4 to BTEX compounds, error bars indicate standard deviations for triplicate determinations of bioluminescence; Fig. 15 illustrates the bioluminescent response of biosensors PpF1G4 and TVA8 to multicomponentNAPLs where Brent, Isthmus, Maya, and Menemota are varieties of crude oil, and in which error bars indicate standard deviations for triplicate determinations of bioluminescence; and

Fig. 16 illustrates the effect of growth media on luminescent response of biosensors PpF1 G4 and TVA8 exposed to 400mg/L toluene.

DETAILED DESCRIPTION OF THE INVENTION

A new gene cluster, designated sepABC and a divergently transcribed sepR, was found downstream of the two-component todST phosphorelay system that regulates toluene degradation (the tod pathway) in Pseudomonas putida F1 (PpF1 ). The deduced amino acid sequences encoded by sepABC show a high homology to bacterial proteins known to be involved in solvent efflux or multidrug pumps. SepA, SepB and SepC are referred to be periplasmic, inner membrane and outer membrane efflux proteins, respectively. Effects on growth of various PpF1 mutants compared to that of the wildtype in the presence of toluene indicate a possible protective role of the solvent efflux system in a solvent-stressed environment. The sepR gene encodes a 260-residue polypeptide that is most similar to a hypothetical transcriptional regulator in Mycobacterium and appears to be a member of the E. coli IclR repressor protein family. The repressor role of SepR was established by conducting tests with a sep-lac transcriptional fusion in E. coli and PpF1 , expression of SepR as a maltose-binding fusion protein, and mRNA analysis. Southern hybridization experiments indicated that the sepR gene is rather unique among degraders of aromatic compounds compared to the distribution of sepABC homologs.

Experimental procedures Organisms and culture conditions

Bacterial strains used in this application are listed in Table 1. Table 1

List of constructs or strains used herein

Simon, R., et al., Biotechnol Y. 784-789, 1983; Tabor, S. (1990) Expression using the T7 RNA polymerase/promoter system. In: Ausubel, F.A., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K. (eds) Current protocols in molecular biology. Greene- Publishing and Wiley-lnterscience, New York, pp 16.2.1- 16.2.11 ; ^c Droiet, M., et al., Proc Nail Acad Sci USA 92: 3526-3530, 1995; ^d Gibson, D.T., et al., Biochemistry 7: 2653-2662, 1968; ^e Applegate, B.M., et aql., Appl Environ Microbiol 64: 2730-2735, 1998; ^f Lau, P.C.K., et al., Proc NatlAcad Sci USA 94: 1453-1458, 1997; ^s Wang, Y., et al., Mol Gen Genet 246: 570-579, 1995; ^h Parales, R. and Harwood, C.S. Gene 133: 23-30, 1993; ¹ Kokotek, W. and Lotz, W. Gene 84: 467-471 , 1989; and > Frackman, S., et al., J Bacteriol 172: 5767-5773, 1990.

£. coli strains were routinely cultured at 37°C, while PpF1 derived strains were grown at 30°C. Media components were purchased from Difco. Throughout this application, bacterial cells were grown in LB (10 g Bacto Tryptone, 5 g Bacto Yeast Extract, 5 g NaCI per liter solution) or M9 media supplemented with glucose (0.2%) and an appropriate trace metal solution. Solid media contained 1.5% agar. When needed, medium was supplemented with antibiotics at the following concentrations: ampicillin, 100 μg/ml; gentamycine, 20 μg/ml; streptomycine, 100 μg/ml; kanamycine, 50 μg/ml. Cultures were shaken on an orbital shaker at 250 rpm. For the catabolic repression tests, the other media used were TB (10 g Bacto Tryptone, 5 g NaCI per liter) and Terrific Broth (12 g Bacto Tryptone, 24 g Bacto Yeast Extract, 4 mL glycerol per liter). Chemicals

The regular unleaded gasoline (octane number 87) and diesel were used in the following experiments. The JP-4 jet fuel was obtained from the Dorval International Airport in Montreal. The coal tar creosote was obtained from Kopper Industries, Carbon Materials and Chemicals Division (Follansbee, WV). The following three crude oils were obtained from a Refinery: Brent Blend originating from the North Sea, Isthmus Maya, a blend of Isthmus and Maya crude oils from Mexico, and Menemota from Venezuela. The concentrations of BTEX compounds in these petroleum products are reported as mass fractions in Table 2.

Table 2

BTEX Content (Weight percent) in pure petroleum fractions of multicomponent NAPLs

Multicomponent Benzene Toluene Ethylbenzene Xylenes Total BTEX

NAPL

Gasoline 2.11 8.65 2.02 6.54 19.32

Diesel 0.07 0.26 0.1 0.29 0.72

JP-4 jet fuel n/a n/a n/a n/a n/a

Creosote n/a n/a n/a n/a n/a

Crude Oils

Menemota n/a n/a n/a n/a n/a

Venezuela

Brent Blend 0.27 0.82 0.33 1.04 2.46

Isthmus 0.04 0.99 0.17 0.69 1.9

Maya 0.05 0.15 0.08 0.31 0.59

BTEX analysis for gasoline and diesel was performed on a Perkin-Elmer Sigma 2000 gas chromatograph, DB-petro100™ capillary column of dimensions 100 m x 0.25 mm x 0.25 μm, through headspace analysis. Typical compositions for Brent Blend, Isthmus, and Maya crude oils were obtained from an online database for the properties of crude oils and oil products, maintained by the Environmental Technology Centre, a division of Environment Canada. The test organic compounds that were used in the β-galactosidase and bioluminescence assays were purchased from Sigma-Aldrich and were of reagent grade purity or better. DNA techniques

