AEM
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Norman, A.
Right arrow Articles by Sørensen, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Norman, A.
Right arrow Articles by Sørensen, S. J.
Agricola
Right arrow Articles by Norman, A.
Right arrow Articles by Sørensen, S. J.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, May 2005, p. 2338-2346, Vol. 71, No. 5
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.5.2338-2346.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Construction of a ColD cda Promoter-Based SOS-Green Fluorescent Protein Whole-Cell Biosensor with Higher Sensitivity toward Genotoxic Compounds than Constructs Based on recA, umuDC, or sulA Promoters

Anders Norman, Lars Hestbjerg Hansen,* and Søren J. Sørensen

Department of Microbiology, University of Copenhagen, 1307 Copenhagen K, Denmark

Received 19 October 2004/ Accepted 30 November 2004


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Four different green fluorescent protein (GFP)-based whole-cell biosensors were created based on the DNA damage inducible SOS response of Escherichia coli in order to evaluate the sensitivity of individual SOS promoters toward genotoxic substances. Treatment with the known carcinogen N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) revealed that the promoter for the ColD plasmid-borne cda gene had responses 12, 5, and 3 times greater than the recA, sulA, and umuDC promoters, respectively, and also considerably higher sensitivity. Furthermore, we showed that when the SOS-GFP construct was introduced into an E. coli host deficient in the tolC gene, the minimal detection limits toward mitomycin C, MNNG, nalidixic acid, and formaldehyde were lowered to 9.1 nM, 0.16 µM, 1.1 µM, and 141 µM, respectively, which were two to six times lower than those in the wild-type strain. This study thus presents a new SOS-GFP whole-cell biosensor which is not only able to detect minute levels of genotoxins but, due to its use of the green fluorescent protein, also a reporter system which should be applicable in high-throughput screening assays as well as a wide variety of in situ detection studies.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
As a result of modern life, we are in daily contact with an ever increasing myriad of substances. This has necessitated the ability to screen large numbers of samples for harmful properties like carcinogenicity. Short-term bacteriological mutagenicity assays have shown great consistency with traditional rodent bioassays and are popular due to the relative ease with which multiple substances can be screened, with only modest requirements for laboratory equipment and space.

In addition to the classic Ames test (2), used as a standard mutagenicity assay since the 1970s, many assays based on the Escherichia coli DNA repair SOS response have been developed. This has been done in order to provide even cheaper and faster alternatives and to accommodate high-throughput screening (13, 33, 35, 36). SOS-based genotoxicity assays function via simple SOS promoter/reporter gene fusions, introduced into strains of E. coli or Salmonella enterica serovar Typhimurium. SOS induction is thereby an indicator of a genotoxic presence (measured by a subsequent increase in reporter gene levels).

SOS genes are controlled by a single repressor, LexA, and are derepressed when cells undergo DNA damage, due to the accumulation of single-stranded DNA-RecA complexes which act as coproteases in a LexA self-cleaving mechanism (22, 28, 29). Expression of a given SOS gene depends on the specific LexA-binding properties of its promoter. These properties are determined by the presence of LexA-binding sequences (SOS boxes), which share a great deal of homology but are distinct from gene to gene (14, 27). The response of an SOS-based genotoxicity assay is therefore greatly influenced by the employed promoter. The preferred promoters have typically been the recA, umuDC, and sulA promoters, as was the case with the rec-lac test (32), the umu test, and the SOS chromotest, respectively. A relatively unstudied promoter from the ColD plasmid gene cda (16, 17) has also been applied with success in the SOS lux and SOS-LUX-LAC-FLUORO assays (34, 38). Although recent studies have performed genome-wide gene expression profiling of E. coli following DNA damage (9, 37), no candidates better than the promoters described here have been found. It is, however, remarkable that they have never been compared directly as biosensor "triggers" in a single study. A priority of this study was therefore to determine which SOS promoter is best suited for the detection of genotoxins.

Early genotoxicity assays have used the well-established lacZ reporter gene from Escherichia coli for the determination of mutagenicity via ß-galactosidase activity measurements in subsequent colorimetric assays. More recent assays have instead relied on light emission (18, 23, 25, 46) from reporters like the lux operons from Vibrio fischeri or Photobacterium leiognathi or the green fluorescent protein (GFP) from Aequorea victoria. These reporters can be almost instantly quantified and also add the possibility of continuous monitoring, since measurements do not require cell lysis. GFP is ideal for in vivo or in situ applications since it has a very stable signal, requires the presence of only oxygen to function, and needs no addition of substrates or cofactors. Studies in our laboratory have enabled in situ detection of specific compounds in a variety of complex environments by using a fluorescence-activated cell sorter-optimized version of GFP as reporter in combination with flow cytometry (4, 7, 19). In this study, we therefore decided to construct a genotoxicity biosensor based on GFP.

A drawback of the GFP reporter is a significant lack in sensitivity compared with constructs using the lux or lac reporters (20, 24). Improvements in biosensor sensitivity toward hydrophobic substances have, however, been obtained, simply by replacing wild-type host strains with mutant strains of E. coli or S. enterica serovar Typhimurium, either with increased membrane permeability and disabled nucleotide excision repair mechanisms or with limited efflux capabilities (10, 40, 42).


