Existing methods for site-directed plasmid mutagenesis are restrained by the small spectrum of modifications that can be introduced by mutagenic primers and the amplicon size limitations of in vitro DNA synthesis. As demonstrated here, the combined use of Red/ET recombination and unique restriction site elimination enables extensive manipulation regardless of plasmid size and DNA sequence elements. First, a selectable marker is PCR-amplified with synthetic primers attaching 50-bp homology target flanks for Red/ET recombination and an arbitrary restriction site absent in the substrate plasmid. The resulting cassette is co-electroporated with substrate plasmids in Red/ET-proficient Escherichia coli cells. Following isolation of recombinant plasmids, linear nonselectable DNA replaces the cassette and introduces the desired mutation(s) in a second Red/ET recombination step. Upon selective digestion of parental plasmids and retransformation, a 38% mutation efficiency was achieved using a synthetic 97-nucleotide oligonucleotide to cure a 17-bp deletion within lacZα of pUC19 (2,686 bp). A PCR fragment was used with similar efficiency to co-replace mouse Cdkn1b codons 9 and 76 in gene-targeting vector pGTC (13,083 bp).

Introduction

Site-directed mutagenesis (SDM) allows DNA modifications by extending oligonucleotide primers which contain the desired mutation(s) flanked by bases complementary to target sequences (1). Several protocols exist in which PCR-and non–PCR-based methods are used (2,3,4,5,6,7,8). However, SDM is restrained by the fidelity and template size limitations of in vitro DNA synthesis and the success rate for mutagenesis decreases with the number of mismatches introduced by the mutagnenic primer(s). Hence, an amplicon size–independent method with a broad mutational spectrum would be very useful.

Red/ET recombination has come to be recognized as superior technology for the size and sequence independent manipulation of DNA in Escherichia coli (reviewed in Reference 9). The phage λ–derived technology relies on ssDNA or dsDNA fragments with ~50-base tails homologous to the target region. No specific recombination sites are required and homology tails of any sequence can easily be integrated into synthetic oligos either used to PCR-amplify dsDNA flanked by homology or for oligonucleotide-directed recombination. Oligonucleotides complementary to the replicon's lagging strand recombine most efficiently (see Reference 9, and references therein). The Red/ET machinery is composed of exonuclease Redα, which processes linear dsDNA and provides 3′ ss overhangs, and Redβ, which mediates strand annealing and exchange reactions starting from ssDNA extremities. To stabilize dsDNA substrates in the cell, Red/ET recombination is assisted by Redγ, an inhibitor of the host ExoV (see Reference 9, and references therein). For ease of stringent expression control and transfer between strains, we fused the red genes to the arabinose-inducible pBAD promoter on thermoreplicative plasmid pRed/ET, which is maintained at ~30°C and lost at temperatures ≥37°C (10).

In genetically stable E. coli cells, the recombineering efficiency is ~10⁻⁴ for dsDNA [e.g., drug resistance (drug^R) cassettes and ~10⁻³ for oligonucleotides complementary to the replicon's lagging strand] (9). Hence, laborious colony screening is required to detect such an event. Counter-selectable (CS) cassettes have been successfully used in single copy replicons to select for seamless mutants that have undergone recombination leading to its loss (11,12,13,14). However, recombineering (recombination-mediated genetic engineering) in a multiple-copy situation results in a mixture of mutated and parental plasmids (15), and phenotypic CS markers are not reliably applicable because a CS gene on a single parental plasmid copy can be sufficient for cell death. In addition, plasmid recombineering facilitates the formation of multimers (15,16,17). Co-electroporation of modifying linear DNA and substrate plasmids into Red/ET proficient cells was shown to minimize, but not to eliminate, the formation of higher-weight species (15,16,17).

Here we report a method based on recombineering and unique restriction site elimination (USE), which allows for the efficient and extensive modification of plasmids regardless of size or sequence requirements. drug^R cassettes were PCR-amplified with primers attaching (i) linearization sites (LS) absent in the substrate plasmids, and (ii) specific 50-bp target flanks for recombination. Substrate plasmids and drug^R-LS cassettes were co-electroporated in Red/ET-proficient E. coli cells. Upon isolation of recombinant plasmids, modifying ssDNA and dsDNA was used to replace the drug^R-LS cassettes in a second recombineering step, rendering the mutated plasmids immune to cleavage by USE. Following selective in vitro digestion of parental plasmids and retransformation, a ~38% mutation efficiency was achieved for target plasmids of ≤13 kb.

Materials and methods

E. coli strain and culture conditions

E. coli HS996 (Table 1) was aerobically propagated on Luria-Bertani (LB) broth and agar. As necessary, ampicillin (Ap), chloramphenicol (Cm), kanamycin (Km), and tetracycline (Tc) were added to final concentrations of 100, 50, 50, and 3 µg/mL, respectively. X-gal (5-bromo-4-chloro-3-indolyl-β-D-galacto-pyranoside; 40 µg/mL) was added to the corresponding plates. Electrocompetent cells for retransformation were prepared as described (18).

Table 1. Strains, Plasmids, and Oligonucleotides Used in This Study (Click to enlarge)

DNA manipulations

Standard protocols were used for conventional in vitro cloning (19). HPLC-purified oligonucleotides were obtained from BioSpring (Frankfurt, Germany); drug^R cassettes A002 (km^R) and A006 (cm^R) were purchased from Gene Bridges (Heidelberg, Germany). PCR reactions were performed with Triple Master DNA polymerase according to the manufacturer's recommendations (Eppendorf, Hamburg, Germany). Prior to electroporation, PCR fragments were treated with DpnI and purified using the MinElute PCR purification kit (Qiagen, Hilden, Germany). The QIAprep Spin Miniprep Kit (Qiagen) was used for plasmid isolation. DNA sequencing was done at GATC Biotech (Konstanz, Germany). λ-Red recombination was performed as described by Gene Bridges with minor modifications. In brief, 1.4 mL Red/ET-proficient culture was grown at 30°C to an OD_{650 nm} of ~0.3. Transient expression of the pRed/ET-encoded red genes was induced by adding 50 µL of 10% (w/v) L-arabinose followed by a temperature increase to 37°C. After 25 min, the cells were washed twice with ice-cold 10% (v/v) glycerol and electroporated with DNA in a chilled 1-mm cuvette.