An Effective Strategy for a Whole-Cell Biosensor Based on Putative Effector Interaction Site of the Regulatory DmpR Protein

Introduction and Rationale The detection of bioavailable phenol is a very important issue in environmental and human hazard assessment. Despite modest developments recently, there is a stern need for development of novel biosensors with high sensitivity for priority phenol pollutants. DmpR (Dimethyl phenol regulatory protein), an NtrC-like regulatory protein for the phenol degradation of Pseudomonas sp. strain CF600, represents an attractive biosensor regimen. Thus, we sought to design a novel biosensor by modifying the phenol detection capacity of DmpR by using mutagenic PCR. Methods Binding sites of ‘A’ domain of DmpR were predicted by LIGSITE, and molecular docking was performed by using GOLD to identify the regions where phenol may interact with DmpR. Total five point mutations, one single at position 42 (Phe-to-Leu), two double at 140 (Asp-to-Glu) and 143 (Gln-to-Leu), and two double at L113M (Leu-to- Met) and D116A (Asp-to- Ala) were created in DmpR by site-directed mutagenesis to construct the reporter plasmids pRLuc42R, pRLuc140p143R, and pRLuc113p116R, respectively. Luciferase assays were performed to measure the activity of luc gene in the presence of phenol and its derivatives, while RT-PCR was used to check the expression of luc gene in the presence of phenol. Results Only pRLuc42R and pRLuc113p116R showed positive responses to phenolic effectors. The lowest detectable concentration of phenol was 0.5 µM (0.047 mg/L), 0.1 µM for 2, 4-dimethylphenol and 2-nitrophenol, 10 µM for 2, 4, 6-trichlorophenol and 2-chlorophenol, 100 µM for 2, 4-dichlorophenol, 0.01 µM for 4-nitrophenol, and 1 µM for o-cresol. These concentrations were measured by modified luciferase assay within 3 hrs compared to 6–7 hrs in previous studies. Importantly, increased expression of luciferase gene of pRLuc42R was observed by RT-PCR. Conclusions The present study offers an effective strategy to design a quick and sensitive biosensor for phenol by constructing recombinant bacteria having DmpR gene.


