A novel allelic donkey b-Lg I protein isoform generated by a non-AUG translation initiation codon is associated with a non-synonymous SNP

β-lactoglobulin I (β-Lg I) is one of the most important whey proteins in donkey milk. However to our knowledge, there has been no study focusing on the full nucleotide sequences of this gene (BLG I). Current investigation of donkey BLG I gene is very limited with only two variants (A and B) characterized so far at the protein level. Recently, a new β-Lg I variant, with a significantly higher mass (+1915 Da) than known variants has been detected. In this study, we report the whole nucleotide sequence of the BLG I gene from two donkeys, whose milk samples are characterized by the β-Lg I SDS-PAGE band with a normal electrophoretic mobility (18,514.25 Da, β-Lg I B1 form) the first, and by the presence of a unique β-Lg I band with a higher electrophoretic mobility (20,428.5 Da, β-Lg I D form) the latter. A high genetic variability was found all over the two sequenced BLG I alleles. In particular, 16 polymorphic sites were found in introns, one in the 5’ flanking region, three SNPs in the 5’ untranslated region (UTR) and one SNP in the coding region (g.1871G/A) located at the 40th nucleotide of exon 2 and responsible for the amino acid substitutions p.Asp28>Asn in the mature protein. Two SNPs (g.920-922CAC>TGT and g.1871G/A) were genotyped in 93 donkeys of two Italian breeds (60 Ragusana and 33 Amiatina, respectively) and the overall frequencies of g.920-922CAC and g.1871A were 0.3065 and 0.043, respectively. Only the rare allele g.1871A was observed to be associated with the slower migrating β-Lg I. Considering this genetic diversity and those found in the database, it was possible to deduce at least five different alleles (BLG I A, B, B1, C, D) responsible for four potential β-Lg I translations. Among these alleles, B1 and D are those characterized in the present research, with the D allele of real novel identification. Haplotype data analysis suggests an evolutionary pathway of donkey BLG I gene and a possible phylogenetic map is proposed. Analyses of mRNA secondary structure showed relevant changes in the structures, as consequence of the g.1871G>A polymorphism, that might be responsible for the recognition of an alternative initiation site providing an additional signal peptide. The extension of 19 amino acids sequence to the mature protein, corresponding to the canonical signal peptide with an additional Alanine residue, is sufficient to provide the observed molecular weight of the slower migrating β-Lg I encoded by the BLG I D allele