An equilibrium thermodynamic model of the sequestration of calcium phosphate by casein micelles and its application to the calculation of the partition of salts in milk

An equilibrium thermodynamic model of the interaction of calcium, phosphate and casein in milk is described in which the micellar calcium phosphate is assumed to be in the form of calcium phosphate nanoclusters. A generalized empirical formula for the nanocluster is used to define the molar ratios of small ions (Ca, Mg, P_i and citrate) to a casein phosphorylated sequence (phosphate centre, PC). From this model, a method of calculating the partition of milk salts into diffusible and non-diffusible fractions is obtained. No arbitrary assumptions are made, no fitting of adjustable parameters is done and the PCs in the caseins are defined by inspection of their primary structures. In addition to the salt partition, the mole fractions of the individual caseins not complexed to the calcium phosphate through one or more of their PCs are computed and a generic stability rule for milks is derived. The use of the model is illustrated by calculations of the partition of salts in a standard milk and by comparison with experimental data on the partition of salts in the milk of individual cows. The generic stability rule is applied to the individual milks to determine whether the micellar calcium phosphate is thermodynamically stable. According to the calculations, compositions that might lead to pathological calcification in the lumen of the mammary gland were seldom found in primiparous healthy cows in early or mid lactation but occurred more often in multiparous animals, in late lactation and during mastitic infection.Abbreviations ACP amorphous calcium phosphate - Cit citrate - CN casein - CPN calcium phosphate nanocluster - DCPD dicalcium phosphate dihydrate - HA hydroxyapatite - IAP ion activity product - MCP micellar calcium phosphate - MWCO molecular weight cut-off - OCP octacalcium phosphate - PC phosphate centre - TCC tricalcium citrate     

The minerals of milk

The salt of milk constitutes a small part of milk (8-9 g.L(-1)); this fraction contains calcium, magnesium, sodium and potassium for the main cations and inorganic phosphate, citrate and chloride for the main anions. In milk, these ions are more or less associated between themselves and with proteins. Depending on the type of ion, they are diffusible (cases of sodium, potassium and chloride) or partially associated with casein molecules (cases of calcium, magnesium, phosphate and citrate), to form large colloidal particles called casein micelles. Today, our knowledge and understanding concerning this fraction is relatively complete. In this review, the different models explaining (i) the nature and distribution of these minerals (especially calcium phosphate) in both fractions of milk and (ii) their behaviour in different physico-chemical conditions, are discussed.     

Composition and ultrastructure of calcium phosphate-citrate complexes in bovine milk systems

The present studies show that the colloidal calcium phosphate of cow's milk has a (Ca + Mg)/P_i ratio of 1.67 (± 0.10; n = 22) and contains citrate, Mg and Zn at molar ratios to Ca averaging 0.05, 0.03 and 0.003, respectively. The composition of the natural colloidal phosphate of milk is similar to the precipitates formed by neutralization of ultrafiltrates obtained from acidified milks, and to that of the calcium phosphate-enriched fraction produced by extensive enzymic hydrolysis of the casein micelles in milk. Examination by electron microscopy of these artificial preparations of milk calcium phosphate revealed in both a very fine and uniform substructure which consisted of granules having an average, true diameter of approx. 2.5 nm. The size and shape of these tiny granules closely resemble the morphologies reported for the colloidal phosphate particles in native casein micelles, as well as for the subunits of amorphous calcium phosphate observed during calcification in other biological systems such as mitochondria and bone.     

