GB2364307A - Bioluminescent reporters - Google Patents

Abstract

A whole-cell biosensor harbours DNA encoding a selectively inducible light-emitting reporter, particularly the <I>P. luminescens lux CDABE</I> gene, such DNA comprising a selectively inducible promoter operatively linked to the reporter gene which is both constitutively and inducibly expressible. The promoter may particularly be the <I>katG</I> promoter. Such a biosensor provides a low-level background luminescence to indicate that levels of toxicants are not lethal.

Description

2364307 Document #: 507795 BIOLUNINESCENT REPORTERS The present invention

relates to whole-cell biosensors harbouring DNA encoding selectively inducible light-emitting reporters.

Many detection methods are used in the field of analytical chemistry and biochemistry, and the use of enzymes and other biological markers, or reporters, is extremely common. However, such in vitro systems have limitations, such as when they are exposed directly to unpurified sample, especially for long periods. Accordingly, more robust systems are required in such circumstances.

In this respect, whole-cell biosensors have several advantages. The range of reporter genes now available makes signal detection easier in response to a given inducing agent. Cells, once suitably engineered, are cheap, reliable and easy to -grow, particularly where E. coli is used as the host. In addition, not only do such whole-cell biosensors report the quantity of the inducing agent present, but also the bioavailability of the agent.

A further advantage of in vivo biosensors is that they both maintain the reporting system and protect it from harmful aspects of the surrounding environment. Thus, the sensing proteins can remain functional for a longer period of time in a harsh environment and are accumulated intracellularly, if suitable conditions are chosen. In comparison with in vitro approaches, whole cells can be exposed directly to the sample and require no complex membrane systems to separate the sensing element from the target chemicals (or other damaging agents) in the sample of interest and, so, lend themselves to easier and less expensive operation.

Initially, work in this area has concentrated on using lux genes from the marine bacterium Vibriofisheri. Cloning and manipulation of the lux cassette has further facilitated its application as a reporter, and has often been used to detect toxic insults. Where cells I 2 showing sufficient expression of the lux product are exposed to toxic insult, the resulting impairment of cellular physiology and/or cell death is reflected in a decease in luminescence.

V. fisheri only grows at temperatures up to about 30'C, so that there are commercial problems when incorporating the lux genes in E. coli, for example, as it is normal practice to grow cultures of E. coli at optimal temperature, i.e. around 37C.

Recently there has been interest in Green Fluorescent Protein [Cubitt, A. B., et aL, (1995), TIBS, 20, 448-455] and its variants, for example GFPuv, a form demonstrating enhanced fluorescence and optimised for prokaryotic expression [Crameri, A., et al., (1996), Nature Biotechnology, 14, 315319]. Also the lux CDABE operon encoding the luciferase of the bacterium P. luminescens, has recently been cloned and characterised [Winson, M. K., et al., (1998), FEMS Microbiology Letters, 163, 193-202].

Where suitable, tightly regulated promoter sequences are known, these can be utilised to construct a reporter system that is inducible in the presence of a selected subset of toxicants. E. coli harbouring V. fisheri lux genes and the nah promoter has been shown to be sensitive to naphthalene, for example. Judicious selection of appropriately sensitive regions of DNA confers selectivity on such recombinant reporter systems and is able to provide a means of discriminating the molecular nature of the damage caused by the presence of the toxicant.

A particular problem with providing a commercial product based on such reporters, even when associated with selective promoters, is that lack of a signal from the reporter is not necessarily associated with only an absence of toxicant - it may also be associated with lethal levels of toxicant, such that the organism is killed before a signal from the reporter can be detected. An extra test, therefore, is needed, in order to establish which condition applies.

It has now, surprisingly, been found that P. luminescens lux CDABE genes, in particular, can be cloned into a suitable organism such that they are both constitutively and inducibly expressed.

Accordingly, in a first aspect, there is provided a whole-cell biosensor harbouring DNA encoding a selectively inducible light-emitting reporter, characterised in that the DNA comprises a selectively inducible promoter operatively linked to the reporter gene and wherein the reporter gene is both constitutively and inducibly expressible.

As used herein, the term "biosensor" relates to any organism which can be used in an assay as an indicator. The organism may not necessarily be useful in all conditions, and it will be apparent to one skilled in the art which organisms may be used for which assays, E. coli is often the organism of choice, as so much is known about it, but other organisms, such as B. subtilis, or Pseudomonads may be used, and even simple eukaryotes, especially the yeasts, may also be used.

