Association of the quadruply phosphorylated beta-casein from human milk with the nonphosphorylated form

The primary structure of a low-Mr multiphosphorylated variant ofβ-casein in equine milk

Highly phosphorylated casein with a low molecular mass was isolated from Haflinger mare's milk by RP-HPLC. It accounts for 4.0% of the casein content. Its mass was determined by LC-ESI-MS before and after treatment by alkaline phosphatase. The molecular mass found for the apo-form (10,591 +/- 2 Da) is in agreement with its primary structure, which was established by ESI-MS/MS from tryptic peptides. It appeared that this short protein (94 amino acid residues) is an internally truncated form of the full-length equine beta-casein (226 residues). This low-Mr variant of equine beta-casein displays a large deletion (residues 50-181), due to a cryptic splice site usage occurring within exon 7 during the course of primary transcripts processing. The phosphorylation pattern of this equine beta-casein variant was investigated by LC-ESI-MS and 2-DE. Seven phosphorylation forms were identified with one to seven phosphate groups with pIs ranging between 4.67 and 4.01. The major isoforms carry five and six phosphate groups.

Determination of the phosphorylation level and deamidation susceptibility of equine β-casein

beta-Casein was isolated from Haflinger mare's milk by RP-HPLC, and displayed microheterogeneity by urea-electrophoresis and 2-DE probably due to a variable degree of phosphorylation. To investigate the degree of phosphorylation, the primary structure of equine beta-casein was determined by tryptic hydrolysis and MS of peptides released and by MS of the protein treated by alkaline phosphatase. The molecular mass found for the apo-form of Haflinger mare's beta-casein (25 514 +/- 3 Da) was close to the theoretical mass of the reported sequence (GenBank AAG43954) modified by insertion of a region (residues 27-34) encoded by an exon sometimes out-spliced (25 511.40 Da). Hence, the beta-casein isolated from Haflinger mare's milk corresponded to a variant of 226 amino acid residues. The latter was composed by highly multi-phosphorylated isoforms with three to seven phosphate groups, and pIs, determined by 2-DE, ranging from 4.74 to 5.30. Moreover, the equine beta-casein was able to deamidate spontaneously, at the level of Asn in the potential deamidation motif (135)Asn-Gly(136). Approximately 80% of the protein was deamidated after 96 h of incubation under physiological conditions.

Dependence of the Interfacial Behavior of β-Casein on Phosphoserine Residues

The role of the phosphoserine residues on the dynamical and structural properties of beta-casein was studied by molecular dynamics of the protein in water/lipid interfacial regions. The initial protein structure adopted in the modeling was that proposed for bovine beta-casein A2, where the five phosphoserine residues, originally present in its primary structure, were partially or totally substituted by serine residues. The simulations revealed a dependence of the interfacial behavior of beta-casein on the phosphorylation grade. When only partially dephosphorylated, the protein showed a similar behavior as that observed for the original beta-casein reported in previous work. During dynamics, the protein migrated from the aqueous environment towards the lipid medium, and remained attached to the interface separating both media. Quite different was the dynamics of the totally dephosphorylated beta-casein, that did not perceive the interface and immersed incessantly into lipid medium. The results suggest that the phosphoserine residues appear to be, in fact, intrinsically related to the mechanisms of beta-casein emulsion stabilization.

The use of lithium chloride to study human milk micelles

Various methods have been used to study the dissociation of milk micelles in attempts to determine their structure and the interactions that stabilize them. These include the addition of urea, cooling to alter hydrophobic bonding, the addition of EDTA to sequester calcium, and changes in pH to alter molecular charge. For this study, the mild chaotropic agent LiCl was added to human milk micelles, and measurements were made on the relative percentages of the six different phosphorylation levels of beta-casein (CN) at various LiCl concentrations for different lengths of time and at different temperatures. Added LiCl had little effect at 37 degrees C but caused maximal dissociation, mainly of the beta-CN species with higher phosphorylation levels, at 23 degrees C and 4 degrees C between 1 and 2 M concentration. Comparison was made with 2-M additions of NaCl, MgCl2, and KCl at 4 degrees C, with LiCl showing the only appreciable change. The results suggest that Li+ may displace Ca2+ in protein-Ca2+-protein or protein-colloidal calcium phos+ phate-protein salt bridges and that the nonphosphorylated form of human beta-CN may change its conformation and mode of interaction upon phosphorylation. Lithium chloride may be useful to study the dissociation of the different CN in bovine milk micelles.

Association of mixtures of the two major forms of beta-casein from human milk

Human milk beta-casein (CN) is unique in that it may be phosphorylated at any level from zero (beta-CN-0P) to five (beta-CN-5P) organic phosphates per molecule. The 2P and 4P forms are the major components, with about 30 to 35% each. Here, we present the association properties of mixtures of these two moieties of human beta-CN. The aggregation patterns, as functions of temperature and ionic strength of these mixtures, generally follow those for the individual components. However, the mixtures yielded polymers with slightly different properties, which indicates extensive interaction between the two. Some properties of the mixtures were more like those for the 2P form, such as association in low salt buffer to give a peak with a sedimentation coefficient, s20,w, of approximately 11 S, in contrast to the 2P form alone with a peak of approximately 13 S and 4P alone with only a small amount of material with s20,w greater than 2 S at 27 degrees C. The solubility and interactions in the presence of Ca2+ ions were intermediate but more like the 4P form. A protein-concentration dependence for s20,w was seen, and laser light scattering indicated that there was an increase in size and/or a change in shape as the protein concentration increased. From the results, it is apparent that submicellar oligomers are probably formed by rapidly established equilibrium association reactions. The presence of an equal amount of the 2P form along with the 4P form does not appear to be a disadvantage in casein micelle formation and function.