A dual lethal system to enhance containment of recombinant micro-organisms

Begoña Torres1,{dagger}, Susanne Jaenecke2,{ddagger}, Kenneth N. Timmis2, José L. García1 and Eduardo Díaz1

1 Department of Molecular Microbiology, Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Científicas, Ramiro de Maeztu 9, 28040 Madrid, Spain
2 Division of Microbiology, GBF-National Research Centre for Biotechnology, Mascheroder Weg 1, D-38124 Braunschweig, Germany

Correspondence
Eduardo Díaz
ediaz{at}cib.csic.es


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Active containment systems based on the controlled expression of a lethal gene are designed to increase containment of recombinant micro-organisms used for environmental applications. A major drawback in containment is the existence of mutations that generate surviving cells that cease to respond to the toxic effect of the lethal function. In this work the authors have developed for the first time a strategy to reduce the problem of mutations and increase the efficiency of containment based on the combination of two lethal functions acting on different cellular targets of major concern in containment, DNA and RNA, and whose expression is under control of different regulatory signals. To engineer the dual gene containment circuit, two toxin–antitoxin pairs, i.e. the colicin E3–immunity E3 and the EcoRI restriction–modification systems, were combined. The genes encoding the immunity E3 and the EcoRI methyltransferase proteins (antitoxins) were stably inserted into the chromosome of the host cell, whereas the broad-host-range lethal genes encoding the colicin E3 RNase and the EcoRI restriction endonuclease (toxins) were flanking the contained trait in a plasmid. This dual lethal cassette decreased gene transfer frequencies, through killing of the recipient cells, by eight orders of magnitude, which provides experimental evidence that the anticipated containment level due to the combination of single containment systems is generally achieved. Survivors that escaped killing were analysed and the mutational events involved were characterized.


{dagger}Present address: Department Molecular and Cell Biology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Campus Universidad Autónoma, Cantoblanco, 28049 Madrid, Spain.

{ddagger}Present address: EuMeCom, Warburgstr. 4, 20354 Hamburg, Germany.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Horizontal gene transfer is a source of concern when genetically engineered micro-organisms are intended to be released in large quantities to the environment for biotechnological applications (Ramos et al., 1995; Wackett, 2000). Gene spread could be also undesirable in contained environments, e.g. a bacterial consortium in a fermentation tank, for process protection and process optimization reasons (Díaz et al., 1999). To make more predictable the behaviour of a genetically modified organism (GMO) introduced into a target habitat, its ability to spread new genetic information to potential recipients has to be reduced to avoid the appearance of undesired novel genetic combinations (gene containment). Moreover, the survival of the GMO has to be limited in time and space, i.e. engineering a controlled life cycle to reduce its dissemination and impact on the indigenous population of organisms (biological containment) (Molin et al., 1993; Ramos et al., 1995). Several gene and biological containment systems have been designed and they are based on a lethal function that, through a transcriptional and/or post-translational control element, becomes active in response to specific environmental signals (Molin et al., 1993; Díaz et al., 1999; Torres et al., 2000; Ronchel & Ramos, 2001; Davison, 2002). A general feature of the containment systems so far described is that a surviving subpopulation of cells, in the range of 10-7 to 10-3 per cell and generation depending on the particular system under study, ceases to respond to the toxic effect of the lethal function, which may be significant in environmental applications that use large quantities of cells (Molin et al., 1993; Szafranski et al., 1997). Analysis of the survivors has revealed that mutations are the main drawback in containment. Mutations that inactivate the containment system have been located either in the lethal function (Jensen et al., 1993; Munthali et al., 1996a; Torres et al., 2000) or in the control element (Bej et al., 1988; Knudsen et al., 1995).

A way to reduce the problem of mutations and increase the efficiency of containment is to engineer genetic circuits with more than one lethal function (Davison, 2002). When the gef gene was used in a biological containment system responding to the absence of benzoate effectors, the rate of escape from killing decreased from 10-6 (one copy of gef) to 10-8 (two copies of gef) (Jensen et al., 1993). Duplication of the relF gene also increased plasmid containment by about three orders of magnitude (Knudsen & Karlström, 1991). Nevertheless, the reduction in the number of survivors by duplicating a lethal function is always smaller than that expected by theoretical calculations (Knudsen et al., 1995). This reduction in the efficiency of containment can be due to several factors such as: (i) homologous recombination and gene conversion between the two copies of the suicide function, (ii) presence of mutations that inactivate the regulatory element controlling the expression of both lethal genes, (iii) existence of mutations in the cellular target of the lethal function (Knudsen et al., 1995). To circumvent these limitations, the use of non-identical suicide functions with different cellular target sites and whose expression is under control of different regulatory circuits may be a suitable strategy.