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Bacterial strains and plasmids.
E. coli strains and plasmids used in this study are listed in Table 1. All derivatives of pJBA27 were kept in host strain XL1-Blue (6). Luria-Bertani (LB) agar plates containing 100 µg/ml of ampicillin were used for the selection of ampicillin-resistant plasmid-containing strains. In addition to ampicillin, N43 tolC::Tn10 (15) strains were grown on LB plates with 20 µg/ml of tetracycline added. LB broth containing appropriate antibiotics was used for growth of overnight culture.


View this table:
[in this window]
[in a new window]
  		TABLE 1. List of strains and plasmids used in this study

 
Molecular cloning and construction of plasmid pANO1 and biosensor strains.
All PCRs were performed using the DyNAzyme EXT kit (FINNZYMES, Espoo, Finland) and primers listed in Table 2. Plasmid purifications from XL1-Blue overnight cultures were done with the QIAprep spin MiniPrep kit, and DNA extractions from agarose gels were done with QIAEX II; both kits were purchased from QIAGEN Inc. (Valencia, CA). All restriction endonucleases and T4 DNA ligase were purchased from New England Biolabs Inc. (Beverly, MA).


View this table:
[in this window]
[in a new window]
  		TABLE 2. List of primers used for PCRs in this study

 
Transformations were performed either by electroporation of electrocompetent host strain XL1-Blue (with ligation reactions) using a Micropulser from Bio-Rad Laboratories (Hercules, CA) or by heat shocking CaCl2-prepared competent cells (41) (with purified plasmid) in all other strains.

Plasmid pJBA27 (3) was used as vector template. It is a pUC18-derived high-copy-number vector containing the gene (labeled gfpmut 3*) encoding a fluorescence-activated cell sorter-optimized variant of GFP (8) where bases adjacent to the start codon have been changed to incorporate an SphI site. The LacI-repressible promoter PA1/04/03 and the optimized synthetic ribosome binding site (RBS II) situated between the multiple cloning site (MCS) and the SphI site were both removed in subsequent promoter fragment insertions and are therefore not described in this study.

Plasmid pANO1 was constructed by PCR amplification of pJBA27 using primer sets P1 with P2 and P3 with P4. Set P1 with P2 led to a ~2,100-bp ClaI-NotI PCR fragment (fragment A), while P3 with P4 led to a ~2,500-bp NotI-ClaI fragment (fragment B). Combined, fragments A and B contained all of pJBA27 except the lac promoter. Both fragments were restricted (sequentially) with ClaI and NotI, run on an agarose gel, purified, and subsequently ligated. The ligation reaction was then transformed into XL1-Blue, and transformants were selected using LB plates containing 20 µg/ml chloramphenicol to avoid selection of any possible religations of fragment B. Removal of a 117-bp fragment, as well as the introduction of two G->T mutations in the MCS NotI site, was later verified by sequencing as described below.

SOS primers were designed to clone regions of approximately 300 bp in length directly upstream of the start codon in each of the E. coli chromosomal genes recA (329-bp fragment), umuD (316-bp fragment), and sulA (318-bp fragment) (5) and the pColD-CA23 cda gene (318-bp fragment). A complete sequence for the whole pColD-CA23 plasmid was not available, so primers were designed using the sequence from a close homologue, the pColD-157 plasmid (21). An EcoRI site was introduced into each of the 5' primers and an SphI site in each of the 3' primers so that inserts would be placed directly upstream to the start codon of the gfpmut 3* gene when inserted into pANO1. This meant that the start codon methionine from each SOS gene replaced the start codon of the gfpmut 3* gene. Promoter inserts recA', umuDC', and sulA' were each amplified from an E. coli MG1655 cell lysate (obtained by boiling 100 µl overnight culture for 10 min) using 1 µl of lysate as template, while cda' was amplified from pColD-CA23 using purified plasmid as template.

Each of the PCR fragments and pANO1 were digested (sequentially) with EcoRI and SphI and subsequently separated from unwanted fragments by agarose gel electrophoresis. Relevant fragments were subsequently cut out from the gel and purified. Inserts were then ligated into EcoRI- and SphI-digested pANO1 and transformed into XL1-Blue. Plasmids were purified and transformed into relevant host strains. The E. coli K-12 strain MG1655 was used as a host for the promoter comparison assay, since the promoters themselves (with the notable exception of cda') originated from the chromosome of this strain.

Sequencing and DNA analysis.
Sequencing was performed in an ABI PRISM 310 genetic analyzer from Applied Biosystems (Foster City, CA) using the sequencing primers listed in Table 2. Comparisons with template DNA sequences were done using the National Center for Biotechnology Information BLAST 2 (Basic Local Alignment Search Tool) algorithm (1).

Assay procedure for reporter-construct comparison.
To ensure maximum consistency in experiments, large batches of overnight culture were prepared and subsequently frozen: 100 ml of LB medium containing 0,4% (wt/vol) glucose and 100 µg/ml of ampicillin were inoculated with a single colony of the relevant biosensor and grown overnight at 30°C and 200 rpm. The high glucose concentration and low growth temperature were both used to prevent any unwanted induction of the SOS response before assays were performed. Cells were harvested by centrifugation (4,000 x g, 5 min), washed twice in 100 ml cold autoclaved 20% glycerol, and finally resuspended in the same volume of 20% glycerol. Washed cells were transferred to 1.5-ml Eppendorf tubes in aliquots of 1 ml and kept frozen (–18°C) until needed.