Plasmids used in this application, as well as their detailed construction, are listed in Table 1. The plasmid pPF9SX, containing the sequence that encompasses the todS -todT genes, was obtained by performing a Sac\-Xho\ digestion of genomic DNA from strain PpF1 (todS::Kmr) and by generating a library in plasmid pBluescriptKS(-), digested with the same enzymes. The resulting plasmid pPF9SX was obtained after kanamycin resistance (Km^r) selection of the corresponding clones transformed in E. coli. This plasmid was then used as a probe to obtain the subsequent overlapping clone pGEM-3.8PB. Cloning procedures, including digestions with restriction enzymes, electrophoresis, Southern hybridizations, and tranformations were carried out using standard methods (Sambrook J., Fritsch, E.F., and Maniatis, T. (1989) Molecular cloning: A laboratory manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press). E. coli DH5α was routinely used for recombinant plasmid maintenance and isolation, unless otherwise specified. Enzymes used in cloning procedures were purchased from Pharmacia or New England Biolabs. Chromosomal DNA was isolated using a Genomic-tip System (Qiagen), plasmid DNA was purified with a QIAprep Spin Miniprep kit (Qiagen), and DNA fragments from agarose gels were purified using a Qiaexll gel extraction system (Qiagen). Nucleotide sequencing reactions were performed with purified double strand plasmid DNA using ABI PRISM^® dRhodamine Terminator Cycle Sequencing Kit, as recommended by the supplier, and the products were detected using an automated DNA sequencer (model 377, Applied Biosystems, Inc.). To analyze sequence homologies, nucleotide sequences were compared with the BLAST program, available from the National Center for Biotechnology Information server. Labelling of DNA probes for southern and RNA dot-blot analysis was performed using a Random Primed DNA Labelling kit (Roche Molecular Biochemicals) with [ ³²P]-dCTP. PCR was performed on a Perkin-Elmer DNA Thermal Cycler™ 480, using Taq™ DNA polymerase. Expression of sep genes with T7 polymerase

E. coli strain K38 harboring the T7 polymerase gene on plasmid pGP1-2 was transformed with pT7-5 or pT7-6 derived plasmids carrying the individual sep genes (Table 1 ). Proteins were labelled with [³⁵S]methionine and separated in conventional sodium dodecyl sulfate

(SDS)-10% polyacrylamide gel.

Distribution of sepABC and sepR genes in other bacterial isolates

Southern transfer of DNA was done on a positively charged nylon-based membrane (GeneScreen™ Plus from Dupont). The aqueous hybridization buffer protocol suggested by the company using ³²P-labelled DNA probes was followed. The probes used for the detection of the individual genes were probe A1 for Sep A, probe B for Sep B, probe C for SepC (Fig. 1). Detection of SepR was done using the BamYW insert fragment from pMBP-sepR as a probe. Prehybridization and hybridization were done at 65°C, and then samples were washed at 60°C in 2X SSC (Sambrook et al, 1989, supra), 1 % SDS. Construction of strains

To construct null mutants of sepR, sepB, and sepC, a Km resistance cassette from pUC4K was introduced in unique restriction sites within the genes of interest, cloned on pUC13-or/T derived plasmids. The resulting plasmids were then conjugated into PpF1 from E. coli S17-1. Km^r transconjugants were first selected on M9 glucose plates, allowing counterselection against E. coli. Transconjugants resulting from double cross-over recombinations were obtained as Km^r and ampicillin sensitive (Ap^s) mutants. Since PpF1 is already resistant to 100 μg ml^"1 of ampicillin, transconjugants resulting from integration of pUC13 into the PpF1 chromosome were ruled out by screening with 750 μg/ml of ampicillin. To confirm that the wild-type genes had been replaced by the Km^r "knocked- out "allele, selected transconjugants were verified by polymerase chain reaction (PCR) amplification of the genomic DNA and by Southern hybridization.

The luxCDABE operon cloned from the terrestrial bacterium Photorhabdus luminescens, formerly Xenorhabdus luminescens, was employed for the construction of the sep-lux biosensor. Use of the complete lux cassette permitted measurement of bioluminescence without the addition of an aldehyde substrate. CaCI₂ treated E. coli DH5α cells were transformed with the ligation mixture containing plasmid pBB5.23- oriT-lux (Table 1 ) and plated on Luria-Bertani plates containing 100 μg/mL ampicillin. Plates were inspected in the dark for the production of light. Plasmid DNA was isolated from positive clones and analyzed with restriction endonucleases to confirm the presence of the sep-lux construct. This recombinant DNA was then introduced into CaCI₂ treated E. coli S17- 1 cells and transferred into PpF1 by conjugation. Transconjugants were plated on minimum M9 glucose medium with 100 μg/mL ampicillin in order to counterselect for the recipient against the donor E. coli S17-1 which cannot grow on minimal media alone.

Since the delivery plasmid cannot be maintained in PpF1 , it became integrated into the bacterial chromosome. The PpF1 cells were grown in liquid minimum M9 glucose medium without selective pressure, in order to ensure loss of the plasmid. After several generations of growth in order to ensure construct stability, the culture was plated on minimum M9 glucose medium containing 750 μg/mL ampicillin. Since wild-type PpF1 is naturally resistant to 100 μg/mL ampicillin, a higher concentration of the antibiotic was required to select for colonies where the plasmid DNA, including increased ampicillin resistance, had been integrated into the chromosome. The viable colonies were grown overnight with and without the presence of toluene vapour, a known inducer of the sep promoter, and * then tested for light production. One clone in particular, designated "F1G4", was chosen because it had low background light levels in the absence of an inducer, and was the brightest when exposed to toluene. The advantages to having a chromosomally-encoded reporter element as opposed to a plasmid based system include the fact that the transcriptional fusion is more stable since selective pressure is not required to prevent plasmid loss and there are no copy number effects. Since the growth rate of the biosensor strain PpF1 G4 was identical to that of the parent strain, PpF1 , it was assumed that the insertion of bioluminescent genes did not affect the fitness of the biosensor.

A similar approach was used to construct PpF1 strain derivatives with a sepABC-lacZ fusion inserted in their chromosomes. PpF1APR1 was obtained by conjugation with E. coli S17-1 , containing the pPsep-/acZ-Km^r plasmid. This plasmid has a lacZ-Krrf cassette excised from pKOK6.1 and inserted in the sepA gene of pBB5.23-or/T (Fig. 1 ). The transconjugants resulting from single cross-over were selected as Km^r and Ap^r clones, after many passages without selective pressure, in order to assure proper stable insertion in the chromosome. Strains PpF1L2.1 and PpF1(toofS::Km^r)L2.1 were constructed in a similar way, but using E. coli S17-1 , containing plasmid pPsep-lacZ, as the donor in the conjugation. In this plasmid a lacZ cassette is inserted into the sepA gene (Fig. 1), thus allowing only an ampicillin resistance (Ap^r) selection for clones presenting a single cross-over.