Colicin E3 is a RNase encoded by the colE3 gene that specifically cleaves the 16S rRNA, causing cell death (Zarivach et al., 2002). The immE3 gene encodes the immunity E3 protein which binds stoichiometrically to colicin E3, preventing its RNase activity (Yajima et al., 1993). The EcoRI restriction endonuclease is encoded by the ecoRIR gene of Escherichia coli (O'Connor & Humphreys, 1982). Protection of the host DNA from attack by this endonuclease is provided by the EcoRI methyltransferase encoded by the ecoRIM gene (Pingoud & Jeltsch, 1997). The colE3/immE3 genes and the ecoRIR/ecoRIM genes have been previously used to design single containment systems that were shown to be efficient in a broad range of Gram-negative bacteria (Díaz et al., 1994; Munthali et al., 1996a; Torres et al., 2000). Since colicin E3 and the EcoRI endonuclease have different cellular targets, RNA and DNA respectively, and their corresponding antitoxins act at different levels, i.e. the toxin itself and the cellular target, respectively, they constitute ideal candidates for combination in a dual containment approach. In this work we present such a dual containment circuit that significantly reduces gene spread to generally achieve the anticipated level of containment.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Bacterial strains, plasmids, growth conditions, and transformation experiments.
The E. coli K-12 strains used were HB101Rif (Torres et al., 2000), XL-1 Blue (Sambrook et al., 1989), CC118{lambda}pir and S17-1{lambda}pir (de Lorenzo & Timmis, 1994). E. coli HB101immE3 is a derivative of strain HB101 that expresses the immE3 gene constitutively from the chromosome and is therefore immune to colicin E3 (Díaz et al., 1994). Plasmid pVTRB is a low-copy-number chloramphenicol-resistance cloning vector that drives expression of cloned genes with the Ptrc promoter (Pérez-Martín & de Lorenzo, 1996). pVTRE is a derivative of pVTRB that expresses the ecoRIR gene under the control of the Ptrc promoter (Torres et al., 2000). pUC18Sfi-colE3 is a pUC18Sfi derivative containing the colE3 and immE3 genes (Díaz et al., 1994). pHP45{Omega} is a pBR322 derivative harbouring the {Omega} interposon which contains the transcription-translation termination signals of gene 32 of phage T4 flanking the gene that confers streptomycin/spectinomycin resistance (Smr/Spr). Plasmid pSJ201 contains the ecoRIM gene cloned under control of the Pr promoter from {lambda} phage in a mini-Tn5Smr/Spr transposon (Torres et al., 2000). The RK2-Mob+ RK2-Tra+ plasmid pRK600 was used as helper in mobilization experiments (de Lorenzo & Timmis, 1994). Bacteria were grown in Luria–Bertani (LB) medium (Sambrook et al., 1989) at 37 °C. When required, antibiotics were added at the following concentrations: ampicillin, 100 µg ml-1; chloramphenicol, 35 µg ml-1; rifampicin, 50 µg ml-1; spectinomycin, 100 µg ml-1; and kanamycin, 50 µg ml-1. E. coli cells were transformed by using the RbCl procedure as described by Sambrook et al. (1989). Electroporation of E. coli cells was performed as previously described (Díaz et al., 1994). Transformants were selected on chloramphenicol-containing LB medium, except with plasmid pVTRECA, for which ampicillin-containing LB medium was also used.

DNA manipulations and sequencing.
Plasmid DNA was prepared by the rapid alkaline lysis method (Sambrook et al., 1989). Genomic DNA was prepared as previously reported (Richards, 1987). DNA manipulations and other molecular biology techniques were essentially as described by Sambrook et al. (1989). DNA fragments were purified using the Gene-Clean Kit (BIO-101). Oligonucleotides were synthesized on an Oligo-1000M nucleotide synthesizer (Beckman Instruments). Nucleotide sequences were determined directly from plasmids by using the dideoxy chain termination method (Sanger et al., 1977). Standard protocols of the manufacturer for Taq DNA polymerase-initiated cycle sequencing reactions with fluorescently labelled dideoxynucleotide terminators (Applied Biosystems) were used. The sequencing reactions were analysed using a model 377 automated DNA sequencer (Applied Biosystems). Nucleotide sequence similarity searches were made by using the BLAST program (Altschul et al., 1990) via the National Institute for Biotechnology Information server (http://www.ncbi.nlm.nih.gov/BLAST/).

