The acid phosphatase of bovine milk was further purified to yield enzyme with an activity of about 2 units/mg. This was almost 105 times the activity present in milk and enabled a detailed study of heat inactivation to be made, together with further measurements on binding to casein substrates.

The effectiveness of caseins as inhibitors of the hydrolysis of p-nitrophenyl phosphate by acid phosphatase paralleled the phosphate content of the casein molecules, so that αs1-casein A was a more potent inhibitor with a K1 of 1·7 mM than β-casein A1A2 (K1 = 4·3 mM), which in turn was more inhibitory than κ-casein A (K1 = 5·9 mM).

The heat inactivation of acid phosphatase followed first-order kinetics at pH 4·9, 5·2 and 6·7 and values of E, the activation energy, were between 2·4×105 and 3·0×105 J mole −1 in all cases, consistent with simple protein denaturation. The presence of 1% αs1-casein A, 1% β-casein A1A2, 1% κ-casein A, 1% isoelectrically precipitated ‘whole’ casein and 1% fresh raw milk provided no substrate protection at pH 5·2 or 6·7. Acid phosphatase was somewhat less heat stable at pH 6·7 than at pH 4·9, but may be expected to survive typical milk pasteurization conditions almost completely. However, conventional milk sterilization or ultra-high-temperature (UHT) processes would be expected to give total inactivation.

Gel filtration on Sepharose 6-B was used to study the size distribution of sub-micellar casein aggregates in colloidal phosphate-free milk, and comparisons were made between the elution profiles obtained for fresh, heated, ultra-hightemperature (UHT) treated and stored UHT milks. Increase in length of storage of UHT milk gave rise to an increase in proportion of casein aggregates excluded from Sepharose 6-B, but no corresponding increase in size of included casein aggregates was observed.

The casein micelles of sheep (Ovis aries), goat (Capra hircus) and cow (Bos taurus) milks have been examined by electron microscopy. The ovine micelles were smaller than bovine micelles, most being below 80 nm diam. The caprine micelles were either large (c. 200 nm diam.), easily sedimented and electron-dense, or small (below 80 nm) and similar to the ovine micelles. Examination of the micelles using the freeze-etch technique showed, however, that they were all composed of subunits of similar dimensions (10–15 nm diam.).

Electrophoresis of the caseins showed that there were differences between them, with the caprine casein containing a smaller proportion of its protein in the more mobile components. The mineral contents of the milks were also different, but these were apparently related to the different pH and casein contents of the milks. It is likely that the diiferences in micelle size distribution are determined by the physical and chemical properties of the casein components.

The design and operation of a laboratory-scale ultra-high-temperature sterilizer capable of treating 0·5-l quantities of milk are described. Temperature control, sporicidal efficiency and change of composition of the milk during processing are discussed.

Diacetyl, acetoin, 2,3-butyleneglycol, 2-butanone and 2-butanol concentrations have been determined in ripened Cheddar cheese made using Streptococcus diacetilactis DRC1 or Str. cremoris AM2 or HP as starter culture. Diacetyl content was found to be initially higher in the DRC 1 cheese than in the other 2 cheeses, presumably because this starter readily ferments citrate. The disappearance of acetoin appears to be related to the formation of 2,3-butyleneglycol in DRCl and AM2 cheeses, but not in HP cheese. Only trace amounts of 2-butanone and 2-butanol were observed in the 3 cheeses even though the concentration of a postulated precursor, 2,3-butyleneglycol, was present in considerable concentration in the DRC1 and AM2 cheeses. Analysis of a number of mature AM2 and HP cheeses selected at random indicated the presence of 2-butanone in only a few of the AM2 cheeses. It is suggested that this compound may have arisen from the presence of a micro-organism which entered the cheese vat at the time of cheese-making.

It has been shown that the inhibition caused by heat treatment, of the primary phase of rennin action on casein micelles, is dependent on the presence of β-lactoglobulin. The degree of inhibition increased with increasing amounts of added β-lactoglobulin for both heated casein micelles and heated skim-milk to a constant value. The results are fully consistent with the hypothesis that the inhibition is caused by complex formation between β-lactoglobulin and κ-casein when milk is heated.

The bacteriological formation of 2-butanone and 2-butanol was investigated. The data suggest that 2-butanol is formed in cheese via 2-butanone in 3 steps, each step being carried out specifically by a different microbial species. Suitable species have been isolated from a single cheese. The first step could be carried out (in the present investigation) either by certain starter organisms or by a strain of Pediococcus cerevisiae. This strain formed substantial amounts of 2,3-butylene glycol in milk, presumably from citrate, but could not metabolize this compound further. A second micro-organism, a strain of Lactobacillus plantarum, was isolated which, in Cheddar cheese, could convert 2,3-butylene glycol into 2-butanone. The final reduction of 2-butanone to 2-butanol could be readily carried out in MRS medium by a strain of L. brevis. Results from cheese made aseptically indicate that 2-butanone is formed in Cheddar cheese only when non-starter organisms are present.