Each assay was started by making a 1:100 dilution of thawed biosensor into preheated LB medium containing 0.2% glucose and growing them at 37°C and 200 rpm until the optical density at 600 nm (OD600) reached a value of approximately 0.2. Aliquots of 2.85 ml biosensor culture were added to 25-ml glass test tubes, each containing 150 µl of either N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) in aqueous solution in 20 times the relevant concentration or in double-distilled water and subsequently grown at 37°C and 200 rpm with samples taken at relevant times. Sample tubes were placed on ice for 3 min and then transferred to 1-cm, 4.5-ml fluorescence cuvettes for immediate measurement. Fluorescence intensity was measured in a L50S-B luminescence spectrophotometer from PerkinElmer (Wellesley, MA). Sample measurements were taken by setting excitation to 488 nm (not the excitation maximum of GFPmut 3*, which is 501 nm, but a standard argon laser wavelength) and measuring emission at 513 nm using a slit width of 2.5 nm. A baseline was first made by measuring a 3-ml sample containing 2.85 ml of LB medium with 0.2% glucose and 150 µl of double-distilled water. The fluorescence intensity was measured in the fluorimeter's own arbitrary fluorescence units ranging from 0 to 1,000. OD600 was measured afterward in a BioMate 5 spectrophotometer (Thermo Electron Corporation, Woburn, MA). The specific fluorescence unit (SFU) for each sample was calculated by dividing the relative fluorescence units by the cell density (RFU/OD600). Induction factors (Fi) were calculated as SFUx/SFU0, where SFUx is the sample treated with genotoxin and SFU0 is the control sample at the same time point. The term "basal level" refers to SFU0 in this paper. Relative optical density values (ODx/OD0) were calculated using OD600 values measured in treated and untreated samples at the same time point. All values were means ± standard deviations (SD) for n = 3.

Procedure for dose-response assays with known genotoxins.
The following serial dilutions of chemicals (from Sigma) in aqueous solution were used for dose-response assays: for biosensors using the MG1655 host strain, 4 to 1,000 nM mitomycin C (MMC), 0.05 to 50 µM MNNG, 0.08 to 20 µM nalidixic acid (NA), or 3 to 800 µM formaldehyde (CH2O); for biosensors using the N43 tolC::Tn10 host strain, 3 to 800 nM MMC, 0.02 to 6 µM MNNG, 0.08 to 20 µM NA, 3 to 800 µM formaldehyde, or 8 to 2,000 µM hydrogen peroxide. Assays were performed, and fluorescence was measured as described above with 90 (promoter comparison) or 120 (host strain comparison) min of induction.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Analysis and verification of promoter inserts.
Four different pANO1 derivatives with promoter regions recA', umuDC', sulA', or cda' cloned into sites EcoRI and SphI were created for this study (Fig. 1). Inserts were verified by sequencing both strands of the plasmids in the region between the ClaI site and the gfpmut 3* gene in all four pANO1::insert constructs. Promoter sequencing results are listed in Fig. 2 All promoter sequences contained SOS boxes in close proximity to or overlapping the –35, –10, or +1 regions of the putative RNA polymerase binding sites. Inserts recA' and sulA' contained only a single SOS box, while umuDC' and cda' had two. The two SOS boxes in the cda' promoter overlapped by 1 base in a 31-bp palindromic sequence positioned after the transcription start site (+2 to +32). A TTTT tract on the 5' side of the SOS boxes was also present in the cda' insert. Overlapping SOS boxes and TTTT tracts are both unique features of colicin gene promoters (30, 31). A sequence previously known to enhance translation in E. coli was found in the recA promoter (44).



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 1. Diagram showing the cloning strategy for the creation of pANO1 and the four SOS-GFP reporter vectors containing cloned promoters from the E. coli SOS response. Plasmid pANO1 was created to avoid the lac promoter, which would otherwise interfere with GFPmut 3* expression. To do this, it was necessary to introduce a ClaI site and remove the NotI site in the MCS. Abbreviations: Plac, lac promoter; PA1/03/04, optimized LacI-repressible promoter; RBS II, synthetic ribosome-binding site; T0 and T1, strong transcriptional terminators; cat, chloramphenicol acetyltransferase (Cm resistance); LacZ', truncated ß-galactosidase; bla, ß-lactamase (Ap resistance); ColE1 ori, origin of replication.

 


View larger version (18K):
[in this window]
[in a new window]
  		FIG. 2. Sequences of the 3' ends of sequenced inserts cloned into pANO1. Letters in bold indicate –35 (consensus sequence, TTGACA), –10 (consensus sequence, TATAAT), and +1 (transcription start site) positions of the putative RNA polymerase-binding regions. Dark grey boxes indicate LexA-binding sites (consensus sequence, CTGTATATATATACAG), and black boxes indicate ribosome binding sites (consensus sequence, AGGAGG). The light grey box indicates a sequence (consensus sequence, TGATCC) known to enhance translation in E. coli. {Delta}1 indicates the distance in bp between the –10 and –35 regions, and {Delta}2 indicates the distance in bp between the ribosome-binding site and the translation initiation codon (underlined). Tick marks ({blacktriangleup}) indicate bases in the cda' insert where the cda promoter differed from pColD-CA23 to pColD-157.

 
The relative LexA-binding strength of SOS boxes is indicated by calculating the heterology index (HI), where high homology to the 20-bp consensus sequence TACTG(TA)5CAGTA gives low values (14, 27). HI was calculated for each SOS box along with an alternative HI, where only homology to bases 4 to 17 was considered (Table 3). These bases were chosen because they seemed to reflect the affinity of SOS boxes toward LexA better. Results showed that, when not considering bases 1 to 3 and 18 to 20, the two cda SOS boxes had equal HI values of 4.17, which were below both recA' and sulA' but above the umuDC' which showed very high homology (a HI of 2.77) to the consensus sequence in the second SOS box.


