The composition of casein in Irish milk for manufacturing was examined throughout the year. Large changes in composition were observed in the final quarter of the year as a result of the influence of late lactation when the proportion of αs- and β-caseins declined and the proportion of κ- and γ-caseins and of minor casein components increased. Similar changes in casein composition were induced by treatment of milk with porcine plasmin. The activity of plasmin-like milk proteinase in milk was quantified from the rate of conversion of added [14C]β-casein to [14C]γ-caseins. This activity was considerably higher in a late lactation milk than in a mid lactation milk and was ∼ 20 times higher at 37 °C than at 4 °C. It was concluded that casein compositional changes in late lactation were largely the result of proteolysis of individual caseins by plasmin-like milk proteinase.

A metalloproteinase, isolated from a shaken milk culture of Pseudomonas fluorescens AFT 36 by chromatography in DEAE and CM-cellulose and Sephadex G-150, was very unstable in 0·1 m-phosphate buffer, pH 6·6, being completely denatured above 70 °C in 1 min. It was also unstable in a Ca-containing buffer (synthetic milk salts, SMS) between 50 and 60 °C (minimum at 55 °C), but stability was very high above 80 °C in this buffer. D-values were determined at 10 °C intervals in the range 70–150 °C in SMS from which a Z value of 31·9 °C and an Ea of 8·82 × 104 J mol-1 were calculated; the half-life at 150 °C was 9 s.

Instability at 55 °C was due to autolysis as evidenced by gel electrophoresis, gel filtration and increase in 2,4,6-trinitrobenzenesulphonic acid-reactive amino groups. The extent of inactivation experienced at 80 °C was inversely related to the rate of heating to 80 °C, i.e. length of time spent in the neighbourhood of 55 °C. Addition of increasing concentrations of caseinate substrate reduced inactivation of the enzyme at 55 °C, presumably due to substrate binding. Attempts to stabilize the enzyme at 55 °C by addition of EDTA or by adjusting the reaction pH to 4·2, at which the enzyme has little proteolytic activity, were unsuccessful, although autolysis was prevented.

Unlike the proteinase from Ps. fluorescens MC 60, AFT 36 proteinase did not inactivate itself on cooling to 55 °C from 80, 100 or 150 °C, but did regain autolytic activity on cooling to below 50 °C to an extent dependent on the duration of holding at the lower temperature. It is suggested that on heating to ∼ 55 °C, a conformational change occurs which renders the enzyme susceptible to proteolysis by still active enzyme; at higher temperatures the enzyme, although susceptible to autolysis, is inactive; an active conformation is restored on cooling to below 50 °C.

Bovine caseins from skim-milk (micellar casein and acid precipitated casein) were separated into 4 discrete components by high performance gel chromatography on a TSK-GEL G4000SW columm. The resolution was better than that obtained by conventional gel chromatography which resulted in separation into 2 components at best. Our results showed that the second and third components consisted of αs1- and κ-caseins and αs1- and β-caseins respectively. They are presumed to be submicellar fractions. Samples prepared by removal of Ca phosphate from large and small micelles were rich in the third and the second components respectively. It is suggested that casein micelles consist of αs1- and κ-casein complexes and complexes and polymers of αs1- and β-casein which are situated preferentially on the periphery and in the interior of the micelles respectively.

The effect of homogenization on the heat stability characteristics of milk was examined. The heat stability of homogenized milk, as determined by the time taken for protein clots to form when heated at 140 °C, was reduced with increasing pressure in the range 3·5–34·5 MPa. The heat stability of homogenized milks was greater for samples obtained in the summer months than for those obtained in the winter. The general destabilizing effect of homogenization could be partly offset by 2-stage homogenization (20·7 MPa followed by 3·5 MPa), addition of phosphate stabilizers (0·08% w/v) or homogenization at a high temperature (65 °C). Whilst homogenized and unhomogenized milks reacted similarly to the addition of Ca, phosphate stabilizers, sulphydryl-blocking and oxidizing agents, the effects of season, addition of urea or formaldehyde were different for homogenized milk.

Using the Charm radioactive assay technique as a rapid test for the determination of penicillin in milk in experiments with milks containing known concentrations of penicillin, the coefficient of variation was found to be 15 and 11% for raw whole and reconstituted skim-milks respectively. Storage of reagents for up to 4 d increased the variance of the test by more than 20% and the estimated penicillin concentration 0·4 %/d of storage up to 25d. In comparison with the Bromocresol Purple Test, which was capable of 100% accuracy in defining a threshold of 0·02 i.u. penicillin/ml milk, the Charm Test falsely classified 15·6% of the samples.

