Heat-induced association of β-lactoglobulin (β-lg) and casein micelles, mainly at 85–90 °C, was studied by means of preparative ultracentrifugation, using [3H]-labelled β-lg. Qualitative aspects were studied by gel electrophoresis and electron microscopy. In this association, the formation of intermolecular S–S bonds between β-lg and κ-casein plays a role, but hydrophobic bonds are also involved.

After heating mixtures containing 25 g/kg casein and 4 g/kg β-lg at 90 °C for 20 min, the amount of β-lg sedimented with the casein micelles by ultracentrifugation decreased by approximately 30% when the milk salt buffer system was reduced to 0·25 of its normal concentration; it decreased by about 20% when the pH was increased from 6·8 to 7·3 and increased by 15% when the pH was reduced from 6·8 to 5·8. When the β-lg concentration was decreased from 4 to 2 g/kg, the amount of sedimented β-lg decreased by 25% after heating at 90 °C for 6 min. After heating in milk salt buffer, the amount of sedimented β-lg was 15% higher than after heating in a salt solution which contained neither Ca nor citrate ions.

Although α-lactalbumin (α-la) is involved in the heat-association of β-lg and casein micelles, no influence of α-la on the amount of associated β-lg was found. In addition to the heat-association product of β-lg and whole casein micelles, an association product consisting mainly of β-lg and κ-casein was also formed. This product was observed by electron microscopy as noodle-like particles. The size of these particles depended on the β-lg/κ-casein ratio.

The increase in unesterified fatty acid content of unpasteurized bulk milk in storage at 4 °C in the presence of known effectors of bovine milk lipoprotein lipase originating from bovine serum was studied. Bovine serum and high density lipoprotein (HDL) caused an increase in developed unesterified fatty acid levels whilst lipoprotein-free serum, apo-HDL, all individual apo-HDL tested, and the unfractionated C-peptide fraction were without lipolytic effect. In the presence of HDL-lipids, 2 C-peptides stimulated considerable lipolysis, as did the combination of HDL-lipid with unfractionated C-peptides. These characteristics of unesterified fatty acid development could be duplicated in milk whose endogenous lipolytic activity had been destroyed by heat treatment (75 °C for 5 min) and then restored by addition of purified bovine milk lipoprotein lipase. Radioactively labelled glycerol trioleate in milk was not hydrolysed in the same way as milk fat on the addition of serum liproproteins.

An examination has been made of the heat stability characteristics of skim-milk concentrate prepared by ultrafiltration (UF). Concentrate prepared by UF was found to be more stable than that prepared by conventional evaporation. In contrast to conventional concentrate, the heat stability of UF concentrate was not appreciably affected by forewarming or addition of permitted stabilizers, but the effect of addition of urea was generally the same for both UF and conventional concentrates; an increase in heat stability was obtained if the milk total solids level was less than 14%. As with conventional concentrate, addition of simple aldehydes induced large increases in the heat stability of UF concentrate. It is suggested that a novel range of sterile milk products could be prepared from UF concentrates. Because of the high protein and low lactose contents of these concentrates, the products might be nutritionally more attractive than those prepared from conventional concentrates.

Lamb and kid rennets were prepared by extraction of dried abomasa with 6% (w/v) NaCl-2% (w/v) H3BO3 and activation of the proenzymes at pH 2·0. Each gave one zone of precipitation on casein-agar gel diffusion, enabling them to be differentiated from calf rennet and pig pepsin. After agarose gel electrophoresis, the proteinase activity of lamb rennet occurred in chymosin and pepsin bands only, whereas kid rennet contained an additional proteinase of intermediate mobility. Relative to their milk-clotting activities, lamb and kid rennets contained less pepsin and were less proteolytic on both haemoglobin at pH 1·8 and casein at pH 5·3 than calf rennet. The milk-clotting activities of lamb and kid rennets increased less with decrease in pH and were more stable to storage at both the pH value of maximum stability and lower pH values than that of calf rennet. Neither cathepsin activity nor lipolytic activity on milk fat was detected in any of the 3 rennets, but lamb rennet caused slight hydrolysis of tributyrin.

