The apparent relative viscosity (ηr) of stored homogenized ultra-heat-treated ‘coffee’ cream increased with time (t) at a rate which was a function of mean fat globule diameter (d) and fat volume fraction (ø); the rate was also influenced by Na2CO3 and Na citrate when present as stabilizing additives. Viscosity increases were attributed to the progressive flocculation of fat globules and to the strengthening of the structure formed by the gradual accretion of casein micelles on to globule surfaces and their points of contact (bridging). The expression

where t is in days and d in μm, was found to hold over a 12-week storage period for a batch of creams (ø = 0·143) of different d values prepared from the bulk cream containing carbonate and citrate stabilizers; η0 is the relative viscosity at t = 0 and k is a rate parameter. In the absence of stabilizers the parameter k assumed a smaller value and so appeared to be a function of chemical composition of the cream and of forces of particle interaction; seasonal factors were also suspected of influencing k. The effect of the fat volume fraction on storage behaviour was not fully resolved, but there was some evidence that the proportionality log ηrα1/ø½ was appropriate.

The proportion of materials which adsorb to casein micelles and accelerate milk coagulation, and remain bound on dilution of the micelle suspension with milk dialysate was determined. This was higher than expected from the postulate that equilibration took place between the solution and many equivalent binding sites on the micelle. This suggests that binding involved either or both hydrophobic interactions and multipoint attachment to charged groups. The location of the additive binding sites in the micelle was investigated using models Na caseinate and hydroxyapatite at pH 6·6. Materials which accelerated coagulation bound to either or both models. Those with the greatest effect on milk coagulation probably bound primarily to the casein rather than to the colloidal calcium phosphate moiety of casein micelles.

Chemical and physical properties of plasteins derived from caseins and skim-milk powder have been investigated. Plasteins were very similar in properties regardless of starting material or of the proteinase used in their production. Plastein formed from Na caseinate hydrolysate at pH 5 was stable over the pH range 1·5–9·0 and had the properties of a thixotropic gel. More extreme pH values caused dissolution as did various chaotropic reagents. 8 M-urea disrupted plastein structure almost entirely, as did various organic acids at concentrations of 50% (v/v) and strongly ionic detergents. Non-ionic or weakly cationic detergents and 6 M-guanidine hydrochloride were less effective. Treatment with strong salt solutions, Ca2+ and EDTA had no effect. Gel filtration experiments demonstrated that plasteins were completely disrupted under dissociating conditions and peptide mapping confirmed that all the peptides originally present in the hydrolysate were also present in plastein. Amino group contents were unchanged on plastein formation. These experiments together with gel electrophoresis studies clearly showed that only hydrophobic and ionic bonding mechanisms were involved in plastein formation and no detectable peptide bond formation or transpeptidation occurred. In accordance with this, amino acid analysis suggested that hydrophobic peptides were preferentially incorporated. Implications for use of plasteins in food products are discussed.

The effect of chemical modification of milk protein on the ethanol (EtOH) stability of skim-milk has been investigated. Modification of lysine residues by reaction with anhydrides increased the EtOH stability whereas amidation of carboxyl groups destabilized the milk. Lysine modification also counteracted destabilization induced by Ca2+ addition. By using multiple linear regression techniques these latter effects can be accounted for in terms of the changes they bring about in micellar charge. These results provide a further verification and an extension of the previously proposed mechanism for EtOH-induced coagulation of skim-milk.

A fermented milk with typical yoghurt flavour and aroma was prepared using a human strain of Lactobacillus acidophilus. Growth of the lactobacillus was stimulated when the whey proteins of the ultrafiltered skim-milk substrate were increased by addition of ultrafiltered Cheddar cheese whey. Fermentation, which was carried out at 37 °C, was accelerated by addition of lactase. The resulting product had a smooth consistency, a pH of 4⋅2, acetaldehyde levels up to 15 p.p.m. and no off-flavours. L. acidophilus remained viable at a level of approx. 108 colony forming units/ml, with no impairment of quality, for up to 14 d at 4 °C.

The calcium-induced aggregation of αs1-casein at pH 7·0 has been studied in the presence of citrate and phosphate anions. The effects of the citrate can be reconciled with the suggestion that the citrate, acting as a Ca chelator, reduces the Ca available for the aggregation reaction, but in no other way interferes with the mechanism of that process. In the presence of phosphate, precipitation is induced below the normal critical Ca concentration. This reaction proceeds to completion at low and high concentrations of phosphate. There is, however, an intermediate range of phosphate concentrations, whose limits depend on the protein concentration, where a colloidally stable precipitate is produced. The significance of these observations is discussed in terms of the currently held theories of micellar structure and formation.

