Heat coagulation time-pH curves at 140°C were obtained for 43 blended skim milk samples from Holstein cows to determine the effects of genetic variants of κ-casein and β-lactoglobulin on milk heat stability. The blended milk samples were similar in terms of protein content and milk salts, but were genotypically different for κ-casein (AA, AB) and β-lactoglobulin (AA, AB, BB). Type A curves were obtained for all milks. Maximum heat stability was affected by the κ-casein genotype (AB > AA, P < 0·01) but the influence of the β-lactoglobulin genotype was only significant when the κ-casein AA genotype was present (β-lactoglobulin AA > BB, P < 0·0001). Minimum heat stability was significantly higher (P < 0·0001) for milk genotyped κ-casein AB:β-lactoglobulin BB. The effects of milk genotyped κ-casein BB on maximum and minimum heat stability were determined by analysing individual milks: κ-casein BB:β-lactoglobulin AB (n=8) and reconstituted milks: κ-casein BB:β-lactoglobulin AA, AB and BB (n = 17). Type B curves were obtained on three occasions for individual κ-casein BB:β-lactoglobulin AB milk and on five occasions in the case of reconstituted milks with κ-casein BB:β-lactoglobulin AA, AB and BB. This suggests a relationship between the type B curve and the κ-casein B genetic variant. Comparison of the mean values of heat stability at the pH of maximum heat stability of each individual and reconstituted milk genotype suggested that the best genotype for κ-casein in terms of heat stability was BB.

The antiproteolytic activity of selected proteinase inhibitors was studied in mastitic bovine milk and urokinase-activated normal milk using a caseolytic agar diffusion assay. The inhibition profiles of mastitic milk and urokinase-activated milk were compared with those of purified proteinases. The proteinase inhibition profile of mastitic milk did not resemble that of any of the pure proteinases, indicating a mixed type of proteinase system in mastitic milk. The trypanocidals diminazene (equivalent to Berenil®) and pentamidine (equivalent to Lomidine®), together with aprotinin (Trasylol®), showed most promise when considering possible applications in mastitis to break up the proteolytic cascade within the inflamed udder.

Heat-induced gelation of milk was studied using both rheological and structural techniques. The sample was a conventional skim milk, concentrated with an ultrafiltration membrane, which formed gels when heated at appropriate pH. We investigated some factors that are considered to affect the gelation, such as concentration, pH and rennet treatment. The gelation process was monitored with a high precision oscillatory shear rheometer and the structure of gels was evaluated with quasi-elastic laser light scattering. From these results the gelation and phase separation behaviour were determined. By combining the results for different concentrations a phase diagram was obtained, which indicated that skim milk had a two-phase region on the higher temperature side. The effects of pH and rennet treatment were also evaluated with the aid of this phase diagram. The results were discussed on the basis of concepts of the phase behaviour of polymers, which were successfully developed in polymer physics.

The flux patterns for sweet whey and buttermilk were strongly influenced by pH. Increasing the pH of buttermilk from 6·6 to 8·0 tended to reduce initial flux values and reduce deposit formation on the membrane as indicated by lower values for the fouling coefficient. Flux was mostly controlled by concentration polarization. Reducing the pH below 6·6 increased the flux but caused more deposit on the membrane as indicated by high fouling coefficient values. Almost the opposite was found for sweet whey. The initial flux increased as pH increased, followed by considerable flux decline, which was linked to greater fouling of the membrane. Reducing the pH reduced the initial flux considerably but also reduced further fouling. These changes are thought to be brought about by the combined effects of pH change on the proteins and minerals, and in particular on calcium.

The cleavage of bovine κ-casein at the Phe105–Met106* bond by chymosin or pepsin is the first stage in casein micelle coagulation and casein digestion. The nature of the interaction of the peptide His98–Pro–His–Pro–His–Leu–Ser–Phe105–Met–Ala–Ile–Pro-Pro-Lys111 with chymosin and porcine pepsin was investigated using molecular modelling and energy minimization techniques. This study verified and extended a proposed model that electrostatic binding (involving His98, His100, His102 and Lys111 or Lys112) at either end of the active site cleft of chymosin is important for the positioning of residues 103–108 in the cleft. The peptide conformation remained unchanged in going from solution to binding into the active site cleft, with the exception that optimum binding of substrate to chymosin required the isomerization of the His98–Pro99 peptide bond from the trans to the cis conformation. The study also identified an acidic region in porcine pepsin that is in a position to form strong electrostatic interactions with the histidines at the N-terminus of the peptide.