Introduction
Phenol and its organic compounds are the most important environmental pollutants at the global level that have been released into the environment in substantial quantities by natural events and industrial activities. Phenol is toxic, carcinogenic, mutagenic, and teratogenic even in low concentration and is mentioned in the list of priority pollutants of U.S. Environmental Protection Agency [1].
The detection of bioavailable phenol is a very important issue in environmental and human hazard assessment studies. Currently, chromatography coupled to the mass spectrometer detection is used for the analysis of phenol [2]. However, these methods are not used extensively because of being cost intensive and the chemicals used in these processes may harm the surrounding environment [3]. Biosensors can provide rapid measurement without labor-intensive and time-consuming sample preparation and apparently allow for a promising way to assess the biologically available phenol in the environment. Previously, different biosensors for phenol have been described such as the amperometric [4,5], enzymatic [6], and optical biosensor [7].
Microbes offer attractive targets for the construction of biosensors for monitoring the status of the environment. Wholecell microbial sensors have become one of the latest approaches of molecular tools in environmental monitoring [8]. Microorganisms, for their low cost, lifespan, and range of suitable pH and temperatures, have been widely used as the biosensing recognition elements in the construction of biosensors [9] Currently available whole cell biosensors have many advantages as they can provide an inexpensive and simple way of determining contaminants. As they are living organisms, they give a more accurate response on the toxicity of different compounds. Some stress-induced biosensors report the mutagenic effects of samples with great sensitivity. Biosensors are unexcelled in gene expression and physiological study of bacteria in complex environments. One of the greatest limitations of whole-cell biosensor development is the availability of strong promoters that respond only to relevant stimuli. To surmount this obstacle, more knowledge on gene regulatory networks in bacteria is needed [10]. Numerous whole cell biosensors for phenol have been developed in which different reporter and regulator genes with various lowest detection ranges of up to 291.4 mg/L (highest) [11] and 0.082 mg/L (lowest) [12] were reported. However, there is always a need for less cumbersome and sensitive approaches to monitor phenol levels. A different approach to construct a microbial biosensor is to connect a strictly regulated promoter sequence to a sensitive reporter gene was reported previously [13]. The ability of the bacteria to survive in a contaminated environment is usually based on a genetically encoded resistance system, the expression of which is regulated very precisely.
DmpR (Di methyl phenol regulatory protein), the product of the Pseudomonas sp. strain CF600 DmpR gene [14,15], mediates the expression of Dmp operon to allow growth on simple phenols. Transcription from Pdmp, the promoter of the Dmp operon, is activated when DmpR detects the presence of an inducing phenol [15]. Previous studies of the sensory 'A' domain may provide some insights regarding interaction between phenol and Dmp protein [16]. The natural interaction of DmpR with a subset of phenols suggested that the modification of its sensor domain might result in protein which can detect a broader range of phenolic derivatives [17]. Till date, there is no crystal structure for the DmpR protein.
However, a threading model for N-terminal 'A' domain of DmpR has been reported [18,19].
The aim of this study was to develop a new whole-cell luminescence-based bacterial sensor for highly selective and sensitive detection of bioavailable phenol in the environment. We accomplished this by choosing the DmpR gene with Po/Pr operator/promoter gene from Pseudomonas sp. CF600 as a receptor for phenol as an analyte and luc genes from pGL3 vector (Promega, USA) as a reporter. Pr promoter of DmpR was identified as a s 70 dependent promoter that is regulated by s 54 dependent Po promoter [20]. Previous mutational studies [15,21,22] were based on the random analysis in DmpR gene. In this study, mutant residues were predicted by docking tools instead of being chosen randomly. Mutant forms of DmpR with altered effector specificity were made by a dock-based site directed mutagenesis method. Three mutant plasmids pRLuc42R, pRLuc140p143R and pRLuc113p116R were constructed by creating mutations in the DmpR gene. The pRLuc42R contains one single mutation F42L (Phenylalanine at 42RLeucine), whereas the pRLuc140p143R and pRLuc113p116R possess two point mutations each (a) D140E (Aspartic acid at 140RGlutamic acid), and (b) Q143L (Glutamine at 143RLeucine); (a) L113M (Leucine at 113RMethionine) and (b) D116A (Aspartic acid at 116RAlanine). Mutant with the single mutation F42L (Phe at 42RLeu) was particularly interesting since it showed an increased response to the phenol inducer. 'A' domain of the DmpR is responsible for sensing the phenol and its derivative. Mutations were induced in this region by computational prediction and validated by mutagenic PCR in the wet lab.

Construction of pRLuc42R Biosensor
Whole cell biosensor (pRLuc42R) sensing principle is shown in Figure 1.
2.2.1. Plasmid isolation, cloning and PCR. Genomic DNA of Pseudomonas sp. CF600 was prepared by using a Genomic DNA isolation kit (RBC, Taiwan). Plasmids were isolated with a Qiagen spin column kit (CA, USA). The expression vectors, pGL3 basic and pGL3 promoter vector were purchased from Promega (WI, USA). Cloning, ligation, and transformation were carried out by using standard techniques [25]. PCR was performed by using forward primer 5-CGATCGATGCTCAGCTCGAGGC-CAGGTTAGGCGTAGGACGCA-3 and a reverse primer, 5-GCTAGCATCGGTATAAGCTTCGCACACGGATGTAAC-GAGTGA-3, with XhoI and HindIII sites at the ends. PCR conditions were as follows: 95uC for 5 min followed by 33 cycles of 95uC for 45 s, 55uC for 45 s, and 72uC for 1 min, followed by 72uC for 7 min. Taq DNA polymerase and deoxynucleoside triphosphates were from New England Biolabs (USA). Site-directed mutagenesis and cloning. Mutagenesis was performed by using the mutagenic PCR [26]. The primers used for site-directed mutagenesis are as follows, (a) for F42L (to generate pRLuc42R) forward primer: TGTTGCTGCAGCTTTCAGCGATGG and reverse primer: GCCATCGCTGAAAGCTGCAGCAA (b) for D140E and Q143L (to generate pRLuc140p143R) forward Primer: AC-CGAACTGGGGCTGATGCA and reverse primer: TGCAT-CAGCCCCAGTTCGGT (c) for L113M and D116A (to generate pRLuc113p116R) forward Primer: ACCGAGATGGA-TATCGCCAAGGAA and reverse primer: TTCCTTGGCGA-TATCCATCTCGGT (Altered codons are underlined). Mutagenic PCR product containing mutated DmpR (1692 bp) and the promoters, Pr and Po fragments (total 2377 bp) were gel-isolated and digested with XhoI and HindIII (New England Biolabs, USA), and introduced upstream of the firefly Luciferase (luc) in the pGL3 basic expression vector to generate approximately 7 kb of pRLuc42R plasmid ( Figure 2). The same method was used to generate pRLuc, pRLuc140p143R and pRLuc113p116R. Plasmids were transformed into E. coli DH5a by the CaCl 2 transformation method. The nucleotide sequences of the pRLuc42R were determined by DNA sequencing from the TCGA (The Centre for Genomic Application, New Delhi, India).

