A recent paper in Development by Hahn and colleagues reports that female mice homozygous for a targeted null mutation of the lunatic fringe(Lfng) gene are infertile (Hahn et al., 2005). In 1998, our laboratory and Randy Johnson's laboratory published independent papers on the construction and analysis of Lfngknockout mice (Evrard et al.,1998; Zhang and Gridley,1998). The construct design of the targeted mutant alleles was very similar. In both alleles, the first coding exon of the Lfng gene was deleted, and embryos homozygous for either Lfng mutant allele exhibited severe defects in somite formation. Hahn and colleagues obtained Lfng mice from the Johnson laboratory for studies on oocyte development and fertility. They reported that Lfng-/-female mice did not mate, exhibited disorganized ovarian morphology and were all infertile.
In their paper, Hahn et al. state that we reported previously that the targeted Lfng mutation our group constructed `results in complete embryonic lethality of Lfng-/- offspring'. Indeed, they give that as the reason they obtained the Johnson Lfng mice for their studies on folliculogenesis. However, we stated quite clearly, both in the initial and in subsequent publications on these mice, that a large number of our Lfng-/- null homozygotes are viable into adulthood(Zhang and Gridley, 1998; Zhang et al., 2000; Zhang et al., 2002).
In addition, contrary to the findings of Hahn et al., at least some of our female Lfng-/- null homozygotes are fertile. Although we generally maintain our Lfng line by backcrossing or intercrossing Lfng+/- heterozygous mice, in a number of instances in the past 5 years we have mated Lfng-/- female homozygotes. While not every breeding pair set up becomes productive, many of these matings have proven to be fertile. We have frequently used matings to Lfng-/- females to generate mice for double-mutant studies. For example, we have mated Lfng-/- females to male mice containing mutations in the Notch1, Jag1, Dll3 and Rfng genes. In all of these cases, we had instances of productive matings in which Lfng-/- female homozygotes had multiple litters. Although we do not dispute that there is probably a role for Notch signaling in general, and the Lfng gene in particular, during ovarian follicle development in mice, the paper by Hahn et al. states that Lfng-/- female homozygotes are all infertile. Our results unequivocally demonstrate that, at least with respect to our Lfngmutant allele, this assertion is not true.
What then are the possible reasons for the discrepancy between the fertility of Lfng-/- female mice of the two groups? One possibility is that the two Lfng targeted alleles are functionally different. This is possible, but seems quite unlikely given the similar designs of the two Lfng alleles and the essentially identical somite defects exhibited by the null homozygotes(Evrard et al., 1998; Zhang and Gridley, 1998). We suggest that more likely possibilities include differences in genetic background, and/or differences in mouse husbandry and colony conditions. However, the assertions of Hahn et al. that mice homozygous for our Lfng targeted mutation cannot survive postnatally and that Lfng-/- female mice are obligatorily infertile are incorrect.
