A polymerase chain reaction (PCR)-based procedure without any cloning step was developed for a rapid mutagenesis/deletion of chromosomal target genes in Yersinia. For this purpose, a PCR fragment carrying an antibiotic resistance gene flanked by regions homologous to the target locus is electroporated into a recipient strain expressing the highly proficient homologous recombination system encoded by plasmid pKOBEG-sacB. Two PCR procedures were tested to generate an amplification product formed of an antibiotic resistance gene flanked by short (55 bp) or long (500 bp) homology extensions. Using this method, three chromosomal loci were successfully disrupted in Yersinia pseudotuberculosis. The use of this technique allows rapid and efficient large-scale mutagenesis of Yersinia target chromosomal genes.
The availability of the genome sequences of two strains of Yersinia pestis, one of Yersinia enterocolitica and soon one strain of Yersinia pseudotuberculosis opens new perspectives in the field of comparative genomics of this genus. A functional analysis of the newly identified genes within these species will deserve investigation. The methodology classically used in Yersinia to generate knock-out mutants by chromosomal gene replacement requires several subcloning steps into a suicide delivery vector followed by electrotransformation into Escherichia coli and subsequent transfer into Yersinia by electrotransformation or conjugation. This methodology represents a major limitation for large-scale mutagenesis in post-genomic studies. To circumvent this fastidious approach of reverse genetics, we tested a method of inactivation of chromosomal genes that does not require a cloning step.
The methodology is based on the expression of the λ phage red operon encoded by plasmid pKOBEG  and used to promote recombination between the chromosomal region of interest and a polymerase chain reaction (PCR)-generated transformation marker flanked by regions of homology with the target DNA. This method was successfully used for chromosomal gene replacement in E. coli and Aspergillus nidulans. Although most of the molecular biology techniques used in E. coli are applicable to Yersinia, methodological adaptations are often necessary because of the much lower transformation capacity of the latter. In this paper, we describe an adaptation of the λ phage red operon technology for efficient allelic exchange of genes in pathogenic Yersinia.
Materials and methods
Bacterial strains, growth conditions and plasmids
The XL1Blue E. coli strain was used for propagation of the vector pKOBEG-sacB. The Y. pseudotuberculosis serotype I strains used in this study are IP32637 and IP32953.
Bacteria were grown in Luria–Bertani (LB) broth, on LB agar plates, or on LB agar plates without NaCl and supplemented with 10% sucrose. Y. pseudotuberculosis strains were grown at 28°C for 24 h with shaking (liquid media) or for 48 h (solid media). E. coli strains were grown at 37°C or 30°C for 24 h. When necessary, kanamycin (30 µg ml−1) or chloramphenicol (25 µg ml−1) was added to the media.
Plasmid pKOBEG  provided the λ phage redγβα operon  expressed under the control of the arabinose-inducible pBAD promoter. Construction of the recombinant plasmid pKOBEG-sacB was as follows: the sacB gene was amplified from plasmid pCVD442  by PCR using a pair of primers containing the Nde I restriction site at their 5′ extremity (forward, 5′-TTTCATATGGCAACTTTATGCCCATGCAAC-3′, reverse 5′-TTTAAGCTTATAGTTCATATGGGATTCACC-3′). After purification with a PCR purification kit (Qiagen), the 1.8-kb PCR fragment containing sacB was digested by Nde I and ligated to the Nde I site of plasmid pKOBEG. Yersinia recombinants cured of pKOBEG-sacB were selected on LB agar without NaCl and supplemented with 10% sucrose at 28°C.
The PCR primers used in this study were purchased from Genset as deprotected, desalted and quality-controlled by polyacrylamide gel electrophoresis. The PCR amplification reaction mixtures contained 4 U of mixed 3:1 Taq (Roche) and Pfu (Stratagene) polymerases used with the Pfu supplier buffer, 1 µM of each primer, and 100 µM of dNTPs.
PCR synthesis of a kan cassette with short flanking homologous (SFH) regions of the Yersinia target DNA
A 945-bp PCR fragment was generated using pUC4K  as a template. We used two primers composed of 20 nucleotides homologous to the extremities of the kanamycin cassette (3′ extremity) and 55 nucleotides homologous to the Yersinia target sequence (5′ extremity). The PCR program used for SFH-PCR was 95°C for 3 min, followed by five cycles at (i) 95°C for 10 s, ramp of 0.3°C s−1 down to 60°C, (ii) 60°C for 30 s, and (iii) 72°C for 90 s, and finally 20 cycles at (i) 95°C for 10 s, (ii) 53°C for 30 s, and (iii) 72°C for 90 s. PCR products were maintained at 72°C for 3 min.
PCR synthesis of a kan cassette with long flanking homologous (LFH) regions of the Yersinia target DNA
In order to obtain a PCR fragment formed of a kanamycin resistance cassette flanked by long portions (i.e. ~500 bp) of the regions surrounding the target sequence, a modified version of the protocol described by Wach  and Horton et al.  was developed. A three-step PCR procedure was used (Fig. 1). In the first step, the Yersinia genomic DNA was used as a template to amplify the 500-bp regions flanking the target sequence with primer pairs UpF/UpR-kan and DownF-kan/DownR. In each pair, one primer contained at its 5′ end a 20-bp region homologous to the extremities of the kan gene. Simultaneously, plasmid pUC4K was used as a template to amplify the kanamycin cassette with the primer pair kanF/R. In the second step, 100 ng of upstream and downstream target sequence PCR products were mixed with 100 ng of kan PCR fragments, and with the primer pair UpF/DownR. Because of the presence of multiple bands after the second step, a third step was introduced, where 1 µl of the PCR products obtained from the second step was used as a template for a PCR reaction with the primer pair UpF/DownR. The first step of the PCR program was the same as that used for SFH-PCR. The program used for the second and third steps of PCR was 95°C for 3 min followed by 30 cycles at (i) 95°C for 30 s, (ii) 55°C for 30 s, and (iii) 72°C for 2 min. PCR products were maintained at 72°C for 3 min.