A comprehensive description of the health status of 2406 cows in 28 herds was used to study complex interrelationships between Staphylococcus aureus occurrence and potential risk factors at the cow level over a broad spectrum of production conditions. The objective of the study was to characterize those cows shedding Staph, aureus in milk and to provide indications of possible causal relations that should be further explored. The estimates are expected to aid in providing valid and precise evaluations of results from cross sectional microbiological examination of bovine (quarter) milk samples. Staph, aureus were isolated at microbiological examination of quarter milk from 5·2, 13·5, 17·9 and 22·0% of 1552, 490, 296 and 277 cows in lactations 1, 2, 3 and 4–9 respectively. Simultaneous isolation of other mastitis pathogens was not significantly associated with occurrence of Staph, aureus. Of the positive samples, 95% had a positive California mastitis test reaction. Among the clinical measures of udder health, only visibly abnormal milk combined with normal udder tissue was associated with increased occurrence of Staph, aureus, although not consistently. Prior antibiotic treatment for udder disorders was not significantly associated with Staph, aureus isolation. The occurrence of sole ulcers in multiple digits was consistently associated with Staph, aureus in the first lactation. Milk yield was involved in interactions with other pathogens isolated, prior veterinary disease treatment, body weight and season of calving. Strong herd-year effects were revealed. This, combined with the interactions, indicated that some herd-specific factors were major determinants of Staph, aureus occurrence. The study indicated which cow characteristics should be taken into account in epidemiological analyses and causal interpretations of data from cross sectional microbiological examinations of dairy herds.

Clarification of whey by microfiltration (MF) can be achieved after appropriate pretreatment of the feed. A control pretreatment consists of a physicochemical process comprising increased ionic calcium and pH accompanied by heat (50 °C, 15 min) to cause aggregation of complex lipid–calcium phosphate particles, which are then separated by MF. This pretreatment process was modified by increasing the temperature to 55 °C and by maintaining the pH constant during heat treatment. This modification resulted in larger calcium phosphate particles and a lower content of soluble calcium and phosphate ions. As a consequence, a longer period of MF operation, better whey clarification and lower calcium and phosphate content of the filtrate were achieved. This suggests that a loosely structured deposit was formed on the membrane surface which was less resistant to filtration than that resulting from the control pretreatment. During MF, it was necessary to avoid zones of high shear in the retentate compartment that might cause physical alteration of the aggregates.

A total of 169 Lactobacillus strains from 12 species (Lb. acidophilus, Lb. brevis, Lb. buchneri, Lb. casei, Lb. delbrueckii subsp. bulgaricus, Lb. delbrueckii subsp. delbrueckii, Lb. delbrueckii subsp. lactis, Lb. fermentum, Lb. helveticus, Lb. paracasei subsp. paracasei, Lb. plantarum and Lb. rhamnosus), isolated from raw milk and various milk products, and 9 Lactococcus lactis strains were evaluated for peptidase activities with five chromogenic substrates and a tryptic digest of casein. Within each species, the peptidase activity of the cell-free extracts of the strains varied. Furthermore, differences were observed between the Lactobacilhis species and Lc. lactis. Lb. helveticus had by far the highest hydrolysing activities towards all substrates, indicating the presence of powerful aminopeptidases, X-prolyl-dipeptidyl aminopeptidases and proline iminopeptidases. Lb. delbrueckii subsp. bulgaricus possessed high hydrolysing activities towards substrates containing proline, alanylprolyl–p–nitroanilide and prolyl–p–nitroanilide. On the other hand, Lb. fermentum and Lb. brevis could be considered as weakly proteolytic species. A more detailed study with highly proteolytic Lactobacillus strains indicated that at least three different proteinases or endopeptidases were present. Compared with Lc. lactis, the Lactobacillus strains had a much lower hydrolytic action on glutamyl-glutamic acid, suggesting that glutamyl aminopeptidase was absent in lactobacilli.