Design of the E. coli HB101immE3ecoRIM strain immune to both colicin E3 and the EcoRI endonuclease.
By means of RP4-mediated mobilization (de Lorenzo & Timmis, 1994), plasmid pSJ201 was transferred from E. coli S17-1{lambda}pir into E. coli HB101immE3, a kanamycin-resistant strain. The resulting transconjugant strain, E. coli HB101immE3ecoRIM, containing the Ptrc : : immE3 and the Pr : : ecoRIM fusions stably inserted into the chromosome, was selected for the transposon marker, spectinomycin, on kanamycin-containing LB medium. The resulting E. coli HB101immE3ecoRIM strain constitutively expressed the immE3 and ecoRIM genes as revealed by the cross-streak test and by protection of its chromosomal DNA from EcoRI-mediated cleavage in an in vitro DNA restriction assay (see below), respectively. E. coli HB101immE3ecoRIM did not show any significant difference in growth characteristics from its parental HB101Rif counterpart as revealed by growth rate determinations and competition experiments, suggesting that none of the immunity functions appears to be detrimental to the host cell under the growth conditions used. The growth rate was determined by viable counting and by measuring OD600 of the cultures along the growth curve. To perform the competition experiments, the immune and parental strains were inoculated together in the same medium, samples were taken at the exponential and stationary phase of growth and cell numbers were determined by plating on rifampicin-containing LB medium (selects for HB101Rif and HB101immE3ecoRIM cells) and on kanamycin-containing LB medium (selects for HB101immE3ecoRIM cells). The cell ratio of the two strains remained constant along the growth curve.

Construction of the colE3/ecoRIR-based lethal cassettes.
For the construction of the Pc : : colE3 cassette, the promoterless colE3 gene was PCR-amplified from plasmid pUC18Sfi-colE3 by using primers Col5 [5'-CCGGATCCTTGACAAACCCAGTGGTTTTATGTACAG-3'; the sequence spans from position -71 to -48 with respect to the ATG initiation codon of the colE3 gene (Masaki & Ohta, 1985); the engineered BamHI site is underlined; the engineered -35 consensus box of the Pc promoter is indicated in italics] and Col3 [5'-GGTCTAGACTTCCTCTCAAAGATATTTC-3'; the sequence spans the TGA stop codon (in italics) of colE3; the engineered XbaI site is double-underlined]. The resulting 1·6 kb DNA fragment was digested with BamHI and XbaI and cloned into double-digested BamHI+XbaI pVTRB vector. The resulting plasmid, pVTRCol (Fig. 1), was isolated in E. coli HB101immE3 and expresses the colE3 gene under the control of the tandem Ptrc promoter and a synthetic promoter (Pc). A Smr/Spr DNA cassette flanked by transcription termination signals from phage T4 was cloned upstream of the Pc : : colE3 fusion in plasmid pVTREC, giving rise to plasmid pVTRColT (Fig. 1).



View larger version (34K):
[in this window]
[in a new window]
 
Fig. 1. Schematic representation of the construction of the contained plasmids. The thick and thin arrows show the promoters and the direction of transcription of the genes, respectively. Ptrc, trp/lac hybrid promoter; Pc, synthetic promoter. The lethal ecoRIR and colE3 genes are indicated by black and vertically striped blocks, respectively. The genes encoding ampicillin (Apr), and streptomycin/spectinomycin (Smr/Spr) resistance are indicated with white and dotted blocks, respectively. The gene encoding ampicillin resistance was cloned with its own P3 promoter (Brosius et al., 1982). Cmr, gene encoding chloramphenicol resistance. Primers Col5, Col3, Ap5' and Ap3' are detailed in Methods. T4, transcriptional terminator of gene 32 from phage T4; T, T1/T2 transcriptional terminators of the E. coli rrnB operon; oriVpSC101, origin of replication of plasmid pSC101. Relevant restriction sites shown are: B, BamHI; H, HindIII; Hp, HpaI; N, NotI; X, XbaI. BamHI(Klenow) indicates digestion with BamHI followed by filling-in the protruding ends generated with E. coli DNA polymerase I Klenow fragment.