Native proteolytic enzymes in good quality normal bovine milk readily hydrolyscd the caseins during incubation or storage, producing the γ-caseins, proteosc-peptone components 5 (PP5) and 8-fast (PP8F) and a considerable number of other unidentified fragments, many of which were also subsequently found in the proteose-pcptone fraction. The rate of casein hydrolysis was greater in pasteurized than in raw milk, with β-casein being slightly more susceptible to attack than αs1-easein. Measurements of γ-cascin and proteose-peptone formation have been made and it was found that PP5 was an intermediate product that was subject to further proteolysis while PP8F was a stable end-product. With the exception of component 3 (PP3), virtually all constituents of the proteose-peptone fraction increased during storage and appeared to be products of the action of proteolytic enzymes. Further evidence was obtained from the effects of various inhibitors that the principal proteinase of normal milk is plasmin, but slight differences were apparent between the protein breakdown patterns induced by storage and by added plasmin, which was consistent with the presence of more than one proteinase. Incubations in the presence of soya bean trypsin inhibitor to prevent plasmin action clearly revealed that another enzyme(s) was also involved.

The heat stability of homogenized concentrated milk was found to be more susceptible than that of unhomogenized concentrated milk to changes in the ratio of β-lactoglobulin to κ-casein which were deliberately designed to alter the extent of disulphide-linked complex formation during heating. When sulphydryl group interactions were prevented by the addition to the milk before homogenization of blocking or oxidizing agents, homogenization did not change the heat stability of subsequently concentrated milk.

Fourteen samples of milk fat, representing a wide range of individual fatty acid contents, were subjected to stereospecific analysis to determine fatty acid contents at the sn-l-, sn-2- and sn-3-positions of the triglycerides. Highly significant linear relationships were found between the content of a fatty acid at the 3 positions and the content of the same acid in the intact triglycerides. As the content of an acid increased in the triglycerides there were increases in the content of that acid at the 3 positions. When the content of an acid changed in the triglycerides, the content at the 3 stereospecific positions did not change at the same rate. The positions with the highest slope values corresponded to the positions where the individual fatty acids were preferentially esterified.

Milk obtained from cows which were either infected by clinical mastitis or had been subjected to intramammary infusion of Escherichia coli endotoxin possessed high counts of somatic cells and very high levels of proteinase activity which hydrolysed the caseins almost completely in a few hours at 37 °C. The rate of hydrolysis of β-casein was slightly greater than that of αs1-casein, but in both cases hydrolysis was enhanced by 6 cycles of freezing and thawing to disrupt somatic cell membranes. A study of the relationship between proteinase activity and cell count suggested that only some of the proteinase activity originated in the somatic cells and also that the identity of the cells making up the total cellular population was important. Maximum proteolysis occurred at 50–60 °C, but the temperature-activity curve was a broad peak. Likewise the pH versus activity plot was very broad and was almost flat over the pH range 6–9. Experiments with a number of inhibitors of proteinases failed to give a clear cut pattern of inhibition. All evidence obtained was consistent with the view that several different enzymes with different pH and temperature optima and different specificities contributed to the overall hydrolysis of caseins in these milks. From electrophoretic band patterns one of these enzymes was clearly plasmin, but in high cell count milks other proteinases also became significant.

A study of the inhibitory action of Debaryomyces hansenii (31 strains) on Clostridium tyrobutyricum (5 strains) and Cl. butyricum (2 strains) on laboratory media showed that Deb. hansenii inhibited the growth of these organisms, and that this effect was due not only to competition for nutrients but also to the production of both extra-and intracellular antimicrobial metabolites. The inhibitory effect varied with strain and occurred whether the yeasts were grown aerobically or under reduced O2 tension.

The role of blood serum lipoproteins in stimulating milk lipolysis was investigated by adding blood serum and heparin to the milk of normal cows and cows in which elevated somatic cell counts were induced by intramammary infusion of Escherichia coli endotoxin and Staphylococcus aureus. There was considerable variation between individual milks in the extent to which lipolysis was stimulated. In some milks there was almost no response. For values obtained from 29 cows during the first 6 months of lactation a relationship was observed between free fatty acid (FFA) levels in the untreated and serum-stimulated samples (r = 0·776). FFA values were higher after 18 h at 4 °C in milks from infused quarters than in those from control quarters, but this difference was not entirely due to higher values from the infused quarters at 0 h. Both blood serum and heparin stimulated lipolysis in high cell count milks and control milks. The response to heparin was greater than that to blood serum, but the response to both was highly correlated with FFA levels in the untreated milks. There were no differences in the ability of skim-milks prepared from infused and control quarters to promote lipolysis in recombined milks containing sterilized homogenized milk fat globules as a source of substrate. Incubation of blood serum with 2 μg/ml trypsin for 1 h at 37 °C reduced the ability of the serum to promote lipolysis. The importance of lipase activation by blood serum lipoproteins in relation to milk lipolysis is discussed.

