Multiple mRNA species code for two non-allelic forms of ovine alpha s2-casein.

The two non-allelic forms of alpha s2-casein, occurring in ovine milk, differ by an internal deletion of nine amino acid residues, including both cysteine residues at positions 34 and 42 in the mature chain. Sequencing of several alpha s2-casein cDNA, isolated from the mammary cDNA library of a single lactating ewe, showed three new types which differed from that previously studied. In addition to the expected deletion of codons +34 to +42 affecting 30-40% of mRNA, another structural difference involving an internal stretch of 44 nucleotides in the 5' untranslated region, was found. S1-nuclease protection assays confirmed the existence of several types of the relevant mRNA and sequencing of in-vitro-amplified genomic DNA demonstrated the presence of the 44-nucleotide stretch in the alpha s2-casein transcriptional unit, thus ruling out the possibility of a cloning artefact. The different alpha s2-casein mRNA, four in terms of deletion and two in terms of nucleotide substitutions for a given ewe, can be readily explained by partial exon skipping and allelic differences, respectively. This assumption is well supported by the following observations: 5' and 3' ends of both deleted DNA fragments are similar to those of exons; sequences neighbouring the 44-nucleotide stretch of the genomic DNA perfectly match consensus sequences described for 3' and 5' ends of introns; the rather simple patterns observed on Southern blots of different enzymatic digests of genomic DNA strongly suggest the occurrence of only 1 copy alpha s2-casein gene/haploid genome. During the course of evolution, the alpha s2-casein-encoding gene has undergone many mutations and some of them might have occurred in regions corresponding to consensus splicing regions of the pre-mRNA. Thus, complete skipping of some exons might be responsible for the shorter sizes of rat and mouse alpha s2-casein mRNA. If so, the overall organization of the alpha s2-casein gene in the different species might be more similar than expected from structural comparisons of the cognate mRNA or caseins.