View this table:
[in this window]
[in a new window]
  		TABLE 3. HIa of SOS boxes present in promoter regions to recA, umuD, sulA, and cda

 
Both the recA' and cda' promoters had –35 and –10 regions very close to the consensus sequences, indicating that they bind very effectively to the RNA polymerase, while umuDC' and sulA' had identical –35 and –10 regions that were less similar to the consensus sequences.

The distance between the –35 and –10 sites as well as the distance between the ribosome-binding Shine-Dalgarno (S/D) sequence and the ATG start codon varied from promoter to promoter. While the –35 to –10 distance varied by only 2 bases (16-18), the S/D-to-ATG codon varied by as much as 6 bases (4-10).

Inserts recA', umuDC', and sulA' all showed 100% homology to regions in the E. coli chromosome from where they were amplified (not counting the primer modifications incorporating restriction sites). The cda' insert, amplified from pColD-CA23, showed only 99% (three mismatches) homology to the pColD-157 plasmid cda promoter region, which was the sequence used to design the primers. This was a T->C in the –1 position, a C->T in the +7 position, and a G->A in the +27 position. The +7 and +27 substitutions were both within the palindromic sequence of the promoter and were in fact mirror bases in the palindrome. The pColD-CA23 SOS boxes were therefore closer to the consensus sequence than pColD-157 and were expected as a consequence to bind LexA more tightly, which is also indicated by the lower HI values in Table 3.

Comparison of SOS-GFP biosensor constructs following MNNG treatment.
SOS induction of the four biosensors was compared by measuring GFP fluorescence over time in both untreated samples and samples treated with 50 µM of the known carcinogen MNNG (Fig. 3). Most notably, measurements showed that the GFP expression controlled by each SOS promoter was distinct in both treated and untreated samples. The untreated recA' and sulA' biosensors both had high levels of GFP to start with (recA', SFU > 350; sulA', SFU > 85), while the umuDC' and cda' biosensors were both below 20 SFU. These values subsequently decreased slightly but later started to increase again in recA', cda', and sulA' biosensors, most notably in the sulA'. The umuDC' biosensor also increased in SFU, but only after 2.5 h (data not shown), indicating that the SOS response is induced overall when going to the stationary phase.



View larger version (15K):
[in this window]
[in a new window]
  		FIG. 3. Development in specific fluorescence over time of biosensor strains of E. coli MG1655 hosting pANO1 with promoter inserts recA', umuDC', sulA', or cda' fused to gfpmut 3*, following treatment with a high dose (50 µM) of MNNG ({blacktriangleup}) compared to control samples ({triangleup}). Values are means ± SD for n = 3.

 
MNNG treatment led to significant increases in SFU over untreated samples in all biosensors but was more dramatic in the sulA' (1,037 ± 1 SFU) and cda' (1,322 ± 13 SFU) biosensors than in the recA' (556 ± 46 SFU) and umuDC' biosensors (103 ± 4 SFU).

Figure 4 shows the development of the calculated Fi over time in the same experiment. The relative strength of each promoter was apparent already after 60 min, but Fi values continued to increase further (at a reduced rate) until finally reaching maximum values after 90 to 120 min. None of the biosensors improved in Fi by being incubated more than 2 h, and Fi in all biosensors had declined after 3 h (data not shown). The cda' biosensor's response clearly surpassed the other three biosensors by reaching a final Fi of 32 ± 2, which was 2.7 times higher than the maximum Fi reached by the umuDC' biosensors (12 ± 0.4), 5 times that by the sulA' biosensors (6.3 ± 0.2, after 90 min), and 12 times that by the recA' biosensors (2.7 ± 0.3).



View larger version (13K):
[in this window]
[in a new window]
  		FIG. 4. Induction over time of four biosensor strains of E. coli MG1655 hosting pANO1 with promoter inserts recA' ({blacktriangleup}), umuDC' ({blacktriangledown}), sulA' ({diamondsuit}), or cda' (•) fused to gfpmut 3*, following treatment with a high dose (50 µM) of MNNG. Values are means ± SD for n = 3.

 
To demonstrate that the cda' biosensor was also superior with regards to sensitivity, an additional experiment was performed in which the recA', sulA', and cda' biosensors were compared at different concentrations of MNNG after 90 min of induction (Fig. 5). The cda' biosensor had much higher Fi values than the sulA' biosensor above 1.5 µM of MNNG and was distinguishable from the other biosensors at 0.8 µM MNNG. As with the previous experiment, the recA' biosensor showed a very limited response due to high basal levels. Based on the performance of the cda' promoter to MNNG, it was deemed a far better choice for detection of genotoxins than the other promoters, and subsequent experiments were performed only with strains harboring the pANO1::cda' plasmid.



View larger version (13K):
[in this window]
[in a new window]
  		FIG. 5. Induction of four biosensor strains of E. coli MG1655 hosting pANO1 with promoter inserts recA' ({blacktriangleup}), umuDC' ({blacktriangledown}), sulA' ({diamondsuit}), or cda' (•) fused to gfpmut 3*, following treatment for 90 min with different concentrations of MNNG. Values are means ± SD for n = 3.

