Sequence and phosphorylation level determination of two donkey b-caseins by mass spectrometry Vincenzo Cunsolo, Elisa Cairone, Rosaria Saletti, Vera Muccilli and Salvatore Foti * Dipartimento di Scienze Chimiche, Universita ` degli Studi di Catania, Viale A. Doria 6, I-95125 Catania, Italy Received 29 January 2009; Revised 8 April 2009; Accepted 12 April 2009 Two coeluting components, with experimentally measured M r values of 25529 and 24606 Da, were identified by reversed-phase high-performance liquid chromatography (RP-HPLC) and mass spectro- metric analysis in the dephosphorylated casein fraction of a milk sample collected from an individual donkey belonging to the Ragusano breed of the east of Sicily. By coupling enzymatic digestions, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and RP-HPLC/nano-electrospray ionization tandem mass spectrometry (nESI-MS/MS) analysis, the two proteins were identified as donkey b-CNs and their sequences characterized completely, using the two known b-CNs from mare as references. The two donkey b-CNs, showing a mass difference of 923 Da, differ by the presence of the domain E 27 SITHINK 34 in the full-length component (M r 25529 Da). In comparison with the mare’s b-CNs used as reference, they present nine amino acid substitutions: L!S 37 ,R!H 52 ,S!N 81 ,P!V 84 ,L!V 91 ,R!Q 203 ,P!L/I 206 ,L!F 210 and A!P 219 . Together, these substitutions account for the increase of 18 Da in the M r of the donkey b-CNs with respect to the counterparts from the mare. The molecular mass determination by ESI-MS for the phosphorylated proteins showed that the full-length component was composed of highly multi- phosphorylated isoforms with five to seven phosphate groups. By analogy with the homologous mare’s b-CNs, the full-length (226 amino acids) b-CN was termed variant A, whereas the shorter (218 amino acids) b-CN was termed variant A D5 . Copyright # 2009 John Wiley & Sons, Ltd. Several clinical studies suggest that donkey’s milk could represent a safe and alternative food to cow’s milk for infants affected by cow’s milk protein allergy. 1–3 Although the molec- ular basis of the reduced hypoallergenicity of donkey’s milk with respect to cow’s milk is still unknown, it is reasonable to hypothesize that the hypoallergenic properties of donkey’s milk can be related to structural differences between their protein components with respect to the counterparts of bovine milk. Equine milk presents remarkable differences when com- pared to bovine milk. The main difference concerns the whey proteins:caseins (CNs) ratio in the mature milk of both species. In fact, in equine milk this ratio is close to 50:50, 4,5 whereas, in bovine milk, whey proteins represent only 20% of the protein fraction. Unfortunately, at the present time equine milk (Equus caballus and Equus asinus, Perissodactyla) has been less studied than bovine milk and limited data are available for its genetic polymorphism. The major informa- tion is available for whey proteins, 6–8 whereas the investi- gation of casein components is still at a relatively early stage of progress. In particular, the complete amino acid sequence of a k-CN 9 and small fragments identified as potential equine a s2 -CN regions 7,10 have been characterized from mare’s milk. In addition, the complete cDNA sequences of mare’s a s1 -CNs (GenBank Accession Nos. AAL05435; AAK83668) and k-CN (AAK83669) have been achieved. 11,12 The a s1 -CN primary structure displays polymorphic patterns, due to alternative splicing processes leading to casual exon skipping events involving exons 7 and 14. 7 More recently, the primary structure of b-CN from mare’s milk has been characterized (nrNCBI Acc. N. Q9GKK3). 13 By comparison with the previously reported cDNA deduced sequence (GenBank Acc. No. AAG43954), 12 direct charac- terization showed that it is constituted of 226 amino acid residues and contains the insertion of a region encoded by exon 5 (residues 27–34: ESITHINK), which is skipped in the case of the AAG43954 protein by alternative splicing. 7 The full-length (226 residues) b-CN was named A variant, whereas the b-CN lacking the region encoded by exon 5 was termed A D5 variant. Conversely, to date knowledge of donkey’s CNs is scant and only the primary structures of k- casein (GenBank Acc. No. EU448385) and a s2 -casein (GenBank Acc. No. FM946022), deduced from corresponding cDNA, have been reported. 14,15 With the aim of improving the knowledge of the com- ponents of donkey’s caseins, we report here the identification and direct sequence characterization of the primary structure of donkey’s b-CN by coupling reversed-phase high-per- formance liquid chromatography (RP-HPLC) analysis, RAPID COMMUNICATIONS IN MASS SPECTROMETRY Rapid Commun. Mass Spectrom. 2009; 23: 1907–1916 Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.4087 *Correspondence to: S. Foti, Dipartimento di Scienze Chimiche, Universita ` degli Studi di Catania, Viale A. Doria 6, I-95125 Catania, Italy. E-mail: sfoti@dipchi.unict.it Contract/grant sponsor: MIUR PRIN 2006; contract/grant num- ber: 2006075417. Contract/grant sponsor: FIRB ‘Italian Human ProteomeNet’; contract/grant number: RBRN07BMCT. Copyright # 2009 John Wiley & Sons, Ltd.