Sequence of normal canine COL1A1 cDNA and identification of a heterozygousalpha 1(I) collagen Gly208Ala mutation in a severe case of canine osteogenesis imperfecta

The sequence of canine COL1A1 cDNA was determined from four overlapping COL 1A1 RT-PCR products generated from canine fibroblast RNA. In the translated region, nucleotide identity between canine and human COL1A1 cDNA was 93.2% , although the canine sequence lacked nucleotides 204 to 215 in the region coding for the N-propeptide. Amino acid identity was 97.7%. Total RNA and t ype I collagen were collected from cultured skin fibroblasts of a 12-week-o ld male golden retriever with pathologic fractures suggestive of osteogenes is imperfecta (OI) and dentinogenesis imperfecta. Sequential, overlapping s imilar to 1000-bp fragements of COL1A1 and COL1A2 cDNA were each amplified by RT-PCR using primers containing 5' T7 polymerase sites. These PCR produc ts were transcribed with T7 RNA polymerase, hybridized into RNA duplexes, a nd cleaved at mismatch sites with RNase. The proband had an unique cleavage pattern for the fragment of COL1A1 mRNA spanning nucleotides 709 to 1531. Sequence analysis identified a G to C point mutation for nucleotide 1276, p redicting a codon change from glycine (GGA) to alanine (GCA) for amino acid 208. This change disrupts the normal Gly-X-Y pattern of the collagen tripl e helix. Restriction enzyme digestion of the RT-PCR product was consistent with a heterozygous COL1A1 mutation. Type I collagen was labeled with H-3-p roline, salt precipitated, and analyzed by SDS-PAGE. Pepsin digested cw cha ins were over-hydroxylated, and procollagen processing was delayed. Thus, c anine and human OI appear homologous in terms of clinical presentation, eti ology, and pathogenesis. (C) 2000 Academic Press.