 
Comparison of a wild-type host strain to a tolC mutant host strain and their response toward chemical genotoxins.
The effect of changing the host strain from a wild type to a tolC mutant, with limited efflux capability, was also tested. Biosensors E. coli N43 tolC::Tn10/pANO1::cda' (N43c) and E. coli MG1655/pANO1::cda' (MG1655c) were treated with different concentrations of MMC, MNNG, NA, or formaldehyde, and Fi values and relative optical densities were compared (Fig. 6). Fi values in N43c were considerably higher than those in MG1655c, especially at mid to low concentrations of genotoxin, but N43c was also more sensitive toward the cytotoxic effects of genotoxin in MNNG, NA, or formaldehyde, as indicated by the significantly lower relative optical density values. MMC did not seem to elicit any increased cytotoxicity at the concentrations tested but still had increased response. A dose-response curve was also made for hydrogen peroxide in N43c to clearly demonstrate the detrimental effects on Fi at cytotoxic concentrations (Fig. 7). The cytotoxic effects of hydrogen peroxide on N43c were illustrated by a continuous drop in optical density with rising concentrations, which ended at 0.19 ± 0.01 at 2,000 µM, meaning that almost no cell growth had occurred during the 120 min at this concentration. Fi increased in a dose-dependant matter until 500 µM of hydrogen peroxide was reached, where it dropped dramatically from 27.9 ± 2.0 to 2.2 ± 0.1 at 1,000 µM and just about remained above 2 at 2,000 µM. The biosensor was, however, still able to function near a relative optical density of 0.3 (at 500 µM), but not at 0.2 (1,000 to 2,000 µM), indicating that quite severe cytotoxicity or growth inhibition is required to disable N43c. Taken together with the result from Fig. 6, the functionality of N43c was not threatened beyond final relative optical densities above 0.5. The minimum detection limits were subsequently determined, based on linear plots of Fi versus genotoxin at low concentrations of MMC, MNNG, NA, hydrogen peroxide, or formaldehyde (Table 4). N43c had minimum detection limits lower than MG1665c with all compounds tested. This confirms that a tolC mutant host strain raises sensitivity considerably, despite being more prone to cytotoxic effects caused by raised internal concentrations of damaging agents.



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 6. Induction factors (Fi, left panels) and relative optical densities (ODx/OD0, right panels) of biosensors MG1655/pANO1::cda' ({blacktriangleup}) and N43 tolC::Tn10/pANO1::cda' ({triangleup}) treated with MMC, MNNG, NA, or formaldehyde. Values are means ± SD for n = 3.

 


View larger version (16K):
[in this window]
[in a new window]
  		FIG. 7. Induction factor ({blacktriangleup}, left axis) and relative optical density ({triangleup}-right axis) of biosensor N43 tolC::Tn10/pANO1::cda' treated with hydrogen peroxide. Values are means ± SD for n = 3.

 

View this table:
[in this window]
[in a new window]
  		TABLE 4. Lower detection limit of genotoxic compounds for biosensor strains N43 tolC::tn10/pANO1::cda' (N43c) and MG1655/pANO1::cda' (MG1655c)

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The need for rapid, sensitive, and versatile genotoxicity screening assays has become increasingly apparent, and while very few assays can compete in sensitivity with the Ames test, developers of new biosensor constructs should always strive toward maximum sensitivity as well as exploring new areas of application. Four different biosensor strains using the green fluorescent protein variant GFPmut 3* as a reporter were constructed to evaluate the response of individual SOS promoters toward chemical genotoxins in order to find the best candidate for genotoxin screening in an SOS-GFP construct. Genotoxicity assays often operate with the ratio of induction (or Fi) of samples treated with genotoxins compared to that of control samples as a quantitative expression of the genotoxicity of a given agent and a minimal detection limit for each compound.

Responses of the four different biosensors to treatment with MNNG over time, and subsequently to different concentrations, revealed that the Fi values of the four biosensors developed in a relatively similar fashion but were very different in strength due to differences in both individual basal levels and responses to induction. The optimal assay time for the constructs was between 90 and 120 min, which, when combined with results in Fig. 3, seemed to depend on the time in each biosensor, where basal levels began to rise significantly. High basal levels were particularly inhibitive for Fi values in the recA' and sulA' biosensors, which meant that they were surpassed by the well-repressed umuDC' biosensor despite having much higher responses to induction. The data presented in this study thus show that the cda promoter is superior with regards to both response and sensitivity, due to a combination of very low basal levels and high response to induction.

Overall, basal levels could have been amplified due to the dilution in the thousands of the repressor/SOS box ratio that occurs when a high-copy-number vector containing an SOS promoter is introduced into a strain with only a single chromosomal lexA gene. Since basal levels were deemed sufficiently low in both cda' and umuDC' strains, they were not investigated any further, but a low-copy-number reporter vector approach might give more favorable results due to lower basal levels.

The relative basal levels observed in this study correspond with earlier studies of SOS gene expression (26) and were apparently influenced by the specific LexA-binding properties of each promoter. The SOS boxes with low HI values bind LexA more tightly and were present in the umuDC' and cda' inserts (Table 3). The cda' insert even had two strong SOS boxes (in an overlapping palindromic sequence), which should account for the uniquely low basal levels of such a strong promoter. Inserts recA' and sulA' both had only one SOS box, with higher HI values. The particularly high basal levels of the recA' strain seemed to be caused by both inefficient binding of LexA and the presence of an expression-enhancing sequence (Fig. 2). It can be concluded that lower basal levels in an SOS-GFP construct can be obtained by using promoters with strong LexA-binding properties. This of course also opens up the possibility of designing synthetic promoter sequences with optimal LexA binding, for example, by introducing consensus SOS boxes into "leaky" promoters or by removing SOS boxes from too strongly repressed promoters. Such modifications have previously been attempted with success in other biosensor constructs, but so far, only with modified recN and sulA promoters (11).