Mineralisation of soft and hard tissues and the stability of biofluids

Evidence is provided from studies on natural and artificial biofluids that the sequestration of amorphous calcium phosphate by peptides or proteins to form nanocluster complexes is of general importance in the control of physiological calcification. A naturally occurring mixture of osteopontin peptides was shown, by light and neutron scattering, to form calcium phosphate nanoclusters with a core–shell structure. In blood serum and stimulated saliva, an invariant calcium phosphate ion activity product was found which corresponds closely in form and magnitude to the ion activity product observed in solutions of these osteopontin nanoclusters. This suggests that types of nanocluster complexes are present in these biofluids as well as in milk. Precipitation of amorphous calcium phosphate from artificial blood serum, urine and saliva was determined as a function of pH and the concentration of osteopontin or casein phosphopeptides. The position of the boundary between stability and precipitation was found to agree quantitatively with the theory of nanocluster formation. Artificial biofluids were prepared that closely matched their natural counterparts in calcium and phosphate concentrations, pH, saturation, ionic strength and osmolality. Such fluids, stabilised by a low concentration of sequestering phosphopeptides, were found to be highly stable and may have a number of beneficial applications in medicine.     

Nature of micellar calcium phosphate in cows' milk as studied by high-resolution electron microscopy

The nature of the inorganic calcium phosphate in the casein micelle of cows' milk has been studied by high-resolution electron microscopy. No periodic lattice spacings could be imaged, and diffraction patterns were of the diffuse amorphous type. Short-range order of less than 15 A may be present, but the results indicate that there is no long-range order in micellar calcium phosphate.     

Nature of micellar calcium phosphate in cows' milk as studied by high-resolution electron microscopy

The nature of the inorganic calcium phosphate in the casein micelle of cows' milk has been studied by high-resolution electron microscopy. No periodic lattice spacings could be imaged, and diffraction patterns were of the diffuse amorphous type. Short-range order of less than 15 Å may be present, but the results indicate that there is no long-range order in micellar calcium phosphate.     

An equilibrium thermodynamic model of the sequestration of calcium phosphate by casein phosphopeptides

Sequestration of calcium phosphate by caseins occurs in the Golgi region of mammary secretory cells during lactation, where it helps to prevent calcification of the gland and to deliver high concentrations of calcium and phosphate to the neonate in the form of milk. Calcium phosphate nanoclusters are formed when a core of amorphous calcium phosphate is sequestered within a shell of casein or casein phosphopeptides. The nanoclusters can form spontaneously from a supersaturated solution or by dispersion of a precipitate of calcium phosphate, demonstrating that they are thermodynamically stable complexes. The average size and chemical composition of the complexes are largely independent of the solution conditions (pH, temperature, peptide concentration, salt composition and rate of reaction) under which they form. Larger, metastable, colloidal particles can form if there is not enough of the phosphopeptide to sequester all the calcium phosphate, or, transiently, if the salt and peptide solutions are mixed together without sufficient care. A thermodynamic model of the sequestration process is presented which makes use of an invariant ion activity product observed in nanocluster-containing solutions. In any given solution that has thermodynamic stability, the extent of the sequestration reaction can be calculated from the empirical formula of the nanoclusters using the criterion that the solution should have the equilibrium value of the invariant ion activity product. Other members of the paralogous group of secretory calcium-binding phosphoproteins to which caseins belong may also be able to sequester calcium phosphate in biological fluids such as saliva and in the extracellular matrix of mineralizing tissues.Abbreviations -PP _s1-casein 5P (f59–79) - -PP -casein 4P (f1–25) - ACP amorphous calcium phosphate - Cit citrate - CPN calcium phosphate nanocluster - CPP commercial phosphopeptide - IAP ion activity product - MWCO molecular weight cut-off - PP phosphopeptide - SAXS small-angle X-ray scattering - SCPP secretory calcium-binding phosphoprotein - UF ultrafiltrate     

Physicochemical characterization of casein phosphopeptide-amorphous calcium phosphate nanocomplexes