Proposed Binding Sites
LIGSITE algorithm [27] was used to predict the active site of 'A' DmpR model for phenol and phenolic derivatives. Thereafter the binding pocket which were predicted by LIGSITE, were given an input to GOLD software for docking analysis.

Treatment of Phenolic Compounds and Activity Assays for Luciferase
Various phenolic compounds (Sigma and Aldrich Chemical Co., MO, USA) were dissolved in ethanol and added to media at final concentrations ranging from 10 nM to 1 mM. The E. coli DH5a cells harboring the plasmid were grown overnight at 37uC in 4 ml of LB medium containing 100 mg/ml ampicillin. Cells were sub-cultured and grown to the log phase of 0.2 in OD at 600 nm [28] followed by supplementation with phenolic compounds at various concentrations. After 1 h, 50 ml of the cells were withdrawn, and kept frozen at 270uC with the addition of 5 ml of 1 M KH 2 PO 4 and 20 mM EDTA (pH 7.8). Cells were lysed by incubating at room temperature for 10 min with the addition of 150 ml of lysis solution (1.25 mg/ml lysozyme, 2.5 mg/ml BSA, 1X CCLR) (Promega, WI, USA). Supernatants were obtained by centrifugation. For the luciferase activity, 20 ml of supernatant was mixed with 20 ml of firefly luciferin solution (Promega, WI, USA). The bioluminescence was measured for 20 s by luminometer Berthold Detection System (Germany).

Calculations
Induction of the pRLuc42R sensor by phenol and phenolic derivatives was expressed as normalised luminescence (NL) calculated as follows [9]:-.
Where SL S is the luminescence of the sensor in the dilution of the phenolic compound, SL B is the background luminescence of the sensor bacteria, and CF is the correction factor.
Where L B is the background luminescence of the control bacteria, and L S is the luminescence of the control bacteria in phenol/ phenolic compound. The detection limit for the sensor bacteria for different phenols was defined as a concentration of the compound which induced the sensor twice above the background level, i.e. NL $2. Linearity curve for phenol induction is presented in Figure 3.

Chemical Analysis of Phenolic Compounds
Chemical analysis of phenolic compounds was performed according to the standard method recommended by American Public Health Association [29]. Synthetic water was filtered through a 0.2 mm membrane (Sartorius, Germany) and then serially diluted in 1 ml volume. Subsequently, 25 ml of 0.5 N NH 4 OH, 10 ml of 2% 4-aminoantipyrine (Sigma, MO, USA), and 10 ml of 8% K 3 Fe(CN) 6 (Sigma, MO, USA) were added in order and mixed thoroughly. OD at 500 nm was measured.  Real time RT-PCR was carried out for 30 min at 60uC following initial 2 min incubation at 50uC to prevent carryover reactions. The reaction was terminated by dissociation step by heating at 95uC for 15 s, 60uC for 1 min and finally 95uC for 15 s. The PCR amplification was then performed for 40 cycles with each cycle at 95uC for 15 s and 60uC for 1 min. All reactions were carried out in triplicate using the ABI 7500 Real Time PCR. The threshold cycle, C T , values were averaged from the values obtained from each reaction, and the mRNA expression was calculated as a relative level of expression.

Statistical Analysis
The data represents mean6SD of three independent experiments. The data was analyzed by one-way ANOVA followed by Newman-Keuls Multiple Comparison Test. The p value of ,0.05 represents statistical significance in different groups.