Yersinia strains harboring the pKOBEG-sacB plasmid were grown in LB with chloramphenicol at 28°C until OD600~0.2. Expression of the red genes carried by pKOBEG-sacB was induced by adding 0.2%l-arabinose to the medium for 2–4 h. Bacteria that were concentrated 200-fold were rendered electrocompetent by three washing steps in ice-cold 10% glycerol. Electroporation was done as described previously  using various concentrations of PCR products. Shocked cells were transferred to 1 ml LB for 1.5 h at 28°C and selection of transformants was achieved on LB kanamycin agar plates. After primary selection, mutants were grown on LB agar plates without NaCl containing 10% sucrose in order to cure the pKOBEG-sacB plasmid. Mutants were verified for correct disruption by PCRs. The location of the three pairs of primers used to ensure that the linear DNA was inserted at the correct position on the chromosome is shown in Fig. 2.
Results and discussion
It has been recently shown that gene replacement by homologous recombination between a chromosomal gene and its linear copy introduced by electroporation could be efficiently achieved in E. coli and A. nidulans if the Redα, Redβ and Redγ functions were expressed in trans on a plasmid [1,2]. In this study, this technique was adapted for efficient allelic exchange in Yersinia.
In E. coli, plasmid pKOBEG , which carries the λ phage red operon, has a thermosensitive origin of replication, allowing a high yield of pKOBEG-cured bacteria at temperatures ≥37°C. In contrast, this plasmid was not thermosensitive in Yersinia, even when the bacteria were grown at temperatures as high as 42°C. In order to easily recover pKOBEG-cured Yersinia clones, the sacB gene was introduced on this replicon, downstream of the Red function. When grown on agar plates containing sucrose, plasmid-cured Yersinia derivatives were easily selected.
To test the efficiency of allelic exchange between a kanamycin cassette flanked by short sequences of a gene and its chromosomal target, the pKOBEG-sacB plasmid was electroporated into two different Y. pseudotuberculosis strains (IP32637 and IP32953). Four genes corresponding to three different loci were chosen as targets for disruption: (i) recA, (ii) int encoding the integrase of the high pathogenicity island  and (iii) gmd and fcI from the LPS operon  encoding a GDP-mannose-4,6-dehydratase and a GDP-fucose synthetase, respectively. Short flanking regions of 55 bp of these three target sequences were added on each side of the kan cassette by PCR in order to remove most of the coding sequence of each gene or region. Upon repeated electroporations of the SFH-PCR fragment into Y. pseudotuberculosis IP32953 (pKOBEG-sacB) strain gave each time single recombinant bacteria disrupted for the LPS locus (Table 1). The proper insertion of the kan cassette and subsequent deletion of the gmd and fcI region were verified by PCR. Despite several attempts, no recombinant clones could be obtained for the int and recA genes of Y. pseudotuberculosis IP32637 strain. Therefore, although the SFH-PCR technique may work in some instances in Y. pseudotuberculosis, the recovery of recombinant clones is gene- or strain-dependent and is not systematically achieved.
|Target gene or region||Number of confirmed mutants (number of kanamycin-resistant colonies)|
|Target gene or region||Number of confirmed mutants (number of kanamycin-resistant colonies)|
Of 170 kanamycin-resistant recombinants, 10/10 were confirmed to have the correct disruption of the gmd-fcI region.
To increase the chances of homologous recombination between the regions flanking the kanamycin cassette and the target gene, we decided to increase the sizes of the flanking regions. For this purpose, a modified version of a previously described technique [6,7] was developed. In addition to the two sequential PCR steps described in the original technique, a third step was performed to increase the purity and concentration of the PCR product (Fig. 1). Approximately 500 bp of the regions flanking the int, recA and gmd-fcI loci were added on each side of the kan cassette by this three-step PCR procedure. Upon electroporation of the LFH-PCR products into Y. pseudotuberculosis, the three chromosomal targets were successfully disrupted with a transformation efficiency in the range of 9–90 transformants µg−1 of linear DNA. The number of recombinants is given in Table 1. Therefore, the use of a three-step PCR procedure to get a longer homologous region (500 bp) on each side of the kan cassette made it possible to systematically delete the target chromosomal gene and to get a significant increase in the number of recombinants as compared to the SFH-PCR.
However, the PCR technique is known to introduce point mutations in the sequences during the amplification process. This may pose a problem if the amplified sequences flanking the target gene contain open reading frames that may consequently become non-functional. Since the linear product used for LFH-PCR is long and results from several steps of amplification, the risk of introducing unwanted mutations is even higher. A proofreading pfu polymerase was used to prevent this problem. However, to ensure that this enzyme efficiently prevented the appearance of mutations in the region surrounding the target sequence, 500 bp of the chromosome regions, located on each side of the deleted int gene, were sequenced. These sequences were identical to those of the original regions.
These PCR-based procedures for rapid mutagenesis of chromosomal target genes were optimized for Y. pseudotuberculosis because this species does not give a high yield of transformants following electroporation . In Y. pestis, a species that has a higher level of electrocompetence , SFH-PCR and LFH-PCR gave even better results. Several chromosomal genes could be successfully deleted (data not shown).
Altogether, our results demonstrate that large-scale mutagenesis/deletion of Yersinia chromosomal genes can be rapidly and efficiently achieved with the LFH-PCR technique.
This study was funded in part by Contract DGA 99 01 110004709450 from the French Defence Ministry. B.L. had a fellowship from the French Research Ministry.