 
To develop a dual lethal cassette, the BamHI+XbaI-digested Pc : : colE3 fragment was cloned into the double-digested BamHI+XbaI pVTRE plasmid. Using E. coli HB101immE3ecoRIM as recipient strain, a chloramphenicol-resistant transformant containing plasmid pVTREC, which expresses the Ptrc : : ecoRIR and Pc : : colE3 fusions in tandem in a chloramphenicol-resistant low-copy-number vector (Fig. 1), was identified. As revealed by growth rate determinations and competition experiments, pVTREC did not provide a significant growth disadvantage to the host cell HB101immE3ecoRIM when compared with the control plasmid pVTRB. The growth rate was determined by viable counting and by measuring OD600 of the cultures along the growth curve. To perform the competition experiments, HB101immE3ecoRIM(pVTREC) and HB101immE3(pVTRB) cells were inoculated together in the same medium, samples were taken at the exponential and stationary phases of growth and cell numbers were determined by plating on LB medium containing chloramphenicol and kanamycin [selects for both strains] and on LB medium containing chloramphenicol and spectinomycin [selects for HB101immE3ecoRIM(pVTREC) cells]. The cell ratio of the two strains remained constant along the growth curve.

To construct an ampicillin resistant (Apr) cassette, the bla gene encoding the {beta}-lactamase from transposon Tn3 was PCR-isolated from plasmid pUC19 by using primers Ap5' [GGGGATCCGTTTCTTAGACGTCAGGTGGCAC, the engineered BamHI site is underlined; it hybridizes upstream of the P3 promoter of the bla gene (Brosius et al., 1982)] and Ap3' [GGGGATCCGGTCATGAGATTATCAAAAAGG, the engineered BamHI site is underlined]. The resulting 1·0 kb Apr DNA cassette was digested with BamHI and cloned into BamHI-digested pVTREC. The resulting plasmid, pVTRECA (Fig. 1), was isolated in E. coli HB101immE3ecoRIM cells on ampicillin-containing LB plates.

Preparation of crude cell extracts and EcoRI endonuclease assay.
To prepare crude extracts from E. coli HB101immE3ecoRIM containing plasmids pVTRB, pVTREC and pVTRECA, cells were grown in chloramphenicol-containing LB medium to an OD600 of about 2. Cell cultures were then centrifuged (3000 g, 10 min at 20 °C), and cells were washed and resuspended in 0·01 vol. 50 mM Tris/HCl (pH 7·6) containing 10 % sucrose and 2 mM dithiothreitol, prior to disruption by passage through a French press (Aminco) operated at 20 000 p.s.i. The cell debris was removed by centrifugation at 26 000 g for 30 min at 4 °C. The clear supernatant fluid was decanted and used as crude cell extract. Protein concentration was determined by the method of Bradford (1976) using bovine serum albumin as standard. EcoRI restriction endonuclease assays were performed as described previously (Torres et al., 2000). Crude extracts from cells containing pVTREC and pVTRECA, but not those from cells containing pVTRB, showed a pattern of restriction fragments indistinguishable from that obtained with purified EcoRI enzyme from a commercial source.

Assay for colicin E3 and immunity E3 activity.
The production of immunity E3 protein in E. coli HB101immE3ecoRIM was checked qualitatively by the cross-streak test (Takagaki et al., 1973). Colicin E3 production by plasmids pVTRCol, pVTRColT, pVTREC and pVTRECA was assayed by the spot test (Masaki & Ohta, 1985) in which the degree of sensitivity to the colicin E3 present in crude extracts was determined as the highest dilution of the extract required to make a clear inhibition spot on the top agar inoculated with a colicin-sensitive indicator strain such as E. coli HB101.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Principles and construction of a dual containment system
The gene containment system developed in this work is based on a double toxin–antitoxin genetic circuit. In the donor cell, two lethal functions (toxins) encoded by genes closely associated with the target DNA whose containment is required are inhibited by the cognate immunity functions (antitoxins). The genes encoding the immunity functions are not linked to the target DNA, and therefore co-transfer of toxin and antitoxin is a highly unlikely event. On the contrary, transfer of the target DNA to a non-immune recipient cell will be accompanied by transfer of the lethal genes, thus causing cell death and preventing further spread of this particular trait. To engineer such a containment system, we have used the colicin E3/immunity E3 (Masaki & Ohta, 1985) and the EcoRI endonuclease/EcoRI methyltransferase (Pingoud & Jeltsch, 1997) toxin–antitoxin pairs.