Little attention has otherwise been paid to specific properties of SOS promoters with regards to genotoxicity screening assays, but as our brief analysis of the inserts revealed, strong up-regulating factors like near-consensus –35, –10, and S/D regions with optimal placement were also present in the cda' insert. The optimal distance between the S/D region and the translation initiation codon has previously been shown to be 7 bp in E. coli (45), and this was the case in the cda promoter (Fig. 2). In the pColD-CA23 cda promoter is therefore a good indication of what makes a "perfect" promoter for genotoxin screening: a combination of strong promoter factors and equally strong LexA-binding properties.

A number of highly sensitive host strains like the deep, rough Salmonella enterica serovar Typhimurium TA1535 (used in the Ames test), with increased membrane permeability, are available for the screening of genotoxins. E. coli strains with truncated tolC genes have a limited efflux capability in part due to the lack in functionality of the Mar/AcrAB pump (15) and thus lead to higher internal concentrations of hydrophobic agents, which should also increase the detection sensitivity (10). Treatment of biosensors MG1655/pANO1::cda' and N43/pANO1::cda' with MMC, MNNG, NA, or formaldehyde revealed that this was indeed the case. Detection limits were lowered considerably (Table 4) toward these genotoxins, but increased cytotoxicity was also observed (Fig. 6 and 7). This should also be taken into account when considering application of this biosensor, since resistance toward cytotoxicity could outweigh the need for high sensitivity in a given experiment. The N43/pANO1::cda' biosensor should therefore be considered a very sensitive supplement rather than a replacement for the MG1655/pANO1::cda' biosensor.

The detection limits obtained in this study compare very well with other studies investigating the response of SOS constructs to the same genotoxins (25, 34) and are in fact lower than the more established umu test and SOS chromotest, but still not in the area of sensitivity as the Ames test. Detection limits were just below those for the original SOS lux test, which demonstrates that the use of the tolC mutant host strain compensates for the use of the GFP reporter.

In light of our findings, traditional SOS-based assays like the umu test or the SOS chromotest could therefore seem dated in their present form because they use less sensitive promoters. However, the high degree of standardization and substantial volume of data already available from these assays (39, 47) will no doubt make sure that they will be around, along with the Ames test, for many years to come. The strength of our particular construct lies in the use of GFP as a reporter, which has been applied while retaining high sensitivity, and should therefore enable us to perform a variety of in situ studies.

A concern with SOS biosensors, when considering in situ studies, is that the SOS response is known to be induced by factors like changes in pH and starvation of nutrients (12, 43), which are typical conditions in soil environments. Such factors should be investigated to evaluate their influence on the sensitivity and response of our SOS-GFP biosensors in situ, but preliminary studies in our lab (not published) have already shown that our biosensor is able to detect mitomycin C down to 2.5 ppb (using the MG1655/pANO1::cda' biosensor) in spiked soil microcosms, which should indicate that basal levels do not ruin the sensitivity of our construct to a great extent.

A highly sensitive SOS-GFP construct is therefore now available and should be applicable in a wide range of in situ studies.


   ACKNOWLEDGMENTS
 
We extend sincere gratitude to Joachim Frey (Institute of Veterinary Bacteriology, University of Berne, Berne, Switzerland), who kindly contributed to this study by supplying us with the pColD-CA23 plasmid.


