The goal of this experiment is to detect homologous recombination between a transgenic sequence and its endogenous chromosomal counterpart. My experimental strategy was to have the desired recombination generate a wild-type
let-2 gene so that
let-2 +recombinants could be selected from a large population of
let-2 mutants. To do this, initial transformant lines were established by transforming
let-2 (
g37ts)animals with either of two
let-2 plasmids (
plet-2 or
plet-2d) using
prol-6 as a transformation marker.
plet-2 and
plet-2d contain only the 3' end of the
let-2 gene and span the site of the
g37 mutation;
plet-2d is additionally marked with a silent polymorphism that removes a DraIII site near the site corresponding to the
g37 mutation but does not alter the
let-2 protein coding sequence. As expected, these Rol transformants carried multiple copies of the injected sequences as extrachromosomal arrays, and they had a temperature-sensitive Let phenotype because neither the array nor the chromosomal locus carried a wild-type
let-2 copy. Large populations were grown from the initial transformants at permissive temperature, exposed to gamma-irradiation (to induce recombination, 6000 rad), and
let-2 +F1 animals were isolated by selecting for growth at the restrictive temperature. Two types of recombinants were found at a combined frequency of 5x10 +E-5. The first type results from the homologous integration of most or all of the extrachromosomal array at the
let-2 locus; there are five strains of this type. These recombinants have a Let-2 +Rol phenotype. Genetic mapping experiments show that the Let-2 +and Rol activities in these animals are located at or near the
let-2 locus on the X chromosome. Southern blotting provides molecular evidence that there has been an insertion at the
let-2 locus; these strains have a novel band instead of the wild-type band when
let-2 flanking sequence is used as a probe. All five strains have integrated most or all of the extrachromosomal array, since quantitative Southern blots show that there are about 500 plasmid copies contained in the insertion. The simplest model that explains these results is that the extrachromosomal array is circular, and that homologous recombination with the
let-2 locus results in the formation of a wild-type
let-2 copy and the integration of the entire extrachromosomal array. The second type of recombinant results in the replacement of the
let-2 chromosomal sequence with
let-2 transgenic sequence; there are six strains of this type. These recombinants have a Let-2 +non-Rol phenotype. The genetic position of the Let-2 +activity in these strains is located at or near the
let-2 locus. Southern blotting analysis of the four recombinants derived from
plet-2d transformation shows that the
plet-2d transgene was used as a template for replacing the chromosomal
let-2 sequence, because the chromosomal locus contains a DraIII polymorphism that originated from the
plet-2d transgene. A likely explanation for the second type of recombinant is that it arises from gene conversion, in which the chromosomal
let-2 locus copies sequence information from the
let-2 transgene. A process similar to gene conversion is likely to be responsible for healing the double stranded broken ends left by transposon excision (Plasterk, EMBO vol.10 p.1919, 1991). The template used for healing can originate from an extrachromosomal array (Plasterk and Groenen, WBG, 12 44, 1991). The recombination observed in this report does not require a transposon insertion at the target locus, so that the strategy adopted here may be more generally applicable. Although the frequency of homologous recombination is low (about 5x10 +E-5),it should be possible to devise a selection strategy that will identify these rare animals. With this selection strategy, it should be possible to use targeted gene disruption and site-directed mutagenesis to study the function of cloned genes.