To design an E. coli host cell immune to both colicin E3 and the EcoRI endonuclease, the ecoRIM gene expressed under control of the strong Pr promoter from the {lambda} phage (Torres et al., 2000) was stably inserted into the chromosome of E. coli HB101immE3, a strain expressing the immE3 gene constitutively, and therefore immune to colicin E3. Since the resulting E. coli HB101immE3ecoRIM strain did not show any significant difference in growth characteristics from its parental HB101Rif counterpart (see Methods), none of the immunity functions appears to be detrimental to the host cell under the conditions used.

On the other hand, we have previously cloned and expressed the ecoRIR lethal gene under the control of the broad-host-range Ptrc promoter in the low-copy-number pVTRB vector, giving rise to the contained pVTRE plasmid (Fig. 1) (Torres et al., 2000). To clone and express the colicin E3 lethal function in an analogous system, plasmid pVTRCol was constructed (Fig. 1). To develop a dual lethal cassette controlled by two different regulatory elements, we designed plasmid pVTREC. This plasmid harbours a 4·5 kb NotI cassette containing the ecoRIR gene under control of the LacI-repressed Ptrc promoter and the colE3 gene under control of the constitutive Pc promoter (Fig. 1). As anticipated, whereas crude extracts from HB101immE3ecoRIM(pVTREC) showed colicin E3 activity by the spot test and EcoRI endonuclease activity in DNA restriction assays (see Methods), extracts from the control strain HB101immE3ecoRIM(pVTRB) did not. As shown with pVTRE (Torres et al., 2000), plasmids pVTRCol and pVTREC did not provide a significant growth disadvantage to the host cell HB101immE3ecoRIM when compared with the control plasmid pVTRB under the conditions used (see Methods). To demonstrate the functionality of the Pc promoter, transcriptional termination signals from the {Omega} interposon were cloned upstream of the Pc : : colE3 fusion in plasmid pVTREC, giving rise to plasmid pVTRColT (Fig. 1). E. coli HB101immE3ecoRIM(pVTRColT) cells showed colicin E3 production, confirming that the Pc promoter was indeed driving the expression of the colE3 gene.

Single containment versus dual containment
To assess whether a dual containment system based on two different lethal functions acting on different cellular targets and controlled by different regulatory elements is significantly more efficient than containment systems based on each single lethal function, we compared the transformation frequencies of plasmids pVTREC, pVTRE, pVTRCol, pVTRColT, and the control plasmid pVTRB using a recipient strain, E. coli XL-1 Blue, that lacks the immE3 and ecoRIM genes. Containment efficiency was measured as the reduction in plasmid transfer with respect to the transfer frequency of the pVTRB control plasmid. Electrotransformation was used as mechanism of plasmid transfer since it allows a high efficiency of DNA uptake in E. coli and it has been previously used to measure the efficacy of containment systems, showing containment levels similar to those obtained using conjugation as the mechanism of plasmid transfer (Díaz et al., 1994; Torres et al., 2000). As shown in Table 1, the efficacy of plasmid containment was about 104 with the EcoRI-based containment system and 105 with the colicin E3-based containment system. Since plasmid pVTRColT behaved similarly to plasmid pVTRCol (Table 1), the expression of the colE3 gene from the Pc promoter seems to be as efficient as that from the tandem Ptrc–Pc promoters. The efficacy of containment of plasmid pVTREC (about 106) was higher than that of plasmids pVTRE and pVTRCol (Table 1), demonstrating that the combination of the ecoRIR and colE3 genes enhances gene containment.