In accordance with the present invention, the biosensor organism will contain DNA encoding at least one light-emitting reporter. The P. luminescens lux CDABE gene has been found to be particularly suitable for use in the present invention, and its product is luminescent, but it will be appreciated that other light-emitting reporters may be used, which may be fluorescent, for example.

Where the P. luminescens lux CDABE gene is specified, for example, it will be understood that any variant, mutant or engineered derivative thereof may also be used, provided that the reporter product fulfils the requirement of being light-emitting. This applies to any genes specified herein, as well as to other sequences, such as promoters.

The reporter is constitutively expressible such that the live organism expresses the reporter, even when the inducer is not present. This provides a background level of radiation. The reporter is also inducibly expressible so that, on exposure to an appropriate level of inducer, the reporter is expressed at measurably higher levels than the background. Thus, the promoter does not actually block expression of the reporter gene, but actively promotes expression when induced.

Thus, biosensors are provided which can give a continuing indication that the sample on test does not contain lethal levels of toxicant, so that no extra test is necessary, provided that sub-lethal levels of light emission are maintained.

It will be appreciated that the promoter used may be selected for any particular analyte in a sample. In one particular use, it may be selected for a specific toxicant in field surveys, for example. One promoter that has been found to be useful is the katG promoter, which is responsive to oxidative stress.

The katG promoter has previously been fused with the V fischeri lux cassette and was used in comparative studies [Belkin, S., et al., (1996), Applied and Environmental Microbiology, 62, 2252-2256]. The E. coli katG gene product is a catalase and its r6le in the biochemistry of the oxidative stress response has been well-defined. Induction of the katG promoter is tightly regulated by the oxidative stress sensing protein, OxyR. Thus, when the katG promoter region is fused to a reporter gene, the resulting biosensor is able to sense oxidative stress.

It will be appreciated that more than one biosensor may be used in an assay, or that a biosensor may comprise more than one reporter. However, the more differingly selectable reporters there are present, the more difficult it is to establish which reporter is reacting with which analyte, or toxicant.

The green fluorescent protein (GFPuv) fluoresces under UV and has also been found to be useful in the present invention, in addition to the lux gene, for example. This particular reporter is not generally constitutively expressed, but is expressed at higher levels of toxicant than can be tolerated by the lux-based system. In fact, the pkatG GFPuv E. , coli biosensor remains responsive, up to 5.2 mM, the upper limit for E. coli survival. In addition, as the lux product is luminescent and GFP is fluorescent, then this can provide a wider range of indication of levels of toxicant, and this is particularly useful in the present invention, and forms a preferred embodiment.

Apparatus exists and is readily commercially available which is able to detect both luminescence and fluorescence, simultaneously, and is advatageously used in connection with the above embodiment.

The present invention provides assays using bionsensors of the invention, the use of such biosensors, and assay kits containg; such biosensors.

EXAMPLES

Both the GFPuv and P. luminescens ha CDABE genes were fused to the katG promoter and tested. GFPuv encodes green fluorescent protein derived from Aequorea victoria. The luxCDABE gene encodes bacterial luciferase from Photorhabdus luminescens. Both reporter systems produced stable signals. The lux construct was more sensitive at lower concentrations of hydrogen peroxide and the response time to inducing agents was shorter when compared with GFPuv. The latter, however, was better able to sense oxidative stress at concentrations that impaired signal output in the E. coli lux system.

The E. coli lux system was additionally responsive to redox cycling agents and organic peroxides, again producing a stable signal output. Low level non-induced bioluminescence was observed using the P. luminescens reporter system and this was utilised to measure EC5().

Plasmids and E. coli strains.

The lux cassette of pSB417 (18) was excised and ligated into EcoRl (Promega) digested pUC 19 to give the vector, plux 19.

The katG promoter region was obtained via PCR amplification (Stratagene Optiprime Kit) from a K12 chromosomal DNA preparation (7) yielding a 688 bp fragment. PCR primer sequences (5'-GTCACCCGGGGTTCAGATC and 5'GAATATTCCCCGGGATATGGTG) (Sigma-Genosys) were based on GenBank (http://www. ncbi.nlm.nih.gov/web/search) sequence data and the published sequence analysis of E. coli katG promoter region (10). Both primers include sequences that contain SmaI restriction sites. PCR products were cloned 6 using the pGem T-vector cloning kit (Promega) and constructs used to transform DH5cc competent cells (Sigma-Aldrich). Cells were grown on LBampicillin (100 gg/ml) and clones were identified by insertional inactivation of lacZ in presence of Xgal (Gibco BRL) and confirmed by sequencing (Applied Biosystems 373A DNA Sequencer). A suitable clone, with a verified sequence, was digested using SmaI and the excised 688 bp fragment ligated into Smal -digested plux 19, to give the promoterreporter fusion, pkatGlux.