Milk caseins stabilize calcium and phosphate ions and make them available to the neonate. Tryptic digestion of the caseins yields phosphopeptides from their polar N-terminal regions that contain clusters of phosphorylated seryl residues. These phosphoseryl clusters have been hypothesized to be responsible for the interaction between the caseins and calcium phosphate that lead to the formation of casein micelles. The casein phosphopeptides stabilize calcium and phosphate ions through the formation of complexes. The calcium phosphate in these complexes is biologically available for intestinal absorption and remineralization of subsurface lesions in tooth enamel. We have studied the structure of the complexes formed by the casein phosphopeptides with calcium phosphate using a range of physicochemical techniques including x-ray powder diffraction, scanning electron microscopy, transmission electron microscopy, and equilibrium binding analyses. The amorphous nature of the calcium phosphate phase was confirmed by two independent methods: x-ray powder diffraction and selected area diffraction. In solution, the ion activity product of a basic amorphous calcium phosphate phase was the only ion product that was a function of bound phosphate independent of pH, consistent with basic amorphous calcium phosphate being the phase stabilized by the casein phosphopeptides. Detailed investigations of calcium and calcium phosphate binding using a library of synthetic homologues and analogues of the casein phosphopeptides have revealed that although the fully phosphorylated seryl-cluster motif is pivotal for the interaction with calcium and phosphate, other factors are also important. In particular, calcium binding and calcium phosphate stabilization by the peptides was influenced by peptide net charge, length, and sequence.     

High pressure-induced changes in bovine milk proteins: a review

High pressure (HP)-induced changes in the proteins of bovine milk have become an area of considerable research interest in recent years; as a result, there is now a detailed understanding of the effects of HP on casein micelles and whey proteins. HP treatment at pressures >400 or >100 MPa denatures the two most abundant whey proteins, alpha-lactalbumin (alpha-la) and beta-lactoglobulin (beta-lg), respectively. The majority of denatured beta-lg in HP-treated milk associates with the casein micelles, although some denatured beta-lg remains in the serum phase or is attached to the milk fat globule membrane; HP-denatured alpha-la is also associated with the milk fat globules. Casein micelles are disrupted on treatment at pressures >200 MPa; the rate and extent of micellar disruption increases with pressure and is probably due to the increased solubility of calcium phosphate with increasing pressure. On prolonged treatment at 250-300 MPa, reassociation of micellar fragments occurs through hydrophobic bonding; this process does not occur at a pressure >300 MPa, leading to considerably smaller micelles in such milk. As a result of HP-induced changes, the size, number, hydration, composition and light-scattering properties of casein micelles in HP-treated milk differ considerably from those in untreated milk.     

The evolution of milk casein genes from tooth genes before the origin of mammals

Caseins are among cardinal proteins that evolved in the lineage leading to mammals. In milk, caseins and calcium phosphate (CaP) form a huge complex called casein micelle. By forming the micelle, milk maintains high CaP concentrations, which help altricial mammalian neonates to grow bone and teeth. Two types of caseins are known. Ca-sensitive caseins (α(s)- and β-caseins) bind Ca but precipitate at high Ca concentrations, whereas Ca-insensitive casein (κ-casein) does not usually interact with Ca but instead stabilizes the micelle. Thus, it is thought that these two types of caseins are both necessary for stable micelle formation. Both types of caseins show high substitution rates, which make it difficult to elucidate the evolution of caseins. Yet, recent studies have revealed that all casein genes belong to the secretory calcium-binding phosphoprotein (SCPP) gene family that arose by gene duplication. In the present study, we investigated exon-intron structures and phylogenetic distributions of casein and other SCPP genes, particularly the odontogenic ameloblast-associated (ODAM) gene, the SCPP-Pro-Gln-rich 1 (SCPPPQ1) gene, and the follicular dendritic cell secreted peptide (FDCSP) gene. The results suggest that contemporary Ca-sensitive casein genes arose from a putative common ancestor, which we refer to as CSN1/2. The six putative exons comprising CSN1/2 are all found in SCPPPQ1, although ODAM also shares four of these exons. By contrast, the five exons of the Ca-insensitive casein gene are all reminiscent of FDCSP. The phylogenetic distribution of these genes suggests that both SCPPPQ1 and FDCSP arose from ODAM. We thus argue that all casein genes evolved from ODAM via two different pathways; Ca-sensitive casein genes likely originated directly from SCPPPQ1, whereas the Ca-insensitive casein genes directly differentiated from FDCSP. Further, expression of ODAM, SCPPPQ1, and FDCSP was detected in dental tissues, supporting the idea that both types of caseins evolved as Ca-binding proteins. Based on these findings, we propose two alternative hypotheses for micelle formation in primitive milk. The conserved biochemical characteristics in caseins and their immediate ancestors also suggest that many slight genetic modifications have created modern caseins, proteins vital to the sustained success of mammals.     

