Mutagenesis of Herpesvirus BACs by Allele Replacement

For mutagenesis of BAC-cloned herpesvirus genomes, we have adapted a two-step replacement procedure originally described by O’Connor et al. ( 1 ) for the manipulation of large DNA segments in Escherichia coli (E. coli) . In principle, the mutant allele to be introduced is provided on a so-called shuttle plasmid. The mutant allele has to be flanked by regions homologous to the desired integration site of the mutation in the BAC (regions A and B, see Fig. 1 , step 1). By homologous recombination in E. coli , the shuttle plasmid will completely insert into the BAC, leading to a cointegrate ( see Fig. 1 , step 2). As a consequence, the cointegrate carries the wild-type, as well as a mutant allele, and a duplication of the homologous sequences flanking the mutation and the wild-type locus. The cointegrate can spontaneously resolve by homologous recombination via regions A or B giving rise to either the wild-type herpesvirus BAC plasmid or a mutant BAC, depending on which of the homologous regions (A or B) is used for recombination ( see Fig. 1 , step 3). The propensity for resolution of the cointegrate goes up with increasing sizes of the homologous sequences. However, even if homologous sequences of 2 or 3 kbp are provided, resolution of the cointegrates remains a rather rare event. Therefore, we have introduced a negative selection marker into the shuttle plasmid (the sacB gene) that allows to select against the bacteria still containing a nonresolved cointegrate ( see Fig. 1 , step 4) and to identify bacterial clones harboring the mutant or the parental BAC.

---