Growth of PPF1 derivatives on toluene and survival in response to toluene shock

Survival of PpF1 strain derivatives in response to toluene shock was determined in LB media. When cells were preexposed to toluene via the gas phase, toluene was introduced into the glass bulb of a central vessel placed in the culture flask, in such a way as to avoid direct contact of the solvent with the liquid media. Cells were pregrown with or without the presence of toluene until the cultures reached an OD₆oo of 1.0. Toluene was then added at a concentration of 0.3% (v/v) and the number of viable cells was determined right before and after toluene addition over time. Antimicrobial susceptibility of PpF1 and sepP::Km^r and sepC::Km^r mutants

The minimal inhibitory concentration (MIC, i.e. the lowest concentration of antibiotic inhibiting visible cell growth after overnight incubation at 30°C) of the following antibiotics was determined: tetracyclin, ampicillin, chloramphenicol, streptomycin and novomycin. The tests were performed in LB, as well as in minimal glucose media. To test the effect of adaptation to toluene on the antibiotic resistance, cells of PpF1 strain derivatives were taken from logarithmic phase after growth in minimal glucose media in the presence or absence of toluene (0.05% v/v). The cells were diluted and plated on LB plates containing different concentrations of antibiotics (chloramphenicol, 50-200 μg/mL; tetracycline, 1-5 μg/mL; ampicillin, 400-1000 μg/mL; streptomycin, 50-100 μg/mL), and the number of surviving cells at each concentration was evaluated. Overexpression and purification of SepR

SepR was overproduced as an MBP fusion protein in E. coli UT5600 cells harbouring the plasmid pMBP-sepP (Table 1), using the pMAL system (NEB). Primers used in PCR amplification of sepR for its cloning in pMal-c2X were: solreg-5' (ATGGATCCAT GAGCGATTCG GAAGAAAG) and solreg-3' (ATGGATCCTC TAATCAACCC G CAAACTC). The addition of Bam \ sites at the 5' and 3' ends allowed cloning into the corresponding sites of the vector. These measures were taken so that the ATG translational start site of sepR would be in the same reading frame as the MBP protein. Expression of the fusion protein was induced by addition of 0.3 mM isopropyl-β-D-thiogalactopyranoside (IPTG) for two hours. The protein was purified using the amylose resin as affinity ligand, as recommended by the manufacturer. The only modification to the original protocol was the use of protease inhibitor (Complete, EDTA-free, Roche Diagnostics) during all the purification steps preceding the binding to the resin. Protein concentration was determined with the Bradford reagent (BioRad Protein Assay), using bovine serum albumin (BSA) as a standard. Cell extracts of E. coli transformants grown with or without IPTG, as well as the purified protein, were subjected to 7.5% SDS- polyacrylamide gel electrophoresis (SDS-PAGE), after denaturation by boiling in 2X SDS sample buffer. The proteins were stained with Coomassie brilliant blue R-250. Gel mobility shift assays

To perform the gel shift assays, DNA fragments originating from the intergenic region of sepR-sepABC were synthetized by PCR, using plasmid pGEM-3.8PB as a template. The following oligonucleotides were used to generate the different DNA probes described in Fig.7: probe C, opsep-5' (TCTCACCGTT CGTCTCCTGG) and opsep-3' (TTCTGATCCA GGCCACCGTG) ; probe L, opsep-5' and opL-3' (TCACGCATGG CATGAACGGC); probe R, opR-5' (CGTGACTGAC CTGCACCCAG) and opsep-3'. Probe N was obtained by digesting probe C with Nsi\. The corresponding fragments were isolated from a 1.6% agarose gel and then labelled with [γ-³²P]-ATP with T4 polynucleotide kinase (Ausubel et al., 1990, supra), and purified on a CentriSep™ column (Princeton Separations, Inc). Binding between labelled DNA (0.1 ng) and varying amounts of purified MBP-SepR protein was carried out for 30 minutes at room temperature in 20 μl of binding buffer (20mM tris-CI pH 7.4, 2 mM MgCI₂, 2 mM EDTA, 10 mM KCI, 0.3 mM DTT, 300 μg/ml BSA, 50 μg/ml poly(dl-dC)^»poly(dl-dC), glycerol 2.5%(v/v)). Half of the mixture was run on a 4.5% polyacrylamide gel (acrylamide/bisacrylamide : 30/0.8) made in TAE buffer. Gels were dried at 80°C under vacuum and subjected to the autoradiography. β-qalactosidase activity measurements

The β-galactosidase activity of the various lacZ transcriptional fusions used throughout this application was determined according to Miller (Miller, J.H. (1992) A short course in bacterial genetics: A laboratory manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press). Cells were permeabilized with chloroform and sodium dodecyl sulfate (0.002%) prior to β-galactosidase measurements.

Determination of the activity of the plasmid-encoded sep-lacZ fusion (pPsep-/acZ-Sm^r) in E. coli RFM443 strain derivatives was performed as follows: cells were grown in LB glucose (0.4%). Since the strains carried a second plasmid (either pTrc99A or psepP-pTrc), antibiotics for the marker of each plasmid were added to the media. When the cells reached an OD₆oo of 0.5, IPTG was added to a final concentration of 1mM. Cells were incubated further for 140 minutes prior to permeabilization of the cells. Activity of the control lacZ plasmids (pMR149 or pHRP311 ) was measured in a similar way.

PpF1 strain derivatives carrying chromosomal sepABC-lacZ fusions (PpF1APR1 and PpF1 L2.1 ) were grown to an OD₆oo θf 0.5 in LB, at which point toluene or other inducers were added, and then cells were grown for an additional 2 hours at 30°C prior to β-galactosidase activity measurements. The same approach was used when PpF1 strain derivatives had been mobilized with plasmids, except that the appropriate antibiotics were added to the media. RNA isolation and analysis

PpF1 derived strains grown in M9 glucose were induced with liquid toluene (2.17mM) when they reached an ODβoo of 0.6. A volume of 3 mL of cell culture was sampled at 10-minute intervals for 30 minutes, centrifuged for 30 seconds in 1.5 mL microtubes, and quickly frozen at -80 °C, until RNA was extracted. Total RNA was isolated using an Rneasy™ extraction Kit (Qiagen), according to the manufacturer's instructions, except for the following modifications : lysosyme final concentration in TE buffer was increased to 12 mg/mL and the volume of lysis buffer was doubled. Samples were on-column digested with Rnase Free Dnase (Qiagen) prior to elution from the column. The integrity of purified RNA was checked by rapid electrophoresis in 1 % agarose TBE gel. RNA was mixed with 2 volumes of denaturing buffer, consisting of 9.25% formaldehyde (w/v), 75% (v/v) formamide, and heated 10 minutes at 60°C prior to electrophoresis. For dot-blot analysis, 3 μg of purified RNA was immobilized on a positively charged nylon membrane (Roche Molecular Biochemicals), using a Bio-Dot apparatus (Bio-Rad Laboratories). RNA blotting, as well as its denaturation prior to its aplication to the membrane, were done according to Sambrook et al. (supra). After UV crosslinking, the membrane was hybridized with a sepA ³²P-randomly labelled probe (probe A2, Table 1 ) using standard protocol. Washes were performed at 55°C and the final high stringency wash was done in 0.2X SSC, 0.1 % SDS. Bioluminescence assays