Binding of calcium and phosphate ions to dentin phosphophoryn

The concomitant binding of calcium and inorganic phosphate ions by the highly phosphorylated rat dentin phosphophoryn (HP) was measured in the pH range of 7.4-8.5 by an ultrafiltration procedure. HP binds almost exclusively the triply charged PO4(3-) ion, and for each PO4(3-) ion bound, the protein binds about 1.5 additional Ca2+ ions. Therefore, the protein-mineral ion complex can be described as a protein with two different ligands, Ca2+ ions and calcium phosphate clusters having a stoichiometry of about Ca1.5PO4. Empirically the binding of calcium and phosphate can best be described as a function of a neutral ion activity product in which 2.5-10% of the phosphate is HPO4(2-). The stoichiometry of the bound clusters is similar to that of amorphous calcium phosphate, and it is clear that the protein does not sequester crystal embryos of octacalcium phosphate or hydroxyapatite. The protein-mineral ion complex is amorphous by electron diffraction analysis and does not catalyze the formation of a crystalline phase when aged in contact with its solution. About 15% of the bound phosphate is buried in protected domains, and it is stable with respect to dissociation for extended periods in phosphate-free calcium buffers. The buried mineral maintains the protein in an aggregated state even at calcium ion concentrations which are too low for the aggregation of unmineralized HP. In vivo HP should be ineffective in the nucleation of a crystalline mineral phase, if it is secreted in a mineralized aggregated state similar to casein and the bivalve phosphoprotein.     

Bioavailability of zinc and its binding to casein in milks and formulas

Differences in zinc bioavailability among milk and formulas may be attributed to binding of zinc to various ligands. We determined the distribution of zinc and protein at different pHs and zinc and calcium concentrations. We used radiolabelled cow's milk, human milk, whey-predominant (WPF) and casein-predominant (CPF) infant formula. Lowering the pH changed zinc and protein distribution: zinc shifted from pellet (casein) to whey in cow's milk, from fat to whey in human milk and from fat and pellet to whey in formulas. Protein shifted from whey to pellet in human milk and from whey and pellet to fat in formulas. Increasing zinc and calcium concentrations shifted protein and zinc from pellet to whey for cow's milk and from whey and pellet to fat for the formulas. Protein distribution was not affected by calcium or zinc addition in human milk or CPF, while zinc shifted from whey to fat in human milk and from fat and pellet to whey in CPF. Zinc and calcium binding to isolated bovine or human casein increased with pH. At 500 mg/L of zinc, bovine casein bound 32.0 +/- 1.8 and human casein 10.0 +/- 0.9 mg zinc/g protein. At 500 mg/L of calcium, calcium was preferentially bound over zinc. Adding calcium and zinc resulted in 32.0 +/- 1.8 mg zinc/g bound to bovine casein and 17.0 +/- 0.8 mg zinc/g to human casein, while calcium binding was low. Suckling rat pups dosed with 65Zn labelled infant diets were killed and individual tissues were gamma counted. Lower zinc bioavailability was found for bovine milk at pH = 4.0 (%65Zn in liver = 18.7+1.4) when compared to WPF (22.8 +/- 1.6) or human milk (26.9 +/- 0.8). Lowering the pH further decreased zinc bioavailability from human milk, but not from cow's milk or WPF. Knowledge of the compounds binding minerals and trace elements in infant formulas is essential for optimizing zinc bioavailability.     