The reporter gene construct, pkatGuv incorporates a 300 bp PCR fragment containing the katG promoter. PCR primers (5'-AGCACAGCATGCTGCCTCGAA and 5'-GATATCGTCTGCAGGCGCTCAT) (Gibco BRL) were based on GenBank (http://www. ncbi.nlm.nih.gov/web/search) sequence data and the published sequence of E. coli katG promoter region (10). The sense and antisense oligonucleotides contained SphI and Pstl restriction sites respectively. PCR products were blunted with Klenow fragment and cloned into Smal digested pUC19. Insertion was confirmed by digestion with Sphl and Smal and this fragment ligated into the mcs of SphI- and PstI-digested pGFPuv to give pkatGuv.

The control plasmid pLTV was constructed by amplifying the pGFPuv plasmid using PCR primer sequences (GIBCO BRL) based on sequence data (Clontech) (sense; 5'ACAGCTATGACCATGATTACGCC, antisense; 5'-GGAGAGGAAGCTTGCGTATTGG).

The pGFPuv product contained a novel Hind III site (shown in bold on the primer) upstream of the lac promoter that is fused to the GFPuv gene. As shown in Figure 1, after PCR, pGFPuv possesses a Hind III site either side of the promoter. Hence digestion of the PCR product with Hind III and religation produced the promoterless construct pLTV. Removal of 216 bp containing the promoter region was confirmed by sequencing. The plasmids, pkatGlux, pUV and pkatGuv were used to transform into Esherishia coli K12 (NCTC W1 10611).

Insert veriflcation The plasmids pkatGlux and pkatGuv were sequenced (Applied Biosystems 373A DNA Sequencer) using sequence primers designed to show the orientation of the insert with respect to the reporter gene and to sequence the whole of the promoter region. Sequencing 7 showed correct orientation of insert in both vectors with respect to the reporter genes with no base mismatches. Nine stop sequences in all three reading frames preceded the promoterreporter fusions.

Experimental methods All E. coli strains were grown in Luria-Bertani medium (LB) at 30 OC. To assay for luminescence cells were grown in the presence of ampicillin (Sigma) at 100 gg/ml to mid exponential phase (OD 600 nin = 0.5 - 0.6) (Pharmacia Biotech Ultrospec 2000 UVNisible Spectrophotometer). Cells were then diluted 50 times in 25 ml LB (with ampicillin) in baffeled flasks and shaken at 30 OC for 90 mins at 200 rpm followed by toxic challenge. 50 gl samples (n=2) were taken from each flask and luminescence measured using a Lumac Biocounter (Celcis-Lumac BV) for 10 S and data presented as relative light units (RLU).

Fluorescence assay was performed on 3 ml samples (n=2) measured using a Jenway fluorescence spectrophotometer (excitation = 420 run, bandwidth = 20 rim, emission = 5 10 nm) and calibrated with I d of rGFPuv (Clonetech) at 1 gg/ml to give the upper limit of fluorescence quantification to derive the measured Relative Fluorescence Units (RFU).

Chemicals All chemicals used were analytical grade and supplied as shown. Hydrogen peroxide (BDH). Ethanol and media components, Na2HP04.7H20, KH2PO4, NaCL, glucose, MgSO4, FeSO4, CaC12 and NH4CL (Fisher scientific). Tryptone and Yeast Extracts (Oxoid). The remaining compounds, cumene hydrogen peroxide, tert- butyl hydrogen peroxide, peroxybenzoate, menadione and paraquat (Sigma- Aldrich).

Results Lux reporter system Responsiveness of E. coli (pkatGlux) to different hydrogen peroxide concentrations is shown in Figure 2. The response ratio was calculated by dividing the luminescence at each time point by that at time zero for each concentration of toxicant (1). A sharp increase in luminescence was observed at 20 min, which continued to increase, at a slower rate during the 8 duration of the experiment. The response was dose dependent up to 1.7 mM hydrogen peroxide.