In preliminary experiments, it was determined that the strongest and most consistent light response occurred when the cells were induced at an ODβoo of 0.5 and incubated for a period of 2 hours. Thus, these conditions were employed for all further experiments. Minimum M9 glucose medium was used for all bioluminescence assays, except for the experiments that evaluated the catabolic repression of biosensors PpF1G4 and TVA8 in rich growth media, where Terrific, Luria-Bertani, and TB broths were used. A fresh overnight culture was used to inoculate flasks containing the growth medium, and cells were grown with shaking at 250 rpm at 30°C to an ODΘOO of 0.5, then centrifuged and resuspended in 1/25 of the original volume. The concentrated cell suspension was diluted 25 times by adding 200 μL to 4.8 mL of growth medium with a known concentration of analyte, thus attaining the original ODeoo for the cells. For the sparingly soluble test compounds used in the screening experiments, dilutions were prepared from near saturated or saturated solutions in minimum M9 medium to attain the desired concentration of the analyte (either 1 mM or a 1 :2 dilution of a saturated solution). For the NAPL tests, 4 μL of either gasoline, JP-4 jet fuel, diesel, coal tar creosote, or the three varieties of crude oil were added to 4.8 mL of minimum M9 glucose medium prior to addition of the cells. The contents of the test vials were well mixed on a vortex, and incubated at 30°C in a shaker with agitation at 250 rpm for 2 hours. A sample from each test vial was diluted to an OD₆oo of 0.3, in order to reduce the light quenching effects related to high cell densities. Light was measured in triplicate, using opaque 96-well plates, in a Dynex™ MLX Microtiter Plate Luminometer. The exact value of ODΘOO after dilution was measured and recorded for each sample, in order to express the light signal in terms of specific RLU (i.e., RLU/ODΘ_OO)- Cloning of sepABC-sepR gene cluster

DNA sequencing of a previously generated clone containing the two-component signal-transduction todS and todT genes that regulate the tod (toluene degradation) pathway in PpF1 indicated the presence of an open reading frame downstream of the todT gene and designated sepC (formerly todil, Wang, Y., et al., Mol Gen Genet 246: 570-579, 1995) with a high sequence identity with the oprM of Pseudomonas aeruginosa PA01 (Table 1).

The plasmid pPF9SX (see Fig. 1), containing the complete sepC gene and downstream region, was obtained and sequenced entirely. This established the sequence of sepC and the accompanying sepB and sepA. By homology with the solvent resistant pump (srp) system of P. putida S12 (Kieboom, J., et al., J Bacteriol 180: 6769-6772, 1998), sepA is missing at least the first two codons. A subsequent clone, pGEM-3.8PB (see Fig. 1 ) allowed the complete sequence determination of sepA as well as the identification of a divergently transcribed regulatory gene, sepR.

The deduced amino acid sequences encoded by sepABC show a high homology to bacterial proteins known to be involved in solvent efflux or multidrug pumps, namely the solvent resistant pump (srpABC) genes from P. putida S12 (Kieboom, J., et al., J Bacteriol 180: 6769-6772, 1998), the mexA-mexB-oprM genes found in P. aeruginosa PA01 (Poole, K., et al., Antimicrob Agents Chemother 40: 2021-2028, 1996), and the toluene tolerance genes (ttgABC) of P. putida DOT-T1 E (Ramos, J.L., et al., J Bacteriol 180: 3323-3329, 1998). The predicted molecular masses of the proteins were authenticated by expression in the E. coli T7 polymerase/promoter system (see Fig. 2). The predicted amino acid sequence of SepR is most closely related (38.2% identity) to a hypothetical protein (Rv1719) present in the Mycobacte um tuberculosis H37Rv genome (emb Z81360). As well, there are two additional homologs, RV2989 (27% identity) and RV1773C (24.2% identity) in the same genome. The closest functional homolog is the E. coli repressor protein (IclR) for the aceBAK operon (M31761 ) at a level of 26.5% identity. Table 3 provides further details about the characteristics of the SepABC and SepR proteins.

Table 3

Characteristics of SepA, SepB, SepC and SepR proteins

Gene Functional name Amino Predicted Experimental % identity acids Mol. Mass Mol. Mass (kDa) sepA Periplasmic efflux 382 41525 42 SrpA (73)^a protein MexA (54.7) TtgA (54.2)^c sepB Inner membrane efflux 1046 113359 116 SrpB (82)^a protein MexB (62)^b TtgB (60)^c sepC Outer membrane efflux 480 52813 52 SrpC (59.6)^a protein OprM (59.8) TtgC (56.5)^c sepR Repressor protein 260 27894 29 ^a Genebank Accession Number AF029405 b Genebank Accession NumberAB11381 and L23839; and ^c Genebank Accession Number AF031417

Since the organization of sepABC with a cognate sepR repressor gene in PpF1 is a rare feature among the solvent efflux pump genes studied thus far, it was decided to probe the distribution of sep genes in various bacterial strains. Southern hybridization experiments indicated that the sepR gene is rather unique among degraders of aromatic compounds compared to the distribution of sepABC homologs (Table 4).

Table 4 Distribution of SepABC and SepR genes in other bacterial isolates Bacterial Strain DNA-DNA hybridization

SepA sepB sepC sepR

P.putida G7+ + + + -

P.putida mt-2 + + + -

P.putida PRS2000 + + + -

P.putida Idaho + + + +

P.pseudoalcaligenes KF707 + + + -

Burkholderia sp. LB400 + + -

Burkholderia sp. JS150 + -

B.cepacia G4 + + -

S.yanoikuyae B1 + + -

S.yanoikuyae Q1 + + -

S.paucimobilis EPA505 - - - -

S.paucimobilis EPA505-47-1 - - - -

Rhodococcus sp. M5 - - - -

Rhodococcus sp. ATCC 55309 - - - -

+ sign means that a visible band was observed on the Southern hybridization when the genomic DNA was hybridized with the respective probes; and - sign means no visible band could be observed.