Casein aggregates built step-by-step on charged polyelectrolyte film surfaces are calcium phosphate-cemented

The possible mechanism of casein aggregation and micelle buildup was studied in a new approach by letting α-casein adsorb from low concentration (0.1 mg·ml(-1)) solutions onto the charged surfaces of polyelectrolyte films. It was found that α-casein could adsorb onto both positively and negatively charged surfaces. However, only when its negative phosphoseryl clusters remained free, i.e. when it adsorbed onto a negative surface, could calcium phosphate (CaP) nanoclusters bind to the casein molecules. Once the CaP clusters were in place, step-by-step building of multilayered casein architectures became possible. The presence of CaP was essential; neither Ca(2+) nor phosphate could alone facilitate casein aggregation. Thus, it seems that CaP is the organizing motive in the casein micelle formation. Atomic force microscopy revealed that even a single adsorbed casein layer was composed of very small (in the range of tens of nanometers) spherical forms. The stiffness of the adsorbed casein layer largely increased in the presence of CaP. On this basis, we can imagine that casein micelles emerge according to the following scheme. The amphipathic casein monomers aggregate into oligomers via hydrophobic interactions even in the absence of CaP. Full scale, CaP-carrying micelles could materialize by interlocking these casein oligomers with CaP nanoclusters. Such a mechanism would not contradict former experimental results and could offer a synthesis between the submicelle and the block copolymer models of casein micelles.     

Photolysis of Aryl Glycosides with Ultraviolet Irradiation

Bovine casein components (α_sl-, β-, and κ-caseins) were chemically phosphorylated and the properties of the modified components were compared with those of the native to clarify the function of the intrinsic phosphate groups of casein components in casein micelle formation. The calcium binding ability of casein components increased after chemical phosphorylation. The concentrations of calcium chloride required to precipitate modified α_sl- and β-caseins were higher than those for native components. However, phosphorylation of α_sl- and β-caseins did not affect their properties of forming micelles through interaction with κ-casein. The stabilizing ability of κ-casein for α_sl- and β caseins was impaired by its phosphorylation, but the stability was recovered by treating phosphorylated κ-casein with phosphoprotein phosphatase. The results show that the phosphate content of κ-casein must be low to form a stable casein micelle. The results also explain why the specific phosphorylation of casein components in the mammary gland is required.     

Preparation and characterization of milk calcium salts by using casein phosphopeptide

Milk calcium salt solution was prepared by the following procedures using casein phosphopeptides (CPP). To CPP solution, 1 M citric acid, 1 M CaCl2 and 1 M K2HPO4 were added with stirring, while adjusting the pH to 6.7. The prepared solution was left for 12 hr at 25 degrees C and then used for subsequent experiments, or lyophilized. The concentrations of organic phosphate of CPP, calcium, inorganic phosphate, and citrate in the typical milk salt solution were 7, 30, 22, and 10 mM, respectively, which were close to those in bovine milk. The lyophilized sample was easily dissolved in water. No crystal structure of hydroxyapatite was shown in the lyophilized milk calcium salts by X-ray diffraction analysis, although the pattern of KCl crystal was observed. The X-ray diffraction pattern of commercial whey mineral, which was prepared by precipitation at alkaline pH from rennet whey, was similar to that of hydroxyapatite. It was confirmed by high-performance gel chromatographic analysis that the form of calcium phosphate in the milk calcium salts was similar to that of casein micelles.     