Results and Discussion
Among the most abundant environmental pollutants, the aromatic compounds (such as phenol or phenolic compounds) are of major concern because of their persistence and toxicity. The environmental problems caused by such coherent pollutants have increased the demand for the development of sensitive pollutant and toxicity detection methods. The fusion of reporter genes to promoters that are induced when cells are exposed to chemicals is one promising approach that has been used to formulate biosensor for such application [30].

Construction of Whole Cell Biosensor pRLuc42R
For confirmation of DmpR protein binding to phenol, E. coli whole cell luminescence biosensor (pRLuc42R) was constructed. The mutated sequence (mutation was suggested by docking and computational study) of 2377 bp (includes DmpR, Pr and Po) were inserted upstream of luc in pGL3 basic expression vector to generate pRLuc42R as a final construct to form whole cell biosensor for phenol. Reporter enzymes such as chloramphenicol acetyltransferase, firefly luciferase, and b-galactosidase from E. coli are frequently used for studies of gene regulation, gene activity, and expression in eukaryotic cells [31]. Firefly luciferase, in particular, has been extensively applied in molecular and cell biology studies because of its negligible background, less incubation time, high sensitivity, and the relative simplicity of the assay [32,33]. A previous study [34] has shown that firefly luciferase can also be used for analysis in prokaryotic cells. Therefore, firefly luciferase was used as a reporter in this study.

Active Site Prediction
In our previous study [35], we performed the docking procedure for phenol and N-DmpR model with GOLD by using both scoring functions (GOLD Score and Chemscore). The GOLD Fitness score and Chemscore for the selected phenolic compounds are shown ( Table 2). In selected phenolic derivatives, phenol showed highest GOLD Fitness score of 33.42 whereas the lowest Chemscore was 19.98. Figure 4 shows the docked phenol with 'A' domain-DmpR model. Image was produced by using PyMOL software [36]. Experimental mutated residues and active site pockets (in blue colour) of N-DmpR are also shown in this figure. Predicted positive key catalytic residues are (in green colour) Phe 42, Asp 140, Arg109, Arg 184, Phe 122, Cys 137, Leu 113, Asp 116 etc; while negative catalytic residues are (in red colour) Cys 149, Glu 133, Ala124, Trp193, and Leu83.Out of these residues our study was based on mutations in Phe42, Asp140,Glu143, Leu 113, and Asp 116 respectively. Apparently, this figure shows that the predicted ligand binding site is located in and near the V4R domain of N-DmpR. Previously, It has been reported that the V4R domain is a small molecule binding domain, which may bind to hydrocarbons [37].
The current study shows the validation of mutation sites and altered amino acids, provided by random mutations in previous studies [38,39]. Based on Gold/Chemscore, amino acid/acids were opted for site directed mutagenesis. Since there is no 3D structure available for DmpR till date, the only option for us was to predict mutations in the dry lab and then try to prove these in wet lab experiments.
On the other hand, 2, 4-dinitrophenol, 2-methyl-4, 6-dinitrophenol, pentachlorophenol, 4-chloro-3-ethylphenol, and Xylene were not detected by the system. The detection range and maximum responses of luciferase activity varied according to the compounds tested. The strong bioluminescence responses were caused by phenol, 2,4-dimethyphenol, and 2-chlorophenol. It has been known that phenol and methyphenol compounds had strong effects on the regulatory proteins involved in phenol degradation. The effective concentration of phenolic compounds for the regulatory proteins is also one of the major concerns. However, only a single concentration of effectors, usually around 2-3 mM was used in the measurement for DmpR in previous reports [12, 15, 18, and 34]. In the present studies, phenolic compounds in a range from 0.01 mM to 1 mM were examined. These results suggested that the residue L42 (Leucine at 42), M113 (Methionine at 113) and A116 (Alanine at 116) might be involved in inducer binding while the inducer binding might not be occurring in case of E140 (Glutamic acid at 140) and L143 (Leucine at 143). It is noteworthy that our system shows very low concentration ranges of phenolic compounds for detection. This aspect of the bacterial biosensors described here could be useful in detecting phenolic compounds at lower concentrations.