A similar result could not be achieved, however, with the pkatGuv construct in E. coli K12 using LB as the medium. Due to background fluorescence from media components, no quantifiable measurement could be undertaken, so a qualitative indication of GFPuv expression was achieved using a fluorescent microscope. Furthermore, in comparing induction profiles of samples in the presence of hydrogen peroxide with those that had none, it was noted that only those with a particularly high concentration of hydrogen peroxide (1.76 to 5.29 mM) showed an increase in fluorescence over time (data not shown).

To eliminate effects from fluorescent media components, pkatGuv fluorescent measurement was carried out in M9 minimal media, and the response is shown in Figure 3a. The optical density versus time is shown in Figure 3b. Although an increase in fluorescence was observed over time, concurrent changes in the optical density (OD 600 run) of the cell suspension indicated that the biomass was actually decreasing in cultures where the inducing agent was present.

Only cells unexposed to hydrogen peroxide demonstrated both a stable OD600 and fluorescence reading that increased with time. Only when signal output and biomass is corrected for dose, does a dose dependent response become apparent (Figure 3c). The response is clear when the response ratio is plotted against time (Figure 3d). From this plot, at time 240 min, a calibration line can be drawn (Figure 3e).

Media development The biochemistry of the oxidative stress response suggests that, in stationary phase, then stationary phase sigma factor, RpoS, and not OxyR, is involved in the regulation of katG, and expresses catalase in a constitutive fashion (13). It has been suggested that OxyR control of katG may only occur during cell growth. If the cells are not growing, as suggested by the decrease in biomass shown in Figure 3b, then RpoS may be controlling expression from the katG promoter in a peroxide non-inducible manner.

The construct pkatGlux was exposed to hydrogen peroxide in M9 media. A response was also observed in a comparable concentration range (1.76 to 3. 52 mM), though this was not as great as when LB was used (data not shown).

E. coli (pkatGlux) shows strong induction when grown in LB medium. However LB broth contains endogenous components that fluoresce. Hence to use pkatGuv required development of a medium suitable for both cell growth and fluorescence measurement. Of the constituents of LB, tryptone showed minimal background fluorescence and was used to supplement M9 media (3g/1). The fluorescence signal from E. coli (pkatGuv) grown in this medium was measured over time with increasing concentrations of hydrogen peroxide. The OD (600mn) showed that biomass was increasing exponentially (data not shown). It was noticed that as the OD increased so did the fluorescence, both in samples without hydrogen peroxide and those with. When the fluorescence is corrected by the OD, with no hydrogen peroxide present, and plotted against time, a similar plot results with both pkatGuv and pLJV, the promoterless GFPuv control, indicating no expression of GFPuv from the katG promoter. When the plot is repeated with pkatGuv and pUV with hydrogen peroxide above 3.52 MM, the RFU/OD value becomes significantly higher in pkatguv compared with the promoterless GFPuv gene (Figure 4). The percentage increase of pkatGuv compared to pLTV is shown in Table 1.

Figure 2 demonstrates that pkatGlux (0.22 - 1.76 m.M H202) has a greater apparent sensitivity than pkatGuv (1.76 - 5.29 mM H202, Figure 3d) whereas the latter is sensitive over a wider concentration range. Thus further experiments were carried out to study the response of pkatGlux to other agents. Figure 2 also shows the response of pkatGlux to a concentration range of hydrogen peroxide. To ensure cell growth was proceeding throughout the time course, the OD was measured with and without hydrogen peroxide (Figure 5). The inducing agent was added at 20 min and the concurrent luminescence readings are shown in Figure 6. Although the OD is increasing linearly during the time course, the luminescent response increases in a sigmoidal plot 20 min after the addition of hydrogen peroxide and peaks at 80 min when the rate of increase declines but the signal appears constant.

Response of pkatGlux to redox cycling agents Various pollutants produce free radicals via redox cycling and, thus, the response of such agents was investigated by using pkatGlux. Menadione (dissolved in ethanol) was added to pkatGlux at concentrations between 0 and 100 tg/ml. The response (Figure 7a) shows dose dependency and an increasing signal at all concentrations for the duration of the experiment. When the response at 20 min is plotted against concentration, a linear calibration curve results (Figure 7b).

A background level of luminescence is evident in the system, and is useful to provide an EC50, measured at time zero immediately after the addition of toxicant (Figure 7c). The EC50 value is close to the concentration of menadione that produces the peak luminescence (25 gg/ml).

The responsiveness of E. coli (pkatGlux) to another redox cycling agent, paraquat, is shown in Figure 8. The maximal response was achieved at up to I mg/ml at 60 min where an unambiguous dose dependent response can be observed. The response was less than with menadione and, so, higher concentrations of reagent were required.

