In five cows that were regularly milked before parturition, cholesteryl esters were continuously released into the mammary fluid; their concentration in the fluid was initially high, but decreased a few days before parturition when mammary secretion of fluid and triglyceride was increasing. The composition of fatty acids in the cholesteryl esters of mammary fluid and in blood plasma was different, suggesting mammary synthesis of cholesteryl esters.

Experiments were designed to study the effects of catecholamines on oxytocin responses and milk removal in dairy cows. Adrenalin, noradrenalin, dopamine, isoproterenol (a β-adrenoceptor agonist), phentolamine (an α-adrenergic blocker) and propranolol (a β-adrenergic blocker) were infused intravenously. In addition, adrenalin was infused together with phentolamine and/or propranolol. Infusions started 8 min before milking and lasted until the end of milking. In some cases electroshocks (for 5 s) were applied immediately before milking in the absence and presence of phentolamine and propranolol. Adrenalin, noradrenalin and dopamine reduced milk removal, but only if administered in supraphysiological amounts. The effect of adrenalin and electroshocks on milk removal could be inhibited by α-, but not by β-adrenergic blockade. The effect of dopamine could be inhibited only partly by phentolamine. Inhibition of milk removal was not mediated by reduced oxytocin responses. Enhanced local release of catecholamines from sympathetic nerves was presumably responsible for lowered milk removal in response to electroshocks. Milk removal was facilitated during α-adrenergic blockade and during β-adrenoceptor stimulation.

Cows were fed on diets supplemented with soyabean oil or soyabean fatty acids which in some cases were protected from rumen hydrogenation. The fat-containing diets reduced the output of short- and medium-chain fatty acids in milk. Associated with this fall in short- and medium-chain fatty acids was a decrease in the concentration of 2-oxoglutarate and an increase in that of isocitrate and citrate. Protection of polyunsaturated fat from rumen hydrogenation had no significant effect. Milk yields were unaffected by diet, but the variation in milk yield among cows correlated positively with the concentration of glucose in milk.

Milk permeate was separated at various temperatures by means of a hollow fibre ultrafiltration unit coupled to a stainless steel heat exchanger. Milk samples conditioned at 4°C were heated to 20, 40, 60, 80, 85 or 90°C prior to ultrafiltration. Ca, P, Mg, Na, K and citrate concentrations were measured in the permeate samples. Ca and P contents of the permeate decreased as the temperature increased. The pH was measured after cooling the permeate to room temperature. Smaller losses of Mg and citrate were also observed with increase in temperature. Na and K levels were not affected. A two-step time-concentration relationship was apparent for the species under study. An initial sharp decrease in concentration occurred in the first minute of holding time and was followed by a slower reaction. The possible occurrence of a two-step mechanism in the heat-induced salt balance changes is discussed. Dicalcium phosphate precipitation is believed to be coupled with tricalcium citrate precipitation upon heating.

Milk which had been heated to 85°C for 40 min was examined for changes to milk salts taking place upon cooling to 4, 20, 40 and 60°C. The time-concentration curves showed a two-step reaction for Ca, Pi and pH. The effect of the heat treatment in milk heated to 85°C for 40 min was to produce greater changes in the second step of the reaction. Because of the variability of the results for other species (Mg and citrate), the reversibility was calculated only for Ca and P. It was 90–95% and 93–99% respectively depending on the cooling temperature. The reversibility of the heat-induced changes and especially of the possible dissolution equilibrium between Ca phosphate and citrate precipitates on cooling are discussed.

A model heat-exchange apparatus was used to investigate the factors affecting deposit formation from milk on a stainless steel surface at 100 °C. The structure and composition of the deposits were determined by scanning electron microscopy, energy-dispersive X-ray spectroscopy, and chemical analysis after solution in alkali. The effects of changing the pH, preheating and skimming of the milk were similar to those observed in a small-scale continuous ultra high temperature plant. The time course of deposit formation showed that a lag phase did not occur, but the deposit which formed after more than 45 min was more porous than that formed after shorter times. Most (50–90%) of the fresh deposit was readily removed by sonication, leaving a sublayer richer in minerais than the original. The results provide evidence for the two-layer model for deposit formation proposed by Tissier & Lalande (1986).