Luciferase Assay and Real Time PCR
E. coli cells harbouring pRLuc42R plasmid were grown with the addition of various concentrations of phenol, and their luciferase activities were measured. Luciferase shows stronger signals in almost all ranges of phenol and phenolic derivative concentration. The pattern of responses began to appear around 10 nM and continued to increase up to 1 mM. These results indicate that the DmpR/phenol complex binds to Po operator to activate luciferase genes most effectively, owing to the fact that the cloned DmpR is an activator gene implicated in the metabolism of phenol [14]. In addition, it was shown that the insect luciferase is stably expressed when they are grown to the log phase of 0.2 (instead of 0.3 or 0.4) in OD at 600 (p,0.001) ( Figure 6). As a result, the time consumption for the assay was decreased by 2-3 hours, as compared to 6-7 h in previous studies [12,34] which made it a relatively faster luciferase assay for phenol detection. The total time taken to complete the assay includes the incubation time for reaching the desired OD and induction with phenol. The quantitation of the luminescence is possible with a liquid scintillation counter, a luminometer, or even X-ray film.
Real time PCR was performed to confirm the responses to phenol with pRLuc42R, pRLuc and pRLuc140p143R. The threshold cycle (C T ) values for expression of luc gene in pRLuc42R, pRLuc, pRLuc140p143R, pRLuc113p116R and pGL3 were 8.0260. 21, 10.8060.63, 14.6360.78, 9.2460.27   Method), when compared to unmodified system pRLuc and pRLuc140p143R, respectively. Folds change has been compared with 1 mM phenol treated and untreated cells of all construct (Figure 7). In luciferase assay, decrease in the response occurring at high concentration of phenolic compounds appears to be due to the toxic effect of phenolic compounds on the cells. It has been suggested that these phenolic compounds may interact with specific components of the electron transport chain, thus inhibiting electron transport and oxidative phosphorylation [41,42]. Even though it cannot be excluded that cell viability may affect the total activities at high concentration of phenolic compounds, major increase in responses seem to occur in harmless or sub-lethal ranges. Taken together, the results indicate that the pRLuc42R protein has different properties in effecter specificity from other relevant regulators for catabolic pathways of phenolic compounds.

Assessment of Phenolic Compounds in Synthetic Water
It is shown above that E. coli cells with DmpR could be used as whole cell biosensor to monitor various pure phenolic compounds. The results show that this system can also detect phenolic compounds in synthetic water. The total amount of phenolic compounds in water measured by a chemical analysis exceeded the phenol amount obtained. The whole cell luminescent biosensor is similar to the chemical analysis method in a sense that they can detect the total phenolic compounds (data not shown). However, its results must be considered in a more complex way due to the differential effects of different phenolic derivatives. Moreover, the concentration of each phenolic compound in synthetic water might be achieved by further refined instrumental analysis.
The biosensor described here is very simple, economical and fast in measurement, possibly with sustaining sensitivity and  reproducibility. It might reflect various aspects of pollutants such as bioavailability and toxicity during the measurement [43,44]. So far the whole cell biosensors have offered unique alternative methods in identifying and quantifying phenol and its derivatives [12,15,18,34]. However, the present sensor does not to detect some phenolic derivatives out of eleven priority pollutant phenols (http://www.epa.gov/) but this is capable in detection of seven priority pollutants phenol and o-cresol. In summary, our data shows that the DmpR and its promoters may be of use for the development of whole cell biosensor, to specifically monitor the contamination of phenolic compounds in the environment.

Conclusions
Microbial biosensor is a simple, rapid, and effective analytical device that combines a biological sensing element with a transducer. As a sensing element, microorganism has not only advantages of easy manipulation and adaptation to the environment but also disadvantages of poor selectivity, low sensitivity and cytotoxicity. Here, we present one promising strategy to address the problems by tailoring microorganism through mutagenesis to the DmpR regulatory gene. Single mutant (pRLuc42R) was constructed at residues 42 of 'A' domain in the DmpR regulatory gene, using docking/computational studies. Replacement of amino acids resulted in positive effects on inducer responses. The pRLuc42R shows increased sensitivity. Thus, the mutagenesis method could open endless possibilities of microorganisms for excellent biosensing elements by overcoming the drawback of low sensitivity and cytotoxicity, eventually resulting in the development and application of microbial biosensors for prospective future advances.