View this table:
[in this window]
[in a new window]
 
Table 1. Testing the efficiency of single and dual containment systems

 
Nevertheless, the level of containment achieved by the combination of two different lethal functions with different regulatory signals did not reach the level expected by theoretical calculations, which predict that the efficiency of a dual containment system should be the product of the containment efficiencies due to each individual lethal function (Knudsen et al., 1995). To investigate the reason for the reduced efficiency of the dual containment system, we examined several E. coli XL-1 Blue transformants that were selected from three different electrotransformation experiments and that survived acquisition of plasmid pVTREC. Electrophoretic analyses revealed that plasmids isolated from ten different survival clones were smaller than the parental pVTREC plasmid (7·9 kb) (Fig. 2). Moreover, sequence analyses showed the existence of deletions that included both lethal genes and, in most cases, the lacIq regulatory gene (Fig. 2). Interestingly, inactivation of the lethal functions by integration of insertion sequences (IS) (Mahillon & Chandler, 1998), the most frequent mechanism of survival of the EcoRI-based gene containment system in E. coli (Torres et al., 2000), was not observed with the dual containment circuit developed in this work. In contrast, deletions involving both lethal genes in pVTREC led to the formation of clones that escaped killing. The different mechanism of mutation in the host cells to inactivate a single lethal function versus a dual lethal cassette reflects the fact that whereas IS-dependent inactivation of two lethal functions controlled by different regulatory signals would require two mutational events, simultaneous deletion of the two lethal functions requires only a single mutational event. Our results show that the frequency of deletion events in a particular organism will determine the efficiency of containment when using a lethal cassette harbouring more than one killing gene. A genetic circuit such as the one presented here can be also of great utility to study the role of mutations in microbial adaptation to adverse conditions.



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 2. Analyses of the plasmids isolated from E. coli clones that have acquired the pVTREC plasmid. Genes are indicated with blocks. The Ptrc and Pc promoters are also indicated. Black blocks and continuous lines represent DNA regions that are present in the analysed plasmids; striped blocks and discontinuous lines represent deleted regions. The number of amino acids encoded by the 3' end of the truncated colE3 gene present in plasmids isolated from clones M3 to M8 is also indicated. oriV, origin of replication of plasmid pSC101.

 
Inactivation of a dual containment system via a single deletion event involves the loss of the DNA region located between the lethal genes. Therefore the cloning and expression of a particular trait flanked by two lethal genes would assure that the spread of the trait to potential recipients will be reduced even in case of inactivation of the dual containment system via deletion. To demonstrate this, we have cloned the bla gene, encoding the {beta}-lactamase of the Tn3 transposon as a model trait (Brosius et al., 1982), in the ecoRIR–colE3 intergenic region of plasmid pVTREC. The expression of the colE3 and ecoRIR genes in the resulting plasmid, pVTRECA (Fig. 1), was confirmed by the spot test and by in vitro DNA restriction assays, respectively. Plasmid pVTRECA was transferred to E. coli XL-1 Blue cells by electroporation, and transformants were selected by growth on rich medium containing either chloramphenicol (selects for the Cmr marker of the plasmid, which is not flanked by the lethal genes) or ampicillin (selects for the contained bla trait). As observed with plasmid pVTREC, chloramphenicol-resistant transformants with pVTRECA appeared with a frequency of about 10-6 compared to that of the control plasmid pVTRB (Table 1), and all harboured plasmids of smaller size than pVTRECA. The fact that none of these chloramphenicol-resistant transformants was able to grow in ampicillin-containing medium, and none showed colicin E3 and EcoRI endonuclease activities, is in agreement with the existence of deletions embracing both lethal genes and the target trait in plasmid pVTRECA. When three different electrotransformation experiments carried out with plasmid pVTRECA were plated on ampicillin-containing LB medium, we only observed the appearance of two transformants, indicating that the frequency of transfer of the target trait (bla gene) was as low as 10-8 (Table 1). Thus, these data reveal that the horizontal spread of a target gene flanked by the ecoRIR and colE3 genes in a dual containment cassette is reduced by a factor (108) that is close to the anticipated containment level due to the combination of colicin E3 (provides a containment of 105) and the EcoRI endonuclease (provides a containment of 104).

When the two E. coli XL-1 Blue ampicillin-resistant transformants that survived acquisition of plasmid pVTRECA were analysed, we observed the existence of a similar plasmid whose size was larger than 8·9 kb. Sequence analysis of this plasmid showed the presence of two IS1A insertion sequences (Umeda & Ohtsubo, 1991) that had integrated in both the colE3 and ecoRIR genes, but that maintained the bla gene unmodified. The existence of two IS sequences in the mutated pVTRECA plasmid caused high instability, and different plasmid rearrangements were observed after several rounds of cultivation of the ampicillin-resistant clones that escaped killing (data not shown). In summary, these data show that two different mutational events, e.g. integration of IS elements within the lethal genes, rather than a single mutation, e.g. deletion of both lethal genes, are necessary to inactivate a dual containment cassette designed to prevent gene spread of a target trait flanked by the lethal genes.