Organic peroxides pkatGlux was treated with three organic peroxides, cumene hydrogen peroxide, tertbutyl hydrogen peroxide and peroxybenzoate (Figures 9a and 9b). Cumene elicited the best response, peaking at 60 min, whereas the remaining two produced a delayed response at 140 min, when the response achieved was higher than that of the background luminescence.

The more rapid responsiveness of the lux systems compared with the GFPbased system is due to the time lag in protein accumulation required for observed or quantification (5). Also, the signal resulting from GFPuv expression has no means of amplification. With lux, signal amplification can occur; the expression of one set of lux genes can result in the turnover of many molecules and production of many photons of light. In P. luminescens, for example, a dose-respondent result can be observed within 20 min (Figure 7b). This is faster than systems based on the V. fisheri lux operon and much faster than the pkatGuv where one may have to wait for 120 min (Figure 4) to obtain a measure of the amount of oxidative stress present.

Both pkatGlux and pkatGuv demonstrate enhanced signal stability over systems that use lux genes derived from V. fisheri. With the latter system signal for hydrogen peroxide peaks around 45 min and declines immediately. With GFPuv and the lux P. luminescens based biosensing systems described in this paper, experiments indicate that the signal continues to accumulate throughout the experiment. Thus, if a sample time is missed or a reading delayed the luminescence can still be measured and an estimate of toxicant concentration made.

The background obtained with P. luminescens is useful to provide an EC50 value obtained at t--O with other measurements of inducing toxicants at later time points (Figure 7c).

Fluorescence was measured indirectly by dividing the observed RFU by the OD to standardise for differences in cell density. With luminescence, the response was faster and greater and thus the changes in cell density have less impact on signal output.

The biochemistry of the katG system suggests that regulation by OxyR occurs during the exponential growth phase. Thus the response in M9 media of pkatGuv to hydrogen peroxide may be erroneous, as the cell density is not increasing (Figures 3a -3e). When the cells are growing (Figure 4), the measure of RFU per unit of OD has been shown to increase in a dose dependent manner. Correcting RFUs per unit of OD has been used previously to express data from GFPuv (2). However, the RFU/OD also increases in a promoterless pGFPuv control with increasing hydrogen peroxide concentrations, although the increase is significantly less than with the pkatGuv construct. This could be due to inner filter field effect (IFFE) (4). This occurs when cells scattering the light affect the fluorescence. Thus light may be reflected and refracted affecting the fluorescence measured.

Other compounds in the cells may have fluorophores and again affect the fluorescence. Thus in order to distinguish GFPuv fluorescence over and above that from bacterial cells may require significant expression only achievable from agents that are intensely oxidising and I 12 strongly induce the katG gene, as seen here, for high hydrogen peroxide concentrations. This may be useful as at these high concentrations luminescent output from lux reduces dramatically. Direct measurement of extracted protein from pkatGuv cells is an option, but this would not fulfil the desired qualities of a biosensor; most notably, easy and rapid signal measurement, in the minimum time with minimal technical input. GFPuv may, therefore, only be a good reporter when there are high levels of inducing agent present or a strong inducible promoter is used.

The lux gene products that catalyse bioluminescence are well suited to whole-cell biosensor applications due to ease of measurement and rapidity of signal output. However, inhibition of metabolic activity by the same toxic chemical one is trying to measure becomes significant at relatively low concentration levels. Using the GFPuv reporter system provides a useful adjunct where potentially high levels of inducing toxicant might be present; with the different signal outputs allowing concurrent or sequential measurement of the two systems.

These two systems can be used in tandem to address a wider range of toxicant concentrations than would be otherwise possible. This, together with concuffent monitoring of non-induced luminescence at t=O for EC50 measurement, forms a preferred.embodiment of the present invention.

13 References 1. Belkin, S., Smulski, D. R., Vollmer, A. C., Van Dyk, T. K. and LaRossa, R. A.

1996. Oxidative stress detection with Escherichia coli haboring a katG'::Iux fusion.

Applied and Envirorunental Microbiology 62: 2252-2256.

2. Cha, H. J., Srivastava, R., Vakharia, V. N., Rao, G. and Bentley, W. E. 1999.

Green Fluorescent Protein as a noninvasive stress probe in resting Escherichia coli cells.

Applied and Environmental Microbiology 65: 409 - 414.