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Microbiology, University of Copenhagen, Sølvgade 83H, 1307 Copenhagen K, Denmark. Phone: 45 35 32 20 53. Fax: 45 35 32 20 40. E-mail: hestbjerg{at}bi.ku.dk. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402.[Abstract/Free Full Text]
  2. Ames, B. N., F. D. Lee, and W. E. Durston. 1973. An improved bacterial test system for the detection and classification of mutagens and carcinogens. Proc. Natl. Acad. Sci. USA 70:782-786.[Abstract/Free Full Text]
  3. Andersen, J. B., C. Sternberg, L. K. Poulsen, S. P. Bjørn, M. Givskov, and S. Molin. 1998. New unstable variants of green fluorescent protein for studies of transient gene expression in bacteria. Appl. Environ. Microbiol. 64:2240-2246.[Abstract/Free Full Text]
  4. Bahl, M. I., L. H. Hansen, T. R. Licht, and S. J. Sørensen. 2004. In vivo detection and quantification of tetracycline by use of a whole-cell biosensor in the rat intestine. Antimicrob. Agents Chemother. 48:1112-1117.[Abstract/Free Full Text]
  5. Blattner, F. R., G. Plunkett III, C. A. Bloch, N. T. Perna, V. Burland, M. Riley, J. ColladoVides, J. D. Glasner, C. K. Rode, G. F. Mayhew, J. Gregor, N. W. Davis, H. A. Kirkpatrick, M. A. Goeden, D. J. Rose, B. Mau, and Y. Shao. 1997. The complete genome sequence of Escherichia coli K-12. Science 277:1453-1474.[Abstract/Free Full Text]
  6. Bullock, W. O., J. M. Fernandez, and J. M. Short. 1987. Xl1-Blue—A high-efficiency plasmid transforming recA Escherichia coli strain with beta-galactosidase selection. BioTechniques 5:376-379.
  7. Burmølle, M., L. H. Hansen, G. Oregaard, and S. J. Sørensen. 2003. Presence of N-acyl homoserine lactones in soil detected by a whole-cell biosensor and flow cytometry. Microb. Ecol. 45:226-236.[CrossRef][Medline]
  8. Cormack, B. P., R. H. Valdivia, and S. Falkow. 1996. FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33-38.[CrossRef][Medline]
  9. Courcelle, J., A. Khodursky, B. Peter, P. O. Brown, and P. C. Hanawalt. 2001. Comparative gene expression profiles following UV exposure in wild-type and SOS-deficient Escherichia coli. Genetics 158:41-64.[Abstract/Free Full Text]
  10. Davidov, Y., R. Rozen, D. R. Smulski, T. K. Van Dyk, A. C. Vollmer, D. A. Elsemore, R. A. LaRossa, and S. Belkin. 2000. Improved bacterial SOS promoter:lux fusions for genotoxicity detection. Mutat. Res. 466:97-107.[Medline]
  11. Dreier, J., E. B. Breitmaier, E. Gocke, C. M. Apfel, and M. G. P. Page. 2002. Direct influence of S9 liver homogenate on fluorescence signals: impact on practical applications in a bacterial genotoxicity assay. Mutat. Res. 513:169-182.[Medline]
  12. Dri, A. M., and P. L. Moreau. 1994. Control of the LexA regulon by pH: evidence for a reversible inactivation of the LexA repressor during the growth cycle of Escherichia coli. Mol. Microbiol. 12:621-629.[CrossRef][Medline]
  13. el Mzibri, M., M. P. De Meo, M. Laget, H. Guiraud, E. Seree, Y. Barra, and G. Dumenil. 1996. The Salmonella sulA-test: a new in vitro system to detect genotoxins. Mutat. Res. 369:195-208.[CrossRef][Medline]
  14. Fernandez De Henestrosa, A. R., T. Ogi, S. Aoyagi, D. Chafin, J. J. Hayes, H. Ohmori, and R. Woodgate. 2000. Identification of additional genes belonging to the LexA regulon in Escherichia coli. Mol. Microbiol. 35:1560-1572.[CrossRef][Medline]
  15. Fralick, J. A. 1996. Evidence that TolC is required for functioning of the Mar/AcrAB efflux pump of Escherichia coli. J. Bacteriol. 178:5803-5805.[Abstract/Free Full Text]
  16. Frey, J., P. Ghersa, P. G. Palacios, and M. Belet. 1986. Physical and genetic analysis of the ColD plasmid. J. Bacteriol. 166:15-19.[Abstract/Free Full Text]
  17. Ghersa, P., R. Lurz, B. Dobrinski, J. Deshusses, M. Belet, and J. Frey. 1988. Genetic evidence for the existence of a repressor that modulates colicin D expression on plasmid ColD-CA23. Mol. Gen. Genet. 213:36-41.[CrossRef][Medline]
  18. Gu, M. B., J. Min, and R. A. LaRossa. 2000. Bacterial bioluminescent emission from recombinant Escherichia coli harboring a recA::luxCDABE fusion. J. Biochem. Biophys. Methods 45:45-56.[CrossRef][Medline]
  19. Hansen, L. H., B. Ferrari, A. H. Sørensen, D. Veal, and S. J. Sørensen. 2001. Detection of oxytetracycline production by Streptomyces rimosus in soil microcosms by combining whole-cell biosensors and flow cytometry. Appl. Environ. Microbiol. 67:239-244.[Abstract/Free Full Text]
  20. Hansen, L. H., and S. J. Sørensen. 2000. Versatile biosensor vectors for detection and quantification of mercury. FEMS Microbiol. Lett. 193:123-127.[CrossRef][Medline]
  21. Hofinger, C., H. Karch, and H. Schmidt. 1998. Structure and function of plasmid pColD157 of enterohemorrhagic Escherichia coli O157 and its distribution among strains from patients with diarrhea and hemolytic-uremic syndrome. J. Clin. Microbiol. 36:24-29.[Abstract/Free Full Text]
  22. Janion, C. 2001. Some aspects of the SOS response system—a critical survey. Acta Biochim. Pol. 48:599-610.[Medline]
  23. Justus, T., and S. M. Thomas. 1998. Construction of a umuC'-luxAB plasmid for the detection of mutagenic DNA repair via luminescence. Mutat. Res. 398:131-141.[Medline]
  24. Justus, T., and S. M. Thomas. 1999. Evaluation of transcriptional fusions with green fluorescent protein versus luciferase as reporters in bacterial mutagenicity tests. Mutagenesis 14:351-356.[Abstract/Free Full Text]
