Common structural and expressional properties of vertebrate caldesmon genes.

We have determined the genomic structure of chicken caldesmon (CaD) gene. The gene, 100-150 kilobases long, is composed of 17 exons. Exons 1a-1, 1a-2, and 1a-3 encode the 5'-terminal sequence specific to the mRNAs for CaDs expressed in gizzard. Exon 1b encodes the 5'-terminal sequence of the brain l-CaD and locates downstream of exons 1a-1, 1a-2, and 1a-3. The genomic construction of the chicken CaD resembles with that of the human CaD. Exon 3 of chicken CaD gene possesses the unique structure similar to that of human CaD gene; the common domain in both h- and l-CaDs (amino acid residues 74-199 for h-CaD and residues 66-191 for l-CaD) and the central repeating domain specific to h-CaD (amino acid residues 200-419) are encoded in exons 3a and 3b, respectively. Of particular interest is that the two consensus 5'-splice sites are found in the borders between exons 3a and 3b, and exon 3b and intron. Therefore, the expressional regulation between h- and l-CaDs can be explained by selection of these 5'-splice sites. Alternative 3'-splice sites also exist at intron/exon junction of exon 14 and the difference in selection of the sites would induce the specific Ala-508 insertion in the brain l-CaD.