The dynamic viscosities of non-homogenized and homogenized (2-stage) whole and half skim-milk and cream, of unhomogenized skim-milk and of water-diluted milks were measured at temperatures between 70 and 135 °C; this range includes those temperatures used in ultra-high-temperature (UHT) treatments. A tube viscometer was used at the outlet of a continuous flow indirect heater. Above 85 °C the mean residence time of a milk or cream at the test temperature was 3–6 s.

An equation relating viscosity μ and temperature θ of the form In μ = aθ2+bθ+c was derived for each product.

It proved possible to relate μ, θ and fat content (0·03–15·5%) in a single equation which was appropriate for all the homogenized milks and creams as well as the non-homogenized skim-, half skim- and whole milk; the equation is considered valid for these products under conditions of indirect UHT sterilization. Viscosity values calculated with the equation, even those relating to temperatures as low as 50 °C, were comparable with those reported in the literature.

The caseinate complex in bovine milk was partitioned by differential centrifugation at both 20 and 4 °C into 4 micellar fractions and a fraction representing serum casein, and the protein composition of the fractions determined. At both temperatures the relative amount of κ-casein in the micellar caseins increased markedly and that of β-casein decreased appreciably with decreasing micelle size. The relative amount of αs2-casein also tended to decrease with decreasing micelle size, but the relative amounts of αs1- and γ-caseins, and an unidentified casein fraction, showed little systematic variation. The serum casein differed appreciably in composition from the micellar caseins, being very rich in β-casein and comparatively poor in αsl- and αs2-caseins, and the amount present at 4 °C was considerably greater than at 20 °C, with the increase being due almost entirely to β-casein, but with γ-casein also making a significant contribution. The changes in the composition and distribution of micellar and serum caseins induced by cooling milk at 4 °C were completely reversible when the milk was re-equilibrated at 20 °C for 18 h.

The concentrations of dihydroxyacetone phosphate (DHAP), glycerol- 3-phosphate (GP), glucose and alkylglycerols were measured in bovine colostrum and milk. Concentrations of GP were higher in colostrum than in milk, and corresponded to intracellular concentrations at the onset of lactation. This suggests that GP equilibrates over the apical membrane at onset of lactation, which is believed to occur also in full lactation. The ratio of DHAP to GP showed a peak in colostrum that was more than twice that in milk, indicating that concentration changes in colostrum/milk were not caused by dilution. DHAP was positively correlated to metabolical compounds such as glucose and alkylglycerols. As DHAP is a required precursor for alkylglycerol synthesis, the positive correlation suggests that DHAP may be limiting for alkylglycerol synthesis and thus could explain the elevated levels of alkylglycerols found in colostrum.

The extracellular acid lipase produced by Penicillium roqueforti strain 1173 was purified by precipitation with 50 % ethanol, DEAE Sephadex A 50 ion exchange chromatography and 2 successive treatments with SP Sephadex C 50. An ~ 400-fold purification with a 35 % yield was obtained. The enzyme was active from pH 1·7–10·5, with an optimum activity pH of 6·5 at 20 °C and 6·0 at 30 °C. A second minor optimum appeared close to pH 2·5 and there was a minimum activity at pH 3·6. In the presence of 0·1 M-CaCl2 at 30 °C, the optimum was at pH 5·5, also the pH value where inhibition by CaCl2 was least (9%). The pH stability curve at 30 °C showed no discontinuities and had a maximum between pH 3·7 and 6·0. At pH 6·5 activity at 5 °C represented 37 % of the maximum value, which was reached at 35–40 °C and the apparent activation energy was about 5·42 kcal mol-1. Between pH 4·0 and 7·0, the ratio of hydrolysis velocity at 30 °C compared to that at 20 °C ranged from 2·1 to 1·2. At pH 6·0, the enzyme began to be inactivated at temperatures above 35 °C with decimal destruction times (D values) of 150, 29 and 4 min for respective temperatures of 40, 45 and 50 °C. In terms of substrate specificity the maximum activity was obtained with tricaproin whilst with tributyrin, tricaprylin, butter oil and triolein relative velocities were 22, 6·7, 3·0 and 1·5% respectively of that for tricaproin whilst Tween 20 and triacetin were 2·5 and 2% respectively.