The turbidity increase on storage of milk dialysate was shown to be due to Ca phosphate precipitation. The rate of precipitation at 40 °C could be decreased by shortening the duration of dialysis or by the addition of small amounts of EDTA. The dialysate was stabilized against Ca phosphate precipitation by the addition of the whey protein fraction of milk. Individual whey proteins varied in their stabilizing ability: α-lactalbumin was found to be more effective than β-lactoglobulin which gave approximately the same effect as an immunoglobulin fraction. BSA and lysozyme had a destabilizing effect and ovalbumin possessed a slight stabilizing ability.

Measurements have been made of the coagulation times (CT) of concentrated and diluted milks when treated with chymosin. After dilution, milks showed a large increase in CT: conversely, concentrated milks showed a similar CT to unconcentrated milk. The results are discussed in terms of a mechanism which reduces to a form similar to that proposed by Holter (1932).

Casein composition in Friesian herd milks was examined by quantitative column chromatography on hydroxyapatite. Whole casein was resolved into 5 fractions: (1) γ-casein + para-κ-casein; (2) κ-casein + unidentified chymosin-resistant components; (3) unidentified minor protein; (4) β-casein; (5) αs1,0- + minor αs-caseins. Protein in fractions 1, 2, 4, 5 was measured by ultraviolet spectrometry, using appropriate specific absorption coefficients. The quantitative contribution of chymosin-resistant components to the κ-casein fraction was measured by a method involving hydroxyapatite chromatography of chymosin-treated whole casein. Compositional analysis was carried out throughout a lactation cycle on a total of 24 casein samples from the milks of a winter/spring calving and autumn-calving herd. Average composition (expressed as % total casein), which was similar for both herds, was 48·7% αs-, 33·8% β-, 4·9% γ-, 9·5% κ- and 3·2% unidentified components. Lactational effects on composition were also similar for both herds and were most pronounced in late lactation where increases were recorded in γ-casein and unidentified components, and decreases in β- and αs-caseins. Fluctuations in κ-casein proportions were small and slightly higher values were obtained in early and late lactation. Casein composition also appeared to be affected to some extent by season, as shown by differences in the proportions of αs-caseins and unidentified components between mid-lactation milks from the 2 herds. Compositional variations resulted partly from the action of milk proteinase on β-casein and also, apparently, from differences in relative rates of protein biosynthesis.

The content and composition of protein in milk samples from creameries in south-west Scotland were determined over a period of 12 months. The composition of the whole casein was expressed in terms of αsl-, β-, κ-, αs2- and γ-caseins, and that of the total milk serum protein in terms of β-lactoglobulins (β-lg), α-lactalbumins, bovine serum albumin, and a mixture of immunoglobulins, proteose-peptone component 3 and lactoferrin (IPL). Concentrations of the individual caseins varied appreciably and for most, concentration was closely correlated with and showed the same seasonal pattern as total casein concentration. Concentrations of the milk serum proteins also varied but only those of β-lg and the IPL fraction were closely correlated with that of total milk serum protein and seasonal trends were not marked. Relative amounts of the individual proteins, on the other hand, showed smaller variations and so throughout the experimental period the milks contained a protein complex of comparatively constant composition. Because of this comparative constancy it would appear that seasonal variations in milk properties in south-west Scotland are unlikely to be determined to a major extent by milk protein composition, but could be more affected by protein concentration.

Dialysed whey and dialysed ultracentrifugate from skim-milk were found to be satisfactory dilution media for light scattering studies on bovine casein micelles at 20 °C. Both milk dialysate and milk ultracentrifugate were only useful at fairly low levels of dilution and even then only for short periods of time. Whey, pH-adjusted to that of the original milk, was found to be an unsuitable diluent medium. Most of the instability of casein micelles to dilution could be attributed to some form of Ca phosphate precipitation or Ca phosphate–casein micelle aggregation.