These results indicate that the sepR gene of PpF1 is rather unique except for a possible homolog in P. putida strain Idaho. Phenotypic characterization of null mutants of sepABC, sepR genes

In order to determine the possible involvement of the sepABC genes and their putative regulatory gene sepR in solvent tolerance, null mutants of some of these genes were constructed and these mutants were characterized. PpF1 and PpF1(sepC::Km^r) were grown in minimal medium at three concentrations of toluene (0.1 %, 0.5% and 1 % v/v) as sole carbon source for 24 hours at 30°C, and the absorbance of each culture was measured. Both strains were found to grow similarly (optical density at 600nm, OD₆₀o: 0.86 and 1.05, respectively) at 0.1% toluene. Growth at 1% was minimal (ODeoo: 0.05 and 0.02, respectively) whereas at 0.5% there was an appreciable difference in growth of the wildtype (ODeoo: 0.47) versus the mutant (ODeoo: 0.08). This result suggests the involvement of sepC in conferring toluene tolerance to PpF1. To further evidence the difference in growth between the wild- type and sep mutants, a time study, in which the growth of PpF1 , PpF1(sepB::Km^r), PpF1(sepC::Km^r) and PpF1(sepR::Km^r) was monitored, was conducted in minimal medium with toluene as sole carbon source, provided in mixed liquid phase (0.1%v/v). Fig. 3 shows that the sepR mutant consistently exhibited a shorter lag period in its growth compared to the wildtype. In order to correlate the possible role of the sep genes with toluene tolerance in PpF1 , the short-term survival of the wild-type and the sep mutants (sepP::Km^r, sepC::Km^r, sepB: :Km^r) was evaluated in response to toluene shock.. After pregrowing cells in Luria-Bertani (LB) at 30°C, with or without toluene supplied in the gas phase, the cultures were spiked with toluene (0.3% v/v), and then the number of viable cells was assessed over time. When cells were preexposed to toluene, both the sepC::Km^r and sepβ::Km^r mutants were less tolerant than the wildtype PpF1 (Figs. 4A and 4B). Moreover, when cells were not pregrown with toluene, the sepP::Km^r mutant had higher viable cell counts compared to the wildtype. The survival of the sepR::Km^r mutant may be linked to increased expression of the sep genes, indicating that the solvent efflux system of PpF1 seems to play a protective role in regard to solvent tolerance.

To determine if the sep genes are involved in antibiotic efflux, the antimicrobial susceptibility of PpF1 and the PpF1(sepP::Km^r) and PpF1(sepC::Km^r) mutants were compared. In tests conducted in LB and minimal glucose media, the minimal inhibitory concentration (MIC) of tetracyclin, ampicillin, chloramphenicol, streptomycin, and novomycin was determined for the wildtype and the mutants. Both sepR and sepC mutants showed no difference in their antibiotic susceptibility compared to the wildtype. The survival of all the strains in the presence of chloramphenicol and tetracycline was seen to increase after they had been preexposed to 0.05% toluene. Evidence that SepR is a repressor of sepABC gene expression

Based on its sequence homology with other DNA-binding proteins known to act as repressors, the inventors wanted to ascertain that the sepR gene, located next to sepABC and transcribed in the opposite direction, codes for a protein that regulates the expression of those genes. In order to show that SepR can repress sepABC gene expression, the β-galactosidase activity of a sepABC-lacZ transcriptional fusion (pPsep- /acZ-Sm^r, Table 1) was compared to that of plasmid pMR149, containing a todX-lacZ transcriptional fusion, and to that of the promoterless lacZ fusion vector pHRP311 , in E. coli strain RFM443 (lac^'). In each of these strains, either SepR was overexpressed in trans (psepR-pTrc), or the plasmid p7rc99A was used as a negative control. Fig. 5 shows that when SepR is provided in trans, sepABC gene expression is completely repressed in E. coli (i.e., β-galactosidase activity returns to the basal level of a promoterless lacZ fusion), while the todX-lacZ fusion is unaffected by the overexpression of SepR. Similar results were obtained when SepR was overexpressed as a maltose binding protein (MBP) fusion protein. In Fig. 5, the plasmid containing the sepR gene (psepP-pTrc) and control plasmid pTrc99A were co-expressed in the same strain.

In order to give direct in vitro evidence for the binding of sepR to the promoter region of sepABC, gel mobility shift assays were performed. Fig. 6 shows the overexpression and purification of MBP-SepR. More particularly, Fig. 6 illustrates a SDS 7.5%-polyacπylamide gel stained with Coomassie blue showing MBP-sepP induction and purification. The molecular mass (in kDa) of standard protein markers (lane 7) are shown on the right. Also shown are the crude extract obtained from UT5600 cells harboring the control plasmid pMAL-c2X under uninduced conditions (lane 1) and induced with IPTG (lane 2), the fusion plasmid pMBP-sepR under uninduced conditions (lane 3) and induced with IPTG (lane 4). The purified MBP-SepR protein eluted from amylase column with maltose can be seen in lane 5 (4μg) and lane 6 (20μg).

Furthermore, Fig. 7C demonstrates the DNA binding activity of MBP-SepR to a 120-bp DNA fragment that is proximal to the sepA gene (Fig. 7A). In a separate assay, MBP alone was found not to bind to any of the fragments. One of the two 6-bp hairpin structures is an operator site (Fig. 7B). These results indicate that SepR binds specifically to the sepABC-sepR intergenic region, thus confirming its role as repressor. In Fig. 7A, converging arrows indicate the two 6-bp inverted repeats that could form the putative hairpin structures and serve as operator sequence. The location of the Λ/s/l restriction site is indicated in bold and the SD (Shine-Dalgamo) sequence is boxed. Promoter elements corresponding to the putative -10 and -35 are indicated by a boxed line above the sequence. In Fig. 7B, the different segments of the DNA region that were used as probes in the mobility shift assays are depicted with the corresponding binding efficiency of MBP-Sep-R. The relative position of the two putative hairpin structures is indicated. Fragment C corresponds to the complete intergenic region, fragment N extends to the Nsi\ restriction site located in the second hairpin structure, while fragments L and R each correspond to one half of fragment C. Fig. 7C illustrates the specific binding of MBP-SepR to probes L and R. [³²P]-labelled probe L (lane 1 to 5), and probe R (lane 6 to 10) were incubated with varying amounts of purified MBP-SepR protein and run on TAE non-denaturing 4.5% polyacrylamide gel. The concentrations of MBP-SepR used were: lanes 1 and 6, no protein; lanes 2 and 7, 1 μg; lanes 3 and 8, 2 μg; lanes 4 and 9, 3 μg; and lanes 5 and 10, 4 μg. The positions of the free probes and bound protein-DNA complexes are indicated on the left.