Caprine skim-milk exhibits markedly lower ethanol (EtOH) stability than bovine skim-milk but can still be characterized by a sigmoidal pH profile. As with bovine milk, the position of this profile along the pH-axis was found to be sensitive to available Ca levels. Manipulation of salt levels, either by serum interchange, addition or diminution did not result in any significant increase in the EtOH stability high pH asymptote, Smax, and reduction of the colloidal calcium phosphate of caprine milk also had no significant effect. EtOH stability/pH profiles similar to those of bovine milk were achieved only by chemical modification of the caprine milk protein by reaction with aldehydes and anhydrides. It is concluded that the low EtOH stability of caprine milk as compared with bovine milk is due to the different proportions of the individual caseins present, in particular the lack of an αsl-casein homologue in caprine milk.

Commercial bulk milk contained, in its acid-soluble nucleotide fraction, small quantities of CMP and AMP as well as a high proportion of orotate (503±16 μg/g of total solids). These components were altered by thermal treatment during processing. Standard sterilization and spray drying of milk lowered significantly the CMP, AMP, and orotate concentrations. HTST pasteurization caused a slight decrease in CMP and AMP concentrations, but the orotate content remained unchanged. The UHT sterilization seemed to be the thermal treatment in which the acid-soluble nucleotides of milk underwent least change. It is therefore concluded that the decrease in concentration of the nucleotides of milk is more directly related to the time for which the milk remains at the processing temperature than to the temperature itself.

The heat stability of skim-milk powder varied throughout the season, but was highest during the period April–August. Addition of urea increased heat coagulation times during March–August, but the extent of increase was dependent upon urea addition either before or after preheating for 30 min at 90 °C during manufacture. During March–June, higher heat stabilities were recorded when urea was added after preheating, while in July–August urea added before preheating was more effective. Partial ultrafiltration before dehydration decreased the heat stability of reconstituted milks due to a loss of natural urea in the permeate together with a disruption in the salt balance.

Effects of pH, added Ca, rennet concentration, temperature, concentration and dilution of milk, on rennet clotting and coagulum development have been investigated. Rennet clotting time (RCT) was decreased by reduced pH, by increased temperature and by increased concentrations of rennet or added Ca. Only extremes of concentration or dilution of milk increased RCT. Strength of milk coagulum at 1 h after RCT was increased by reduction in pH and by increased concentrations of added Ca. Increased temperatures decreased coagulum strength whereas increased rennet concentrations had little effect. A curvilinear relationship between total casein concentration and coagulum strength was found for Friesian and Jersey milks. Two phases of coagulum development after clotting were observed. Changes in pH particularly affected the first phase whereas changes in temperature particularly affected the second. Effects of rennet, casein and Ca were similar for both phases.

The acid-soluble ribonucleotides of human milk were measured by enzymic and ion-exchange chromatographic procedures at different stages of lactation. Human colostrum and milk contained CMP, AMP, GMP, UMP, GDP-mannose, UDP-N-acetyl-glucosamine, UDP-N-acetyl-galactosamine, UDP-glucose, UDP-galactose, UDP and other minor nucleotides. Cytidine and adenosine derivatives were present in relatively higher amounts than in milk from ruminant species. The nucleotide concentration decreased with advancing lactation. Human milk at 3 months of lactation contained about 10 µmol/100 ml nucleotides, which represented about 75% of the amount of nucleotides present in human colostrum. Human milk did not contain orotic acid. The UDP-glucose/UDP-galactose ratio was constant during lactation and was similar to that of milks from ruminant species.

Continuous measurements of curd-firming using a commercial vibratingreed viscometer and a pressure transmission system (PTS) developed at this Institute were compared both in the laboratory and during Cheddar cheesemaking. The coefficient of variation about the mean of the curd-firmness measured in reconstituted skim-milk 40 min after renneting was 3·5% with the viscometer and 2·6% with the PTS. The effects of changes in conditions on the curd-firming rate of raw whole milk were determined. As the rennet concentration was raised the rates increased linearly as measured by the viscometer, but non-linearly as measured by the PTS. Addition of CaCl2 increased the rates measured by both instruments rapidly at first then more slowly at higher levels. Raising the temperature increased the rate exponentially in the case of the viscometer but more slowly with the PTS. With both instruments the curd-firming rates of milk at pH 5·8 were 6·5 times those at pH 6·7 but the rates fell slightly at pH 5·6. In experimental cheesemaking, the mean readings of both instruments at the subjectively determined cutting time of the curd had the same coefficient of variation.

The commercial viscometer is probably more suitable for laboratory investigations of the earlier stages of the coagulation of milk by rennet whilst the PTS may prove the more useful in indicating the optimum cutting time of the curd during cheesemaking.