A quantitative method for rapid routine analysis of individual free fatty acids (FFA) in milk was developed. Lipid was extracted from milk in ether and FFA were recovered by shaking the extract with anion exchange resin Amberlyst 26. The resin-bound FFA were methylated directly and the individual acids quantified, using internal standards, by gas-liquid chromatography. The properties of the resin were measured. The validity of the method was established by extraction of FFA mixtures and milk. Individual acids were, on average, found to be within 6% of the actual concentration present in the mixture. An average coefficient of variation of 4·3% was achieved for the major individual fatty acids on repeated extraction of a single milk sample.

In order to examine the influence of micelle size on the formation of soluble casein brought about by heating concentrated milk, casein micelles were separated into 3 fractions by differential centrifugation. The micelles were dispersed in the ultrafiltrate of skim-milk and the preparations were concentrated to 1/2·5 of the original volume. When the concentrated milk samples were heated at 135–140 °C, a larger amount of soluble casein was formed from smaller micelles than from larger micelles. Only a slight effect on the formation of soluble casein on heating was observed when colloidal calcium phosphate (CCP) was removed from the large micelles to a level lower than that associated with the small micelles or when the CCP in the small micelles was increased to a level higher than that of the large micelles. The soluble caseins formed on heating from different sizes of micelles were similar in composition. The fact that the formation of soluble casein on heating was affected by micelle size may be ascribed to variation in the κ-casein content of micelles rather than to variation in CCP content.

Aeration increased the growth and lipase production in milk by Pseudomonas fluorescens strain AFT 36, isolated from refrigerated bulk milk. A heat-stable lipase was isolated from a shaken milk culture of this microorganism by DEAE-chromatography and gel filtration in Sepharose 6B. The lipase-rich fraction from DEAE cellulose contained 3 lipases that were separated by gel filtration; only the principal lipase, which represented ∼ 71 % of total lipolytic activity, was characterized. The purified enzyme showed maximum activity on tributyrin at pH 8·0 and 35 °C; it had a Km on tributyrin of 3·65 mM. and was inhibited by concentrations of substrate > ∼ 17 mM. The enzyme was very stable over the pH range 6–9; it was relatively heat-labile in phosphate buffer in the temperature range 60–80 °C, where it was stabilized significantly by Ca2+. It was, however, very stable at 100–150 °C: the D values at 150 °C were ∼ 22 s and 28 s in phosphate buffer and synthetic milk serum respectively; the corresponding Z values in the temperature range 100–150 °C were ∼ 40 and ∼ 42 °C and the Ea for inactivation were 7·65 × 104 J mol-1 and 6·97 × 104 J mol-1 respectively.

The rates of coagulation of completely renneted casein micelles have been measured as functions of ionic strength, temperature, and concentration of Ca2+. At 25 °C and below, the rate constants for the coagulation were found to be low, but increased with temperature so that at 60 °C the particles were coagulating at almost maximum rate permitted by diffusion. This maximal rate at 60 °C was achieved at nearly all of the ionic strengths and concentrations of Ca2+ used. At lower temperatures the rate constant decreased with increasing ionic strength, the dependence being more marked at lower temperatures. Increasing concentration of Ca2+ also increased the rate at low and moderate temperatures. The implications of these results are discussed in terms of specific and non-specific ionic interactions and of hydrophobic bond formation.

An extract of Kluyveromyces lactis 416 and a β-galactosidase preparation (Maxilact 40000) contaminated with proteinase, showed similar pH profiles of caseinolytic activity. Similar modes of casein hydrolysis (κ-, > αs-, ≥ β-) were observed at pH 5·0 (the pH of Cheddar cheese), without detection of bitterness. The contaminated Maxilact preparation contained similar proteinase types to those detected in an autolysate of K. lactis. Both the autolysate and the Maxilact preparation contained acid endopeptidase (proteinase A), serine endopeptidase (proteinase B) and serine exopeptidase (carboxypeptidase Y) activities. Some aminopeptidase activity was also detected in both preparations. There were some differences in apparent molecular weight and charge properties between proteinase A and B and carboxypeptidase Y from the 2 proteinase sources.