Complete CSN1S2 characterization, novel alleles identification and association with milk fatty acids composition in river buffalo

The αs2-casein is one of the phosphoproteins secreted in all ruminants milk and it is the most hydrophilic of all caseins. However, this important gene (CSN1S2) has not been characterised in detail in buffaloes with only two alleles detected (reported as alleles A and B) and no association studies with milk traits have been carried out unlike what has been achieved for other species of ruminants. In this study, we sequenced the whole gene of two Mediterranean river buffaloes homozygotes for the presence/absence of the nucleotide C (g.7539G>C) realised at donor splice site of the exon 7 and, therefore, responsible for the skipping of the same exon at mRNA level (allele B). A high genetic variability was found all over the two sequenced CSN1S2 alleles. In particular, 74 polymorphic sites were found in introns, 6 in the promoter and 3 SNPs in the coding region (g.11072C>T, g.12803A>T and g.14067A>G) with 2 of them responsible of amino acid replacements. Considering this genetic diversity, those found in the database and the SNP at donor splice site of the exon 7, it is possible to deduce at least eight different alleles (CSN1S2 A, B, B1, B2, C, D, E and F) responsible of seven different possible translations of the buffalo αs2-casein. Haplotype data analysis suggests an evolutionary pathway of buffalo CSN1S2 gene consistent with our proposal that the published allele CSN1S2 A is the ancestral αs2-CN form and the B2 probably arises from interallelic recombination (single crossing) between the alleles D and B (or B1). The allele CSN1S2 C is of new identification, while CSN1S2 B, B1, B2 are deleted alleles because all characterized by the mutation g.7539G>C. Two SNPs (g.7539G>C and g.14067A>G) were genotyped in 747 Italian buffaloes and major alleles had a relative frequency of 0.83 and 0.51, respectively. An association study between these SNPs and milk traits including fatty acids composition was carried out. The SNP g.14067A>G showed a significant association (P