Results Isolation and characterization of Capn2 targeted ES cell clones Two independent Capn2+/- ES cell lines, designated ES27 and ES36, were isolated from a screen of 305 drug-resistant clones. Correct targeting of the Capn2 locus was established both by Southern blot hybridization and PCR analysis. A probe located outside the short (upstream) arm of homology hybridized to a 3.5-kb BamHI fragment of the wild-type allele and 5.3-kb BamHI fragment of the mutant allele as predicted from genomic maps (Figure 2A). The same probe also detected the expected 7.2-kb wild-type and 6.4-kb mutant BglII fragments, 4.9-kb wild-type and 5.7-kb mutant NcoI fragments, as well as 7.2-kb wild-type and 4.9-kb mutant BglII/AgeI fragments (not shown). A probe derived from the PGK-Neo cassette recognized only the 5.3-kb BamHI fragment in Capn2+/- ES cells, suggesting that the targeting vector had integrated solely at the Capn2 locus (not shown). A PCR screening method was also established that generated a wild-type product of 2,749 bp and a 2,711 bp product from the mutant allele. The 2,711 bp product was only evident in the two targeted cell lines (Figure 2B). Figure 2 Characterization of targeted ES cell lines. (A) Targeted disruption of the Capn2 locus was detected initially by Southern blotting. Membranes were blotted with BamHI-digested genomic DNA extracted from ES cells and hybridized with a DIG-labeled 823 bp BamHI/HindIII fragment located immediately upstream of the short arm of the targeting vector (Figure 1). A 3.5-kb BamHI fragment corresponding to the wild-type allele was present in all cells, whereas a 5.3-kb fragment from the mutant allele was detected in two targeted cell lines, designated ES27 and ES36. (B) PCR genotyping was carried out with two separate reactions designed to amplify either a 2,748 bp segment from the wild-type allele or a 2,711 bp segment from the mutant allele. Both reactions used a common sense primer located in intron 4, outside the short arm of the targeting vector, and distinct allele-specific antisense primers. The reaction to detect the wild-type allele used an antisense primer located in exon 7 while the amplification of the mutant sequence was done with an antisense primer in the PGK-Neo cassette. The results confirm the presence of the wild-type allele in all cells, whereas the mutant allele signal was observed only in the two targeted clones. (M) denotes the molecular weight marker. Generation of chimeric mice and germline transmission of the mutant Capn2 allele Eight chimeric male mice were produced in morula aggregation experiments using the Capn2+/- ES27 cell line. Two of these males transmitted the Capn2+/- ES27 genotype through the germline into the F1 generation. Heterozygous Capn2+/- animals appeared normal, with no obvious defects in gross anatomy, reproduction, or life span. Out of 199 weanlings from heterozygous intercrosses, no Capn2-/- progeny were detected (Table 2). We did not observe high rates of perinatal death, and no Capn2-/- stillborns were observed. This suggested that Capn2-/- animals perished at some stage during embryonic development. In an attempt to determine if embryonic death occurred at a post-implantation stage, embryos were harvested for genotyping at different times between E10.5 and E18.5. No Capn2-/- embryos were observed and no signs of embryo resorption were detected (Table 2). This indicated that the Capn2-/- embryos might be dying prior to implantation. Embryos were then flushed from the oviducts of pregnant females at E2.5 or E3.5, and genotyped by means of a nested PCR strategy (Figure 4). Two of 90 successfully genotyped pre-implantation embryos were Capn2-/-, (Table 2; Figure 5). Both of these Capn2-/- embryos were isolated at the 8-cell stage and did not display any obvious morphological defects. None of the 46 successfully genotyped blastocyst-staged embryos were Capn2-/-. The scarcity of Capn2-deficient embryos surviving to the 8-cell stage suggested that the loss of m-calpain activity must fatally compromise the viability of early embryos. Furthermore, it is possible that persistence of some maternally derived mRNA transcript or protein might have allowed a small number of Capn2-/- embryos to survive to the morula-stage. Table 2 Genotype distribution of offspring derived from Capn2 transgenic mice. Cross (N) Age Genotype +/+ +/- -/- ND* (+/-)♀ × (+/-)♂ 199 3 weeks 23 176 0 0 (+/+)♀ × (+/-)♂ 122 3 weeks 50 72 0 0 (+/-)♀ × (+/+)♂ 103 3 weeks 28 75 0 0 (+/-)♀ × (+/-)♂ 9 E18.5 0 7 0 2 (+/-)♀ × (+/-)♂ 5 E17.5 0 5 0 0 (+/-)♀ × (+/-)♂ 9 E14.5 2 7 0 0 (+/-)♀ × (+/-)♂ 10 E11.5 3 7 0 0 (+/-)♀ × (+/-)♂ 7 