Conclusion
We have demonstrated here that the combination of different lethal functions acting on different cellular targets and controlled by different regulatory signals is a valuable strategy to increase containment. Engineering the multiple containment system in mini-transposon vectors, which themselves exhibit non-detectable transfer frequencies when integrated into the host chromosome (de Lorenzo & Timmis, 1994), should decrease further any spread of cloned traits to ecologically insignificant levels that cannot be detected even under optimal gene transfer conditions (Munthali et al., 1996b). This work also reveals that in a biological containment system the lethal genes should be placed at different locations to avoid their inactivation through a single deletion event. It was shown previously that the efficiency of a biological containment system can be reinforced by using a host strain with a genetically engineered background (Ronchel & Ramos, 2001). Therefore, the combination of such a strategy with a multiple lethal system like the one described here may be a suitable approach to achieve efficient programmed suicide for increasing containment of novel recombinant micro-organisms.


   ACKNOWLEDGEMENTS
 
The help of E. Aporta with oligonucleotide synthesis, A. Díaz, G. Porras and S. Carbajo with sequencing, and the technical assistance of E. Cano, M. Carrasco and F. Morante are gratefully acknowledged. This work was supported by EU contract QLRT-2001-02884; by the Red del CSIC de Biorremediación y Fitorremediación, and by grants BMC2000-0125-CO4-02 and GEN2001-4698-C05-02 from the Comisión Interministerial de Ciencia y Tecnología.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990). Basic local alignment search tool. J Mol Biol 215, 403–410.[CrossRef][Medline]

Bej, A. K., Perlin, M. H. & Atlas, R. M. (1988). Model suicide vector for containment of genetically engineered microorganisms. Appl Environ Microbiol 54, 2472–2477.[Medline]

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248–254.[CrossRef][Medline]

Brosius, J., Cate, R. L. & Perlmutter, A. P. (1982). Precise location of the two promoters for the {beta}-lactamase gene of pBR322. S1 mapping of ribonucleic acid isolated from Escherichia coli or synthesized in vitro. J Biol Chem 257, 9205–9210.[Abstract/Free Full Text]

Davison, J. (2002). Towards safer vectors for the field release of recombinant bacteria. Environ Biosafety Res 1, 9–18.

de Lorenzo, V. & Timmis, K. N. (1994). Analysis and construction of stable phenotypes in Gram-negative bacteria with Tn5 and Tn10-derived mini-transposons. Methods Enzymol 235, 386–405.[Medline]

Díaz, E., Munthali, M., de Lorenzo, V. & Timmis, K. N. (1994). Universal barrier to lateral spread of specific genes among microorganisms. Mol Microbiol 13, 855–861.[Medline]

Díaz, E., Chithila-Munthali, M. T., Jaenecke, S., de Lorenzo, V. & Timmis, K. N. (1999). Design of genetic circuits for restricting gene and biocatalyst dispersal. In Molecular Microbial Ecology Manual, pp. 1–18, unit 6.1.14. Edited by A. D. L. Akkermans, J. D. van Elsas & F. J. de Bruijn. Dordrecht: Kluwer.

Jensen, L. B., Ramos, J. L., Kaneva, Z. & Molin, S. (1993). A substrate-dependent biological containment system for Pseudomonas putida based on the Escherichia coli gef gene. Appl Environ Microbiol 59, 3713–3717.[Abstract]

Knudsen, S. & Karlström, O. H. (1991). Development of efficient suicide mechanisms for biological containment of bacteria. Appl Environ Microbiol 57, 85–92.[Medline]

Knudsen, S., Saadbye, P., Hansen, L. H., Collier, A., Jacobsen, B. L., Schlundt, J. & Karlström, O. H. (1995). Development and testing of improved suicide functions for biological containment of bacteria. Appl Environ Microbiol 61, 985–991.[Abstract]

Mahillon, J. & Chandler, M. (1998). Insertion sequences. Microbiol Mol Biol Rev 62, 725–774.[Abstract/Free Full Text]