When the sepABC-lacZ transcriptional fusion (pPsep-/acZ-Sm^r) was mobilized in the PpF1 strain derivative PpF1(sepP::Km^r), measurement of β-galactosidase activity showed a 1.5-fold increase, compared to the wildtype, as shown in Fig. 8, thus confirming the repressor function of SepR.

Measurement of specific sepA mRNA by dot-blot analysis indicates that sepA is not expressed constitutively in absence of toluene (Fig. 9). Moreover, sepA gene expression is increased in the PpF1(sepR::Km^r) mutant, providing further evidence that sepR functions a repressor. These results also show that upon toluene induction, SepA levels increase very rapidly (within 10 minutes), reaching a maximum level which is maintained during the next 20 minutes. In Fig. 9, 3 μg of purified RNA was used and hybridized with [³²P]-labelled probe A2 (Fig. 1), which corresponds to an internal fragment within sepA. Denatured DNA from unlabelled probe A2 was used as a positive control on the membrane. Inducibility of the sep genes

According to the mRNA study, expression of sepABC in PpF1 is under positive regulation in response to toluene. The relationship between toluene concentration and sep gene induction was determined by measuring the β-galactosidase activity of a chromosomally encoded sepASC-/acZ-Km^r fusion (pPsep-/acZ-Km^r) in PpF1APR1. Fig. 10 shows that toluene can induce expression of sepABC-lacZ-K ^r at a very low concentration (0.005% v/v). This strain was also exposed to a variety of possible inducers, general stress conditions, heavy metals, and antibiotics, in order to investigate the effect on sep gene expression. The results presented in Fig. 11 show that the antibiotics chloramphenicol and tetracycline and the heavy metals ZnCI₂ and CdCI₂ had no apparent effect on sepABC gene expression in PpF1APR1. While there was a strong response to the aromatic compounds tested, no induction by aliphatic solvents or alcohols was observed, unlike what was reported for the solvent efflux genes srpABC of Pseudomonas putida S12 (Kieboom, J., et al., J Bacteriol 180: 6769-6772, 1998). In Fig. 11 , concentrations of the inducers were the following: toluene 0.1 %, styrene 3mM, benzene 6mM, ethylbenzene 3mM, p-xylene 3mM, m-xylene 3mM, p-cymene 3mM, p- cumate 1mM, ethanol 3%, 1-butanol 3mM, 2-propanol 3mM, hexane 1mM, sodium salicylate 1 mM, NaCI 50g/l, sodium acetate 60mM, chloramphenicol 150μg/ml, tetracycline 2.5μg/ml, zinc chloride 1mM, cadmium chloride 1 mM. Inducers were added at mid-log phase and β- galactosidase activity was measured 2 hours later.

Since the tod operon and the sepABC genes are both induced by toluene and TodS is required for tod gene induction, it was further investigated as to whether TodS signaling is required for sepABC induction by toluene. Strains PpF1 L2.1 and PpF1 (-o S::Km^r)L2.1 were created by introducing the sepABC-lacZ fusion (pPsep-lacZ) into the chromosome of PpF1 and PpF1 (todS::Km^r), respectively (Fig. 1 ). When the β-galactosidase activity was measured in both strains, TodS did not appear to be required for sepABC induction by toluene (Fig. 12). The inventors wished to confirm this in a mutant derivative of PpF1 lacking todT, since TodT is the response regulator of the sensor TodS.

Preliminary tests with the sep-lacZ fusion showed that sep gene expression was induced by a variety of aromatic compounds, providing the incentive to further investigate the substrate specificity of sepR. To this end, a sep-lux biosensor was developed, since bioluminescence assays are rapid and easily quantifiable. PpF1G4 is a chromosomally-based whole-cell biosensor that was created by placing a \uxCDABE cassette under the control of the promoter element for the sep genes.

The specific bioluminescent response of biosensor strain PpF1G4 exposed to a variety of environmentally significant organic compounds was assessed, in order to determine which of these analytes were inducers of the sep genes. Results of this screening process are shown in Fig. 13, where test compounds are listed in order of their hydrophobicity, as expressed by their log P values, where P is the octanol- water partition coefficient. The log P values for the compounds were either obtained from Howard and Meylan (Howard, P.H. and Meylan, W.M. (1997) Handbook of physical properties of organic chemicals. CRC Press, Inc. USA) or calculated from hydrophobic fragmental constants with the software program KowWin (Meylan, W.M. and Howard, P.H. J Pharm Sci 84: 83-92, 1995). The results indicate that the regulator protein SepR has a relatively broad effector specificity since it is activated by a wide range of aromatic compounds. The trend indicates a quasi-normal distribution, with peak bioluminescent responses occurring for compounds with log P values in the range of about 1.5 to 3.7. With some exceptions, very hydrophobic (log P > 4.0) and very hydrophilic (log P < 1.5) compounds are not recognized as effectors by SepR. Since BTEX compounds are ubiquitous environmental pollutants, the biosensor's response to these compounds was investigated further. It was determined that the intensity of the bioluminescent response of PpF1G4 to BTEX compounds is concentration-dependent, as evidenced by Fig. 14.

A number of ubiquitous multicomponent non-aqueous phase liquids (NAPLs), including gasoline, JP-4 jet fuel, diesel, coal tar creosote, and three varieties of crude oil (Brent, Isthmus Maya, and Menemota), were tested as possible inducers of the sep and tod genes. For these tests, bioluminescent assays were performed with both PpF1G4 and TVA8, a PpF1 derivative containing a modified mini-Tn5 chromosomal insertion of a tod-lux fusion (Table 1 ). All the NAPLs tested produced a bioluminescent response in both biosensors (Fig. 15). Although BTEX compounds are known inducers of both the sep and tod genes, the magnitude of the light response to a given NAPL cannot be correlated to its BTEX content. Since these multicomponent NAPLs are complex mixtures, the biosensors' light response may have been enhanced or suppressed by the presence of other unidentified components present within the NAPLs.

