A portable toxicity biosensor using freeze-dried recombinant bioluminescent bacteria

A portable biosensor has been developed to meet the demands of field toxicity analysis. This biosensor consists of three parts, a freeze-dried biosensing strain within a vial, a small light-proof test chamber, and an optic-fiber connected between the sample chamber and a luminometer. Various genetically engineered bioluminescent bacteria were freeze-dried to measure different types of toxicity based upon their modes of action. GC2 (lac::luxCDABE), a constitutively bioluminescent strain, was used to monitor the general toxicity of samples through a decrease in its bioluminescence, while specific toxicity was detected through the use of strains such as DPD2540 (fabA::luxCDABE), TV1061 (grpE::luxCDABE), DPD2794 (recA::luxCDABE), and DPD2511 (katG::luxCDABE). These inducible strains show an increase in bioluminescence under specific stressful conditions, i.e. membrane-, protein-, DNA-, and oxidative-stress, respectively. The toxicity of a sample could be detected by measuring the bioluminescence 30 min after addition to the freeze-dried strains. In an attempt to enhance the sensitivity of the freeze-dried cells, glucose and Tween 80 were tested as additives. It was found that the addition of glucose had a negative effect on the viability of the freeze-dried cells, while samples having Tween 80 showed an increase in their viability. On the other hand, the addition of either Tween 80 or glucose decreased the final bioluminescent response of DPD2540 in response to 4-chlorophenol. Using these strains, many different chemicals were tested and characterized. This portable biosensor, with a very simple protocol, can be used for field sample analysis and the monitoring of various water systems on-site.     

Enhancing the sensitivity of a two-stage continuous toxicity monitoring system through the manipulation of the dilution rate.

Optimization of the dilution rates has been studied to provide an enhanced sensitivity to toxicity by several recombinant bioluminescent Escherichia coli strains, TV1061 (grpE::luxCDABE), DPD2794 (recA::luxCDABE) and DPD2540 (fabA::luxCDABE), in the two-stage continuous toxicity monitoring system. It was found that the sensitivity of both TV1061 and DPD2794 to a pulse injection of phenol and mitomycin C increased with a decrease in the dilution rate. The sensitivity, however, for all the strains to step injections of the toxic chemicals was found to increase with an increase in the dilution rate up to a certain dilution rate and then decreased, mainly due to the rapid washing out of the injected chemicals. The response kinetics of the strains were explained by evaluating the mode of action of the recombinant bioluminescent bacteria to toxicity with the dilution rate, the operating parameter of minibioreactors under consideration in this study.     

Acclimation and repair of DNA damage in recombinant bioluminescent Escherichia coli

AIMS: The aim of this study is to understand different adaptive responses in bacteria caused by three different mutagens, namely, an intercalating agent, an alkylating agent and a hydroxylating agent, and the repair systems according to the type of DNA damage, that is, DNA cross-linking and delayed DNA synthesis, alkylation and hydroxylation of DNA. A recombinant bioluminescent Escherichia coli, DPD2794 with the recA promoter fused to luxCDABE originating from Vibrio fischeri, was used in this study. METHODS AND RESULTS: The recombinant bioluminescent E. coli strain DPD2794, containing a recA promoter fused to luxCDABE from V. fischeri, was used to detect adaptive and repair responses to DNA damage caused by mitomycin C (MMC), and these responses were compared with those when the cells were induced with N-methyl-N-nitro-N-nitrosoguanidine (MNNG) and hydrogen peroxide (H2O2). The response ratio between the induced samples and that of the controls decreased suddenly when the induced culture was used in further inductions, indicating a possible adaptive response to DNA damage. DNA damage, or the proteins produced, because of MMC addition does not appear to be completely resolved until the seventh sub-culture after the initial induction, whereas simple damage, such as the base modification caused by MNNG and H2O2, appears to be repaired rapidly as evidenced by the quick recovery of sensitivity. CONCLUSIONS: These results suggest that it takes more time to completely repair DNA damage caused by MMC, as compared with a simple repair such as that required for the damage caused by MNNG and H2O2. Therefore, repair of the damage caused by these three mutagens is controlled by different regulons, even though they all induced the recA promoter. SIGNIFICANCE AND IMPACT OF THE STUDY: Using a bioluminescent E. coli harbouring a recA promoter-lux fusion, it was found that different adaptive responses and repair systems for DNA damage caused by several mutagens exists in E. coli.     