  25. Kostrzynska, M., K. T. Leung, H. Lee, and J. T. Trevors. 2002. Green fluorescent protein-based biosensor for detecting SOS-inducing activity of genotoxic compounds. J. Microbiol. Methods 48:43-51.[CrossRef][Medline]
  26. Kuzminov, A. 1999. Recombinational repair of DNA damage in Escherichia coli and bacteriophage {lambda}. Microbiol. Mol. Biol. Rev. 63:751-813.[Abstract/Free Full Text]
  27. Lewis, L. K., G. R. Harlow, L. A. Gregg-Jolly, and D. W. Mount. 1994. Identification of high affinity binding sites for LexA which define new DNA damage-inducible genes in Escherichia coli. J. Mol. Biol. 241:507-523.[CrossRef][Medline]
  28. Little, J. W. 1991. Mechanism of specific LexA cleavage: autodigestion and the role of RecA coprotease. Biochimie 73:411-421.[Medline]
  29. Little, J. W., D. W. Mount, and C. R. Yanisch-Perron. 1981. Purified LexA protein is a repressor of the recA and lexA genes. Proc. Natl. Acad. Sci. USA 78:4199-4203.[Abstract/Free Full Text]
  30. Lloubes, R., M. Granger-Schnarr, C. Lazdunski, and M. Schnarr. 1991. Interaction of a regulatory protein with a DNA target containing two overlapping binding sites. J. Biol. Chem. 266:2303-2312.[Abstract/Free Full Text]
  31. Lloubes, R., C. Lazdunski, M. Granger-Schnarr, and M. Schnarr. 1993. DNA sequence determinants of LexA-induced DNA bending. Nucleic Acids Res. 21:2363-2367.[Abstract/Free Full Text]
  32. Nunoshiba, T., and H. Nishioka. 1991. ‘Rec-lac test’ for detecting SOS-inducing activity of environmental genotoxic substance. Mutat. Res. 254:71-77.[Medline]
  33. Oda, Y., S. Nakamura, I. Oki, T. Kato, and H. Shinagawa. 1985. Evaluation of the new system (umu-test) for the detection of environmental mutagens and carcinogens. Mutat. Res. 147:219-229.[CrossRef][Medline]
  34. Ptitsyn, L. R., G. Horneck, O. Komova, S. Kozubek, E. A. Krasavin, M. Bonev, and P. Rettberg. 1997. A biosensor for environmental genotoxin screening based on an SOS lux assay in recombinant Escherichia coli cells. Appl. Environ. Microbiol. 63:4377-4384.[Abstract]
  35. Quillardet, P., and M. Hofnung. 1985. The SOS Chromotest, a colorimetric bacterial assay for genotoxins: procedures. Mutat. Res. 147:65-78.[Medline]
  36. Quillardet, P., and M. Hofnung. 1993. The SOS chromotest: a review. Mutat. Res. 297:235-279.[Medline]
  37. Quillardet, P., M. A. Rouffaud, and P. Bouige. 2003. DNA array analysis of gene expression in response to UV irradiation in Escherichia coli. Res. Microbiol. 154:559-572.[Medline]
  38. Rabbow, E., P. Rettberg, C. Baumstark-Khan, and G. Horneck. 2003. The SOS-LUX-LAC-FLUORO-toxicity-test on the International Space Station (ISS). Adv. Space Res. 31:1513-1524.[CrossRef][Medline]
  39. Reifferscheid, G., and J. Heil. 1996. Validation of the SOS/umu test using test results of 486 chemicals and comparison with the Ames test and carcinogenicity data. Mutat. Res. 369:129-145.[CrossRef][Medline]
  40. Rettberg, P., K. Bandel, C. Baumstark-Khan, and G. Horneck. 2001. Increased sensitivity of the SOS-LUX-Test for the detection of hydrophobic genotoxic substances with Salmonella typhimurium TA1535 as host strain. Anal. Chim. Acta 426:167-173.[CrossRef]
  41. Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
  42. Shapiro, E., and F. Baneyx. 2002. Stress-based identification and classification of antibacterial agents: second-generation Escherichia coli reporter strains and optimization of detection. Antimicrob. Agents Chemother. 46:2490-2497.[Abstract/Free Full Text]
  43. Taddei, F., I. Matic, and M. Radman. 1995. cAMP-dependent SOS induction and mutagenesis in resting bacterial populations. Proc. Natl. Acad. Sci. USA 92:11736-11740.[Abstract/Free Full Text]
  44. Thanaraj, T. A., and M. W. Pandit. 1989. An additional ribosome-binding site on mRNA of highly expressed genes and a bifunctional site on the colicin fragment of 16S rRNA from Escherichia coli: important determinants of the efficiency of translation-initiation. Nucleic Acids Res. 17:2973-2985.[Abstract/Free Full Text]
  45. Vellanoweth, R. L., and J. C. Rabinowitz. 1992. The influence of ribosome-binding-site elements on translational efficiency in Bacillus subtilis and Escherichia coli invivo. Mol. Microbiol. 6:1105-1114.[Medline]
  46. Vollmer, A. C., S. Belkin, D. R. Smulski, T. K. Van Dyk, and R. A. LaRossa. 1997. Detection of DNA damage by use of Escherichia coli carrying recA'::lux, uvrA'::lux, or alkA'::lux reporter plasmids. Appl. Environ. Microbiol. 63:2566-2571.[Abstract]
  47. White, P. A., and J. B. Rasmussen. 1996. SOS Chromotest results in a broader context: empirical relationships between genotoxic potency, mutagenic potency, and carcinogenic potency. Environ. Mol. Mutagen. 27:270-305.[CrossRef][Medline]

Applied and Environmental Microbiology, May 2005, p. 2338-2346, Vol. 71, No. 5
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.5.2338-2346.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Norman, A.
Right arrow Articles by Sørensen, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Norman, A.
Right arrow Articles by Sørensen, S. J.
Agricola
Right arrow Articles by Norman, A.
Right arrow Articles by Sørensen, S. J.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
J. Bacteriol. Microbiol. Mol. Biol. Rev. Eukaryot. Cell All ASM Journals