The primary structures of ovine α>s1-casein variants A, C and D (formerly called Welsh variant) were determined. Separation of variants from whole casein was achieved using a fast and reliable reversed-phase HPLC method. Extended structural characterization of the purified proteins using electrospray mass spectrometry, automated Edman degradation and peptide mapping by means of HPLC-fast atom bombardment-mass spectrometry demonstrated that the mature protein was a mixture of two molecular species that differed in the deletion of residues 141–148 and were therefore 199 and 191 residues long respectively. The 199 residue peptide chain, which accounted for ∼ 80% of the entire translated αsl-casein, was as long as its caprine and bovine counterparts, and had a 98 and 89 % degree of identity with those two proteins respectively. Nine serine residues (positions 12, 44, 46, 64 to 68 and 75) were fully phosphorylated in αsl-casein A, whereas Ser115 and Ser41 were phosphorylated by ∼ 50 and ∼ 20% respectively. The differences between the three genetic variants A, C and D were simple silent substitutions, which however involved the degree to which the protein was phosphorylated. Variant C differed from variant A in the substitution Ser13 →> Pro13 which determined the loss of the phosphate group on site 12 of the protein chain, SerP12→>Ser12. A further substitution, SerP68 →> Asn68 caused the disappearance of both phosphate groups in the phosphorylated residues Ser64 and Ser66 in variant D; in this last casein variant there was no evidence of phosphorylation at Ser41.

Cell extracts of various lactobacilli and two Lactococcus strains were investigated for their immunoresponse with monoclonal and polyclonal antibodies raised against various proteolytic enzymes from Lc. lactis. Except for Lactobacillus casei SBT 2233, none of the lactobacilli proteins showed immunoresponse with the monoclonal antibodies. With polyclonal antibodies raised against aminopeptidases N and C and endopeptidase of Lc. lactis an immunoresponse was observed. However, the molecular masses of the reactive bands on the blot were considerably different from those of the corresponding lactococcal peptidases, except for the band that reacted with polyclonal antibodies against aminopeptidase C. The polyclonal antibodies raised against X-prolyl-dipeptidyl aminopeptidase and tripeptidase did not show any immunoreaction. As a control, all antibodies reacted with the lactococcal proteins on the blot, with molecular masses corresponding to those reported for the proteinases and peptidases. The results clearly showed that most of the proteolytic enzymes of lactobacilli were immunologically different from those of lactococci. The proteolytic enzymes in the cell-free extracts were separated by non-denaturing PAGE and visualized by zymogram staining. The electrophoretic pattern of the proteolytic enzymes of lactobacilli was different from that of Lc. lactis. Both experiments indicate that the enzymes of the proteolytic system of lactobacilli are different from those of lactococci.

The effects of pH and added NaCl or CaCl2 on gel development in heated whey protein concentrate (WPC) solutions were studied by oscillatory rheometry using a Bohlin rheometer, and by qualitative visual and tactile observations. The storage modulus (G′) exhibited maxima at pH values of about 4 and 7, while the phase angle (δ) exhibited minima at the same pH values. At pH 5·5, which lies within the range of the isoelectric points of the whey proteins, G′ was a minimum while δ was a maximum. At pH 7 and 8 G′ increased to a maximum and then decreased as added ionic strength was increased. At pH 4 G′ decreased steadily but slowly with increasing added ionic strength. The effects of CaCl2 were essentially the same as those of NaCl, but occurred at much lower added ionic strength. All the results are discussed in physicochemical terms. The main conclusions were, firstly, that changes in pH and salt concentration both affected gel development by altering the balance between attractive and repulsive protein-protein interactions, and secondly, that the effects of pH, salt concentration and salt type were interactive. Interaction both partly explains the wide variation in gel character and rheological properties among commercial WPC and potentially provides means of tailoring the gelling behaviour of a WPC so that it functions in a desired way in a given food system.