Bacterial bioluminescent emission from recombinant Escherichia coli harboring a recA::luxCDABE fusion

This paper describes the quantitative evaluation of a bioluminescence assay for DNA damaging agents with respect to the linearity, sensitivity, specificity and dependence on the cell culture status. A recombinant bacterium, DPD2794, harboring a plasmid with a recA promoter fused to the luxCDABE operon, showed a very sensitive response to DNA-damaging stress. DPD2794 was found to show no noticeable response to non-mutagenic agents, i.e. phenol, except for some false responses appearing soon after injection. DPD2794 also showed a highly sensitive response to Mitomycin C, which was found to be a growth-stage-dependent response, not a growth-rate-dependent response. In addition, the relationship between the bioluminescence emitted in vivo, luciferase activity measured in vitro, and the amount of Lux proteins expressed was determined. The intensity of the bioluminescence emitted was found to be proportional to the luciferase activity in vitro, while the bioluminescence also seems to be correlated with the level of Lux proteins expressed in these Escherichia coli cells, up to 230 min post induction.     

A bioluminescent sensor for high throughput toxicity classification

A high throughput toxicity monitoring and classification biosensor system has been successfully developed using four immobilized bioluminescent Escherichia coli strains, DPD2511, DPD2540, DPD2794 and TV1061, which have plasmids bearing a fusion of a specific promoter to the luxCDABE operon. The bioluminescence of DPD2511 increases in the presence of oxidative damage, DPD2540 by membrane damage, DPD2794 by DNA damage and TV1061 by protein damage. In the developed biosensor these strains are immobilized in a single 96 well plate using an LB-agar matrix, and are able to detect the toxicities of hydrogen peroxide, phenol and mitomycin C in water samples. As the concentration of each chemical was increased, the bioluminescence levels from the corresponding wells, containing either DPD2511, DPD2540, DPD2794 or TV1061, increased. This increase in bioluminescence followed a dose dependent response to the toxic chemicals within a specific concentration range. In particular, each test requires only 4 h to give clear bioluminescent response signature. Storage of the biosensor at 4 degrees C for 2 weeks caused no change in its dose-dependent response. The fast and easy detection of oxidative, membrane, protein and DNA damaging agents in aqueous environments is possible due to the high throughput capability of this biosensor.     

Response of bioluminescent bacteria to sixteen azo dyes

Recombinant bioluminescent bacteria were used to monitor and classify the toxicity of azo dyes. Two constitutive bioluminescent bacteria,Photobacterium phosphoreum andEscherichia coli, E. coli GC2 (lac::luxCDABE), were used to detect the cellular toxicity of the azo dyes. In addition, four stress-inducible bioluminescentE. coli, DPD2794 (recA::luxCDABE), a DNA damage sensitive strain; DPD2540 (fabA::luxCDABE), a membrane damage sensitive strain; DPD2511 (katG::luxCDABE), an oxidative damage sensitive strain; and TV1061 (grpE::luxCDABE), a protein damage sensitive strain, were used to provide information about the type of toxicity caused by crystal violet, the most toxic dye of the 16 azo dyes tested. These results suggest that azo dyes result in serious cellular toxicity in bacteria, and that toxicity monitoring and classification of some azo dyes, in the field, may be possible using these recombinant bioluminescent bacteria.     

Toxicity monitoring of endocrine disrupting chemicals (EDCs) using freeze-dried recombinant bioluminescent bacteria

Five different freeze-dried recombinant bioluminescent bacteria were used for the detection of cellular stresses caused by endocrine disrupting chemicals. These strains were DPD2794 (recA::luxCDABE), which is sensitive to DNA damage, DPD2540 (fabA::luxCDABE), sensitive to cellular membrane damage, DPD2511 (katG::luxCDABE), sensitive to oxidative damage, and TV1061 (grpE::luxCDABE), sensitive to protein damage. GC2, which emits bioluminescence constitutively, was also used in this study. The toxicity of several chemicals was determined on the first four freeze-dried bacteria, while nonspecific cellular stresses were measured using GC2. Damage caused by known endocrine disrupting chemicals, such as nonyl phenol, bisphenol A, and styrene, was detected and classified according to toxicity mode, while others, such as phathalate and DDT, were not detected with the bacteria. These results suggest that endocrine disrupting chemicals are toxic in bacteria, and do not act via an estrogenic effect, and that toxicity monitoring and classification of some endocrine disrupting chemicals may be possible in the field using these freeze-dried recombinant bioluminescent bacteria.     

Gamma-radiation dose-rate effects on DNA damage and toxicity in bacterial cells

In order to investigate the relationship between radiation dose-rate and bacterial DNA damage as well as general cellular toxicity, two recombinant Escherichia coli strains, DPD2794 and GC2 were used. Following gamma-ray irradiation, these bioluminescent bacteria showed quantitative stress responses in terms of DNA damage and general toxicity depending on the dose rates of energy deposition, i.e. dose-rate of radiation. In addition, an inverse relationship was found, at lower dose rates between 0.5 and 1 Gy/h and a parabolic relationship at dose rates between 0.5 and 2.6 Gy/h.     