A method using low resolution NMR spectroscopy is described for investigating whey protein thermal denaturation. The method is based on measuring at 20 °C changes in water proton transverse (T2) relaxation parameter following the denaturing treatment. This parameter is shown to be sensitive to protein denaturation and not to other phenomena such as gelation. Examples are given for the qualitative study of protein thermal denaturation in whey protein concentratc, β-lactoglobulin, α-lactalbumin, bovine serum albumin and immunoglobulins aqueous solutions and for the quantitative determination of thermal denaturation in whey protein concentrate solutions.

The surface energies of highly hydrated casein micelle layers isolated from variously pretreated skim milks have been determined by means of contact angle measurements. The long range Lifshitz-Van der Waals (LW) and the short range hydrogen bonding (SR) components of surface energy were determined using α-bromonaphthalene and water for contact angle measurements. Casein micelles isolated from untreated and heat treated milks showed similar surface energy values of about 63·5 mJ.m-2 with an LW component of 19·2 mJ.m-2 and an SR component of 44·3 mJ.m-2. The calculated attraction potential energy was − 0·7 mJ.m-2. Casein micelles isolated from renneted milk showed a surface energy of 33·0 mJ.m-2 with an LW component of 30·7 mJ.m-2 and an SR component of 2·3 mJ.m-2. The attraction potential energy of renneted micelles was nearly two orders of magnitude higher than those of micelles from other milks ( − 63·3 mJ.m-2). The SR component of interfacial energy accounted for 98% of this attraction potential. The importance of attractive forces in relation to casein micelle stability is discussed.

Zn binding by whole bovine and human casein and by purified bovine caseins and whey proteins was investigated by equilibrium dialysis. Bovine αs1 casein had the greatest Zn-binding capacity (˜ 11 atoms Zn/mol). Protein aggregation was observed as Zn concentration was increased and- the protein precipitated at a free Zn concentration of 1·7 mM. Zn binding increased with increasing pH in the range 5·4–7·0 and decreased with increasing ionic strength. Competition between Zn and Ca was observed for binding to αs1-casein indicating common binding sites for these two metals. Bovine β-casein bound up to 8 atoms Zn/ mol and precipitated at a free Zn concentration of ˜ 2·5 mM, while K-casein bound 1–2 atoms Zn/mol. Whole bovine and human casein bound 5–8 atoms Zn/mol and precipitated at a free Zn concentration of ˜ 2·0 mM. Scatchard plots for Zn binding to caseins showed upward convexity, possibly due to Zn-induced association of caseins. Apparent average association constants (K¯app) for all caseins were similar (log K¯app 3·0–3·2). Enzymic dephosphorylation of αs1- or whole bovine casein markedly reduced, but did not eliminate, Zn binding. Thus, phosphoserine residues appeared to be the primary Zn-binding sites in caseins. With the exception of bovine serum albumin. which bound over 8 atoms Zn/mol, the bovine whey proteins, β-lactoglobulin, α-lactalbumin and lactotransferrin, had little capacity for Zn binding.

About 90% of the Zn in bovine skim milk was sedimented by ultracentrifugation at 100000 g for 1 h. About half of the non-sedimentable Zn was non-dialysable, indicating that it was associated with protein, probably non-sedimented casein micelles. Casein micelles incorporated considerable amounts of Zn added to skim milk as ZnCl2, and at Zn concentrations ≥ 16 mM coagulation of casein micelles occurred. Ca was displaced from casein micelles by increasing ZnCl2 concentration and ˜ 40% of micellar Ca was displaced by 16 mM-ZnCl2. Micellar Zn, Ca and P1 were gradually rendered soluble as the pH of milk was lowered and at pH 4·6 > 95% of the Zn, Ca and P1 were non-sedimentable. These changes were largely reversible by readjustment of the pH to 6·7. About 40% of the total Zn in skim milk was non-sedimentable at 0·2 mM-EDTA and most of the remainder was gradually rendered soluble by EDTA over the concentration range 1–50 mM. This indicates that there are two distinct micellar Zn fractions. No micellar Ca or P1 was solubilized at EDTA concentrations up to 1·0 mM, indicating that both colloidal calcium phosphate (CCP) and casein micelles remained intact under conditions where the more loosely bound micellar Zn fraction dissolved. Depletion of casein micelles of colloidal Ca and P1 by acidification and equilibrium dialysis resulted in removal of Zn, and in colloidal Pi-free milk non-dialysable Zn was reduced to ·-2 mg/1 (˜ 32% of the original Zn). Thus, ˜ 32% of the Zn in skim milk is directly bound to caseins, while ˜ 63% is associated with CCP. Over 80% of the Zn in colloidal Pi-free milk was rendered soluble by 0·2 mM-EDTA, indicating that the casein-bound Zn is the loosely bound Zn fraction in casein micelles. A considerable fraction of the Zn in acid whey (pH 4·6) co-precipitated with Ca and Pi on raising the pH to 6·7 and heating for 2 h at 40 °C, indicating that insoluble Zn phosphate complexes form readily under these conditions. Studies on dialysis of milk against water, or dilution of milk or casein micelles with water, showed that CCP and its associated Zn is very stable and dissolves only very slowly at pH 6·6. The nature of Zn binding in casein micelles may help to explain the lower nutritional bioavailability of Zn in bovine milk and infant formulae compared with human milk.

