In the process of handling roughly 300 independent transformed lines obtained by injecting any of several different marker genes, including,
unc-31,
emb-9,
let-2,
par-2,
dpy-10 and
dpy-2 we found only one line that had become genetically linked to a chromosome. Furthermore all of the lines we had analyzed by southern analysis (n=73) appeared to contain hundreds of copies of the injected DNA (including the integrated line). In summary while injecting plasmids, phage and cosmids into the syncytial gonad of C. elegans we have not observed low copy number integrative transformation. Recently while conducting experiments involving the coinjection of plasmids and single stranded oligos (see abstract in this issue) we were surprised to find roller lines which could be readily homozygosed and which contained between one and ten copies of the plasmid sequences. Initially we were injecting N2 animals with a mixture of 80 g/ml of the
rol-6 plasmid, pRF4, and 500 g/ml of a single stranded (
su1006) 50mer, homologous to the plasmid. From 63 injected animals we obtained 188 F1 rollers which in turn established only 6 F2 lines. Three of these lines had integrated copies of the
rol-6 plasmid. Each integrated line contained tandem copies of the plasmid, for a total of two, three and ten copies respectively. When the plasmid concentration was doubled to 160 g/ml we obtained more germ line transformants, however fewer of these proved to be integrated, (three out of thirty five lines from approximately eighty injected animals). Similarly in experiments involving coinjection of a
lin-29 50mer at ~1mg/ml and the
rol-6 plasmid at 160 g/ml, only two out of fifty three F2 lines were integrated lines. Despite the higher concentration of injected plasmid in these two experiments all five integrated lines contained a low copy number of plasmid, two, two, three, three, and one copy(s) respectively. This preliminary data suggests that, above 80 g/ml, plasmid concentration does not limit the frequency of integration events. It is plausible that as plasmid concentration is increased, the concentration of chromosomal target sequences becomes the limiting factor defining integration frequency. Animals from two different lines which carry two or three integrated copies of the
rol-6 plasmid behave like recessive right rollers. They do not roll when crossed to N2. This is in contrast to the phenotype of the allele (
su1006) which does roll as a heterozygote. Perhaps the integrated copies are not expressed as efficiently as their wild type homologs. We found that in a
rol-6 null background (
n1270e187) these lines exhibited complete penetrance. From this data it seems likely that we might miss single copy integrants in an N2 background. In fact the single copy integrated line mentioned above was obtained by injecting the null. In summary we have found that single stranded oligos (perhaps of any sequence or length) promote low copy number chromosomal integration of coinjected plasmids. We hypothesize that they do so by increasing the concentration of target sites at which integration may occur. This might involve pairing with chromosomal sequences, or simply an activation of general DNA repair and recombination. In these experiments we have observed approximately one integrated line from every twenty to thirty injected animal. So far none of the lines which we have tested at the DNA level have integrated plasmid sequences homologously. We are now attempting to look more directly for homologous integration of plasmids using this approach.