Comparison of the primary sequences of bovine, ovine and caprine αs1-casein shows a deletion of eight amino acid residues in the ovine casein region 141–148, which is identical in the bovine and caprine proteins except for a single difference in position 148 (Q or E). Polyclonal antibodies raised in rabbits against the bovine casein sequence 140–149 (QELAYFYPEL) appeared monospecific for bovine αsl-casein, since no antibody-antigen complex was formed with homologous ovine or caprine proteins. These antibodies remained unable to recognize the caprine sequence in the native protein even after extensive tryptic proteolysis. The lack of immunoreactivity of the antibodies against synthetic caprine αsl-casein peptide 138–149 (VNQELAYFYPQL) suggested that the glutamic acid residue in position 148 is essential for the antigenic character of the bovine peptide. From these observations, the use of these antibodies for the detection and quantitation of bovine milk present in ovine dairy products could be extended to caprine products.

In order to predict the oozing susceptibility of heat-induced milk protein gels such as processed cheeses during storage, ultracentrifugation forces are required to accelerate the expulsion of water from the gel structure. Two predictive methods have been tested. Direct ultracentrifugation of processed cheese was used to study effects of centrifugal compression on the water-holding ability. The water release with optimal parameters (time, temperature and ultracentrifugation force) from a freshly manufactured processed cheese was correlated with visual assessment of oozing after 6 months storage. A second method was based on the susceptibility of gelled proteins to resist the dissociating action of solutions of chemical agents such as SDS, urea, EDTA and 2-mercaptoethanol. Most of the protein sedimented by ultracentrifugation in the presence of SDS represented that still in complexed form. This correlated with optimal water binding by the processed cheese. Response-surface methods, used to optimize the ultracentrifugation variables and the composition of the dissociation solutions, showed that the dissociation test was best with a dispersion of processed cheese in SDS solution (10 g/l) at a ratio of 1:6 (w/v) at 20°C, ultracentrifuged at 86000 g and 20°C for 25 min. Analysis of processed cheese cooked at 115°C and then dissociated in different solutions showed various protein–protein interactions in the gel network, whereas hydrophobic interactions were the most important stabilizers of the protein matrix of cheese cooked at higher temperature.

The whole N fraction of six samples of hard and semi-hard pressed cheeses was analysed using PAGE, polyacrylamide gel isoelectric focusing and immuno blotting with polyclonal antibodies against β- and αs1-casein. The origin of some electrophoretic bands corresponding to peptides produced from the enzymic degradation of the casein fractions was established. A number of these peptides were also present in the in vitro hydrolysates of casein with plasmin and chymosin. Thus, it was also possible to determine which casein was the source of each peptide and which enzymes were active in cheese. Compared with the traditional Coomassie staining procedures, immunoblotting is more sensitive and specific, making the interpretation of each electrophoretic profile easy. Thus, it was also possible to obtain a clear picture of the state of each casein fraction in a cheese variety. Two main peptides were isolated from the pH 4·6-insoluble N fraction of Parmigiano-Reggiano using DEAE-cellulose chromatography and identified, from the amino acid sequence of the N- and C-terminal ends, as γ3-casein (β-casein(f108–209)) and αs1-PL1 (αs1-casein(f80–199)). In both cases, a Lys–X bond was hydrolysed, indicating the action of a trypsin-like enzyme in β- and αs1-casein hydrolysis during the ripening of this variety of hard pressed cheese.

The heterogeneity of ovine β-casein, apparent after discontinuous gel electrophoresis at alkaline pH and gel isoelectric focusing either by staining with Coomassie blue or immunoblotting, was investigated in detail. Electrospray mass spectrometry was used to characterize ovine casein by direct examination of whole casein samples. β-Casein in individual milks consisted of several components which were characterized through the accurate determination of their molecular masses. It was demonstrated that protein species occurring simultaneously in individual ovine milks differed mainly in the number of phosphate groups.