Hydrogenated milk fats have been subjected to a process of fractional melting. In five of the six samples, the procedure yielded two fractions, defined as oil/upper solid and lower solid fractions, whilst in the remaining sample, three fractions (oil, upper solid and lower solid) were obtained. The oil and oil/upper solid fractions were rich in short-chain fatty acids (FA) (4:0 and 6:0), whereas the lower solid fractions contained large amounts of long-chain FA (16:0 and 18:0). Triglyceride distributions reflected these differences in FA composition, the oil and oil/upper solid fractions being dominated by triglycerides with carbon numbers (CN) 36–38; the dominant CN for the lower solid fraction was 48–54. Melting spectra, obtained using differential scanning calorimetry, showed that the ability of the fractionation procedure to separate on the basis of melting properties ranged from excellent to relatively poor, depending on the type of milk fat.

Gels varying in protein content and composition were prepared from whey proteins by a standard method and their compression and impact strengths were determined. No surface friction could be observed between the gels and the compression plates, and the Poisson's ratio was close to 0·5. Thus, the elastic moduli of the gels were determined. Compression and impact strengths and elastic moduli were related logarithmically to the β-lactoglobulin content of the protein powders. The strength and manner of fracture of the gels depended on the gel microstructure. Strong gels had a honeycomb structure consisting of a very tightly bound network, and failed by cracking on impact and by shear along cleavage planes in compression. Weaker gels had a loose open network with no definite means of fracture.

Accelerated ripening was successfully achieved in Feta cheese by the addition of a heat-shocked culture of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus, or neutral or acid microbial proteinase. Optimum levels of microbial proteinases were established at which trichloroacetic acid-soluble N (TCA-N) compounds were increased by about 35% in 40 d. Treated cheeses of this age had flavour intensities and body and texture qualities typical of 80 d-old control cheeses. Any further increase in TCA-N by adding larger amounts of enzyme resulted in bitterness. Heat-shocked culture treated cheeses developed similar characteristics to those prepared with neutral or acid proteinases. In contrast to the control cheese, it was necessary in the experimental cheeses to add minute amounts of lamb lipase (0·1–0·2 g/100 kg milk) to the cheese milk to keep the flavour balance. Best organoleptic characteristics were obtained from cheese treated with heat-shocked culture, followed by those treated with neutral or acid proteinase.

Milk protein genetic polymorphism has a major influence on the composition of milk, and on its processing properties, including yield of cheese (see Schaar et al. 1985; McLean et al. 1984, 1987; McLean, 1987). However, there appears to be little information on the effects of milk protein genetic variants on syneresis of cheese curd. The effect of casein composition on syneresis was studied by Pearse et al. (1986), who found that syneresis was affected only by the level of β-casein. Syneresis is an essential requirement in cheese making from renneted or acidified milk, but is undesirable during the storage of products such as yogurt. Milk for yogurt manufacture is preheated to minimize syneresis and to give maximal firmness of the yogurt coagulum (Tamime & Deeth, 1980). Pearse et al. (1985) showed that the reduction of one-third in the extent of syneresis caused by heating artificial micelle milk (AMM) containing βlactoglobulin (β-lg) in natural concentrations was due to sulphydryl-mediated complex formation between β-lg and micellar κ-casein which appeared to interfere with the micelle–micelle interactions responsible for syneresis. The results presented here were part of a study which investigated the effects of κcasein and κ-lg genetic variants and concentrations on syneresis of curd formed from renneted heated AMM.