It has been recently reported that, in the reporter strain P. fluorescens HK44, which contains a nahG-luxCDABE fusion, non-inducing organic solvents produced a significant bioluminescent response (Heitzer, A., et al., J Microbiol Methods 33: 45-57, 1998). In that study, the analysis of mRNA levels confirmed that certain solvents did not induce lux gene expression, even though they triggered a bioluminescent response. This phenomenon has been termed the "solvent effect". It was determined that exposure of HK44 to organic solvents resulted in membrane perturbation, causing increased fatty acid synthesis, which in turn elevated the aldehyde supply for the bioluminescence reaction. Therefore, basal levels of luciferase present in the cells due to leaky expression were oxidized along with the increased aldehyde supply to produce light. It was confirmed that the cells were aldehyde-limited, since the addition of n-decanal to either non-induced or induced test cultures significantly increased bioluminescence.

In order to verify that the light response of PpF1G4 to test compounds was due to induction of the sep genes, and not to a "solvent effect", the effect of adding π-decanal to biosensor cells exposed to 3 mM of toluene, o-xylene, TCE, and limonene was investigated. Three different concentrations of n-decanal (2 mM, 1 mM and 0.5 mM) were tested by adding the appropriate volume of a 1% (v/v) aqueous solution of n-decanal to test solutions prior to the light reading. A concentration of ~ 2 mM n- decanal has been used for similar bioluminescence assays (Sticher, P., et al., Appl Environ Microbiol 63: 4053-4060, 1997; and Heitzer, A., et al., J Microbiol Methods 33: 45-57, 1998). While at the two lower concentrations of n-decanal (0.5 mM and 1 mM), there was no significant increase in light production, at a concentration of 2 mM, there was actually an inhibition of light production in all cases. This result is not surprising, since it has been reported that n-decanal can be cytotoxic at concentrations above a few hundred micromolar (Blouin K., et al., Appl Environ Microbiol 62: 2013- 2021 , 1996). Furthermore, evidence that the bioluminescent response of PpF1G4 is actually due to induction of the sep genes is provided by the fact that a /acZ-based fusion in PpF1 (pPsep-/acZ-Km^r) was induced by BTEX compounds and styrene (Fig. 11 ). Moreover, it was shown that sepA mRNA levels increase upon exposure to toluene (Fig. 9).

Bacterial regulation of biodegradation pathways is a complex phenomenon that is affected by a variety of environmental signals, especially nutrient availability. Catabolic repression, also termed "post- exponential induction" or "exponential silencing" is an example of a situation where the expression of catabolic genes is influenced by the physiological and metabolic state of the cells. In this phenomenon, while bacteria grow rapidly on a nutrient-rich media, there is a lack of transcriptional activity, even in the presence of an inducing compound. However, once the growth rate of the bacteria subsides as they enter stationary phase, the promoter begins to respond to the effector (Cases, I. and de Lorenzo, V. EMBO J 20: 1-11 , 2001 ). Fig. 16 shows how the type of growth medium affects the bioluminescent response of biosensors PpF1G4 and TVA8 exposed to 400 mg/L (5mM) toluene, a concentration at which the bioluminescence response was found to be usually strong; i.e., > 5000 specific relative light units (RLU). In Fig. 16, error bars represent standard deviations for triplicate determinations of bioluminescence. The trend indicates that the richer the nutrient media, the weaker the bioluminescent response. TVA8 cultures grown in either Terrific or Luria-Bertani broth exhibit catabolic repression and do not produce a bioluminescent response in the presence of the inducer toluene. This phenomenon does not occur to the same extent with PpF1G4, by virtue of the fact that its sensing mechanism is based on a non-catabolic promoter. The fact that PpF1 G4 can sense inducing organic compounds and produce a bioluminescent response in both rich and minimal media illustrates the advantage of employing a biosensor that is based on solvent efflux activity instead of catabolic activity. Nucleotide sequence accession number

The DNA sequence of the fragment has been deposited in the DDBJ, GenBank and EMBL DNA databases under accession number U72354. Discussion

By sequence homology to known solvent efflux pump systems (eg. Srp proteins of P. putida S12, Ttg proteins of P. putida DOT-T1 , and MexAB-OprM of P. aeruginosa PAO1) the SepA, SepB and SepC proteins are predicted to function as periplamic efflux protein (PEP), inner membrane efflux protein (IEP) and outer membrane efflux protein (OEP), respectively. SepA is predicted to provide the connection between the function of SepC, a lipoprotein in the outer membrane and SepB, a large "xenobiotic-exporting" component in the inner membrane.

The protective role of the solvent efflux system of PpF1 in a solvent stressed environment has now been confirmed. This finding is consistent with that reported for the Ttg system in Pseudomonas putida DOT-T1E (Mosqueda, G. and Ramos, J.L. J. Bacteriol. 182: 937-943, 2000). After exposure to toluene shock, the survival of PpF1 cells lacking the sepB or sepC gene, when pregrown in toluene, was greatly impaired. Assuming that the sepABC genes are maximally expressed under induced conditions, it is normal that the lack of these genes would be detrimental to the cells. When strains were not pregrown in toluene, only a de-repressed sepR mutant of PpF1 showed an increased level of survival to toluene shock compared to the wildtype. The presence of a threshold level of the SepABC proteins in the cells at the moment toluene is added, seems to be necessary for the cells to survive a toluene shock. Disruption of sepR in PpF1 rendered the cells unable to grow faster than the wildtype in toluene as sole carbon source, the only difference being that the mutant PpF1 (sepP::Km^r) exhibited a shorter lag period. Recent work showed that PpF1 could tolerate "solvent shock" (up to 10% vol/wt toluene) in soil rather well (Huertas, M.-J., et al., Appl Environ Microbiol 64: 38-42, 1998). The presence of a sep gene cluster described in the present invention most likely provides the molecular basis for this strain characteristic. It has been noted that the ability of a microorganism to degrade a given solvent does not necessarily mean that it is solvent-tolerant. The converse is also the case.

The multidrug efflux system of P. aeruginosa was recently shown to confer tolerance to organic solvents as well as antibiotics (Li, X.Z., et al., J Bacteriol 180: 2987-2991, 1998). Preliminary experiments in accordance with the present invention indicated that the sepABC system has no effect on the level of the tested antibiotic resistance. This is consistent with the study of inducibility of the P. putida S12 srp efflux system by solvents but not by the presence of antibiotics (Kieboom, J., et al., J Bacteriol 180: 6769-6772, 1998). The increased resistance of PpF1 toward antibiotics when bacterial cells had been adapted to toluene was similarly observed in Pseudomonas putida S12 (Isken, S., et al., Appl Microbiol Biotechnol 48: 642-647, 1997). This increase was observed for wildtype PpF1 , as well as for null mutants of sepB and sepC, implying that the sep solvent efflux system is not directly responsible for the enhanced antibiotic resistance.

