Fig. 1

Analysis of Rhbdf1-deficient mice suggests that the targeting strategy might affect the phenotype of homozygous mutant mice. a Schematic representation of exons (filled boxes) and introns (lines) in the Rhbdf1 gene. Previous studies have produced conflicting results regarding the consequences of Rhbdf1 inactivation in mice. Christova et al. reported that Rhbdf1^−/− mice generated using the VelociGene definitive-null gene KO strategy suffer from multi-organ pathologies and die within ~ 6 weeks (A1) and that the Rhbdf1^−/−Rhbdf2^−/− (A2) double KO mice, generated using A1 Rhbdf1^−/− mice, died prenatally with phenotypes that were significantly more severe. In contrast, Li et al. reported that Rhbdf1^−/− homozygous mutant mice generated with a targeted KO-first allele showed no evidence of pathological phenotypes (A3). Moreover, Rhbdf1^−/−Rhbdf2^−/− double mutant mice, generated using A3 Rhbdf1^−/− mice, exhibited less severe disease phenotypes, such as heart valve defects and an “open eyelids at birth” phenotype (A4). b We hypothesized that the targeted KO-first allele (A3), but not the definitive-null allele, generates an N-terminally truncated RHBDF1 protein product from the mutant mRNA that is not degraded by the NMD mechanism, to rescue the overt phenotype observed in RHBDF1-deficient mice (A1). To test our hypothesis, we independently generated Rhbdf1 homozygous mutant mice using two different strategies: (1) delete complete coding sequence (A5) and (2) delete exon(s) containing the start codon (A7). We postulated that whereas the A5 allele would be analogous to the A1 definitive-null allele, the A7 allele would generate a functional truncated mutant RHBDF1 protein, analogous to the targeted KO-first mutant allele A3, rescuing the multi-organ pathology of RHBDF1-deficient mice (A1). c Table indicating the phenotypes of previously generated Rhbdf1 single mutant and Rhbdf1:Rhbdf2 double mutant mice