The proteins of bovine milk are the best-characterized food-protein system. Lactoproteins, which have high biological value, contribute ∼25% of dietary protein in North West Europe, North America and Oceania. However, in protein-rich western diets, milk proteins are frequently more highly valued for their functional properties than for their nutritional qualities. The remarkably high heat stability of the caseinate system permits the manufacture of a range of sterilized, concentrated and dehydrated products while its gelation on very limited proteolysis is the basis of cheese manufacture. Skim-milk powders, caseinates and whey protein concentrates are the most flexible and widely used functional proteins in food processing.

This communication reviews recent studies on milk proteins with respect to molecular and colloidal properties; coagulation by Ca2+, heat and ethanol; and functional properties and their chemical and enzymic modification.

The efficiencies of small capacity homogenizing valves of the high pressure type have been studied in relation to their design. The performance of a valve was strongly affected by the magnitude of the effective inlet diameter D0 of the seating for a given exit diameter D when D0/D > 0·5. The indications are that narrow mating faces situated such that D0/D → 1 should maximize performance. Milk fat globule dispersion varied regularly with a valve's face angle; optimum angles for maximum disruptive effect were about +35° and –30°. Perforated metal insets interposed between flat faces of a valve piece and its seat noticeably improved fat dispersion; reductions in mean globule diameter of up to 25% were obtained.

Variations in the sensitivity of different milks to homogenization were observed but avoided by appropriate experimental procedures.

The effect of aqueous algal extract, yeast extract casein peptone and meat peptone on acid production by 4 strains of lactic acid bacteria was tested manometrically using a milk–bicarbonate medium. A strain-dependent stimulation of the bacteria was detected after 4 h incubation. After fractionation of an aqueous extract of the green alga Scenedesmus obliquus on a Sephadex G25 column, 2 fractions with stimulatory activity were found using Streptococcus lactis as the test organism. One of the peaks of activity, confined to shortening of the lag phase, was due to hypoxanthine which, however, had only a slight stimulatory effect in conventional milk souring tests. The manometric method provides a sensitive and rapid test for detecting compounds with biological activity in µg amounts, but it should be accompanied by milk souring tests for interpretation of the type of stimulation involved.

Destabilization of emulsion droplets is described in terms of flocculation, coalescence and adsorption processes, and then discussed with reference to various technological aspects of dairy emulsions. A semi-quantitative measurement of homogenization-induced flocculation of cream is described using measurements on the time-dependent decrease in viscosity that occurs during shear. The whipping of cream is discussed in terms of the adsorption of fat droplets to air bubbles and the formation of a fat particle network with the fat directly in contact with air. The whipping of homogenized creams is shown to be dependent upon the structure of the newly formed milk fat globule membrane. Since the major protein component of the membrane is casein, any change in casein micelle stability is reflected in the whipping properties of the cream.

The presence of crystalline fat within emulsion droplets is essential for many destabilization processes. Fat crystals pierce the film between 2 approaching interfaces, thus leading to aggregation. The contact angle between fat crystal, oil and aqueous phase can be used to describe the susceptibility of an emulsion to destabilize.

A quantitative measure of emulsion destabilization under shear is described kinetically in terms of reversible flocculation followed by coalescence. The measurement of emulsion viscosity as a function of time at constant shear stress is used to monitor the aggregation processes and the 3 rate constants evaluated by computer curve fitting.

The aspartyl proteinases of Penicillium caseicolum and P. roqueforti acted identically on β-casein; both enzymes split at least 3 bonds: Lys29–Ile30, Lys97–Val98 and Lys99–Glu100. From αsl-casein, these proteinases released 6 main degradation products which arose from the splitting of 4 bonds; P. roqueforti aspartyl proteinase was found to cleave 1 bond at a higher rate than P. caseicolum aspartyl proteinase. A hypothetical sequential hydrolysis mechanism of αsl-casein by these 2 enzymes is proposed from a study of the degradation by isoelectric focusing and by 2-dimensional electrophoresis.

The development of typical flavour in Cheddar cheese was accelerated by the addition of bacterial neutral proteinase to the curds, but not by the addition of fungal acid proteinase which produced a bitter defect even at low concentrations. An optimum level of neutral proteinase addition was established at which general proteolysis was increased by about 20% in 2 months and treated cheeses of this age had flavour intensities typical of 4-month-old untreated cheeses. Enzyme-treated cheeses were softer-bodied and more brittle than untreated cheeses of the same age.

The performance characteristics of a valve accepting a high rate of milk flow and operating at Reynolds’ numbers (Re) of 6000–15000 have been studied. Features of valve design which improved homogenizing efficiency were identified. As with small valves operating at lower Re numbers, efficiency increased as D0/D increased beyond 0·5; for D0/D ≈ 0·7, a 15% reduction in mean globule size was obtained. Changes of face angle up to 45° had virtually no effect on milk fat globule dispersion. The use of impact rings produced only small improvements in dispersion. The incorporation of perforated insets led to substantial globule size reductions of 15–25% or the equivalent in reductions of homogenizer energy consumption of 12–35%.