Methods are described for the solubilization of the nitrogenous portion of cheese in a dissociating solvent system and separation of the solution into fractions on a Sephadex G-100 column under conditions where the proteins and peptides would be expected to be split into monomeric units. The E280 measurement of the fractions, expressed as a percentage of the total material eluted, followed a reproducible pattern, both with duplicate runs on the same sample and comparative runs on non-identical samples of the same type. With cheese curd, a small peak of undissociated material was obtained at the void volume of the column; this was followed by a main peak and a minor third peak, with a small quantity of material being eluted at larger volumes. A qualitatively similar pattern was obtained with 43-day-old cheese as the sample, but the main peak was smaller, the third peak was larger, and a relatively larger amount of material was eluted at greater volumes. There was only partial separation of the proteins on the column; the main peak contained β- and αs-caseins, their larger breakdown products and some para-κ-casein, and the third peak contained smaller breakdown products of the main casein fractions and para-κ-casein. Although the resolution was considerably inferior to that obtained by gel electrophoresis, the method has the advantage over this and other methods in that it gives a quantitative analysis of the sample based on molecular size. Thus, it is suggested that it may find use in analysing protein breakdown in cheese during ripening.

Cheddar cheeses made using single strains AM2 and HP of Streptococcus cremoris as starter were analysed for methyl ketones during ripening at 7°C. The level of odd-numbered C3–C15 methyl ketones increased very slowly in all cheeses and at maturity represented only 15–20% of the level potentially available from the degradation of esterified β-ketoalkanoic acids of milk fat in cheese. The potential for formation of the methyl ketones in unripened cheese curd was similar to the sum of free and potentially available ketones in mature cheese which suggests that the oddnumbered ketones once formed are not reduced or otherwise degraded by the normal cheese microflora. Significant variations were observed in the level of 2-butanone, which, in some cases, far exceeded the quantity which could be formed by breakdown of β-ketovaleric acid. Organoleptic evaluations and comparison of the levels of methyl ketones present in mature cheese with published threshold data suggest that while the methyl ketones may contribute to flavour they do not in their own right impart a typical mature Cheddar flavour to cheese.

The complexing of whey proteins in hydrochloric acid whey by carboxymethyl cellulose (CMC) has been studied. The efficiency of precipitation of CMC (the ability to form insoluble complexes with whey proteins) increased with increasing degree of substitution (DOS) of CMC. At a DOS of greater than 1·4, efficiencies of precipitation of whey protein greater than 90% were achieved without the need for dilution of whey. CMC of high DOS would thus be useful in whey protein reclamation schemes. The amount of CMC necessary for maximum efficiency of precipitation also increased with increasing DOS. The composition of the CMC-protein complex varied little over the range pH 2·8–4·0. Under optimum conditions for binding, the complex contained 61–68% protein and more than 95% of the CMC originally in solution was precipitated in the complex.

A method by which it is possible to quantify with very high accuracy the levels of rennin, bovine pepsin, porcine pepsin and proteases from Mucor miehei (fromase) or M. pusillus (emporase) in a ‘commercial rennet’ containing some or all of these enzymes is described. The method employs chromatography on DEAE cellulose in piperazine buffer (0·02M, pH 5·3) to fractionate the enzymes into 2 groups: (peak A) rennin, emporase, and fromase, eluted at ∼0·2 m-NaCl and (peak B) bovine and porcine pepsins eluted at ∼0·4 m-NaCl. The rennin in peak A is selectively denatured at pH 9·0, 20°C for 60 min and the rennin plus emporase at pH 6·0, 60°C for 20 min, permitting the quantification of emporase and fromase (and rennin by difference) and fromase, respectively. The porcine pepsin content of peak B is quantified by selective denaturation at pH 7·0, 30°C for 3 min. A number of commercial rennet preparations were analysed using this method: full rennet contained 45–60% bovine pepsin and ‘half and half’ contained 20–25% rennin, 25–30% bovine pepsin and 50% porcine pepsin.