Evaluation of a high throughput toxicity biosensor and comparison with a Daphnia magna bioassay

A high throughput toxicity biosensor has been designed and constructed using recombinant Escherichia coli cells, containing stress specific promoters (recA, fabA, or katG) or constitutive promoters (lac) fused to luciferase genes originating from Vibrio fisheri. These genetically engineered cells were immobilized in 96 well plates. By optimizing cell immobilization conditions and the strains' response specificity to toxic chemicals, bioluminescent outputs decreased or increased dose-dependently upon adding test chemicals. However, to date the toxicity data obtained using this biosensor have not been compared with the results of other toxicity tests. Phenolics were chosen to evaluate the correlation between the LD50 and the EC50 (GC2) or EC120 (DPD2540) of Daphnia magna and E. coli, respectively. Toxicity data obtained from constitutive strains by bioluminescent level decrements were compared with the results from D. magna as a standard. LD50 values were used as parameters of D. magna toxicity and EC50 of EC120 values were used for the immobilized biosensor. In the DPD2540 test, phenolics, membrane damaging toxic chemicals, for testing immobilized stress specific bacterial strains trigger dose-dependant bioluminescence increase within specific concentration. Although the stress specific responsiveness from the strains could not be compared with D. magna's LD50 values, these responses offer additional information, such as upon the mode of toxic action in the sample, in addition to the cellular toxicity results as indicated by the EC50. This novel high throughput toxicity biosensor can be implemented to investigate the toxicity of any other soluble materials, and can be used as a standardization tool for the evaluation of toxicity.     

Action of 1,1-dimethylhydrazine on bacterial cells is determined by hydrogen peroxide

Seven different recombinant bioluminescent strains of Escherichia coli containing, respectively, the promoters katG and soxS (responsive to oxidative damage), recA (DNA damage), fabA (membrane damage), grpE, and rpoE (protein damage) and lac (constitutive expression) fused to the bacterial operon from Photorhabdus luminescens, were used to describe the mechanism of toxicity of 1,1-dimethylhydrazine (1,1-DMH) on bacteria, as well as to determine whether bacteria can sensitively detect the presence of this compound. A clear response to 1,1-DMH was observed only in E. coli carrying the katG'::lux, soxS'::lux, and recA'::lux-containing constructs. Preliminary treatment with catalase of the medium containing 1,1-DMH completely diminished the stress-response of the P(katG), P(recA), and P(soxS) promoters. In the strain E. coli (pXen7), which contains a constitutive promoter, the level of cellular toxicity caused by the addition of 1,1-DMH was dramatically reduced in the presence of catalase. It is suggested that the action of 1,1-DMH on bacterial cells is determined by hydrogen peroxide, which is formed in response to reduction of the air oxygen level.     

Distinct responses of a recA::luxCDABE Escherichia coli strain to direct and indirect DNA damaging agents.

The recombinant Escherichia coli strain DPD2794 containing a recA::luxCDABE fusion is used to detect genotoxicity of various chemicals. Genotoxic agents were previously categorized into two groups, Direct DNA Damaging (DDD) agents and Indirect DNA Damaging (IDD) agents; these two groups have been distinguished with this strain. Minimum detectable concentrations of the DDD agents were about one to five orders of magnitude lower than those of the IDD agents. The response patterns of this strain to DDD agents differed from those to IDD agents in terms of kinetics and the forms of the dose-dependent response.     

Enhancement of the multi-channel continuous monitoring system through the use of Xenorhabdus luminescens lux fusions

The enhancement of the multi-channel continuous toxicity monitoring system developed previously was studied. To achieve better and more stable results from the system, the use of thermo-lux fusion strains that express the luxCDABE genes from Xenorhabdus luminescens was evaluated. A total of six recombinant Escherichia coli strains with the promoters from three oxidative-stress responsive genes, i.e. the katG, sodA and pqi-5 genes, fused to either the lux genes from Vibrio fischeri or X. luminescens were characterized and their responses to different chemicals compared. It was found that the basal level bioluminescence (BL) from the thermo-lux fusion strains was always higher while that of the V. fischeri lux strains were always near or below the lower limit of detection of the system. For example, the katG::V. fischeri lux strain, DPD2511, gave no discernible response due to its low level expression while a fusion of the katG promoter with the X. luminescens lux operon was clearly responsive and capable of detecting hydrogen peroxide down to about 1 ppm. The use of the thermo-lux strains found them to be as sensitive as the V. fischeri lux strains while providing a brighter, more stable basal level bioluminescence, making the analysis and monitoring of water-borne toxicity more reliable.     