Masaki, H. & Ohta, T. (1985). Colicin E3 and its immunity genes. J Mol Biol 182, 217–227.[Medline]

Molin, S., Boe, L., Jensen, L. B., Kristensen, C. S., Givskov, M., Ramos, J. L. & Bej, A. K. (1993). Suicidal genetic elements and their use in biological containment of bacteria. Annu Rev Microbiol 47, 139–166.[CrossRef][Medline]

Munthali, M. T., Timmis, K. N. & Díaz, E. (1996a). Use of colicin E3 for biological containment of microorganisms. Appl Environ Microbiol 62, 1805–1807.[Abstract]

Munthali, M. T., Timmis, K. N. & Díaz, E. (1996b). Restricting the dispersal of recombinant DNA: design of a contained biological catalyst. Bio/Technology 14, 189–191.[Medline]

O'Connor, C. D. & Humphreys, G. O. (1982). Expression of the EcoRI restriction-modification system and the construction of positive-selection cloning vectors. Gene 20, 219–229.[CrossRef][Medline]

Pérez-Martín, J. & de Lorenzo, V. (1996). VTR expression cassettes for engineering conditional phenotypes in Pseudomonas: activity of the Pu promoter of the TOL plasmid under limiting concentrations of the XylR activator protein. Gene 172, 81–86.[CrossRef][Medline]

Pingoud, A. & Jeltsch, A. (1997). Recognition and cleavage of DNA by type-II restriction endonucleases. Eur J Biochem 246, 1–22.[Abstract]

Ramos, J. L., Andersson, P., Jensen, L. B., Ramos, C., Ronchel, M. C., Díaz, E., Timmis, K. N. & Molin, S. (1995). Suicide microbes on the loose. Bio/Technology 13, 35–37.[Medline]

Richards, E. (1987). Preparation of genomic DNA from bacteria. Miniprep of bacterial genomic DNA. In Current Protocols in Molecular Biology, pp. 1–2, unit 2.4.1. Edited by F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith & K. Struhl. New York: Wiley.

Ronchel, M. C. & Ramos, J. L. (2001). Dual system to reinforce biological containment of recombinant bacteria designed for rhizoremediation. Appl Environ Microbiol 67, 2649–2656.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74, 5463–5467.[Abstract]

Szafranski, P., Mello, C. M., Sano, T., Smith, C. L., Kaplan, D. L. & Cantor, C. R. (1997). A new approach for containment of microorganisms: dual control of streptavidin expression by antisense RNA and the T7 transcription system. Proc Natl Acad Sci USA 94, 1059–1063.[Abstract/Free Full Text]

Takagaki, Y., Kunugita, K. & Matsuhashi, M. (1973). Evidence for the direct action of colicin K on aerobic 32Pi uptake in Escherichia coli in vivo and in vitro. J Bacteriol 113, 42–50.[Medline]

Torres, B., Jaenecke, S., Timmis, K. N., García, J. L. & Díaz, E. (2000). A gene containment strategy based on a restriction-modification system. Environ Microbiol 2, 555–563.[CrossRef][Medline]

Umeda, M. & Ohtsubo, E. (1991). Four types of IS1 with differences in nucleotide sequence reside in the Escherichia coli K-12 chromosome. Gene 98, 1–5.[CrossRef][Medline]

Wackett, L. P. (2000). Environmental biotechnology. Trends Biotechnol 18, 19–21.[CrossRef][Medline]

Yajima, S., Muto, Y., Morikawa, S., Nakamura, H., Yokoyama, S., Masaki, H. & Uozumi, T. (1993). The three-dimensional structure of the colicin E3 immunity protein by distance geometry calculation. FEBS Lett 333, 257–260.[CrossRef][Medline]

Zarivach, R., Ben-Zeev, E., Wu, N. & 9 other authors (2002). On the interaction of colicin E3 with the ribosome. Biochimie 84, 447–454.[CrossRef][Medline]

Received 1 July 2003; revised 5 September 2003; accepted 9 September 2003.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Torres, B.
Articles by Díaz, E.
Articles citing this Article
PubMed
PubMed Citation
Articles by Torres, B.
Articles by Díaz, E.
Agricola
Articles by Torres, B.
Articles by Díaz, E.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2003 Society for General Microbiology.