The rates of proteolysis during ripening were followed in cheeses made with either calf rennet or swine pepsin and either starter or δ-gluconic acid lactone (GAL) as a replacement for the starter. A gel-filtration column technique and starch-gel electrophoresis were used for analysis, and bacterial counts were made on all samples. Proteolysis was faster in cheeses made using GAL than in those made using starter and also slightly faster in GAL cheeses made with swine pepsin than in those made with rennet. Further, it was considerably slower in starter-containing cheeses made with swine pepsin than in those made with rennet. It is suggested that these differences were due to the much greater rate of development of acidity in cheeses made with GAL than in those made with starter, which resulted in more of the coagulant being incorporated into the curd in an active state. The rate of proteolysis in starter-containing cheeses appeared to follow a characteristic course, being initially slow, then markedly increasing with a later slow decline. It is suggested that the increase in the rate of proteolysis was due to an increase in the total activity of bacterial proteinases released by lysis of the bacteria. Indications were obtained that the coagulants and bacterial proteinases catalysed broadly similar patterns of protein breakdown in cheese, and that medium-sized peptides (mol. wt 9000–14000) were formed as definite intermediates in the process. The results also showed that rennet and swine pepsin remained active for at least 7 months in GAL cheeses, that rennet contributed significantly to proteolysis in starter-containing cheeses, and that swine pepsin was at least extensively inactivated and possibly completely inactivated during cheese-making with starter.

Comparisons are made between the sulphur-containing volatiles of 4 Cheddar cheeses, using a gas chromatograph equipped with a flame photometric detector.

Two of the cheeses were made aseptically from whole milk, with and without starter, the third was made from whole milk with starter in an open vat, and the fourth was made from skim-milk also in an open vat. Only the cheeses made from whole milk with a starter culture developed normal Cheddar flavour; the cheese made without starter and the skim-milk cheese had little or no flavour.

Considerable differences were found in the chromatograms of the 4 distillates and in particular methanethiol was found only in the distillates from the cheeses having Cheddar flavour.

The development of structure in Cheddar cheese was followed using electron microscopy, of both thin sections and freeze-etched preparations. The casein micelles, at first separate, were seen to aggregate into a network, coalesce and finally form a granular mass. The fat globules, also separate at first, were gradually forced into clumps as a result of shrinkage of the casein network and incomplete fusion took place during ripening. Starter bacteria were seen trapped in the casein near the fat–casein interface, which was shown to be the region of highest water content in the mature cheese.

Concentration of milk by reverse osmosis is severely hindered by the formation of deposits on the membrane. Electron micrographs of this deposit show a thin (11nm) continuous electron-dense layer adjacent to the membrane. An electronlucent zone (10–15nm thick) separates this layer from the main bulk of the deposit which is 30 μm thick and consists largely of granules with some vesicles in a diffuse matrix. The density of packing of the granules increases with proximity to the membrane. Their appearance, size and digestion with pronase suggest that these are milk protein granules. The order and nature of these various layers could explain the pattern of flux decline.

Treatment of unfractionated bovine casein with papain produced a digest which was bitter in taste. The bitterness was shown to be a property of a peptide originating only in β-casein and having an amino-acid composition consistent with its derivation from residues 53–79 inclusive of β-casein. The relationship existing between flavour and primary structure in this peptide is discussed briefly.

Arterio-venous (AV) blood plasma concentration differences of amino acids across the mammary glands of 2 lactating goats were measured at intervals throughout a day. One gland of each animal had been transplanted to the neck for experimental purposes. The variation throughout the day in arterial concentration of all amino acids was similar. The variation in AV differences was slight for most essential amino acids, greater for glutamic acid and proline and very marked for aspartic acid, alanine, glycine and citrulline. There was no statistical difference between the AV difference of any amino acid measured simultaneously across the 2 glands of either goat. The arterial concentrations of certain pairs of amino acids were significantly correlated. The implications of the results for estimation of rates of amino-acid synthesis in the mammary gland using the intra-arterial infusion technique are discussed.

A method has been developed for the routine determination of acetic, propionic and butyric acids in cheese. The cheese sample is homogenized with water (1:2) at 40°C for 15 min and the slurry centrifuged at −5°C and 12000 g for 10 min. The free fatty acids in a portion of the aqueous centrifugate are adsorbed on to a column of anion-exchange resin in the hydroxide form. After washing the column with water to remove extraneous material the acids are eluted from the resin with aqueous formic acid. Amino acids in the acidic eluate which interfere with subsequent analyses are removed by a cation-exchange resin column in the hydrogen ion form. The concentrations of acetic, propionic and butyric acids in the cation-exchange resin eluate are determined by isothermal gas chromatography using a glass column (1·8m×2mm i.d.) packed with Chromosorb 101 (100/120mesh). Cheese oil-water partition coefficients are then used to calculate the concentrations of acetic, propionic and butyric acids in the cheese sample.