Fig. 1. Schematic representation of the targeting strategy used to generate Dsc3^+/– ES cells. The map is not drawn to scale. (A) The 5' end of the Dsc3 gene is shown. Exons are represented by vertical boxes. The location of the start codon (ATG) is indicated. Horizontal bars indicate the positions of the external probes used to identify targeted ES cells in Southern blots. (B) Structure of the Dsc3 gene targeting vector. A fragment containing promoter elements, exon 1 and part of intron 1 was deleted and replaced by a neomycin-resistance mini gene (PGK-Neo) which was flanked by loxP sites (triangles). (C) Structure of the recombinant Dsc3 locus after homologous recombination. (D) Dsc3-null allele after excision of the PGK-Neo cassette via Cre-mediated recombination (see text for details). (E) EcoRV-digested genomic DNA from Dsc3 line 1 mice was tested with the external probes. The 5' probe detected a 13.5 kb wild-type fragment and a 4.3 kb mutant fragment in Dsc3^+/– DNA. The fragment sizes for the 3' probe were 13.5 kb (wild type) and 6.7 kb (mutant). The neo probe detected a single band of 6.7 kb. (F) Western blot analysis of ear lysates from wild-type (+/+) and adult Dsc3^+/– (+/–) mice using antibodies against Dsc3, plakoglobin (Pg) and ß-catenin. A quantitative analysis of the Dsc3 signals (using the ß-catenin signal as a loading control) revealed that Dsc3 synthesis was reduced by half in the Dsc3^+/– sample. This was particularly obvious when comparing the intensities of the Dsc3b bands in wild-type and mutant mice. (G) QPCR analysis of Dsc3 mRNA levels in wild-type (WT) and adult Dsc3^+/– (Het) ear. Two pairs of heterozygous mutants and wild-type littermates were used in this analysis. Dsc3 expression was approximately halved in ear.