Modeling and measurement of a whole-cell bioluminescent biosensor based on a single photon avalanche diode

Whole-cell biosensors are potential candidates for on-line and in situ environmental monitoring. In this work we present a new design of a whole-cell bioluminescence biosensor for water toxicity detection, based on genetically engineered Escherichia coli bacteria, carrying a recA::luxCDABE promoter-reporter fusion. Sensitive optical detection is achieved using a single photon avalanche photodiode (SPAD) working in the Geiger mode. The present work describes a simple mathematical model for the kinetic process of the bioluminescence based SOS toxin response of E. coli bacteria. We find that initially the bioluminescence signal depends on the time square and we show that the spectral intensity of the bioluminescence signal is inverse proportional to the frequency. We get excellent agreement between the theoretical model and the measured light signal. Furthermore, we present experimental results of the bioluminescent signal measurement using a SPAD and a photomultiplier, and demonstrate improvement of the measurement by applying a matched digital filter. Low intensity bioluminescence signals were measured after the whole-cell sensors were exposed to various toxicant concentrations (5, 15 and 20ppm).     

An investigation of the reaction kinetics of luciferase and the effect of ionizing radiation on the reaction rate

The bioluminescence produced by luciferase, a firefly enzyme, requires three substrates: luciferin, ATP and oxygen. We find that ionizing radiation, in the form of a proton beam from a cyclotron, will eliminate dissolved oxygen prior to any damage to other substrates or to the protein. The dose constant for removal of oxygen is 70 ± 20 Gy, a much smaller dose than required to cause damage to protein. Removal of oxygen, which is initially in excess, leads to a sigmoidal response of bioluminescence to radiation dose, consistent with a Michaelis–Menten relationship to substrate concentration. When excess oxygen is exhausted, the response becomes exponential. Following the irradiation, bioluminescence recovers due to a slow leak of oxygen into the solution. This may also explain previous observations on the response of bioluminescent bacteria to radiation. We have studied the dependence of the reaction rate on enzyme and substrate concentration and propose a model for the reaction pathway consistent with this data. The light output from unirradiated samples decreases significantly with time due to product inhibition. We observe that this inhibition rate changes dramatically immediately after a sample is exposed to the beam. This sudden change of the inhibition rate is unexplained but shows that enzyme regulatory function responds to ionizing radiation at a dose level less than 0.6 Gy.     

Analysis of the stress effects of endocrine disrupting chemicals (EDcs) on Escherichia coli

In this study, three of the representative EDCs, 17beta-estradiol, bisphenol A, and styrene, were employed to find their mode of toxic actions in E. coli. To accomplish this, four different stress response genes, recA, katG, fabA, and grpE genes, were used as a representative for DNA, oxidative, membrane, or protein damage, respectively. The expression levels of these four genes were quantified using a real-time RT-PCR after challenge with three different EDCs individually. Bisphenol A and styrene caused high-level expression of recA and katG genes, respectively, whereas 17beta-estradiol made no significant changes in expression of any of those genes. These results lead to the classification of the mode of toxic actions of EDCs on E. coli.     

Monitoring and classification of PAH toxicity using an immobilized bioluminescent bacteria

An immobilized recombinant bioluminescent Escherichia coli strain, harboring a lac::luxCDABE fused plasmid, which shows lower bioluminescence levels when cellular metabolism is inhibited, was used to monitor the cellular toxicity of polycyclic aromatic hydrocarbons (PAHs). PAHs, classified as pericondensed (PCPAHs) or catacondensed (CCPAHs) according to their molecular structures, were differentiable according to the response of this biosensor. Only CCPAHs were found to cause cellular toxicity, resulting in a dose-dependent decrease in the bioluminescent output. The induction of cellular toxicity by CCPAHs and PCPAHs was compared with acute toxicity predictions obtained using the quantitative structure-activity relationship (QSAR) model. A good relationship was obtained between the toxicities determined with the bioluminescent response of the immobilized bacterium GC2 and the QSAR model. It was also found that the present study offers a new method of predicting the cellular toxicities of CCPAHs or PCPAHs using this biosensor.     

Hypochlorous acid activates the heat shock and soxRS systems of Escherichia coli.

A series of plasmids, containing fusions of different stress promoters to lux reporter genes, was used in an attempt to monitor the defense circuits activated upon exposure of Escherichia coli to sublethal doses of free chlorine. A significant level of activation was exhibited by promoters of three heat shock genes (grpE, dnaK, and lon), in an rpoH-dependent manner. The promoter of micF, a gene under the control of the soxRS regulon, was also strongly induced, but not in a soxR mutant. This induction was not affected by sodA and sodB mutations, implying that it did not involve oxygen radical activity. Free-chlorine activation of both heat shock and soxRS regulons required an exposure of less then I s in duration. The oxyR or the SOS regulons were apparently not induced by free chlorine (as judged by lack of activation of katG and recA, respectively), and neither was the universal stress (uspA) protein.