Bovine milk systems were analysed for adenosine 5′-triphosphate (ATP) using the luciferase-ATP reaction in a liquid scintillation counter. Approximately 0·2 µmol ATP/1 milk serum were evident both in whole milks and the corresponding skim-milks. ATP was not detectable in skim-milk ultrafiltrates. These findings indicated that ATP was present in a non-dialysable portion of skim-milk. Centri-fugation of whole milks from individual cows at 5500 g for 15 min at 10°C yielded skim-milks essentially devoid of somatic cells and bacteria. However, the ATP in the skim-milks decreased by less than 20% compared with the whole milks indicating that the calcium phosphate-citrate (CPC)–caseinate micelles were the source of the ATP. ATP was not detectable in colloidal phosphate-free milk, from which CPC had been removed, confirming that the ATP was sequestered in the constituent CPC. Likewise, the occurrence of significant amounts of Mg, another potent stabilizer of amorphous calcium phosphate (ACP) in other biological systems, was confirmed in the colloidal phosphate of milk. From 0·13 to 0·31 µmole ATP/1 (mean, 0·23) was found in the 9 milk samples studied. The discovery of small but appreciable levels of ATP in the CPC of milk provides further evidence for the analogy previously shown to exist between the CPC complex of milk and the ACP which accumulates in mitochondria. The latter has been postulated to provide an essential precursor for crystalline bone salts to form in ordered calcification processes. The implications of these findings in the biosynthesis of milk are briefly discussed.

A method has been developed for quantitative determination of para-κ-casein, involving spectrophotometric scanning of stained protein bands following polyacrylamide gel electrophoresis. The rate of hydrolysis of κ-casein in skim-milk at pH 6·6 and 30 °C was compared with that in EDTA-treated skim-milk under the same conditions. This showed that at the visually observed clotting time, at least 90% of the total κ-casein in milk had been hydrolysed. The time course of the reaction was consistent with all the κ-casein molecules being hydrolysed with the same efficiency. The results strongly suggest that essentially all of the κ-casein in milk is equally accessible to rennet action. This is consistent with the casein micelle being porous, or having all the κ-casein on the surface.

Casein micelles in milk-salts solution adsorbed charged detergents and highly-charged polypeptides strongly, neutral detergents less strongly and low molecular-weight amines weakly. A tetra-amine was adsorbed more strongly than a tri-amine. The extent of adsorption of proteins tended to rise as the molecular weight increased. Glycerol and lactate were adsorbed to a limited extent but dextran and α-ketoglutarate were not adsorbed at all. Proline was partly adsorbed, indicating that hydrophobic binding sites were available, and caused some disruption of the casein micelles. Additives were bound to approximately the same extent by casein micelles and rennet coagula. The proportions adsorbed were constant over at least 10-fold ranges of concentration. Additives which increased the rennet clotting time (RCT) acted by binding Ca2+. Most additives decreased the RCT, the extent increasing with the amount adsorbed and the positive charge on the additive. The greatest reduction in RCT was observed with those additives which had positively-charged and hydrophobic moieties and bound most strongly to casein micelles. Of the additives tested, only sodium dodecyl sulphate affected the enzymic action of rennet. The reduction in RCT may have resulted from the neutralization of the negative charge of the micelles or enhancement of their hydrophobicity, favouring hydrophobic interactions between the particles.

The 4 major caseins, A1, A2, B1, B2, from rat milk have been isolated and analysed. From molecular weight determination, amino acid and phosphorus analyses and N-terminal sequence determination, A1 and A2 are concluded to possess similar peptide chains as do B1 and B2, with the individual fractions within each of these 2 groups differing only in their sialic acid content.