Insertion of a casein kinase recognition sequence induces phosphorylation of ovine beta-lactoglobulin in transgenic mice

We have shown that the cellular mechanisms of the mammary gland can be used to produce a phosphorylated form of a normally unphosphorylated milk protein. This was achieved by the insertion of a beta-casein DNA sequence coding for a group of mammary gland casein kinase recognition sites into ovine beta-lactoglobulin. Transgenic mice carrying this modified gene were generated and lactating females were shown to produce a novel beta-lactoglobulin in their milk. The infrared spectrum, reactivity to antiphosphoserine antibody and reduction of electrophoretic mobility on treatment with alkaline phosphatase showed that the novel protein recovered from the milk whey (serum) was phosphorylated and molecular mass determination by mass spectrometry was consistent with the phosphorylation of one or two residues. A similar level of phosphorylation was measured by quantitative infrared spectroscopy. Centrifugation of the milk to pellet the casein micelles showed that most of the phosphorylated beta-lactoglobulin was in the whey and hence not incorporated into casein micelles.     

Three kinetically different inorganic phosphate entities in bovine casein micelles revealed by isotopic exchange method and compartmental analysis

Much controversy exists concerning the way calcium phosphate is linked to milk phosphoproteins including caseins. Homoionic exchange of inorganic phosphate between micellar calcium phosphates (MCP) of casein micelles and solute phosphates in cows' milk was investigated using H(32)PO(4)(2-) as radiotracer. Compartmental analysis and modelling revealed the presence of three MCP-related inorganic phosphate compartments each representing a separate phosphate entity. The relative phosphate quantities per compartment, i.e. the quantities of kinetically identical phosphate ions per MCP-ion cluster, and their mean residence times are 2:1:1 and 818, 0.24 and 23 h, respectively. Hence each MCP-ion cluster comprises four inorganic phosphate ions divided over three intra-MCP binding sites each characterised by a mean residence time for homomolecular phosphate exchange at solution/MCP interface.     

Characterisation of the potential SNARE proteins relevant to milk product release by mouse mammary epithelial cells

Casein micelles and fat globules are essential components of milk and are both secreted at the apical side of mammary epithelial cells during lactation. Milk fat globules are excreted by budding, being enwrapped by the apical plasma membrane, while caseins contained in transport vesicles are released by exocytosis. Nevertheless, the molecular mechanisms governing casein exocytosis are, to date, not fully deciphered. SNARE proteins are known to take part in cellular membrane trafficking and in exocytosis events in many cell types and we therefore attempted to identify those relevant to casein secretion. With this aim, we performed a detailed analysis of their expression by RT-PCR in both whole mouse mammary gland and in purified mammary acini at various physiological stages, as well as in the HC11 cell line. The expression of some regulatory proteins involved in SNARE complex formation such as Munc-13, Munc-18 and complexins was also explored. The amount of certain SNAREs appeared to be regulated depending on the physiological stage of the mammary gland. Co-immunoprecipitation experiments indicated that SNAP-23 interacted with syntaxin-6, -7 and -12, as well as with VAMP-3, -4 and -8 in mammary epithelial cells during lactation. Finally, the subcellular localisation of candidate SNAREs in these cells was determined both by indirect immunofluorescence and immunogold labelling. The present work provides important new data concerning SNARE proteins in mammary epithelial cells and points to SNAP-23 as a potential central player for the coupling of casein and milk fat globule secretion during lactation.     

Role of Ca2+ in the secretion of milk caseins in lactating rabbit mammary epithelial cells

Prolactin and arachidonic acid increase milk casein secretion in mammary gland slices. These effects do not necessitate Ca2+ in the incubation medium. Prolactin does not modify the influx or the efflux of 45Ca2+. The Ca2+ channel blocking agent D600 (6 micrograms/ml) decreases the stimulatory effect of prolactin on casein secretion, but does not interfere in the stimulatory effect of arachidonic acid. The calmodulin inhibitor trifluoperazine (100 microM) inhibits stimulation of casein secretion by both prolactin and arachidonic acid. From these data, it is concluded that a flow of Ca2+ from the outside into the cell is not a requisite for the stimulation of casein secretion. However, stimulation by prolactin, but not stimulation by arachidonic acid, requires Ca2+ movement through calcium pathways. Intracellular transport of Ca2+ seems necessary for the stimulation of secretion.