Two polysaccharides produced concurrently by Lactococcus lactis subsp. cremoris strain LC330 have been identified. One had a high molecular mass (> 1 × 106 Da) and was neutral. The second was smaller (∼ 10000 Da), charged and had a high phosphorus content. Sugar composition also differed. In chemostat culture the neutral polysaccharide was influenced by temperature and by nitrogen limitation. This polysaccharide was branched with terminal galactose moieties and contained galactose, glucose and glucosamine. The phosphopolysaccharide was more complex with glucose, rhamnose, galactose and glucosamine in an approximate ratio of 6:5:4:1.

The effects of pH and added salts or chelating agents on the gel strength and dynamic rheological properties of acid–heat-induced gels made from reconstituted skim milk (200 g solids/l) were investigated. Gel strength increased as pH was lowered between 5·75 and 5·15 except in the range pH 5·45–5·25 where a local maximum in gel strength was obtained at pH 5·35. Gel characteristics were affected by addition of salts or chelating agents but each of their effects was different, depending on the final pH of the milk gel. The addition of CaCl2 or chelating agents (Na2HPO4, disodium citrate or the disodium salt of EDTA) which affected micellar calcium phosphate, non-sedimentable casein and Ca2+ activity in different ways all resulted in decreased gel strength at pH 5·5. The addition of CaCl2 or MgCl2 caused a decrease in tanδ (ratio of the viscous modulus G″ to the elastic modulus G′) whereas disodium citrate or the disodium salt of EDTA addition caused an increase and Na2HPO4 addition did not cause a change. The addition of NaCl (up to 50 mM), which causes an increase in ionic strength but has no effect on non-sedimentable casein and Ca2+ activity, decreased gel strength but did not change tanδ. The addition of a range of other salts (KCl, NH4Cl, NaSCN, NaNO3 or Na2SO4) also decreased gel strength at pH 5·5.

Different electrophoretic techniques have been applied to the study of ovine whey protein polymorphism. Isoelectric focusing (IEF) over the pH gradient 3·5–9·5 was a suitable method for resolving both genetic variants of ovine β-lactoglobulin (β-lg); using this type of IEF, ovine β-lg A seemed to be defined by a major and a minor ‘satellite’ band. Approximate isoelectric point, relative molecular mass, amino acid composition and N-terminal sequence of this minor band were determined. A good separation of ovine β-lg was achieved using ultrathin layer isoelectric focusing (UTLIEF) over the pH gradient 2·5–7·0. However, addition of urea in UTLIEF gels led to some differences in the patterns of the ovine whey proteins when compared with those obtained by IEF on gels not containing urea. Two dimensional electrophoretic techniques provided a considerable separation of ovine whey proteins. Immunoblotting applied to a two dimensional separation permitted the identification of the bands belonging to β-lg and α-lactalbumin fractions.

The water-soluble extract of Cheddar cheese was fractionated by diafiltration using 10 kDa cut-off membranes. Peptides were isolated from the diafiltrate retentate by chromatography on DEAE-cellulose with a linear NaCl gradient in 50 mM-Tris-HCl, pH 8·6, and reversed-phase HPLC or electroblotting from urea-PAGE gels. Peptides were identified by determining N-terminal amino acid sequences and mass spectrometry. Most (45) of the total 51 peptides identified in the diafiltrate retentate originated from β-casein, especially from a short region in the N-terminal half of the molecule. Only six peptides originated from αs1-casein; β-lactoglobulin was also identified in the retentate. The origin of most of these peptides could be explained on the basis of known specificities of lactococcal cell envelope proteinases.

A method for isolation, detection and quantification of cholesterol oxidation products based on solid phase extraction in combination with preparative HPLC and gas chromatography–mass spectrometry selected ion monitoring has been developed for dairy products. The isolation procedure had a high recovery and artifact formation was minimal, as shown by isotope labelling. The limits of detection ranged from 0.3 to 35 pg/μl of the isomeric forms of 7-hydroxycholesterol, 20α-hydroxycholesterol, the isomeric forms of cholesterol-5,6-epoxides, cholestanetriol, 25-hydroxycholesterol and 7-ketocholesterol corresponding to a limit of quantification of 2–6 ng oxysterol/g lipid in the dairy product, depending on the nature of the cholesterol oxidation product.