Mol. wts of approximately 22000 and 38000 were found for A1–A2 and B1–B2 respectively. The amino acid sequence of the first 13 residues of A1–A2 has been partly established. Components A1 and A2 appeared to be homologous with bovine β-casein, whereas B1 and B2 were different from any known casein, especially in their molecular weight.

Two stages of gelling by the milk-clotting enzymes chymosin and pepsin were studied at different milk protein concentrations, using milk retentates. In each case, enzymic velocity versus protein concentration described a standard hyperbola. According to the experimental conditions such as pH and type of enzyme, which changed the enzyme concentration used, either the quasi-linear part of the hyperbola was observed, or the velocity hardly increased and tended to a limiting value. Gelling occurred with a lower degree of proteolysis of κ-casein when the protein concentration increased but a minimum proteolysis (1% of total nitrogenous matter content) was required for aggregation to take place. Gelling time varied with the protein concentration, the pH and the enzyme concentration. The final degree of proteolysis of κ-casein was the same whatever the substrate concentration used.

The influence of storage time and temperature on the distribution of individual milk proteins between the micellar and soluble phases has been examined. Storage at 4 or 7 °C is accompanied by a dissociation of micellar caseins, particularly β-casein, into the soluble phase during the first 48 h, but on further storage there is a partial reversal of this process. At higher storage temperatures (10–20 °C) the contents of all the individual caseins in the soluble phase decrease throughout storage. During cheese–making, losses of fat and curd fines in whey were greater with increased soluble phase casein and clotting times were prolonged. Curd structure was weaker, curds were more moist and slightly lower cheese yields were obtained in stored milks with elevated soluble–phase casein. When milks were stored for up to 3 d at 4 °C cheese gradings were virtually unaffected by storage, but higher temperatures (10–15 °C) led to cheeses being downgraded, largely for body and texture defects but also for flavour after 3 d.

Attempts to reverse the pattern of casein dissociation and to minimize the detrimental effects of milk storage on subsequent cheese-making showed that a heat treatment of 30 min at 60 °C was the most beneficial.

In assessing the performance of milk for cheese-making, the influence of a number of factors in addition to cold storage time and temperature of the milk on the equilibria between the soluble and micellar phases and on cheese-making parameters has now been examined. An important factor was pH and although not apparent in fresh unstored milk, after storage at 4 °C there was a minimum in the soluble casein concentration at about pH 6·6, largely due to variations in β-casein concentration. As expected, addition of Ca reduced soluble casein concentrations while treatment with EDTA caused increases, and in most cases the cyclic pattern of casein dissociation from the micellar phase followed by a partial reassociation was again observed during storage at 4 °C over a 3-d period. Addition of low levels of urea (up to 75 mg/100 ml milk) led to some dissociation of all casein components, Ca and phosphate into the soluble phase while treatment of milk with urease caused small decreases. Additions of purified casein components to milk were accompanied by only small alterations to the composition of the soluble phase and added casein was largely incorporated into the micellar phase. In spite of this, most added casein was not recovered in curds, and cheese yields were not greatly enhanced by quite considerable additions of caseins to the milk. The effects of pH adjustment, addition of Ca, EDTA, urea, cyanate, purified αs1-, β- or κ-casein and urease treatment on rennet clotting time, curd strength, curd moisture and cheese yield are presented.

Binding isotherms for the calcium ion–αsl-casein system have been measured, as functions of ionic strength, temperature, and pH, and the isotherms have been analysed in terms of binding constants modified by substitution effects. The results demonstrate that the strength of binding is increased with increasing temperature and decreased by increasing ionic strength or decreasing pH, all of which may be explained semi-quantitatively. Parallel studies on the precipitability of the αsl-casein–Ca2+ complexes showed that there is considerable variation in the extent of calcium binding required to initiate precipitation of the protein, and in the calcium concentration necessary to achieve the required extent of ligand binding.