E10.5 2 5 0 0 (+/-)♀ × (+/-)♂ 46 E3.5 7 39 0 0 (+/-)♀ × (+/-)♂ 48 E2.5 8 34 2 4 * ND, not determined Figure 4 Nested PCR strategy for genotyping of pre-implantation embryos. Due to the scarcity of extractable genetic material, a nested PCR strategy was developed in order to genotype pre-implantation embryos. Separate reactions were used to amplify a 429 bp fragment from the wild-type allele and a 389 bp segment from the mutant allele, both spanning the 3' end of the short (upstream) arm of the targeting vector. In both reactions, a 213 bp sequence located within the short arm was co-amplified with the 'diagnostic' products as an internal control. The same sense primers were used to amplify 'diagnostic' sequences in both reactions, whereas the antisense primers were allele-specific. The primers, represented by triangles, are depicted in two (nested) sets for each of the three reactions. Exons are represented by open, vertical rectangles, the PGK-Neo cassette by an open, horizontal rectangle, while thin vertical lines denote the boundaries of the short arm and the 5' end of the long (downstream) arm. A grey, horizontal rectangle delineates the segment of the wild-type allele that is replaced by the PGK-Neo cassette in the mutant allele. Figure 5 Genotyping of pre-implantation embryos. A nested PCR strategy was used to genotype embryos prior to implantation as described in Figure 4. Capn2+/- mice were mated and the date of fertilization established by the appearance of vaginal plugs. Blastocyst (E3.5) or 8-cell embryos (E2.5) were flushed from the oviducts and then digested with proteinase K. In separate reactions segments found exclusively in either the wild-type or mutant alleles were co-amplified with an internal control sequence, located in the short (upstream) arm of the targeting vector, which is found in both alleles. The final products were 429 bp for the wild-type allele, 389 bp for the mutant allele, and 213 bp for the internal control. (A) A representative example of the genotyping of blastocyst stage embryos. Embryos #1, 2, 4, 5, and 6 were Capn2+/- whereas embryos #3 and #7 were Capn2+/+, denoted by the absence of the 389 bp mutant signal. (B) An example of the genotyping of 8-cell embryos is shown. Embryos #1, 2, 3, 5, and 6 were Capn2+/- while embryo #4 was Capn2-/-, marked by the absence of the 429 bp wild-type signal. (M) denotes the molecular weight marker. The genotyping results for weanlings, post-implantation, and pre-implantation embryos are shown in Table 2. Curiously, the ratio of Capn2+/+ to Capn2+/-animals from inbred heterozygous intercrosses was substantially less than the predicted 1:2 Mendelian ratio. In a group of 199 animals derived from heterozygote breeding (33 litters), 23 Capn2+/+ (11.6%) and 176 (88.4%) Capn2+/- animals were observed. The ratio of Capn2+/+ to Capn2+/- animals among males (14% to 82%) or females (13% to 90%) was essentially the same is it was for the combined population, and there were an average of six pups per litter, which is normal for this background strain. A larger than expected proportion Capn2+/- animals was also evident in both post-implantation and pre-implantation embryos (Table 2). Interestingly, a similar over-representation of heterozygous animals was also reported in one of the Capn4 transgenic lines, though the genotype skewing was not as extreme [32]. Crosses between Capn2+/+ and Capn2+/- animals also resulted in a greater than expected proportion of Capn2+/- animals (Table 2). An even higher degree of Capn2+/- over-representation was seen when the mutant allele came the mother (73%) compared to when it came from the father (59%). In these crosses the ratios of Capn2+/+ to Capn2+/- animals among males or females compared well with the ratio in the combined populations; 77% of males and 69% of females were Capn2+/- when the mutant allele came from the mother, and 62% of males and 55% of females were Capn2+/- when the mutant allele came from the father. Attempted generation of Capn2-/- ES cells Capn2+/- ES cells were subjected to clonal selection in the presence of 2 mg/mL G418 in attempts to generate homozygous mutant cells by gene conversion. This procedure has been extensively applied to targeted ES cells and was readily achieved in the case of Capn4+/- ES cells [31]. In this case, however, no Capn2-/- ES cells were isolated in screens of 126 drug-resistant clones. The inability to isolate Capn2-/- ES cells, in concert with the absence of Capn2-/- embryos beyond the 8-cell stage, suggested that m-calpain activity might be essential for cell viability or the establishment of viable ES cell clones.