3. Crameri, A., Whitehorn, E. A., Tate, E. and Stemmer, W. P. C. 1996. Improved Green Fluorescent Protein by molecular evolution using DNA shuffling. Nature Biotechnology 14: 315-319.

4. Li, K. and Humphrey, A. E. 1992. Factors affecting culture fluorescence when monitoring bioreactors. J. Fermentation and Bioengineering 74: 104-111.

5. Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E., Gross, L. A. and Tsien, R. Y. 1995. Understanding, improving and using green fluorescent proteins. TIBS 20: 448-455.

6. Methods in Biotechnology 6. Enzyme and Microbial Biosensors. Techniques and Protocols. Edited by Ashok Mulchandani and Kim R. Rogers. Humana Press, Totowa, New Jersey. 1998.

7. Pitcher, D. G., Saunders, N. A. and Owen, R. J. (1989). Rapid extraction of bacterial genomic DNA with guanidium thiocyanate. Letters in Applied Microbiol. 8:

151-156.

8. Reddy, S. M and Vadgama, P. M. 1997. Ion exchanger modified PVC membranes- selectivity studies and response amplification of oxalate and lactate enzyme electrodes.

Biosensors and Bioelectronics 12: 9-10,1003-1012.

9. Sousa S., Duffy, C., Weitz, H., Glover, A. L. and Bar, E. 1998. Use of a lux- modified bacterial biosensor to identify constraints to bioremediation of BTEX contaminated sites. Environment Toxicology and Chemistry 17: 6, 1039-1045.

10. Tartaglia, L. A., Storz, G. and Ames, B. N. 1989. Identification and Molecular analysis of oxyR-regulated promoters important for the bacterial adaptation to oxidative stress. J. Mol. Biol. 210: 709 - 719.

11. The Metablism of Drugs and Other Xenobiotics. Edited by Bernard Testa and John Caldwell. Biochemistry of Redox Reactions. Academic Press. Harcourt Brace and Company, Publishers. 1993.

12. Toledano, M. B., Kullik, I., Trinh, F., Baird, P. T., Schneider, T. D. and Storz, G. 1994. Redox-dependant shift of OxyR-DNA contacts along an extended DNA-binding site: A mechanism for differential promoter selection. Cell 78: 897 - 909.

13. Schellhom, H. E. 1994. Regulation of hydroperoxidase (catalase) expression in Escherichia coli. FEMS Microbiology Letters 31: 113 - 119.

14. Van Dyk, T. K., Majarian, W. R., Konstantinov, K. B., Young, R. M., Dhudati, P. S. and LaRossa, R. A. 1994. Rapid and sensitive pollutant detection by induction of heat shock gene-bioluminescence gene fusions. Applied and Environmental Microbiology 60: 1414 - 1420.

15. Van Dyk, T. K., Reed, T. R., Vollmer, A. C. and LaRossa, R. A. 1995. Synergistic induction of the heat shock response in Escherichia coli by simultaneous treatment with chemical inducers. Journal of Bacteriology 177: 6001 - 6004).

16. Van Dyk, T. K., Smulski, D R., Reed, T. R., Belkin, S., Vollmer A. C. and LaRossa, R. A, 1995. Responses to toxicants of an Escherichia coli strain carring a uspA '::lux genetic fusion and an E. coli strain carrying a GrpE'::Iux fusion are similar. Applied and Environmental Microbiology 61: 4124 - 4127.

17. Winson, M. K., Swift, S., Fish, L., Throup, I P., Jorgensen, F., Chhabra, R. S., Bycroft, B. W., Williams, P. and Stewart, G. S. A. B. 1998. Construction and analysis of luxCDABE-base plasmid sensors for investigating N-acyl homoserine lactone-mediated quorum sensing. FEMS Microbiology Letters 163: 185-192.

18. Winson, M, K., Swift, S., Hill, P. J., Sims, C. M., Griesmayr, G., Bycroft, B. W., Williams, P. and Stewart, G. S. A. B. 1998. Engineeringthe luxCDABE genes from Photorhabdus luminescens to provide a bioluminescent reporter for constitutive and promoter probe plasmids and mini-Tn5 constructs. FEMS Microbiology Letters 163: 193202.

16

Claims (2)

Claims

1. A whole-cell biosensor harbouring DNA encoding a selectively inducible light emitting reporter, characterised in that the DNA comprises a selectively inducible promoter operatively linked to the reporter gene and wherein the reporter gene is both constitutively and inducibly expressible.

2. A biosensor according to claim 1, wherein the light-emitting reporter is encoded by the P. luminescens lux CDABE gene.