The sepR gene of PpF1 is rather unique. A second efflux system has been described in Pseudomonas putida S12. The ttgDEF genes show a high degree of homology with the sepABC genes. No evidence for the presence of a transcriptional repressor for the ttg genes has been presented. In the srp system of P. putida S12, the sequences of two regulators, srpS and srpR have been determined (GenBank AF061937), indicating that SrpS is 60% identical to sepR.

SepR is predicted to function as a repressor. It has been shown that in vitro, SepR can bind specifically to the sepR-sepABC intergenic region proximal to sepA. Elevated mRNA levels observed in PpF1 (sepR::Km^r) indicate that sepR functions as a repressor of sepABC. Using a lacZ transcriptional fusion, the role of SepR is now confirmed in vivo as a repressor. Overexpression of SepR in E. coli completely abolished the expression of the sepABC-lacZ-Sm^r fusion. It was also observed that in a de-repressed sepR mutant of PpF1 , expression of the same lacZ fusion is increased, although only by 1.5-fold. This low level of increase may be due to the fact that the sepR repressor gene is present in monocopy on the chromosome, while the sepABC-/acZ-Sm^r fusion is expressed from a low-copy pHRP309-derived plasmid, resulting in some titration of the SepR repressor.

PpF1G4 is a chromosomally-based whole-cell biosensor for aromatic compounds whose bioluminescent response is based on solvent efflux activity instead of catabolic activity. PpF1G4 represents a second- generation biosensor that is not based on a catabolic promoter but is nonetheless inducible by aromatic pollutants.

PpF1 is capable of degrading toluene and 4-isopropyl-toluene (p-cymene) via two independently regulated pathways (tod regulated by the two-component todST system and cym/cmt regulated by a repressor CymR). It is interesting to note that p-cymene (p-isopropyltoluene) and its benzoic acid derivative (p-cumate), are not inducers of the sep genes, while the structurally related p-isopropylbenzene does indeed act as an effector.

It is noteworthy that of all the compounds that act as inducers for PpF1G4, TCE is the only effector that does not contain an aromatic ring in its molecular structure. However, it has been previously reported that TCE induces the tod operon in TVA8 and its parent strain PpF1 (Shingleton, J.T., et al., Appl Environ Microbiol 64: 5049-5052, 1998). In addition, TVA8 produces a bioluminescent response when exposed to phenol, JP-4 jet fuel, benzene, toluene, ethylbenzene, m-xylene, and p-xylene, but not to o- xylene (Applegate, B.M., et aql., Appl Environ Microbiol 64: 2730-2735, 1998). The inventors have now confirmed that TVA8 does not respond to o-xylene, while PpF1 responds to all three isomers of xylene. Thus, the specificity of the regulatory protein SepR encompasses and extends beyond that of the regulatory proteins for the tod operon, TodS and TodT.

While both the tod pathway and the sep genes are induced by toluene and many of the organic solvents tested, the results presented herein indicated that sepABC inducibility could be independent of tod, since the β-galactosidase activity of a sepABC-lacZ fusion is the same in PpF1 L2.1 and PpF1 (foc/S::Km^r)L2.1. Results using a sepR-lacZ seem to indicate that sepR expression is not affected by the addition of BTEX compounds.

The existence of a sep (solvent efflux pump) gene cluster, most likely an operon, immediately downstream of the tod pathway in PpF1 , and identification of a new transcriptional regulator, SepR, provide new challenges and opportunities to the study of metabolism of environmental pollutants by a single microorganism. The development of the broad- specificity bioluminescent biosensor PpF1G4 represents a means of sensing aromatic compounds that is not based on a catabolic promoter. This feature permits sensing in both rich and minimal media, making it a more versatile tool.

While the invention has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth, and as follows in the scope of the appended claims.

Claims

WHAT IS CLAIMED IS:

1. A biosensor for detecting at least one target effectors selected from the group consisting of aromatic compounds, a hydrocarbon mixture containing trichloroethylene and a hydrocarbon mixture containing limonene, said biosensor comprising a first nucleic acid molecule including an operator/promoter sequence of sepR and a sequence encoding sepR, or a functional fragment thereof, and a second nucleic acid molecule including a sequence encoding a reporter molecule having a detectable activity, said second nucleic acid molecule being under the control of said first nucleic acid molecule acting as an inducible regulator whereby sepR, in absence of the target effector binding thereto, binds to its operator/promoter sequence preventing transcription of said second nucleic acid molecule encoding said reporter molecule, and whereby in presence of the target effector binding to sepR, sepR detach from its operator/promoter sequence allowing transcription of said second nucleic acid molecule encoding the reporter molecule.

2. The biosensor of claim 1 , wherein the detectable activity of the reporter molecule is generation of light.

3. The biosensor of claim 1 , wherein the operator/promoter sequence of sepR is sepABC.

4. The biosensor of claim 1 , wherein the second nucleic acid molecule is the lux gene or a functional fragment thereof.

5. The biosensor of claim 4, wherein the lux gene or functional fragment thereof is obtained from a lux operon of a bioluminescent microorganism.

6. The biosensor of claim 5, wherein the microorganisms belong to the genera Vibrio, Xenorhabdus, Photorhabdus or Photobacterium.

7. The biosensor of claim 6, wherein the microorganism is Photorhabdus luminescens.

8. The biosensor of claim 1 , which is inserted into a host cell.

9. The biosensor of claim 8, wherein the biosensor is incorporated in a plasmid to be inserted in the host cell.

10. The biosensor of claim 8, which is integrated in the chromosome of said host cell.

11. The biosensor of claim 8, wherein said host cell is a bacterial cell, a yeast cell, a fungal cell, a plant cell or an animal cell.

12. The biosensor of claim 11 , therein the host cell is a bacterial cell.

13. The biosensor of claim 1 , wherein the biosensor is capable of detecting all three isomers of xylene.

14. The biosensor of claim 1 , wherein the aromatic compound is selected from the group consisting of benzene, toluene, ethylbenzene, and xylenes (BTEX), naphthalene, and complex mixtures of hydrocarbons.

15. The whole-cell construct PpF1 G4.

16. A cell containing the biosensor as defined in any one of claims 1 to 15.